1
00:00:04,960 --> 00:00:10,300
DR. ANGELA ODOMS-YOUNG: Hello and welcome
to day two of the NIH Food Insecurity, Neighborhood

2
00:00:10,300 --> 00:00:16,550
Food Environment, and Nutrition Health Disparities:
State of the Science Workshop. Yesterday,

3
00:00:16,550 --> 00:00:23,070
we had several wonderful presentations that
focused on food insecurity innovations and

4
00:00:23,070 --> 00:00:30,529
strategies to inform policy and practice to
specifically prevent diet-related health disparities,

5
00:00:30,529 --> 00:00:37,760
as well as promote health equity. And I just
want to do a brief recap. I'm Angela Odoms-Young.

6
00:00:37,760 --> 00:00:42,430
I'm one of the co-chairs of the workshop.
I'm an associate professor in the Division

7
00:00:42,430 --> 00:00:51,170
of Nutritional Sciences at Cornell University.
So, we had, as I mentioned, a series of wonderful

8
00:00:51,170 --> 00:00:57,190
presentations that really focused in on the
causes and consequences of food insecurity

9
00:00:57,190 --> 00:01:03,840
and strategies and solutions to address food
insecurity. So, just to hit the highlights,

10
00:01:03,840 --> 00:01:12,210
several presentations focused on the multiple
level…multi-level 

11
00:01:12,210 --> 00:01:18,550
determinants that impact food insecurity and
the relationship between food insecurity and

12
00:01:18,550 --> 00:01:25,090
health. The presentations highlighted the
limited research and data that we have on

13
00:01:25,090 --> 00:01:30,310
certain populations that are at risk of food
insecurity, such as Indigenous populations,

14
00:01:30,310 --> 00:01:38,870
young adults that are not in college.
Several of the presentations explored methodological

15
00:01:38,870 --> 00:01:44,760
challenges and how we need to look at multiple
methodologies, including qualitative methods,

16
00:01:44,760 --> 00:01:50,960
as well as policy studies. And then, it was
highlighted by several presenters that we

17
00:01:50,960 --> 00:01:59,329
need to further define and identify measures
and metrics related to this concept of nutrition

18
00:01:59,329 --> 00:02:07,640
security or nutrition insecurity. In the
latter part of the session, we focused in

19
00:02:07,640 --> 00:02:14,120
on strategies and solutions to address food
insecurity. And there was a key point about

20
00:02:14,120 --> 00:02:22,209
highlighting the lived experience and really
barriers and facilitators that people experience,

21
00:02:22,209 --> 00:02:27,380
both individuals and families experience, to
food assistance program participation. There

22
00:02:27,380 --> 00:02:33,550
was a discussion about the impact of restrictions
and incentives in federal food assistance

23
00:02:33,550 --> 00:02:39,129
programs and how those relate to diet and
health outcomes and address food insecurity.

24
00:02:39,129 --> 00:02:46,159
Implications of the broader policy context,
not just the food policy context, but also

25
00:02:46,159 --> 00:02:52,690
broader economic policy context that exists
and that may impact solutions or strategies

26
00:02:52,690 --> 00:02:58,260
when we want to address food insecurity.
Again, methodological approaches to understand

27
00:02:58,260 --> 00:03:04,920
program implementation and associated outcomes.
And then what combination of intervention

28
00:03:04,920 --> 00:03:10,840
strategies, particularly in health care settings,
can produce reductions or result in reductions

29
00:03:10,840 --> 00:03:17,480
in food insecurity, but also better diets
and quality of life. There was some challenges

30
00:03:17,480 --> 00:03:22,110
that were highlighted about the heterogeneity
in program implementation that can make it

31
00:03:22,110 --> 00:03:28,650
difficult to evaluate and look at the literature
overall and understand what's happening when

32
00:03:28,650 --> 00:03:33,830
it comes to the impact of these interventions
on food insecurity. Other challenges include

33
00:03:33,830 --> 00:03:42,500
a short-term funding or changes in priority
populations. And then, one area for expansion

34
00:03:42,500 --> 00:03:48,931
related to understanding the impact of structural
interventions and moving more upstream, but

35
00:03:48,931 --> 00:03:56,269
still considering downstream outcomes. And
then lastly, we had a closing speaker that

36
00:03:56,269 --> 00:04:04,769
emphasized the need to decolonize methodologies
and to use theories, like Indigenous theories,

37
00:04:04,769 --> 00:04:10,599
and ways of knowing critical race theory and
using an intersectionality lens to understand

38
00:04:10,599 --> 00:04:15,570
and address food insecurity.
And we had, that was the emphasis of our closing

39
00:04:15,570 --> 00:04:23,540
speaker, but that was highlighted by several
speakers throughout. So, throughout the workshop

40
00:04:23,540 --> 00:04:29,100
today we're, as I mentioned, we're going to
focus on neighborhood food environments. And

41
00:04:29,100 --> 00:04:35,820
you will have the opportunity to address questions
to our speakers and panelists. You can simply

42
00:04:35,820 --> 00:04:41,479
put your questions in the chat box, and please
don't be shy. We want to hear from you. I

43
00:04:41,479 --> 00:04:46,590
know it's, a virtual platform, but we want
to make sure is as interactive as possible.

44
00:04:46,590 --> 00:04:54,320
Now I have the pleasure of introducing next,
Dr. Shannon Zenk, who's the Director of the

45
00:04:54,320 --> 00:05:00,760
National Institute of Nursing Research, who
will provide us with some opening remarks.

46
00:05:00,760 --> 00:05:02,320
And so, Shannon, I'm going to turn it over
to you.

47
00:05:02,320 --> 00:05:08,900
DR. SHANNON ZENK: Thank you. Thank you, Angela.
Good afternoon, everyone. I'm Shannon Zenk,

48
00:05:08,900 --> 00:05:16,190
Director of the National Institute of Nursing
Research. It's so nice to join my longtime

49
00:05:16,190 --> 00:05:22,810
Chicago colleague, Dr. Angela Odoms-Young
and welcoming you day two of the Food Insecurity,

50
00:05:22,810 --> 00:05:27,319
Neighborhood Food Environment, and Nutrition
Health Disparities: State of the Science Workshop.

51
00:05:27,319 --> 00:05:33,410
And I'm so thrilled that NINR co-sponsored
this event with our NIH colleagues, and thank

52
00:05:33,410 --> 00:05:38,560
you to our partners at the CDC and at the
Department of Agriculture for your contributions

53
00:05:38,560 --> 00:05:44,940
as well. Before I turn things over to Dr.
Odoms-Young and Dr. Glanz, I'd like to share

54
00:05:44,940 --> 00:05:50,889
a little bit about why I think this area of
research is so important. Next slide. So,

55
00:05:50,889 --> 00:05:56,940
I became interested in the health impacts
of neighborhood environments while practicing

56
00:05:56,940 --> 00:06:01,630
as a nurse. One of my first jobs as a nurse
was as a home health care case manager. And

57
00:06:01,630 --> 00:06:05,610
in this role, I visited the people in their
homes, typically after a hospitalization,

58
00:06:05,610 --> 00:06:12,190
to help with pain management, provide wound
care, and teach about how to manage chronic

59
00:06:12,190 --> 00:06:16,130
conditions.
But spending time in people's homes and in

60
00:06:16,130 --> 00:06:20,261
different neighborhoods, I was struck by the
tremendous differences in the environments

61
00:06:20,261 --> 00:06:26,110
of patients in my caseload, both in terms
of privilege and poverty. I found it difficult

62
00:06:26,110 --> 00:06:31,789
to talk to some patients about healthy eating,
for example, when what they really needed

63
00:06:31,789 --> 00:06:36,360
to restore their health was far more fundamental:
decent stable housing, a safe environment,

64
00:06:36,360 --> 00:06:43,770
access to affordable healthy foods nearby.
Next slide, please. These observations inspired

65
00:06:43,770 --> 00:06:52,020
me to go back to school to learn more and
contribute to solutions. Ultimately, colleagues

66
00:06:52,020 --> 00:06:58,509
and I conducted some of the early studies
in Detroit on food deserts. And we know that

67
00:06:58,509 --> 00:07:03,810
many of the inequities in neighborhood food
environments that we see can be traced to

68
00:07:03,810 --> 00:07:09,509
redlining and disinvestment from communities
of color. Observational and intervention research

69
00:07:09,509 --> 00:07:15,810
on neighborhood food environments, diet, and
health have grown tremendously and evolved

70
00:07:15,810 --> 00:07:24,729
over the past 15 to 20 years due to work
of so many participating in this meeting. It's

71
00:07:24,729 --> 00:07:25,870
a perfect time to reflect on the state of
the science, what's been learned,

72
00:07:25,870 --> 00:07:33,530
what are the gaps and opportunities, so that
the most impactful science can move forward.

73
00:07:33,530 --> 00:07:38,280
Next slide, please. We know that there are
not only inequities in the neighborhood food

74
00:07:38,280 --> 00:07:43,550
environment, but in food and nutrition security
as well. The economic crisis that accompanied

75
00:07:43,550 --> 00:07:50,710
the COVID-19 pandemic led to spiking rates
of food insecurity as jobs were lost and schools

76
00:07:50,710 --> 00:07:58,479
closed. People of color were among those hardest
hit by this food crisis. And the levels of

77
00:07:58,479 --> 00:08:04,460
food insecurity are down from their peak during
the pandemic. We're not out of the woods yet.

78
00:08:04,460 --> 00:08:09,110
We also need to acknowledge that the economic
crisis we faced in 2020 is only the latest

79
00:08:09,110 --> 00:08:16,820
economic crisis faced by many Americans. The
racial wealth gap in our nation is well established,

80
00:08:16,820 --> 00:08:24,130
with white household wealth at 10 times that
of black households and seven times that of

81
00:08:24,130 --> 00:08:30,500
Hispanic households. Not only does greater
wealth allows some to weather storms like

82
00:08:30,500 --> 00:08:36,510
COVID-19, but greater wealth also comes with
better access to health care, education, child

83
00:08:36,510 --> 00:08:43,719
care, and any number of resources in our society.
While we need research to solve pressing problems

84
00:08:43,719 --> 00:08:49,860
like food insecurity, we can't ignore the
impact of policies that have perpetuated such

85
00:08:49,860 --> 00:08:57,250
fundamental inequalities in our nation. Next
slide, please. So, then how can we right the

86
00:08:57,250 --> 00:09:03,700
historic and contemporary wrongs of racism
that have produced and perpetuated persistent

87
00:09:03,700 --> 00:09:09,070
health disparities, including nutrition health
disparities? It's clear that we urgently need

88
00:09:09,070 --> 00:09:15,640
practice and policy solutions for the challenging
health problems we face and the disparities.

89
00:09:15,640 --> 00:09:22,070
There is now a broader recognition of the
many ways that racial and social injustice

90
00:09:22,070 --> 00:09:28,470
have significant impacts on individual and
community health, and the public is demanding

91
00:09:28,470 --> 00:09:34,560
change. But in this landscape and to inform
action, we need scientific evidence on what

92
00:09:34,560 --> 00:09:40,970
policies and interventions are effective in
increasing resources and improving the environment

93
00:09:40,970 --> 00:09:46,790
in low income and communities of color, and
improving health behaviors and health outcomes

94
00:09:46,790 --> 00:09:51,970
in these communities. That's exactly why we're
holding this event and why thousands of you

95
00:09:51,970 --> 00:09:58,410
registered to attend.
Next slide, please. At NINR we intend to do

96
00:09:58,410 --> 00:10:07,250
our part. A major focus of our current research
is prevention, and a third of NINR’s funding

97
00:10:07,250 --> 00:10:12,880
already focuses on research to eliminate health
disparities. As we look forward, we want to

98
00:10:12,880 --> 00:10:17,890
ensure that our research tackles the nation's
most pressing health challenges and discover

99
00:10:17,890 --> 00:10:24,450
solutions in the wide variety of settings
where nurses practice not only in hospitals

100
00:10:24,450 --> 00:10:30,839
and clinics, but also in people's homes, schools,
worksites, justice settings, and communities.

101
00:10:30,839 --> 00:10:38,690
We want to ensure our research applies a holistic
approach that includes addressing social determinants

102
00:10:38,690 --> 00:10:45,050
of health. And we want to advance health equity.
Next slide, please. Today, I'll share with

103
00:10:45,050 --> 00:10:49,550
you just one example of NINR-supported research
that examines the health impact of the food

104
00:10:49,550 --> 00:10:56,300
environment and other social factors through
a health equity lens. NINR grantee Dr. Suzan

105
00:10:56,300 --> 00:11:02,839
Carmichael studies racial and ethnic disparities
and severe maternal morbidity. Dr. Carmichael

106
00:11:02,839 --> 00:11:10,470
has found severe maternal mortality is increasing
among all racial and ethnic groups studied,

107
00:11:10,470 --> 00:11:14,149
with the greatest increase among non-Hispanic
black women.

108
00:11:14,149 --> 00:11:20,990
She found that pre-pregnancy BMI and gestational
weight gain are related to maternal morbidity.

109
00:11:20,990 --> 00:11:26,519
And her latest NINR-funded study is examining
the impact of neighborhood resources like

110
00:11:26,519 --> 00:11:33,440
walkability, green space, and the food environment
on severe maternal morbidity. I'm really proud

111
00:11:33,440 --> 00:11:39,710
that NINR supports work like this, which is
essential in determining what is driving such

112
00:11:39,710 --> 00:11:46,620
disparities so that policy solutions can be
developed. Before I turn the meeting back

113
00:11:46,620 --> 00:11:54,290
over to Dr. Odoms-Young, I want to invite
you to share your thoughts on NINR’s next

114
00:11:54,290 --> 00:11:59,120
strategic plan. We'll soon post our strategic
plan framework for public comment through

115
00:11:59,120 --> 00:12:03,420
an RFI. We'll publicize this widely, and I
hope you'll provide us with your comments

116
00:12:03,420 --> 00:12:07,580
based on your very important perspective on
economic and environmental factors, nutrition,

117
00:12:07,580 --> 00:12:16,649
and health equity. So, thanks again for inviting
me to help open the session. I also want to

118
00:12:16,649 --> 00:12:24,209
thank our workshop co-chairs, Dr. Karen Glanz
and Dr. Angela Odoms-Young; the workshop organizers,

119
00:12:24,209 --> 00:12:29,560
Dr. Alison Brown and Dr.
Tanya Agurs-Collins; as well as everyone who

120
00:12:29,560 --> 00:12:39,959
worked with them to put together this outstanding
program. Please enjoy today's session. Thanks again.

121
00:12:39,959 --> 00:12:44,930
DR. ANGELA ODOMS-YOUNG: Thank you so much,
Shannon, and thank you for your thoughtful

122
00:12:44,930 --> 00:12:51,520
comments. I have the pleasure...again, I'm Angela
Odoms-Young, faculty member at Cornell University

123
00:12:51,520 --> 00:12:58,860
and co-chair for the workshop. I have the
pleasure of introducing Session 4, where

124
00:12:58,860 --> 00:13:06,750
we will focus on providing an Overview and
Measurement of the Neighborhood Food Environment.

125
00:13:06,750 --> 00:13:11,180
We're going to kick off this session with
a State of the Science, Neighborhood Food

126
00:13:11,180 --> 00:13:17,230
Environment and How It Influences Health by
Dr. Karen Glanz, a faculty member at University

127
00:13:17,230 --> 00:13:24,240
of Pennsylvania and then also co-chair of
this NIH workshop. The objective of this session

128
00:13:24,240 --> 00:13:29,360
is to discuss the current state of the science
and disparities across food environments,

129
00:13:29,360 --> 00:13:35,170
measurement, and the association of food environments
with diet, obesity, and health. And so I'm

130
00:13:35,170 --> 00:13:42,680
going to turn it over to you, Karen.

131
00:13:42,680 --> 00:13:49,839
DR. KAREN GLANZ: Good afternoon. Karen Glanz,
I'm back. Happy to provide an Overview on

132
00:13:49,839 --> 00:13:54,130
Neighborhood Food Environments and Health
as we kick off this part of the workshop.

133
00:13:54,130 --> 00:14:00,709
Going to briefly talk about some of the driving
forces for focusing research and practice

134
00:14:00,709 --> 00:14:06,390
on neighborhood food environments, the complexity
of food environments. Talk about a simple

135
00:14:06,390 --> 00:14:14,310
model and what's changed, how COVID-19 pandemic
is an inflection point in this area and give

136
00:14:14,310 --> 00:14:18,839
you a very brief overview of the state of
the science in observing, changing, and measuring

137
00:14:18,839 --> 00:14:24,500
neighborhood food environments. And I'll finish
with some thoughts about research gaps and

138
00:14:24,500 --> 00:14:32,310
the road forward in this area. The idea that
environments affect behavior is not new, in

139
00:14:32,310 --> 00:14:41,980
fact, it's more than a century old in modern
thinking, as B.F. Skinner developed the Skinner

140
00:14:41,980 --> 00:14:51,180
Box to see if he could get a rat to push a
lever by sounding speaker, signaling lights,

141
00:14:51,180 --> 00:14:58,300
or running the shock generator. And found
in many different ways that people could

142
00:14:58,300 --> 00:15:00,760
be conditioned to respond to their environments
as well.

143
00:15:00,760 --> 00:15:06,550
However, I don't think he was thinking about
the ecological perspective that's become so

144
00:15:06,550 --> 00:15:12,980
prominent in our public health thinking nowadays.
The idea that behavior is affected by and

145
00:15:12,980 --> 00:15:19,060
affects multiple levels of influence and that
there's reciprocal causation between individuals

146
00:15:19,060 --> 00:15:29,070
and our environments. In 2008, Mary Story,
Karen Kaphingst, Ramona Robinson-O'Brien and

147
00:15:29,070 --> 00:15:35,529
I published this ecological framework depicting
the multiple influences on what people eat

148
00:15:35,529 --> 00:15:40,760
in the Annual Review of Public Health. And,
to say the least, it's very complicated. It

149
00:15:40,760 --> 00:15:45,470
focuses on all these different levels and
many different environments. But let's see

150
00:15:45,470 --> 00:15:51,250
what happens if we take away the individual
level factors. Then the picture gets a bit

151
00:15:51,250 --> 00:15:57,610
less complex but still has a lot of components
that maybe we don't want to focus on if we're

152
00:15:57,610 --> 00:16:03,291
focusing on neighborhood environments. So,
let's take out the social environment, and

153
00:16:03,291 --> 00:16:07,639
we still have a pretty complicated picture.
But these are the things that we want to focus

154
00:16:07,639 --> 00:16:14,230
on in today's workshop. Neighborhoods, restaurants,
supermarkets, convenience stores, food pantries,

155
00:16:14,230 --> 00:16:18,310
and many of the factors that may be influencing
them.

156
00:16:18,310 --> 00:16:24,779
Two of the driving forces for understanding
and improving neighborhood food environments

157
00:16:24,779 --> 00:16:30,000
and the burgeoning research in this area—this started about two decades ago—included

158
00:16:30,000 --> 00:16:36,930
the recognition that the obesity and chronic
disease epidemics were not just determined

159
00:16:36,930 --> 00:16:43,300
by biology and psychology, but that multilevel
determinants, as you've just seen, have been

160
00:16:43,300 --> 00:16:50,100
recognized, and that our environments play
a big part in our diet, risk factors, and diseases.

161
00:16:50,100 --> 00:16:55,420
Secondly, there was a recognition of unequal
access to healthful food, disparities, and

162
00:16:55,420 --> 00:17:04,970
the need for more equity, that these also
are linked to health outcomes. A good depiction

163
00:17:04,970 --> 00:17:12,420
of the importance of neighborhoods as they
relate to food and to health outcomes comes

164
00:17:12,420 --> 00:17:20,679
in the form of some really provocative maps
from New York City and also from the Seattle

165
00:17:20,679 --> 00:17:27,049
King County area. And you can see by the variations
in shading in each of these maps that there's

166
00:17:27,049 --> 00:17:34,630
anywhere from a four to an eightfold difference
in rates of obesity in neighborhoods within

167
00:17:34,630 --> 00:17:39,240
these cities.
So, we're not talking about climate differences,

168
00:17:39,240 --> 00:17:44,289
we're not talking about geographic differences,
but really neighborhood-level differences.

169
00:17:44,289 --> 00:17:51,600
And New York has done a particularly good
job of documenting many of the health-related

170
00:17:51,600 --> 00:17:59,309
factors and determinants. And these maps at
the bottom focus on tobacco use, poverty rates,

171
00:17:59,309 --> 00:18:04,840
and the percentage of blacks living in these
different neighborhoods. And isn't it interesting

172
00:18:04,840 --> 00:18:11,460
that they also align pretty well with the
rates of obesity in New York neighborhoods?

173
00:18:11,460 --> 00:18:18,990
So, this is a simplified version of the food
environment hypothesis: that there are factors

174
00:18:18,990 --> 00:18:25,270
that influence the proximal food environments
in stores, restaurants, and other sources,

175
00:18:25,270 --> 00:18:31,600
including online sources, and that these affect
our food intake, risk factors, and disease

176
00:18:31,600 --> 00:18:39,650
profiles. This model of community nutrition
environments overlaps with the ones that you've

177
00:18:39,650 --> 00:18:46,280
already seen, but I wanted to show it because
it's particularly important in clarifying

178
00:18:46,280 --> 00:18:51,059
some distinctions about aspects of neighborhood
nutrition environments.

179
00:18:51,059 --> 00:18:56,679
We differentiate community nutrition environments
from consumer nutrition environments, both

180
00:18:56,679 --> 00:19:04,100
of which may affect organizational nutrition
environments. And specifically, those definitions

181
00:19:04,100 --> 00:19:09,570
are that community nutrition environments
are where people get their food, where they

182
00:19:09,570 --> 00:19:14,480
buy it, where it's prepared, etc., the type
and location of supermarkets, restaurants,

183
00:19:14,480 --> 00:19:22,020
and so on and their accessibility. Whereas
consumer nutrition environments reflect the

184
00:19:22,020 --> 00:19:27,390
availability of healthy food choices, pricing,
promotion, placement, and information availability.

185
00:19:27,390 --> 00:19:33,809
In other words, what is inside the stores and what
is inside the restaurants. That's particularly

186
00:19:33,809 --> 00:19:40,130
pertinent to this area of research because
it gives us two different levels to look at,

187
00:19:40,130 --> 00:19:45,610
two different levels at which interventions
may be possible and at which understanding

188
00:19:45,610 --> 00:19:52,090
occurs differentially. So, what's the biggest
source of food? What's changing? What has

189
00:19:52,090 --> 00:19:57,240
changed? Well, one thing that hasn't changed
is that supermarkets are still the number

190
00:19:57,240 --> 00:20:04,730
one source of retail food and provide the
largest share of household calories.

191
00:20:04,730 --> 00:20:11,050
Over 65 percent of our calories come from
supermarkets and grocery stores. However,

192
00:20:11,050 --> 00:20:16,860
over time, the links between retail and food
service have increasingly become blurred.

193
00:20:16,860 --> 00:20:22,720
And then non-supermarket sources have a large
role, particularly a larger role in disadvantaged

194
00:20:22,720 --> 00:20:28,059
communities, even though they're only this
small little other sliver, they're a growing

195
00:20:28,059 --> 00:20:34,840
sliver and an important sliver for the communities
that are most needy. And online ordering and

196
00:20:34,840 --> 00:20:42,049
purchasing and hybrids have grown, and particularly
during the COVID pandemic. The COVID pandemic

197
00:20:42,049 --> 00:20:46,880
is something of an inflection point for neighborhood
food environments. It may raise new issues

198
00:20:46,880 --> 00:20:53,870
that may be with us for years to come. People
are making fewer shopping trips. Online grocery

199
00:20:53,870 --> 00:20:59,880
ordering has jumped tremendously, yet still
most people do less than a quarter of their

200
00:20:59,880 --> 00:21:04,660
shopping online. There are hybrid models.
Food delivery and takeout from restaurants

201
00:21:04,660 --> 00:21:10,840
more than doubled, as people are eating at
home but they're taking out food from restaurants

202
00:21:10,840 --> 00:21:14,370
at a much higher pace than they were in the
past.

203
00:21:14,370 --> 00:21:21,120
And most recently, in 2021, we see supply chain
issues, some foods in short supply, prices

204
00:21:21,120 --> 00:21:27,730
going up, and shortages of some fresh foods
caused by extreme weather yet another influence

205
00:21:27,730 --> 00:21:34,870
on the food environment. So, what about food
environments and policies? How do they go

206
00:21:34,870 --> 00:21:40,340
together? Policies can shape environments,
things like sugary beverage taxes, catering

207
00:21:40,340 --> 00:21:46,279
policies, price supports and food assistance
policies all can really determine what's

208
00:21:46,279 --> 00:21:51,841
in the environment, where food sources are,
and so forth. But oftentimes, environments

209
00:21:51,841 --> 00:21:58,169
evolve in the absence of specific policies.
They may be market-driven decisions. They

210
00:21:58,169 --> 00:22:05,390
may be decisions made by city planners. They
may be made by school boards. And policies

211
00:22:05,390 --> 00:22:07,760
can be health promoting or not so health promoting.
So, bear that in mind as a backdrop to our

212
00:22:07,760 --> 00:22:17,190
understanding of the food environment. So,
a quick overview of some of the things that

213
00:22:17,190 --> 00:22:25,100
we know and how these have entered into the
consciousness of people interested in understanding

214
00:22:25,100 --> 00:22:32,070
and impacting food environments and health.
Early food environment research was mostly

215
00:22:32,070 --> 00:22:39,630
looking at associations between environments,
behavior, and risk factors. And the key findings

216
00:22:39,630 --> 00:22:46,179
were that supermarkets made for healthier
food environments, that people who shopped

217
00:22:46,179 --> 00:22:52,430
at supermarkets ate more fresh fruits and
vegetables, and that fast food restaurants

218
00:22:52,430 --> 00:23:00,169
were obesogenic, that the food was high in
calories, salt, and fat that people got at

219
00:23:00,169 --> 00:23:08,130
fast food restaurants. So, that was a big
focus. Two pluses and minuses. There was early

220
00:23:08,130 --> 00:23:13,779
recognition of disparities that there are
more fast food restaurants in minority neighborhoods,

221
00:23:13,779 --> 00:23:20,380
that some foods were less available, poorer
quality in minority and low-income areas,

222
00:23:20,380 --> 00:23:27,860
and that supermarkets were less accessible
in poor and black neighborhoods. All this

223
00:23:27,860 --> 00:23:33,510
research gave way to a lot more research.
And one of the best, most in-depth reviews

224
00:23:33,510 --> 00:23:41,110
is a systematic review conducted a few years
ago by Cobb and others. And what do you know?

225
00:23:41,110 --> 00:23:46,090
The most studied exposures were supermarket
availability and fast food, restaurant availability.

226
00:23:46,090 --> 00:23:55,800
But what about the findings? Did the findings
replicate what we learned earlier? Not really.

227
00:23:55,800 --> 00:24:01,640
Associations between the outlets and obesity
was mostly null, and indeed, the review found

228
00:24:01,640 --> 00:24:08,350
an overall low quality of studies, which is
rather concerning. And again looking at supermarket,

229
00:24:08,350 --> 00:24:14,330
fast food and small stores, some stores found
positive findings, some negative, but more

230
00:24:14,330 --> 00:24:21,490
than anything, the findings were null. Along
with all this research there is an increasing

231
00:24:21,490 --> 00:24:27,690
interest in how to measure food environments.
And we talk about measurement often at the

232
00:24:27,690 --> 00:24:31,750
community food environment, where are the
stores? And at the consumer food environment, what's

233
00:24:31,750 --> 00:24:38,850
in the stores? And the measures of what's
in retail store food environments has become

234
00:24:38,850 --> 00:24:45,670
of great interest and focus and a lot of focus
on availability of healthy foods and unhealthy

235
00:24:45,670 --> 00:24:52,110
foods, price, and quality. We increasingly
see reports of reliability and validity of

236
00:24:52,110 --> 00:24:58,980
measures and some intervention studies indications
of whether the measures can pick up change

237
00:24:58,980 --> 00:25:04,010
in the food environment.
The Nutrition Environment Measures Survey

238
00:25:04,010 --> 00:25:12,059
that I and my colleagues developed and first
published in 2007, has been crafted in such

239
00:25:12,059 --> 00:25:18,159
a way that many people could adapt it, use
it, translate it, and a whole suite of tools

240
00:25:18,159 --> 00:25:30,010
has been created. And up to now, 174 publications
have used the NEMS tools. This is a little

241
00:25:30,010 --> 00:25:35,029
social network analysis of the NEMS tools,
starting with the NEMS restaurant and NEMS

242
00:25:35,029 --> 00:25:43,240
store tools, and many spin-offs and many languages
and many environments.

243
00:25:43,240 --> 00:25:50,640
This is an indication of the eagerness for
systematic observational tools to measure

244
00:25:50,640 --> 00:25:56,529
the food environment. And also, this tool…these
tools that were originally developed for research

245
00:25:56,529 --> 00:26:01,500
can be and have been used for community assessment,
advocacy, and intervention. You'll hear more

246
00:26:01,500 --> 00:26:07,090
about the measurement of the food environment
and the consumer food environment from Alison

247
00:26:07,090 --> 00:26:14,990
Gustafson later on. So, what do we know about
interventions? We have a body of intervention

248
00:26:14,990 --> 00:26:23,890
research in supermarkets, small stores, and
restaurants. And we also have some whole-community

249
00:26:23,890 --> 00:26:30,530
interventions that have been conducted to
help us see whether we can improve food environments.

250
00:26:30,530 --> 00:26:38,020
So, the first question is, if it's helpful
to have a supermarket in your neighborhood.

251
00:26:38,020 --> 00:26:44,550
Does adding a new supermarket in a food desert
improve diet and reduce obesity? A couple

252
00:26:44,550 --> 00:26:50,029
of speakers will address this issue and talk
about the Healthy Food Financing Initiative.

253
00:26:50,029 --> 00:26:56,270
Two important published studies by Tamara
Dubowitz and Stephen Matthews have both found

254
00:26:56,270 --> 00:27:02,491
that with a new supermarket, there is increased
awareness of food access, but healthful food

255
00:27:02,491 --> 00:27:07,500
intake and lower BMI were not attributed to
the new stores, and that there are positive

256
00:27:07,500 --> 00:27:10,919
and negative changes in food availability
found.

257
00:27:10,919 --> 00:27:19,570
Perhaps not so surprisingly, Vaughn, Dubowitz,
and others published a paper reporting that

258
00:27:19,570 --> 00:27:24,980
residents of food deserts buy most of their
junk food at supermarkets. After all, it's

259
00:27:24,980 --> 00:27:28,490
probably cheaper in the supermarkets than
in other smaller stores. There's also been

260
00:27:28,490 --> 00:27:37,730
a fair bit of attention to interventions at
supermarkets or in supermarkets as being potentially

261
00:27:37,730 --> 00:27:44,840
efficacious for increasing purchase of healthy
foods. And these are just a few of the reviews

262
00:27:44,840 --> 00:27:50,779
that have been published on the topic. And
you'll be hearing from Allison Karpyn at the

263
00:27:50,779 --> 00:27:59,320
workshop also reporting on her review. We
conducted a study called Healthy Retail Solutions

264
00:27:59,320 --> 00:28:03,770
to use placement and promotion strategies
to try to increase sales of healthier products

265
00:28:03,770 --> 00:28:09,900
in supermarkets in low-income neighborhoods.
This is one example of a beverage cooler, where

266
00:28:09,900 --> 00:28:17,320
we made water more prominent and took the
higher calorie, higher sugar beverages and

267
00:28:17,320 --> 00:28:23,290
put them at the bottom. Results of that study
found that there were greater sales of skim,

268
00:28:23,290 --> 00:28:30,580
1 percent milk, water, and frozen entrees,
but no change in sales in some other food

269
00:28:30,580 --> 00:28:34,970
categories.
So, some changes and not others. We've scaled

270
00:28:34,970 --> 00:28:42,960
up the study and conducted it for a longer
period of time, and the results will be forthcoming

271
00:28:42,960 --> 00:28:50,640
soon. Small food store interventions have
been of particular interest in disadvantaged

272
00:28:50,640 --> 00:28:59,290
and minority neighborhoods, and there have
been a number of studies. A nice review was

273
00:28:59,290 --> 00:29:05,200
conducted and reported in 2012 by Joel Gittlelsohn
and others, and you'll hear more from him

274
00:29:05,200 --> 00:29:10,630
about small store interventions. Some of the
highlights of the findings were that stocking

275
00:29:10,630 --> 00:29:16,279
and availability of healthful foods tended
to increase. The manager said they sold more

276
00:29:16,279 --> 00:29:21,110
promoted or healthful foods, but that there
were limited data on purchase or impact on

277
00:29:21,110 --> 00:29:29,809
shopper consumption. And there are design
and measurement limitations to many of these

278
00:29:29,809 --> 00:29:35,299
studies. There were two large studies of urban
corner store initiatives conducted in Philadelphia

279
00:29:35,299 --> 00:29:43,789
about five, six years ago, and they did find,
similar to what other studies have found,

280
00:29:43,789 --> 00:29:51,320
that the food environment could be made healthier.
But that changes in purchases, which were extensively

281
00:29:51,320 --> 00:29:57,080
studied, did not change from baseline to follow-up
in this natural experiment.

282
00:29:57,080 --> 00:30:04,299
In a randomized trial of the Healthy Corner
Store Initiative around schools, Gary Foster

283
00:30:04,299 --> 00:30:10,450
and colleagues did not find any changes in
energy content per purchase and no changes

284
00:30:10,450 --> 00:30:17,710
in BMI Z-scores or obesity prevalence. So,
some food for thought about why that perhaps

285
00:30:17,710 --> 00:30:23,480
wasn't as successful as we would have hoped.
There have been a lot of restaurant interventions

286
00:30:23,480 --> 00:30:29,929
in recent years, in particular, those focusing
on calorie menu labeling as calorie menu labeling,

287
00:30:29,929 --> 00:30:37,010
policies have become more widespread. Some
of those studies have found reduced energy

288
00:30:37,010 --> 00:30:46,490
intake of orders, but others have found nonsignificant
effect or no effect. Industry response to

289
00:30:46,490 --> 00:30:52,919
the national requirement of calorie menu labeling
in chain restaurants has shown an association

290
00:30:52,919 --> 00:30:58,980
with small decreases in mean calorie and nutrient
content of fast food meals one to two years

291
00:30:58,980 --> 00:31:05,960
after the implementation of nationwide labeling.
And another approach in restaurant interventions

292
00:31:05,960 --> 00:31:11,519
is that of actually changing the menu. You'll
hear a bit more about the Silver Diner study

293
00:31:11,519 --> 00:31:18,559
from Chris Economos later on.
It's been suggested that we need whole-of-community

294
00:31:18,559 --> 00:31:22,960
interventions, not just in stores, not just
in restaurants, to really make a dent in the

295
00:31:22,960 --> 00:31:31,110
obesity problem. And there've been a number
of intervention studies in whole communities,

296
00:31:31,110 --> 00:31:39,870
lasting multiple years, and a couple of nice
reviews of those studies. Wolfenden and others

297
00:31:39,870 --> 00:31:45,600
looked at these programs in the U.S. and in
some Pacific islands, and looked at eight

298
00:31:45,600 --> 00:31:52,490
trials and found that seven of them had positive
effects. However, the mean reduction in BMI

299
00:31:52,490 --> 00:31:59,900
Z-scores were quite small. Ewart-Pierce, Ruiz,
and Gittlelsohn also conducted a review with

300
00:31:59,900 --> 00:32:05,570
some overlap and some different studies, and
they found mixed findings: some significant,

301
00:32:05,570 --> 00:32:10,779
some not significant findings, and concluded
that multilevel, multi-community approaches

302
00:32:10,779 --> 00:32:19,159
show promising results. The Healthy Eating
Active Living Zones program, sponsored by

303
00:32:19,159 --> 00:32:25,860
Kaiser Permanente in California, was a really
ambitious intervention program of three years

304
00:32:25,860 --> 00:32:31,320
in 12 low-income communities, guided by a
very complex logic model that you can see

305
00:32:31,320 --> 00:32:38,510
a simplified version of here, and with more
than 230 different community change strategies.

306
00:32:38,510 --> 00:32:44,679
The main bottom line findings reported by
Cheadle and others in 2018 were that there

307
00:32:44,679 --> 00:32:50,500
were positive population-level results for
higher dose strategies. However, one thing

308
00:32:50,500 --> 00:32:57,399
you can see here is that only a small proportion
of the strategies in what they call strategy

309
00:32:57,399 --> 00:33:04,900
clusters were higher dose. These studies are
marked by research limitations, and it's hard

310
00:33:04,900 --> 00:33:10,271
to assess the food environment changes, and
there's practice limitations as well. We

311
00:33:10,271 --> 00:33:17,090
don't know if these programs are sustainable
or can be scaled up beyond the research. So,

312
00:33:17,090 --> 00:33:25,240
a few points in closing. There's signal and
there's noise. Signals of effectiveness, but

313
00:33:25,240 --> 00:33:32,049
a lot of studies whose results we might consider
disappointing. If environmental strategies

314
00:33:32,049 --> 00:33:37,190
and change strategies aren't found effective,
why is that? It could be research issues,

315
00:33:37,190 --> 00:33:44,010
it could be strategy issues, and it could
be that we're working from the wrong assumptions

316
00:33:44,010 --> 00:33:51,320
and not addressing the key causes of the problems.
We also need to broaden our thinking to think

317
00:33:51,320 --> 00:33:54,809
about food justice, social justice, and unintended
consequences.

318
00:33:54,809 --> 00:34:01,520
I think most of us would agree that everyone
has a right to healthy, affordable food and

319
00:34:01,520 --> 00:34:07,049
that consumers have a right to know what's
in their food when they buy a meal from a

320
00:34:07,049 --> 00:34:13,210
restaurant. And communities can benefit from
some of these strategies and policies. For

321
00:34:13,210 --> 00:34:19,339
instance, a sugary beverage tax that produces
tax revenue to improve preschool education

322
00:34:19,339 --> 00:34:25,630
is a plus. But, if it raises taxes and sends
people outside the city to buy their groceries,

323
00:34:25,630 --> 00:34:34,030
it could result in job losses and net tax
losses. So, we need to work constantly to

324
00:34:34,030 --> 00:34:39,190
prevent the demand for short-term results
from undermining long-term research and policy.

325
00:34:39,190 --> 00:34:45,350
And we need to ask what we're doing that promotes
or prevents change. The food environment is

326
00:34:45,350 --> 00:34:50,770
complex, like the story of the blind man and
the elephant. It's hard to see the whole picture

327
00:34:50,770 --> 00:34:58,510
at once, and we need to remember that the
real elephant is the sum of its parts. We

328
00:34:58,510 --> 00:35:02,490
don't know how much environmental change is
needed to achieve health improvement. We don't

329
00:35:02,490 --> 00:35:08,270
know how long it will take to improve behavior
in health, whether the changes can be sustained,

330
00:35:08,270 --> 00:35:12,180
and who changes.
Do only motivated people change? Do they reduce

331
00:35:12,180 --> 00:35:21,050
health inequity? A few closing thoughts and
a few things we know. Local policies and environments

332
00:35:21,050 --> 00:35:30,020
can be powerful tools for changing behavior.
We need to be cautious when we take associations

333
00:35:30,020 --> 00:35:34,530
and try to turn them into interventions, because
the strategies may not work if they're based

334
00:35:34,530 --> 00:35:41,369
on an incomplete understanding of the causes.
And targeted inequality-reducing strategies

335
00:35:41,369 --> 00:35:47,480
may have more impact than population-based
strategies spread across all groups. And last,

336
00:35:47,480 --> 00:35:52,720
we need to consider the balance of social
justice and goals to improve health and behavior.

337
00:35:52,720 --> 00:36:00,850
I want to thank people whose input is important
to this presentation and who've supported

338
00:36:00,850 --> 00:36:06,630
and collaborated on my work over the years,
and, of course, the funding institutions as

339
00:36:06,630 --> 00:36:15,319
well. Thank you, and we'll have time for some
questions.

340
00:36:15,319 --> 00:36:28,540
DR. ANGELA ODOMS-YOUNG: Great. Thanks, Karen.
That was really a good frame-in for the kick off

341
00:36:28,540 --> 00:36:36,770
today and for the remaining sessions. We have
a few questions that have come into the chat.

342
00:36:36,770 --> 00:36:44,990
The first question is about types of stores.
So, are there any findings regarding ethnic-focused

343
00:36:44,990 --> 00:36:47,830
supermarkets versus large chain supermarkets?

344
00:36:47,830 --> 00:36:56,470
DR. KAREN GLANZ: Sure, that's a great question.
I am not aware of studies that have actually

345
00:36:56,470 --> 00:37:03,119
compared ethnic supermarkets to large chain
supermarkets, but there are quite a significant

346
00:37:03,119 --> 00:37:09,470
number that have been done in ethnic supermarkets,
and particularly in small stores. Joel Gittlesohn

347
00:37:09,470 --> 00:37:16,109
will be talking about interventions in small
stores. There have been a lot of studies in

348
00:37:16,109 --> 00:37:23,819
bodegas that kind of fit in that small store
category, and there have been a lot of case

349
00:37:23,819 --> 00:37:31,970
studies that have been done in a whole range
of ethnic-focused stores, such as Asian supermarkets,

350
00:37:31,970 --> 00:37:38,119
Latino-Hispanic supermarkets in different
regions of the country. So, and you know,

351
00:37:38,119 --> 00:37:46,840
the findings of those are pretty very hard
to summarize in a brief sentence. I would

352
00:37:46,840 --> 00:37:54,310
say generally that in the smaller store studies,
it's been shown that it's possible to increase

353
00:37:54,310 --> 00:38:03,540
the availability of healthier food options,
but that there hasn't been a lot of significant

354
00:38:03,540 --> 00:38:11,440
change in many studies in terms of healthier
diet or weight changes. And I would say from

355
00:38:11,440 --> 00:38:18,180
my point of view, some of that has to do with
the enormous challenge of reducing the plethora

356
00:38:18,180 --> 00:38:22,760
of unhealthy items in some of the smaller
stores.

357
00:38:22,760 --> 00:38:27,480
So, that's a bit of a generalization, but
there's quite a lot of literature out there in ethnic

358
00:38:27,480 --> 00:38:30,780
supermarkets, just not compared to grocery
stores.

359
00:38:30,780 --> 00:38:37,089
DR. ANGELA ODOMS-YOUNG: Thanks, Karen. The
next question is about prepared food sources.

360
00:38:37,089 --> 00:38:42,930
Can you talk about any studies on prepared
food sources and availability, with a focus

361
00:38:42,930 --> 00:38:48,090
on cultural foods and utilization in the neighborhood
food environment?

362
00:38:48,090 --> 00:38:55,240
DR. KAREN GLANZ: So, yeah. There are a number
of studies. I'm not sure if you're referring

363
00:38:55,240 --> 00:39:06,640
to prepared foods that people get in supermarkets.
That's one source. But in terms of small restaurants,

364
00:39:06,640 --> 00:39:12,349
there are a number of studies and there are
particularly some studies in Asian restaurants.

365
00:39:12,349 --> 00:39:18,980
Grace Mai and colleagues in Philadelphia have
done a study where they were aiming to reduce

366
00:39:18,980 --> 00:39:27,500
the sodium content in small restaurants. And
I know that Joel Gittlesohn and his team have

367
00:39:27,500 --> 00:39:33,940
done some work in Baltimore, where they've
worked with reformulating menu choices in

368
00:39:33,940 --> 00:39:40,380
Asian restaurants, and I would refer you to
look at those. They're interesting, and they're

369
00:39:40,380 --> 00:39:47,660
also very well marked by community participation
and retailer participation, which is really

370
00:39:47,660 --> 00:39:52,390
key if we want to try to make an impact in
those environments. There's also quite a bit

371
00:39:52,390 --> 00:39:58,030
in some other regions of the country, but
those are two that come to mind almost immediately.

372
00:39:58,030 --> 00:40:04,599
DR. ANGELA ODOMS-YOUNG: Thanks, Karen. I know
you didn't mention as much about some of the

373
00:40:04,599 --> 00:40:10,510
margin areas. I know many people are interested
in community-supported agriculture and food

374
00:40:10,510 --> 00:40:17,760
hubs, and we have a question that is related
to that. Are you aware of research on the

375
00:40:17,760 --> 00:40:21,920
impact of community-supported agriculture
or food hubs on dietary outcomes?

376
00:40:21,920 --> 00:40:31,180
DR. KAREN GLANZ: I have to admit that I'm
not. I'm aware of some case studies and descriptions

377
00:40:31,180 --> 00:40:39,349
of programs, and I agree those are growing
areas. And there may be some that I've missed,

378
00:40:39,349 --> 00:40:44,720
but I'm not aware of any that have looked
at dietary outcomes per se. They have looked…some

379
00:40:44,720 --> 00:40:52,520
of them have looked at purchases. And of course,
if you put more of something in the community,

380
00:40:52,520 --> 00:40:58,010
there will be purchases. And I think that,
also, that this is a place where things like

381
00:40:58,010 --> 00:41:06,590
GusNIP and fruit and vegetable incentives
and coupons have really infiltrated in large

382
00:41:06,590 --> 00:41:13,510
ways, programmatically. We have less…we
have a lot of information about the use of

383
00:41:13,510 --> 00:41:19,950
these, but less information about their specific
impact in those venues that you asked about.

384
00:41:19,950 --> 00:41:25,100
DR. ANGELA ODOMS-YOUNG: The next question
is about food marketing, where I know you

385
00:41:25,100 --> 00:41:32,250
have some extensive work and have a good handle,
probably on the literature in that area, thinking

386
00:41:32,250 --> 00:41:39,760
about the [inaudible] and the marketing focus. Do
you have any insights into the role of food

387
00:41:39,760 --> 00:41:47,600
marketing? Is there evidence that promotion
and marketing interfere with some of the interventions

388
00:41:47,600 --> 00:41:49,920
in the local food retail space?

389
00:41:49,920 --> 00:41:54,980
DR. KAREN GLANZ: Yeah, well, that could be
a whole day's conference in itself. That’s

390
00:41:54,980 --> 00:42:02,990
a great question. There is…we can look at
food marketing in neighborhoods, so there's

391
00:42:02,990 --> 00:42:09,080
really good evidence like, without even thinking
about TV and media advertising, there's evidence

392
00:42:09,080 --> 00:42:17,980
that there's more advertising for low nutritional
value foods in poor and minority neighborhoods.

393
00:42:17,980 --> 00:42:23,190
So, we know that there's a target audience.
If you really want to read a story about that,

394
00:42:23,190 --> 00:42:30,360
read the story of how Pepsi—there's a whole
book on this—marketed in black neighborhoods,

395
00:42:30,360 --> 00:42:37,770
a bigger foothold in there. Within stores,
there is a range of things that we would consider

396
00:42:37,770 --> 00:42:44,720
marketing and placement types of interventions
that are driven by industry, where they want

397
00:42:44,720 --> 00:42:52,290
certain products to be given more visibility.
And there's quite a bit of research around

398
00:42:52,290 --> 00:42:57,790
that. Not too much that's conclusive, it's
a little bit difficult to do that. Jaclyn

399
00:42:57,790 --> 00:43:03,910
Kerr and colleagues, and we collaborated with
them, did a study where they used grocery

400
00:43:03,910 --> 00:43:12,230
receipts as outcomes and found some associations
between the in-store marketing, placement

401
00:43:12,230 --> 00:43:16,630
of certain kinds of foods and the purchase
of those foods by participants.

402
00:43:16,630 --> 00:43:22,401
So, it doesn't interfere with other interventions.
Sometimes it overwhelms, and I think that

403
00:43:22,401 --> 00:43:28,240
is one of the challenges that we face in trying
to create more healthful food environments.

404
00:43:28,240 --> 00:43:35,540
DR. ANGELA ODOMS-YOUNG: Thanks, Karen. We
have a question about ethnic culture, cultural

405
00:43:35,540 --> 00:43:41,710
norms, and social norms. How do you suggest
the research be designed to examine ethnic

406
00:43:41,710 --> 00:43:48,190
and cultural norms, social norms intersecting
with identifying effective interventions for

407
00:43:48,190 --> 00:43:50,290
the neighborhood food environment?

408
00:43:50,290 --> 00:43:58,819
DR. KAREN GLANZ: Well, I think the best way
to go about that is to look at some of the

409
00:43:58,819 --> 00:44:05,520
wonderful examples that are out there that
have been already created with community involvement

410
00:44:05,520 --> 00:44:13,470
in it. And again, the ones that I think of
first, at least in the U.S., are those in

411
00:44:13,470 --> 00:44:17,859
Hispanic communities in California, such as
the work of Suchi Ayala, who'll be moderating

412
00:44:17,859 --> 00:44:24,599
in our next session, and also in Texas and
Florida and Arizona, there's been a ton of

413
00:44:24,599 --> 00:44:32,910
work in this area adapted to Hispanic communities,
ranging from immigrant communities to more,

414
00:44:32,910 --> 00:44:38,460
you know, longer-term communities that are
kind of more Americanized. There's been a

415
00:44:38,460 --> 00:44:44,480
fair bit in the Asian community. Some of that’s
been in California, but as I mentioned, the

416
00:44:44,480 --> 00:44:55,099
work of people like Grace Mei and Joel Gittlelsohn,
and there has been a lot of work done…oh.

417
00:44:55,099 --> 00:45:17,880
DR. ANGELA ODOMS-YOUNG: OK, we…maybe we
just have a few technological issues. So,

418
00:45:17,880 --> 00:45:25,900
I think Karen froze for a bit. And I think
she's back. Are you back, Karen?

419
00:45:25,900 --> 00:45:35,319
DR. KAREN GLANZ: Yes. Sorry, I had a little
something fall down here in my space. I am

420
00:45:35,319 --> 00:45:45,490
back. I can't be seen. That's unusual, can
you hear me?

421
00:45:45,490 --> 00:45:50,510
DR. ANGELA ODOMS-YOUNG: Yes, we can hear you.

422
00:45:50,510 --> 00:46:02,540
DR. KAREN GLANZ: I don't know what's going
on here. Well, let's continue without my face

423
00:46:02,540 --> 00:46:03,730
on the screen, sorry.

424
00:46:03,730 --> 00:46:08,030
DR. ANGELA ODOMS-YOUNG: Yeah, and we have
a lot of good questions. So, I'm sure that

425
00:46:08,030 --> 00:46:17,050
everybody can hear the responses. And I know
technology is great, and so at least we can

426
00:46:17,050 --> 00:46:26,819
have the pleasure of hearing you. So, the
next question is about community nutrition

427
00:46:26,819 --> 00:46:33,940
education. The attendee said, "Thank you for
the wonderful presentation. I am wondering

428
00:46:33,940 --> 00:46:40,150
if community nutrition education alongside
food environment interventions have been tested.

429
00:46:40,150 --> 00:46:42,230
And if so, what are the results?"

430
00:46:42,230 --> 00:46:55,710
DR. KAREN GLANZ: Sorry, can you repeat that?
I was trying to fix my camera there. About

431
00:46:55,710 --> 00:46:57,301
how nutrition education...

432
00:46:57,301 --> 00:47:00,830
DR. ANGELA ODOMS-YOUNG: Yeah, nutrition education
alongside food environment interventions.

433
00:47:00,830 --> 00:47:07,530
Have those been tested? And if so, what are
the results when you combine community nutrition

434
00:47:07,530 --> 00:47:10,040
education with interventions in the food environment?

435
00:47:10,040 --> 00:47:20,430
DR. KAREN GLANZ: Well, first, I would say
that the dividing line between interventions

436
00:47:20,430 --> 00:47:26,160
in the environment and education writ large.
Education, motivation incentives, next to

437
00:47:26,160 --> 00:47:33,700
things like choice architecture, which would
be changes in the environment, have been tried.

438
00:47:33,700 --> 00:47:41,530
In many cases, we've looked at the environment
without combining the food environment into

439
00:47:41,530 --> 00:47:50,001
it, and I would say that it's often been difficult
to separate out the effect of the education

440
00:47:50,001 --> 00:47:58,069
or the individual or group-focused intervention
strategies from the environmental strategies,

441
00:47:58,069 --> 00:48:04,780
for instance. I mean, many years ago in the
Netherlands, I worked with Ingrid Stenhouse

442
00:48:04,780 --> 00:48:10,910
and others to do supermarket tours and we
found that those led to healthier food choices.

443
00:48:10,910 --> 00:48:19,829
Another study that I worked on was using a
simplified index of nutrient-rich foods in

444
00:48:19,829 --> 00:48:29,589
supermarkets, and that also led people to…we
gave people a bunch of tools and they were

445
00:48:29,589 --> 00:48:35,470
more likely to use less in their supermarkets
and purchased and ate healthier foods, had

446
00:48:35,470 --> 00:48:41,119
higher scores on the healthy eating index.
So, we have a number of these kind of educational

447
00:48:41,119 --> 00:48:47,030
and behavioral strategies in food environments.
When they've been tried together, they've

448
00:48:47,030 --> 00:48:52,309
seldom been separated. And I think that one
of the lessons that we need to think about

449
00:48:52,309 --> 00:48:58,819
is whether some of the environmental change
strategies that have not found significant

450
00:48:58,819 --> 00:49:07,020
positive results on diet or food purchase
might be more effective if we could bolster

451
00:49:07,020 --> 00:49:11,910
them with education and behavioral or motivational
strategies.

452
00:49:11,910 --> 00:49:18,440
DR. ANGELA ODOMS-YOUNG: Thanks, Karen. We
have one last question. In yesterday's talk

453
00:49:18,440 --> 00:49:23,859
by Mariana Chilton, she talked about the one
thing we need to change to address nutrition

454
00:49:23,859 --> 00:49:31,460
equity with everything. What kind of theoretical
framework for PSE change can make this happen?

455
00:49:31,460 --> 00:49:40,349
DR. KAREN GLANZ: Well, you know, you've hit
on a topic that's one of my favorite theoretical

456
00:49:40,349 --> 00:49:44,799
frameworks. But one of the things I say about
theoretical frameworks is that a framework

457
00:49:44,799 --> 00:49:51,299
is like a coffee cup. It's just a framework
until you fill it up with content. So, we

458
00:49:51,299 --> 00:49:58,300
have to apply it. So, I think that, for instance,
an ecological systems perspective is a great

459
00:49:58,300 --> 00:50:04,600
framework, but it depends what you put in
those different levels of the framework and

460
00:50:04,600 --> 00:50:11,840
what we focus on that makes a difference.
With reference to some of the really important

461
00:50:11,840 --> 00:50:22,030
ideas related to structural racism and colonialism
and so forth that Mariana brought up, we perhaps

462
00:50:22,030 --> 00:50:27,090
have been leaving those out. And I think putting
them back in—in thoughtful ways—not just

463
00:50:27,090 --> 00:50:34,290
to kind of throw everything into the bucket
or the coffee cup would be advances in what

464
00:50:34,290 --> 00:50:35,290
we know now.

465
00:50:35,290 --> 00:50:43,230
DR. ANGELA ODOMS-YOUNG: Thank you so much,
Karen. And thank you to attendees for your

466
00:50:43,230 --> 00:50:53,780
thoughtful questions. We're going to move
into Session 5, which is focused on the

467
00:50:53,780 --> 00:50:59,030
state of the science in neighborhood food
environments, and explore research gaps and

468
00:50:59,030 --> 00:51:06,940
opportunities. And I'm going to turn it over…back
over to Karen so can introduce our next

469
00:51:06,940 --> 00:51:07,940
panel.

470
00:51:07,940 --> 00:51:12,990
DR. KAREN GLANZ: Thank you so much and thanks,
everyone, for the thoughtful questions that

471
00:51:12,990 --> 00:51:23,089
you gave. I think I'm still not going to be
visible. Might have to get my camera fixed. But

472
00:51:23,089 --> 00:51:28,819
it's okay right now. I'm just super pleased
to introduce Dr. Guadalupe Xochitl Ayala,

473
00:51:28,819 --> 00:51:36,510
San Diego State University, who's going to
be the moderator for the next panel. I've

474
00:51:36,510 --> 00:51:42,420
made reference to some of her outstanding
work in this area as well, and maybe you'll

475
00:51:42,420 --> 00:51:46,299
all hear bits about that when we get to the
Q&A. So Xochitl, over to you.

476
00:51:46,299 --> 00:51:53,270
DR. QUADALUPE XOCHITL AYALA: Thank you, Karen.
Good day, everybody. So good to be here with

477
00:51:53,270 --> 00:51:59,220
you today. For this session of the Neighborhood
Food Environment, Diet and Health Outcomes,

478
00:51:59,220 --> 00:52:05,480
our first panelist is Dr. Alice Ammerman,
of the University of North Carolina. She's

479
00:52:05,480 --> 00:52:12,420
going to discuss her work examining retail
food environments, diet, and chronic disease.

480
00:52:12,420 --> 00:52:20,640
Our second speaker is Dr. Christina Economos,
from Tufts University. She will review restaurants,

481
00:52:20,640 --> 00:52:27,640
food service environment, diet, and chronic
disease. I just want to remind everybody that

482
00:52:27,640 --> 00:52:33,140
we will have a conversation at the end of
both of these presentations. So I ask that

483
00:52:33,140 --> 00:52:37,440
you put your questions on the chat and, if
possible, direct your questions to one of

484
00:52:37,440 --> 00:52:43,140
the two speakers. Let's get started.

485
00:52:43,140 --> 00:52:51,730
DR. ALICE AMMERMAN: Hello, I'm happy to join
you today to talk about retail food environments

486
00:52:51,730 --> 00:52:56,990
and their impact on diet and chronic disease.
My name is Alice Ammerman, I'm at UNC Chapel

487
00:52:56,990 --> 00:53:02,799
Hill, and I'd like to thank some of my students
who helped me with this presentation, and

488
00:53:02,799 --> 00:53:09,980
thanks to the organizers for this really nice
workshop, I am learning a lot. So today we

489
00:53:09,980 --> 00:53:17,150
will be talking about a number of different
topics. One, why is this important? What is

490
00:53:17,150 --> 00:53:21,930
the retail environment? What is it made up
of? Where do people shop and particularly SNAP

491
00:53:21,930 --> 00:53:27,550
lower income consumers? What do we know
about the quality of different retail venues

492
00:53:27,550 --> 00:53:32,849
and what has been the diet/disease impact?
And then what is some promising research

493
00:53:32,849 --> 00:53:38,940
opportunity for the future? So why is this
important? One is to inform policy, as a lot

494
00:53:38,940 --> 00:53:44,730
of the work we're doing in this workshop is
pointing towards and which will in turn guide

495
00:53:44,730 --> 00:53:50,280
research and funding, and then this will help
with guiding planning and goal setting. And

496
00:53:50,280 --> 00:53:57,500
as an example, this is our 2030 state health
plan in North Carolina, and we pick indicators

497
00:53:57,500 --> 00:54:01,890
that are reasonable to measure to try to guide
certain priorities.

498
00:54:01,890 --> 00:54:09,470
And I was talking to a friend with DHHS, who
mentioned that they're doing a lot of work

499
00:54:09,470 --> 00:54:14,809
on spending resources and time to improve
healthy retail access. And given the data

500
00:54:14,809 --> 00:54:19,440
that and what we're seeing is this is the
place where they should be putting their energy,

501
00:54:19,440 --> 00:54:25,119
so that's a good reason to do this workshop.
So back to this plan, so one of our health

502
00:54:25,119 --> 00:54:33,010
indicators is limited access to healthy food,
and we're trying to move to a smaller percent

503
00:54:33,010 --> 00:54:39,561
of the population who has limited access.
Now how do we define that? We define that

504
00:54:39,561 --> 00:54:44,119
as the percent of people who are low income
that are not in close proximity to a grocery

505
00:54:44,119 --> 00:54:51,660
store, and this is how we define proximity;
less than a mile in more urban and more than

506
00:54:51,660 --> 00:54:56,760
10 miles in rural. And this comes from the
data…the county health rankings. I think

507
00:54:56,760 --> 00:55:02,940
all of us are looking at whether there may
be ways to improve this measure. Definition

508
00:55:02,940 --> 00:55:09,640
of retail. This is from the corporate world:
considered all food, other than restaurant

509
00:55:09,640 --> 00:55:13,559
food, that's purchased by consumers and consumed
off premise.

510
00:55:13,559 --> 00:55:23,400
And the industry estimates $6 trillion in
sales in 2019, so it's a big part of our economy.

511
00:55:23,400 --> 00:55:29,140
Types of retail food environments. We'll mostly
be focusing on these kinds of larger stores,

512
00:55:29,140 --> 00:55:33,620
more traditional grocery stores, but then,
of course, the big box and the club stores.

513
00:55:33,620 --> 00:55:36,960
And then there's a lot of interest in corner
convenience stores, you'll hear more about

514
00:55:36,960 --> 00:55:41,890
that from other speakers, vending machines,
specialty shops, you can get food almost anywhere

515
00:55:41,890 --> 00:55:47,880
these days. And some of the other interesting
areas that we won't be able to touch on all

516
00:55:47,880 --> 00:55:53,680
of these but mobile markets, farmer's markets,
CSA food box, food trucks, home delivery meal

517
00:55:53,680 --> 00:56:00,000
kits, all the Door Dashes and the many different
ways that we can tap into the retail food

518
00:56:00,000 --> 00:56:06,280
environment. Looking at USDA data in terms
of share of household calories by food source,

519
00:56:06,280 --> 00:56:10,630
still the large groceries are the bulk of
it, and of course, we have to take out restaurants

520
00:56:10,630 --> 00:56:15,859
for this discussion, but that's an increasing
part of our food source, maybe adjusted some

521
00:56:15,859 --> 00:56:23,980
during COVID. And then there's the convenience
stores, which are the next largest chunk.

522
00:56:23,980 --> 00:56:29,490
Taking a look at kind of nutrition scores
for household food acquisition by source,

523
00:56:29,490 --> 00:56:37,470
so these again are the different types of
stores or places to get food. Schools being

524
00:56:37,470 --> 00:56:42,730
an exception to stores. And if you look at
it by…basically as you come down this list,

525
00:56:42,730 --> 00:56:47,910
you're increasing income. So starting with
SNAP and then higher income. So for the most

526
00:56:47,910 --> 00:56:54,490
part, it looks like higher income consumers
that the foods they purchase are somewhat

527
00:56:54,490 --> 00:57:00,530
higher nutritional quality, except when you
get to the schools, where the food is provided,

528
00:57:00,530 --> 00:57:07,280
and it's more equity in terms of all students
getting the same food, so that's a place

529
00:57:07,280 --> 00:57:12,100
where we know it's important to reach children
with good nutrition. And then another way

530
00:57:12,100 --> 00:57:17,890
to look at this is in terms of access. This
is another kind of distance-oriented definition,

531
00:57:17,890 --> 00:57:22,490
and I think the main point here is that those
who are more access burdened are more likely

532
00:57:22,490 --> 00:57:28,440
to use some of the convenience dollar stores
where some of the quality of the nutrition

533
00:57:28,440 --> 00:57:35,760
is less good. So early research on the links
between diet and health really focused

534
00:57:35,760 --> 00:57:44,079
on distance.
So increased nutrition with access to supermarkets,

535
00:57:44,079 --> 00:57:50,619
number of supermarkets in the community, and
then there's how they rated their nutritional

536
00:57:50,619 --> 00:57:57,059
quality. So a big focus on how close are you
to a supermarket? And that seemed to be associated

537
00:57:57,059 --> 00:58:03,000
with improved diet. And this has had a big
impact on policy and practice but there are

538
00:58:03,000 --> 00:58:09,920
some concerns now as we do more research in
this area. The results did kind of focus our

539
00:58:09,920 --> 00:58:14,391
attention on this concept of food deserts,
which we’re beginning to, I think, question

540
00:58:14,391 --> 00:58:21,270
more now. And then it released a lot of funding
for supermarket location and construction

541
00:58:21,270 --> 00:58:25,799
like the Healthy Food Financing Initiative.
But there were a lot of assumptions with this

542
00:58:25,799 --> 00:58:32,630
data that say shopping at larger stores with
a wider selection results in healthier choices

543
00:58:32,630 --> 00:58:39,190
and most of the studies didn't really look
at specific data where people shop and what

544
00:58:39,190 --> 00:58:43,240
they purchase. But we do have data now...
a number of researchers have been looking

545
00:58:43,240 --> 00:58:48,260
at the Homescan Nielsen data that provides
household packaged good data.

546
00:58:48,260 --> 00:58:52,369
It doesn't give us a lot about fruits and
vegetables, but there are other ways that

547
00:58:52,369 --> 00:58:59,860
we're getting at that. So here's a look at
nutrient content by store type, and these

548
00:58:59,860 --> 00:59:10,090
are focusing on some of the types of parts
of our food intake that we are trying to modify,

549
00:59:10,090 --> 00:59:15,630
such as calories, sugar, saturated fat. And
you see, in general, it seems like the larger

550
00:59:15,630 --> 00:59:23,079
stores and the…seem to be doing somewhat
better. This also looks at changes over time,

551
00:59:23,079 --> 00:59:31,109
but that the convenience stores, some of the
warehouse, the large club stores where people

552
00:59:31,109 --> 00:59:38,200
buy a lot of packaged goods that you're getting
kind of higher intakes from that, changes

553
00:59:38,200 --> 00:59:43,109
over time are not terribly consistent, although
this may reflect some improvements in some

554
00:59:43,109 --> 00:59:51,870
of the formulation of packaged foods. Another
way to look at this is urban vs. rural. And

555
00:59:51,870 --> 00:59:58,609
again, we're looking at the nutritional quality
of packaged goods. One big distinction here

556
00:59:58,609 --> 01:00:04,670
which is interesting is most of these lines
kind of line up between urban and rural, but

557
01:00:04,670 --> 01:00:12,420
not for the kind of Walmart-type stores, where
you've got a much higher dependence from rural

558
01:00:12,420 --> 01:00:18,920
areas on that.
I think we'll see more going on here over

559
01:00:18,920 --> 01:00:24,690
time in terms of online purchasing. This doesn't
capture, I think, the most recent trends in

560
01:00:24,690 --> 01:00:32,420
that direction. Another approach to this again
with urban and rural and then calories per

561
01:00:32,420 --> 01:00:39,849
person per day for packaged good purchases.
I think things as you might expect in terms

562
01:00:39,849 --> 01:00:44,599
of grocery stores being the main purchasing
point, I think there's a lot of interest in

563
01:00:44,599 --> 01:00:51,290
dollar stores and you see breaking it down
by income that lower income in both urban

564
01:00:51,290 --> 01:00:55,980
and rural areas are more dependent on the
dollar stores. So, what we can learn about

565
01:00:55,980 --> 01:01:02,020
what food is available there will be important.
One more way to look at this…kind of complicated,

566
01:01:02,020 --> 01:01:10,130
but I think showing that the grocery stores
as opposed to say, full-service and quick-

567
01:01:10,130 --> 01:01:14,480
service restaurants are still providing us
with healthier options and schools probably

568
01:01:14,480 --> 01:01:20,250
doing the best in that regard. So how do we
summarize all of this? There have been a number

569
01:01:20,250 --> 01:01:27,809
of big review papers to look at this, and
this is, I think a particularly good one,

570
01:01:27,809 --> 01:01:33,240
it was done in 2015, so we probably have more
current ones as well.

571
01:01:33,240 --> 01:01:41,880
But overall, looking at the local food environment
and associations with obesity, the diet, the

572
01:01:41,880 --> 01:01:49,431
study quality was quite low, and the opinion
of those who did this study was 60 of 71 studies

573
01:01:49,431 --> 01:01:53,819
were cross-sectional. Of course, this is a
hard thing to study, so we have to rely on

574
01:01:53,819 --> 01:02:03,150
that kind of study design often. Most of the
findings were null, there were some trends

575
01:02:03,150 --> 01:02:09,119
inverse associations between supermarket availability
and obesity and then obesity in children.

576
01:02:09,119 --> 01:02:14,240
But if you look at this, the majority of the
studies were null, but there were some that

577
01:02:14,240 --> 01:02:19,440
were positive in there. And the authors do
state that because of the study quality, we

578
01:02:19,440 --> 01:02:26,369
have to interpret these carefully but I think
the underlying message is that proximity to

579
01:02:26,369 --> 01:02:34,650
particular retail outlets is not the whole
story. So and they conclude that despite the

580
01:02:34,650 --> 01:02:39,710
large number of studies, limited evidence
for association between local food environments

581
01:02:39,710 --> 01:02:45,050
and obesity. So why isn't there more of an
association? Well, there are many factors

582
01:02:45,050 --> 01:02:50,470
other than distance, as we've said, accessibility
of a highway, perhaps between you and a store,

583
01:02:50,470 --> 01:02:55,569
even if it's a half mile away, that's going
to make it very difficult if you have a limited

584
01:02:55,569 --> 01:02:57,990
transportation.
Quality of food in the stores, as we talked

585
01:02:57,990 --> 01:03:03,369
about the cost of healthier food options,
is it culturally relevant? Neighborhood ownership

586
01:03:03,369 --> 01:03:08,660
a sense that this is a neighborhood grocery
store that people want to frequent, and it

587
01:03:08,660 --> 01:03:13,070
does seem from some of the research that grocery
stores may bring benefits other than nutrition,

588
01:03:13,070 --> 01:03:19,089
like more jobs to the community, perception
of healthier food access or a better food

589
01:03:19,089 --> 01:03:26,300
environment, although that doesn't always
seem to be correlated with improved diet intake.

590
01:03:26,300 --> 01:03:31,289
So I'll point you towards a publication that
I think could be very helpful going forward.

591
01:03:31,289 --> 01:03:36,640
The Healthy Eating Research Group kind of
spearheaded this national research agenda

592
01:03:36,640 --> 01:03:41,829
to support healthy eating through retail strategies.
I'm not going to go through all of these,

593
01:03:41,829 --> 01:03:47,310
you'll have access to the slides, but 10 priority
areas were identified in terms of both understanding

594
01:03:47,310 --> 01:03:52,740
the current environment and then assessing
the implementation effectiveness of interventions.

595
01:03:52,740 --> 01:03:58,640
And that will be a subject of a whole other
section of this workshop is interventions.

596
01:03:58,640 --> 01:04:03,890
And they did an attempt to look at some of
the high priority areas in terms of how they

597
01:04:03,890 --> 01:04:08,920
would address health equity as well as feasibility,
and some of the biggest ones as you might

598
01:04:08,920 --> 01:04:15,380
guess, were related to SNAP increasing the
amounts, the benefits, frequencies. So these

599
01:04:15,380 --> 01:04:23,210
are all areas of active research and I think
continue to be. Another interesting way to

600
01:04:23,210 --> 01:04:28,510
look at this is a study by Winkler et al.,
in terms of kind of accessibility from anyone's

601
01:04:28,510 --> 01:04:36,420
home, how much does it take to actually get
to the place of the store? And then also the

602
01:04:36,420 --> 01:04:43,099
preparation that's required. So you can see
we're sort of trending towards easy access,

603
01:04:43,099 --> 01:04:47,849
little preparation, food that's delivered
to our doorstep that's half prepared already,

604
01:04:47,849 --> 01:04:54,920
or vending machines at work. So interesting
to see how things go as we move more and more

605
01:04:54,920 --> 01:05:01,221
in the direction of accessibility and whether
or not that involves equity in terms of people

606
01:05:01,221 --> 01:05:09,779
who are able to access those things. Online
shopping, a big area of research. Very recent

607
01:05:09,779 --> 01:05:14,839
Gallup poll showed that a quarter of people
order groceries online; it's double the share

608
01:05:14,839 --> 01:05:20,190
from the pre-pandemic year of 2019.
Traditionally, the demographic has mostly

609
01:05:20,190 --> 01:05:28,349
been younger, with households over 100,000,
and it tends to be resulting in decreased

610
01:05:28,349 --> 01:05:35,250
shopping frequency, of course, some of that
is COVID related. But the corporate food world

611
01:05:35,250 --> 01:05:41,089
predicts that many of these habits will likely
stick in terms of people becoming more comfortable

612
01:05:41,089 --> 01:05:47,510
and at ease with online shopping. For lower
income consumers, again, a very active area

613
01:05:47,510 --> 01:05:53,589
of research, especially as options to pay
online with SNAP increase, which is an important

614
01:05:53,589 --> 01:05:59,190
part of this equation. Historically, there's
been lower participation by the SNAP consumers,

615
01:05:59,190 --> 01:06:06,670
but this varies by retailers, outside pick
up, when that becomes more of an option may

616
01:06:06,670 --> 01:06:12,070
benefit families with young children. Delivery
costs can still be a barrier, and that varies

617
01:06:12,070 --> 01:06:17,589
by retailer. Some of the COVID and other emergencies
have sped this process along a bit. Some of

618
01:06:17,589 --> 01:06:23,099
my colleagues have looked at the notion of
pester power, and whether you can order online

619
01:06:23,099 --> 01:06:29,410
means your children are pestering you to buy
the latest sugar-sweetened cereal.

620
01:06:29,410 --> 01:06:34,500
And multiple interventions are now testing
tailoring…tailored ordering systems and

621
01:06:34,500 --> 01:06:39,660
nudges to see if we can use online to really
move nutrition in a positive direction. So

622
01:06:39,660 --> 01:06:45,089
quickly to wrap up, what's the role for more
qualitative community engaged work to try

623
01:06:45,089 --> 01:06:53,641
to improve opportunities for SNAP-eligible
shoppers? We have a team at our shop looking

624
01:06:53,641 --> 01:06:59,970
at this and supported by some help from Robert
Wood Johnson, Cooking Matters Share Our Strength, this

625
01:06:59,970 --> 01:07:06,359
is the team here. Looking at human-centered
design to test and implement food retail interventions

626
01:07:06,359 --> 01:07:14,220
that promote healthier choices. So the challenge
here is to develop an intervention for SNAP-

627
01:07:14,220 --> 01:07:19,059
eligible caregivers of young children that
meets their needs and advances SNAP-Ed's reach

628
01:07:19,059 --> 01:07:25,480
and impact. As you know, SNAP-Ed is one of
our primary ways of reaching lower income

629
01:07:25,480 --> 01:07:34,990
consumers to address some nutrition improvements.
So we use the human-centered design to approach,

630
01:07:34,990 --> 01:07:43,819
a participatory approach. And it involves
a complicated but very interesting synthesized

631
01:07:43,819 --> 01:07:49,319
journey mapping approach, where you look at,
you won't be able to see the details here,

632
01:07:49,319 --> 01:07:54,020
but before the store, at the store, and after
the store, what are people's feelings?

633
01:07:54,020 --> 01:07:58,250
What are their frustrations? What would make
life easier for them? And the basic findings

634
01:07:58,250 --> 01:08:04,710
we had key insights, were that families would
appreciate ways to reduce the cognitive burden

635
01:08:04,710 --> 01:08:09,690
of shopping since there's a lot to juggle.
So, more understanding of what lower cost

636
01:08:09,690 --> 01:08:16,380
items are, how to find WIC-eligible items,
opportunities for kids and shopping. Again,

637
01:08:16,380 --> 01:08:21,430
this might relate to the pester power, you
know, having kids be engaged in some way—and

638
01:08:21,430 --> 01:08:27,620
also later in cooking. And then flexible,
customizable options to meet their dietary

639
01:08:27,620 --> 01:08:32,290
needs and preferences. Are there ways to build
food boxes and bundled goods and things that

640
01:08:32,290 --> 01:08:39,180
can help that. So in summary, looking at retail
and the neighborhood food environment, it's

641
01:08:39,180 --> 01:08:44,810
more than distance either to or the number
of retail grocery options. More supermarkets

642
01:08:44,810 --> 01:08:49,910
may improve other environmental factors like
job, access, perception of health options

643
01:08:49,910 --> 01:08:55,470
and…but they may not always improve the
nutrition or health, and that's where we're

644
01:08:55,470 --> 01:09:00,759
trying to expand the work. Online ordering
and delivery options hold promise but are

645
01:09:00,759 --> 01:09:05,000
not yet widely studied.
In-store and marketing interventions are under

646
01:09:05,000 --> 01:09:10,549
active study with some promising results,
and you'll hear more about that. And we may

647
01:09:10,549 --> 01:09:15,660
need broader approaches to understand and
address social drivers of health related to

648
01:09:15,660 --> 01:09:22,990
how we can make access to healthy food more
equitable. Thank you for listening to this

649
01:09:22,990 --> 01:09:24,159
talk.

650
01:09:24,159 --> 01:09:32,650
DR. CHRISTINA ECONOMOS: Hi, I'm Chris Economas,
Professor of Nutrition at the Friedman School

651
01:09:32,650 --> 01:09:36,790
of Nutrition, Science and Policy at Tufts
University. And the title of my talk today

652
01:09:36,790 --> 01:09:47,440
is, Restaurant/Food Service Environments,
Diet, and Chronic Disease. So why study restaurants?

653
01:09:47,440 --> 01:09:52,670
Currently, there are over 1 million restaurant
locations in the U.S., and of these, about

654
01:09:52,670 --> 01:09:59,290
70 percent are single-unit operations, so
they're managed and they're run independently,

655
01:09:59,290 --> 01:10:05,900
whereas there are about 195,000 quick service
or fast food restaurant franchises. And despite

656
01:10:05,900 --> 01:10:10,540
the fact that there are more of the single-unit,
you'll see in subsequent slides that more

657
01:10:10,540 --> 01:10:17,400
calories are actually coming from quick service.
About 20 percent of calories across the population

658
01:10:17,400 --> 01:10:22,679
are consumed from restaurants, and food away
from home is associated with higher calorie

659
01:10:22,679 --> 01:10:27,850
and lower nutritional value. There are lots
of different intervention strategies to improve

660
01:10:27,850 --> 01:10:33,370
restaurant eating, for example, density of
restaurants within a community, proximity

661
01:10:33,370 --> 01:10:40,480
from residents to a restaurant establishment,
location, for example, near a park, labeling

662
01:10:40,480 --> 01:10:45,989
policies, supply, looking at the nutritional
value of what's being offered, demand, looking

663
01:10:45,989 --> 01:10:51,810
at ordering patterns of customers and then,
of course, advertising and marketing.

664
01:10:51,810 --> 01:10:56,620
So we know that eating out is the norm, and
any of these intervention strategies alone

665
01:10:56,620 --> 01:11:02,220
or in combination with other environmental
food strategies have the potential for massive

666
01:11:02,220 --> 01:11:09,449
impact. So beginning with food expenditures
over time, you can see that in about 2010,

667
01:11:09,449 --> 01:11:14,449
these lines cross, and this represents the
point in time where we started to spend more

668
01:11:14,449 --> 01:11:22,393
food away from home than on food at home,
coming from retail or grocery. So although

669
01:11:22,393 --> 01:11:26,000
more calories are coming from retail and grocery,
it's more expensive to eat out, and we're

670
01:11:26,000 --> 01:11:33,390
spending more food away from home. And so
in 2017, we were spending approximately $400

671
01:11:33,390 --> 01:11:40,120
billion on food away from home. So looking
at proximity and density, there was a recent

672
01:11:40,120 --> 01:11:45,679
systematic review of longitudinal studies
looking at proximity and density of restaurants

673
01:11:45,679 --> 01:11:52,150
in relationship to health. And we know that
overall, the proximity of residents to restaurants

674
01:11:52,150 --> 01:11:57,670
on things like pediatric weight and pediatric
weight change are predominantly null, but

675
01:11:57,670 --> 01:12:01,600
there are stronger associations in girls than
in boys when they're seen.

676
01:12:01,600 --> 01:12:06,909
There's limited or mixed evidence looking
at the proximity to fast food restaurants

677
01:12:06,909 --> 01:12:12,130
and the effect on increased weight change
over time and the density of fast food showing

678
01:12:12,130 --> 01:12:17,150
some effect on weight trajectories and BMI
Z-score with stronger relationships in girls

679
01:12:17,150 --> 01:12:23,610
in rural areas, children in high income households
and white children. With respect to disparities,

680
01:12:23,610 --> 01:12:29,830
there are some associations with individual
outcomes among some groups…subgroups, including

681
01:12:29,830 --> 01:12:34,870
sugar sweetened beverage consumption. And
we know that unhealthy food is ubiquitous,

682
01:12:34,870 --> 01:12:39,620
coming from lots of different food environments.
And again, 20 percent of calories is coming

683
01:12:39,620 --> 01:12:45,440
from restaurants, so just working on restaurants
alone is unlikely to make a difference in

684
01:12:45,440 --> 01:12:51,290
health outcomes but if we work on restaurants
in combination with other things in the food

685
01:12:51,290 --> 01:12:57,520
environment, we can have impact on health
outcomes. So looking first at the supply side,

686
01:12:57,520 --> 01:13:04,630
these data are from MenuStat looking in that
percent of meals at 73 U.S. restaurants that

687
01:13:04,630 --> 01:13:08,659
are compliant with the American Heart Association
criteria.

688
01:13:08,659 --> 01:13:15,641
So looking at these 73 U.S. restaurants that
are the largest in sales that also have nutritional

689
01:13:15,641 --> 01:13:22,520
information and stratifying by all restaurants
seen in the darkest bar, followed by fast

690
01:13:22,520 --> 01:13:28,739
food restaurants, which is in the next darkest
bar, followed by full-service restaurants

691
01:13:28,739 --> 01:13:33,870
and then finally, by fast casual. So looking
across the different nutrition criteria, you

692
01:13:33,870 --> 01:13:39,630
can see that restaurant meals are doing well
with respect to offering protein and meeting

693
01:13:39,630 --> 01:13:47,889
criteria. They're doing next in line pretty
well providing fiber, followed by trans-fat,

694
01:13:47,889 --> 01:13:56,199
followed by cholesterol and then for things
like calories less than 700, for total fat,

695
01:13:56,199 --> 01:14:02,950
for saturated fat and for sodium less than
50 percent of meals. And this includes an

696
01:14:02,950 --> 01:14:08,340
entree and a side, that are being offered
in the largest restaurants with the most sales

697
01:14:08,340 --> 01:14:12,940
are not meeting these criteria. So there's
a lot of room for improvement in terms of

698
01:14:12,940 --> 01:14:19,520
what's being offered. And if we look at these
same data clustered together by the number

699
01:14:19,520 --> 01:14:25,800
of criteria that the meals are meeting from
zero to one, two to four, five to six, or all

700
01:14:25,800 --> 01:14:30,300
seven, which is the highest score that you
can get when you combine these criteria, you

701
01:14:30,300 --> 01:14:36,750
can see that most meals are falling into the
two to five criteria met, and less than 10

702
01:14:36,750 --> 01:14:40,350
percent are falling into all seven criteria
being met.

703
01:14:40,350 --> 01:14:45,800
So again, lots of room for improvement. And
if we look at who's actually doing the best,

704
01:14:45,800 --> 01:14:50,890
it's the fast casual restaurants, followed
by the fast food restaurants, followed by

705
01:14:50,890 --> 01:14:59,320
full service. So looking more deeply at children's
meals, we conducted a study looking at meal

706
01:14:59,320 --> 01:15:05,650
bundles with fruit or non-fried vegetable
sides, as well as non-sugary beverages. And

707
01:15:05,650 --> 01:15:09,870
we wanted to see over time if restaurants
were offering more of these and if they were

708
01:15:09,870 --> 01:15:15,400
offering them by default. So using another
data set called Technomics and looking at

709
01:15:15,400 --> 01:15:22,630
the top 20 restaurants for sales from 2005
to 2015, you can see on the left panel that

710
01:15:22,630 --> 01:15:28,560
the meal bundles with fruit and vegetable
sides and non-sugary beverages that were offered

711
01:15:28,560 --> 01:15:35,850
actually increased, as well as the sugary
sweetened beverage substitutes or non-sugary

712
01:15:35,850 --> 01:15:42,000
drinks increasing over time. So, in terms
of offering in supply, this is good news.

713
01:15:42,000 --> 01:15:46,210
But when you look more specifically at the
average percentage of meal bundles with healthy

714
01:15:46,210 --> 01:15:51,929
sides and beverages by default, which is a
recommendation you can see it's very low.

715
01:15:51,929 --> 01:15:56,900
So as of 2015, less than 20 percent of these
bundles were offering a fruit and vegetable

716
01:15:56,900 --> 01:16:02,870
side and a non-sugary beverage by default.
And we know the default strategy is very effective,

717
01:16:02,870 --> 01:16:10,350
so getting more restaurants to implement default
is likely to increase what happens with dietary

718
01:16:10,350 --> 01:16:16,250
intake. So on the top fruit and vegetable,
and on the bottom sugar sweetened beverage.

719
01:16:16,250 --> 01:16:22,380
So now turning to what we actually consume
and using NHANES data, this is showing trend

720
01:16:22,380 --> 01:16:31,020
data from 2003 to the most recent NHANES data
2017/2018. And first looking at estimated percentage

721
01:16:31,020 --> 01:16:36,420
of energy intake among U.S. adults 20 or older,
you can see the majority of the calories of

722
01:16:36,420 --> 01:16:42,920
70 percent are coming from grocery or retail,
followed by calories coming from restaurants,

723
01:16:42,920 --> 01:16:49,680
hovering at about 20 percent, not changing
much over time, followed by the "other" category

724
01:16:49,680 --> 01:16:56,770
and then lastly by the work environment. For
children, it's very similar. Again, fairly

725
01:16:56,770 --> 01:17:01,860
stable over time, and you can see that it's
about 20 percent of calories consumed from

726
01:17:01,860 --> 01:17:07,480
restaurants by children with again the highest
category being grocery or retail, in this

727
01:17:07,480 --> 01:17:11,880
case, followed by school and other, which
are tied at about 15 percent.

728
01:17:11,880 --> 01:17:20,580
So now, if we dive into the most recent enhanced
data 2017/2018 and look more deeply at the

729
01:17:20,580 --> 01:17:28,710
restaurant category, these data are actually
showing the types of restaurants divided into

730
01:17:28,710 --> 01:17:36,880
full serve and quick serve, and you can see
that the youngest children are consuming actually

731
01:17:36,880 --> 01:17:43,900
not quite as much from quick-serve or full-
serve restaurants as other children, hovering

732
01:17:43,900 --> 01:17:51,940
here at about 13 percent, and the category
shown with the dots is quick serve or fast

733
01:17:51,940 --> 01:17:58,060
food and the category shown with the gray
shading is full-service restaurants. So together,

734
01:17:58,060 --> 01:18:03,739
this is restaurant eating. This increases
as children age. So you can see in school–aged

735
01:18:03,739 --> 01:18:09,132
children, it gets higher, again, higher in
quick serve than full serve. And then when

736
01:18:09,132 --> 01:18:14,820
you get to teenagers or 12 to 19 year olds,
it's about 28 percent of calories coming from

737
01:18:14,820 --> 01:18:20,420
restaurants with the majority coming from
the quick serve segment. This is stable in

738
01:18:20,420 --> 01:18:26,699
about 20 to 40 year olds, and then it starts
to decline again as people go through the

739
01:18:26,699 --> 01:18:31,610
aging process and it's much lower in those
over 71 years of age.

740
01:18:31,610 --> 01:18:37,630
So the 20 percent is the average across the
population, but it does vary by age group.

741
01:18:37,630 --> 01:18:45,000
So now looking at whether or not those foods
consumed and the calories consumed are high

742
01:18:45,000 --> 01:18:51,750
quality or not, these data are also using
NHANES and looking over time from 2003 to

743
01:18:51,750 --> 01:19:00,580
the most recent 2017/2018, and actually categorizing
the foods consumed from restaurants as poor,

744
01:19:00,580 --> 01:19:05,120
intermediate, or ideal diet quality. And so
the panel on the left are showing children;

745
01:19:05,120 --> 01:19:11,140
the panel on the right are showing adults.
And this is using the American Heart Association

746
01:19:11,140 --> 01:19:16,970
criteria. So you can see that the majority
of foods from restaurants for children and

747
01:19:16,970 --> 01:19:23,040
adults falls into the poor diet quality category,
followed by the blue showing the intermediate

748
01:19:23,040 --> 01:19:28,790
diet quality. And last down at the bottom,
very little is ideal diet quality coming from

749
01:19:28,790 --> 01:19:36,290
restaurants. You can see that the blue line
for children is going up a bit over time, showing

750
01:19:36,290 --> 01:19:42,270
that there's been some improvement and the
red line is going down a bit. So more intermediate

751
01:19:42,270 --> 01:19:46,500
diet quality foods being consumed by children
when they dine out in restaurants.

752
01:19:46,500 --> 01:19:53,640
But there's still a very big room for improvement
here in both children and adults. Looking

753
01:19:53,640 --> 01:20:00,250
here at these radar plots, these are also
from NHANES using the healthy eating index

754
01:20:00,250 --> 01:20:06,060
2015. And looking at food component scores
over time between full-serve and quick-serve

755
01:20:06,060 --> 01:20:11,500
restaurants using, again, the NHANES data.
And on these radar plots, the center point

756
01:20:11,500 --> 01:20:16,710
of the graph represents a score of zero, and
the outer point of each axis represents the

757
01:20:16,710 --> 01:20:22,230
maximum score for each component. So plots
with the most points closer to the outer edge

758
01:20:22,230 --> 01:20:27,460
represent a food pattern that's closer to
meeting the recommendations of the 2015/2020

759
01:20:27,460 --> 01:20:33,670
Dietary Guidelines for Americans. They had
a plot with many points closer to the center

760
01:20:33,670 --> 01:20:39,870
of the graph. So looking specifically at restaurants,
we can see that in full-service restaurants

761
01:20:39,870 --> 01:20:45,330
doing better in total vegetables, greens and
beans, meeting saturated fat requirements,

762
01:20:45,330 --> 01:20:51,250
fatty acid, seafood, plant proteins and total
protein foods where in quick-serve restaurants

763
01:20:51,250 --> 01:20:56,710
for total protein, dairy, and a little bit
for vegetables and saturated fats.

764
01:20:56,710 --> 01:21:01,420
But overall, there's a lot of room for improvement,
particularly for quick service with respect

765
01:21:01,420 --> 01:21:11,210
to food components and meeting the AGI guidance
that is put out for high diet quality. So

766
01:21:11,210 --> 01:21:15,520
when you get to an actual restaurant, there's
an opportunity to make changes as well. And

767
01:21:15,520 --> 01:21:20,280
these data are from a study we conducted with
the Silver Diner, a full-service regional

768
01:21:20,280 --> 01:21:26,409
restaurant chain which made changes to its
kids menu in 2012. And what they did is they

769
01:21:26,409 --> 01:21:31,570
offered more healthy kids meals, they changed
the side dishes to be healthier–fruits and

770
01:21:31,570 --> 01:21:35,881
vegetables. And they removed french fries
and soda from the kids menu. And giving you

771
01:21:35,881 --> 01:21:42,790
a sense of what these things actually looked
like, these were some meals that were offered.

772
01:21:42,790 --> 01:21:49,430
And in total, what they did is they changed
things to be healthier in a way that was more

773
01:21:49,430 --> 01:21:54,429
prominent, more prevalent, and by default.
Three things we think are really important

774
01:21:54,429 --> 01:22:00,680
within a restaurant establishment. And so
what actually happened? These data are showing

775
01:22:00,680 --> 01:22:09,929
the changes that were from pre-menu changes
to shortly after the menu changes. Also 2

776
01:22:09,929 --> 01:22:13,550
years later.
And you can see that for healthy children’s

777
01:22:13,550 --> 01:22:17,680
entrees ordered, it went from 3 percent to
almost half, and that was sustained

778
01:22:17,680 --> 01:22:23,810
over time. For healthy children's side dishes
ordered. it went to about three quarters of orders.

779
01:22:23,810 --> 01:22:30,290
For french fries, it decreased from 50 percent
to 22 percent and soda went down from 35 percent

780
01:22:30,290 --> 01:22:37,500
to about 30 percent immediately and further
to 24 percent. So looking over time was important.

781
01:22:37,500 --> 01:22:42,639
So we assured that it wasn't just a new or
novel change, but something that could be

782
01:22:42,639 --> 01:22:49,120
maintained over time. And also these are pretty
dramatic changes over a chain that has about

783
01:22:49,120 --> 01:22:53,640
12 establishments and suggest that this could
be done in other restaurants as well. We

784
01:22:53,640 --> 01:22:58,270
also looked at the restaurant's annual revenue,
and it continued to grow during this period

785
01:22:58,270 --> 01:23:03,520
by 5.3 percent, exceeding that of similar
family dining chains during the same time

786
01:23:03,520 --> 01:23:09,310
period, which is really important because industry
can't afford to lose money, and they're shy

787
01:23:09,310 --> 01:23:14,100
to make changes to the menu that might do
that. So this demonstrates that even with

788
01:23:14,100 --> 01:23:19,260
healthy changes, you can see an increase in
revenue growth.

789
01:23:19,260 --> 01:23:25,160
We also looked at the healthy adult menu orders
before and after when these new healthier

790
01:23:25,160 --> 01:23:29,760
children's menu changes were made to see if
there was spillover ,and other people ordering

791
01:23:29,760 --> 01:23:36,139
at the table that were on the same check actually
had changes as well. So looking at these millions

792
01:23:36,139 --> 01:23:42,040
of orders, we saw for others who were on the
same check at the table, there was an increase

793
01:23:42,040 --> 01:23:47,820
in healthier entrees, healthier sides, healthier
appetizers, and a decrease in soda and dessert.

794
01:23:47,820 --> 01:23:52,730
Again, suggesting they're made in spillover
as a result of the healthier children's menu.

795
01:23:52,730 --> 01:23:58,610
So taking a look at marketing to children
inside quick-service restaurants and differences

796
01:23:58,610 --> 01:24:05,330
by community demographics, we've recently
examined marketing techniques in 165 quick-service

797
01:24:05,330 --> 01:24:11,560
restaurants from five national chains in socially,
economically, and racially ethnic diverse communities

798
01:24:11,560 --> 01:24:16,620
throughout New England. And approximately
95 percent of quick serve restaurants were

799
01:24:16,620 --> 01:24:22,240
marketing less healthy foods, while only 6.5
percent were marketing healthy options.

800
01:24:22,240 --> 01:24:27,960
Some important differences is that more press
promotion advertisements inside and on the

801
01:24:27,960 --> 01:24:33,390
exterior of the quick-serve restaurants, or
in lower income communities. And more child

802
01:24:33,390 --> 01:24:39,210
directed advertisements with cartoon or TV
movie characters, as well as fewer healthy

803
01:24:39,210 --> 01:24:44,020
entree options and more sugar sweetened beverage
and dessert options were on the children's

804
01:24:44,020 --> 01:24:50,210
menu inside quick-serve restaurants in communities
with higher minority populations. Again, underscoring

805
01:24:50,210 --> 01:24:55,040
some of these differences which is really
important when thinking about interventions

806
01:24:55,040 --> 01:25:02,710
and policy changes. Since federal calorie
labeling has become mandatory, there's been

807
01:25:02,710 --> 01:25:08,700
quite good compliance across U.S. restaurant
chains shown here. And a recent study actually

808
01:25:08,700 --> 01:25:15,060
looked at mean purchase calories per transaction
after franchise labeling was started in April

809
01:25:15,060 --> 01:25:21,640
2017 and nationwide implementation in April
2018. Data were collected in three states

810
01:25:21,640 --> 01:25:26,820
and then modeled to represent what might happen
nationally. And you can see that there was

811
01:25:26,820 --> 01:25:32,140
a slight decrease in calories per transaction,
which is quite promising and suggests that

812
01:25:32,140 --> 01:25:36,440
calorie labeling may have an impact on consumer
behavior.

813
01:25:36,440 --> 01:25:41,639
One of the findings in this study was that
it wasn't even across communities and by census

814
01:25:41,639 --> 01:25:49,190
track, those that were lower income actually
showed less of a decrease in calories ordered

815
01:25:49,190 --> 01:25:55,040
compared to the higher income communities.
So, thank you. I look forward to the question

816
01:25:55,040 --> 01:25:56,650
and answer period.

817
01:25:56,650 --> 01:26:09,000
DR. GUADALUPE AYALA: Thank you. Those were
fantastic presentations. Thank you, Alice,

818
01:26:09,000 --> 01:26:16,270
and thank you, Chris for leading us in this
effort. Really appreciated it. We received

819
01:26:16,270 --> 01:26:22,520
a lot of great questions. So we'll go ahead
and get started. Let's see. The first question,

820
01:26:22,520 --> 01:26:27,270
I will direct this to Alice, although I know
with most of these questions, both Alice and

821
01:26:27,270 --> 01:26:32,989
Chris could probably answer them. For the
first question: Is there a supportive…is

822
01:26:32,989 --> 01:26:39,730
there supportive evidence that demonstrations
of healthy food preparation, meals, providing

823
01:26:39,730 --> 01:26:45,270
in-store education about how to read labels
and interpret nutrition quality and calories

824
01:26:45,270 --> 01:26:53,071
per serving within supermarkets in rural and
urban settings helpful, specifically with

825
01:26:53,071 --> 01:26:58,230
respect to improving self motivation and
health outcomes such as obesity?

826
01:26:58,230 --> 01:27:03,640
DR. ALICE AMMERMAN: Great. Well, I will take
a stab at that and others chime in —and

827
01:27:03,640 --> 01:27:07,060
Suchi, you too—because you've done a lot
of work in this area. So I would say a number

828
01:27:07,060 --> 01:27:13,920
of those things have been tested but not necessarily
in the grocery store or retail setting. Like,

829
01:27:13,920 --> 01:27:18,239
and then sometimes like with many things,
we see change in knowledge, like reading labels,

830
01:27:18,239 --> 01:27:24,340
and we see people's ability to improve their
knowledge, although we're constantly also

831
01:27:24,340 --> 01:27:28,610
trying to improve the label to make it more
readable. But whether or not that translates

832
01:27:28,610 --> 01:27:33,890
into longer term behavior change I think
is a question. I think taste-testing, we all

833
01:27:33,890 --> 01:27:39,909
know is important from a number of interventions,
especially with children, but then maybe referring

834
01:27:39,909 --> 01:27:45,460
question or maybe referring to like sampling
in a grocery store whether or not something

835
01:27:45,460 --> 01:27:51,110
like that actually leads to improved intake.
I don't think that's been studied.

836
01:27:51,110 --> 01:27:56,060
DR. GUADALUPE AYALA: Awesome. Thank you. We'll
move on to the next question just because

837
01:27:56,060 --> 01:28:02,040
we have a few and we have about 10 minutes. OK,
here's a general question that either of you

838
01:28:02,040 --> 01:28:08,190
could answer; Has work life balance, folks
relying on multiple streams of income and

839
01:28:08,190 --> 01:28:14,030
spending less time at home impacted their
increased dependence on restaurant eating?

840
01:28:14,030 --> 01:28:19,309
DR. ALICE AMMERMAN: Chris, I'll take the restaurant,
why don't you start? 

841
01:28:19,309 --> 01:28:27,040
CHRISTINA ECONOMOS: Yeah sure, I can take that. You know, we have demonstrated that
in one of our studies where we were looking

842
01:28:27,040 --> 01:28:35,950
at mother/child diets who were new immigrants
to the U.S., we saw that time stress measured

843
01:28:35,950 --> 01:28:45,420
in a validated way really did translate to
more processed and take-out restaurant food.

844
01:28:45,420 --> 01:28:52,410
And looked at the fact that working multiple
jobs cuts into food preparation time, and

845
01:28:52,410 --> 01:29:00,140
even procurement. And it's much easier to
get something that's prepared. We know that,

846
01:29:00,140 --> 01:29:05,880
in general, prepared takeout restaurant food
contributes less nutritional value to the

847
01:29:05,880 --> 01:29:13,739
diet. But I really want to underscore the
hope and promise that I alluded to and some

848
01:29:13,739 --> 01:29:19,230
of the studies that I showed. Reformulation
is often driven by the science and by policies

849
01:29:19,230 --> 01:29:25,660
that are put into place. So supply and demand
are both important. And if we can work to

850
01:29:25,660 --> 01:29:32,100
make sure the quality of prepared food, particularly
for individuals with low income, even on SNAP

851
01:29:32,100 --> 01:29:38,570
where that is possible, I think in California,
maybe it is but not across the country or

852
01:29:38,570 --> 01:29:45,969
it's going to go a long way, as people are
forced to work multiple jobs and have other

853
01:29:45,969 --> 01:29:50,020
responsibilities like child care and elder
care, etc.

854
01:29:50,020 --> 01:29:56,940
So high quality prepared food shouldn't just
be something that is available to people with

855
01:29:56,940 --> 01:29:57,940
advantage.

856
01:29:57,940 --> 01:30:03,369
DR. GUADALUPE AYALA: Thank you, Chris. OK,
next question. Alice, I'll direct this one

857
01:30:03,369 --> 01:30:10,230
to you: Can you talk more about online ordering
for SNAP participants and what needs to be

858
01:30:10,230 --> 01:30:16,090
done to decrease barriers among participants.
For example, knowing about online options,

859
01:30:16,090 --> 01:30:18,910
navigating online, perceived charges, and
so on.

860
01:30:18,910 --> 01:30:25,960
DR. ALICE AMMERMAN: Yes, of course, to start
with, it's kind of the logistics of how it

861
01:30:25,960 --> 01:30:30,330
works, then, you know, it started with a small
pilot, where just a few of the bigger grocery

862
01:30:30,330 --> 01:30:36,380
chains were able to do the online ordering,
particularly for SNAP. Online ordering has

863
01:30:36,380 --> 01:30:41,350
existed for a while, but often you would have
to pay with a credit card if you wanted to

864
01:30:41,350 --> 01:30:47,940
pick up outside the store. So as online ordering
with SNAP becomes more of a possibility that

865
01:30:47,940 --> 01:30:55,739
removes some of that barrier. I think a lot
of people are interested in whether some of

866
01:30:55,739 --> 01:31:00,489
the ordering could be tweaked in some way
through interventions to direct people towards

867
01:31:00,489 --> 01:31:04,600
healthier options.
I'm not sure that results are out of from

868
01:31:04,600 --> 01:31:09,090
those studies yet. And then I think I mentioned
the delivery…home delivery that could be

869
01:31:09,090 --> 01:31:15,310
a cost related barrier to people as that’s
a part of the online ordering for SNAP participants.

870
01:31:15,310 --> 01:31:18,650
Either of you guys want to add to that?

871
01:31:18,650 --> 01:31:24,950
DR. CHRISTINA ECONOMOS: I guess I just want
to add what I've seen in other studies with

872
01:31:24,950 --> 01:31:29,739
SNAP and farmers markets, for example. And
something that we did here at Tufts showed

873
01:31:29,739 --> 01:31:33,219
social media can be beneficial. So if we can
advance our marketing and advertising of some

874
01:31:33,219 --> 01:31:40,620
of the things that are going on within SNAP
and WIC, by using social media and meeting

875
01:31:40,620 --> 01:31:46,659
people where they are, I think it will go
a long way in awareness and then hopefully

876
01:31:46,659 --> 01:31:47,659
utilization.

877
01:31:47,659 --> 01:31:54,480
DR. GUADALUPE AYALA: Excellent, thank you.
OK, another question. Again, this could be

878
01:31:54,480 --> 01:32:02,540
directed to either of you. This is related
to combining intervention approaches. So can

879
01:32:02,540 --> 01:32:07,360
you talk about the studies, or maybe elaborate
on what you talked about, that offer healthier

880
01:32:07,360 --> 01:32:11,530
meal options at lower costs and combining
that with nutrition education?

881
01:32:11,530 --> 01:32:16,909
So sort of the combination of those two
strategies and the impact that has on changing

882
01:32:16,909 --> 01:32:20,430
behavior in either the restaurant or grocery
environment.

883
01:32:20,430 --> 01:32:28,739
DR. ALICE AMMERMAN: I'll say a little bit
and then Chris can probably say more. I think

884
01:32:28,739 --> 01:32:34,230
it does kind of point to the importance of
combined interventions that offering lower

885
01:32:34,230 --> 01:32:41,600
cost things without it, whether it be in store
purchasing, or meals, without some education

886
01:32:41,600 --> 01:32:48,440
or maybe exposure to being able to taste it
is probably not going to be successful without

887
01:32:48,440 --> 01:32:54,000
kind of coordinating those efforts. So just
because it's low cost doesn't mean people are

888
01:32:54,000 --> 01:32:58,510
going to buy it but if they obviously taste
is really important. If it tastes good, and

889
01:32:58,510 --> 01:33:03,739
people understand that it may be beneficial
for their health, then putting those two together.

890
01:33:03,739 --> 01:33:11,520
Clearly, the cost of food is really important
for low income folks to be able to eat a healthy

891
01:33:11,520 --> 01:33:16,119
diet. So looking for options like that is
really important.

892
01:33:16,119 --> 01:33:21,310
DR. CHRISTINA ECONOMOS: Yeah, and I guess
I would just underscore that combining things

893
01:33:21,310 --> 01:33:27,290
it’s really important. And in some of the
work we did looking at reviews of calorie

894
01:33:27,290 --> 01:33:31,690
labeling, it's clear that action-oriented
messaging in the form of education is really

895
01:33:31,690 --> 01:33:39,260
important to pair with just the numeric strategy
of labeling with calories.

896
01:33:39,260 --> 01:33:45,270
So it's not just the knowledge part, but it's
also making it actionable, and often making

897
01:33:45,270 --> 01:33:51,430
it positive as well is really important. And
the other thing that I would say is when I

898
01:33:51,430 --> 01:33:57,110
mentioned the work on the Silver Diner, I
talked about prevalence, prominence, and by

899
01:33:57,110 --> 01:34:03,130
default. There's a combination of things that
were done there, increasing the number, increasing

900
01:34:03,130 --> 01:34:05,540
where it's shown, and then that default strategy
is really important. So that's what we felt

901
01:34:05,540 --> 01:34:09,920
in that particular study was really important.
And then the last thing that's harder to do,

902
01:34:09,920 --> 01:34:18,960
but it has been beneficial in a number of
studies, are incentives, and kind of, rewards

903
01:34:18,960 --> 01:34:25,520
to get people to start making a behavior change
that hopefully they will maintain.

904
01:34:25,520 --> 01:34:31,040
DR. GUADALUPE AYALA: Excellent, thank you
both. Building on that last comment, Chris,

905
01:34:31,040 --> 01:34:35,460
and this is a question for both of you as
well. So thinking of influential people in

906
01:34:35,460 --> 01:34:41,250
our lives, including primary care physicians,
can you talk a little bit about the state

907
01:34:41,250 --> 01:34:47,060
of the science related to food prescription
programs and their influence on changing

908
01:34:47,060 --> 01:34:51,030
behavior, whether it's grocery shopping or
restaurant consumption?

909
01:34:51,030 --> 01:34:57,770
Yeah, there's quite a bit of work going on
in that area. And some of it is coming out

910
01:34:57,770 --> 01:35:02,880
of Tufts. I would say that the evidence is
emerging in this area. And there's definitely

911
01:35:02,880 --> 01:35:08,560
some promise for specific populations particularly
those that have a diagnosis. But I think in

912
01:35:08,560 --> 01:35:16,090
terms of going to the entire population, there
still needs to be a lot more research to see

913
01:35:16,090 --> 01:35:21,139
if that can be helpful. I would say again,
that has to be paired with affordability.

914
01:35:21,139 --> 01:35:27,070
So prescribing high quality foods without
making them affordable is going to be very,

915
01:35:27,070 --> 01:35:31,460
very difficult. So in combination with our
federal food assistance programs, I think

916
01:35:31,460 --> 01:35:33,170
has a lot of promise.

917
01:35:33,170 --> 01:35:38,610
DR. CHRISTINA ECONOMOS: Just to add a little
bit, we often think of physicians and other

918
01:35:38,610 --> 01:35:44,610
providers as being cheerleaders. They don't
have a lot of time to spend going and sometimes

919
01:35:44,610 --> 01:35:48,860
don't have the background to give a lot of
detailed information. But they can certainly

920
01:35:48,860 --> 01:35:54,960
say this is really important for you. Make
a referral. We've done a small food prescription

921
01:35:54,960 --> 01:35:59,870
program where we use kind of CSA type boxes.
And, of course, you have to do a lot going

922
01:35:59,870 --> 01:36:05,690
along with that with how you use what's in the
box. And we found that getting just the pickup

923
01:36:05,690 --> 01:36:09,790
of the box could be challenging. Sometimes
the prescription programs give you a coupon

924
01:36:09,790 --> 01:36:14,110
to take to a farmer's market. But that may
not be something where people are most comfortable

925
01:36:14,110 --> 01:36:18,619
being people have tried for years to make
farmer's markets more comfortable for a broader

926
01:36:18,619 --> 01:36:22,570
socioeconomic range, and I think we still
have work to do there.

927
01:36:22,570 --> 01:36:24,310
DR. GUADALUPE AYALA: Great, thank you.

928
01:36:24,310 --> 01:36:28,240
DR. ALICE AMMERMAN: Chris, you should add
something there. You've done work in that

929
01:36:28,240 --> 01:36:29,240
area.

930
01:36:29,240 --> 01:36:32,840
DR. GUADALUPE AYALA: That's okay. There’s
a lot of questions there. I could answer many

931
01:36:32,840 --> 01:36:38,130
of them. Here's another question though. And
it was posed actually during session four,

932
01:36:38,130 --> 01:36:42,090
but we never got to it. And I'm curious to
know if anybody has an answer to this one:

933
01:36:42,090 --> 01:36:47,480
Is there any measurement for food preparation
skills, healthy eating knowledge and consumer

934
01:36:47,480 --> 01:36:54,230
purchasing knowledge at the community level,
like county level or census track level?

935
01:36:54,230 --> 01:37:02,710
DR. ALICE AMMERMAN: So food preparation skills?
Food preparation skills, healthy knowledge,

936
01:37:02,710 --> 01:37:04,610
healthy eating knowledge, consumer purchasing
knowledge?

937
01:37:04,610 --> 01:37:09,820
DR. CHRISTINA ECONOMOS: Yeah, I mean, there's
some nutritional literacy scales that have

938
01:37:09,820 --> 01:37:17,260
come out which I think are more comprehensive
than just nutrition education and kind of

939
01:37:17,260 --> 01:37:23,520
bring, I know this one for children that we
worked on here at Tufts. Sarah Amin, URI, was

940
01:37:23,520 --> 01:37:31,949
leading that work. But I think it's really
important. So I would say, looking more into

941
01:37:31,949 --> 01:37:34,060
the food literacy area.

942
01:37:34,060 --> 01:37:38,310
DR. ALICE AMMERMAN: I will say that some of
the students who helped me put together this

943
01:37:38,310 --> 01:37:43,980
presentation are particularly interested
in culinary skills and food prep and how that

944
01:37:43,980 --> 01:37:48,040
can affect diet. So hopefully, we'll have
a future generation of researchers looking

945
01:37:48,040 --> 01:37:49,040
at that.

946
01:37:49,040 --> 01:37:52,980
DR. GUADALUPE AYALA: That's great, always
promoting the next generation, Alice. Thank

947
01:37:52,980 --> 01:38:00,060
you for that. Chris, here's a question for
you: Were you able to control for how marketing

948
01:38:00,060 --> 01:38:05,870
outside of a restroom affected restaurant
food choices by children and families?

949
01:38:05,870 --> 01:38:13,150
DR. CHRISTINA ECONOMOS: In the particular
studies I showed, no. But we actually did

950
01:38:13,150 --> 01:38:21,119
another study where we marketed healthy alternatives,
meaning size and beverages, community wide.

951
01:38:21,119 --> 01:38:30,680
And we did that using traditional things like
billboards, and radio, and lots of other community

952
01:38:30,680 --> 01:38:33,869
advertising.
And in that case, we were able to substitute

953
01:38:33,869 --> 01:38:40,349
some of the traditional advertising that was
going on by buying billboard space and buying

954
01:38:40,349 --> 01:38:46,219
radio time, and within the buses, signs like
that. So you're actually displacing so to

955
01:38:46,219 --> 01:38:52,060
speak, which I think is really important,
instead of just adding something healthy to

956
01:38:52,060 --> 01:38:57,080
the mix without displacing something unhealthy.
And those results are forthcoming. And we've

957
01:38:57,080 --> 01:39:01,949
taken that work and turned it into a social
media campaign for scale because those are

958
01:39:01,949 --> 01:39:06,360
really expensive studies to do with placement
in a community. But using social media, I

959
01:39:06,360 --> 01:39:12,719
think it's a way that we can, again, really
meet people with what they're utilizing on

960
01:39:12,719 --> 01:39:14,030
a daily basis.

961
01:39:14,030 --> 01:39:20,480
DR. GUADALUPE AYALA: Great, thank you. This
is our last question. And it's very much along

962
01:39:20,480 --> 01:39:26,179
the themes of Dr. Chilton’s talk yesterday,
and feel free either of you to answer: So

963
01:39:26,179 --> 01:39:32,000
in terms of policy, is there a push within
nutrition sciences to improve living wages,

964
01:39:32,000 --> 01:39:40,310
as it relates to cost of living across neighborhoods?
Also, what are some, if any, policies being

965
01:39:40,310 --> 01:39:45,860
promoted to address structural racism and
injustice, and marketing low value, nutritious

966
01:39:45,860 --> 01:39:48,210
food, depending on community demographics?

967
01:39:48,210 --> 01:39:53,220
DR. ALICE AMMERMAN: There's a lot packed into
that question.

968
01:39:53,220 --> 01:39:54,220
DR. CHRISTINA ECONOMOS: Yes.

969
01:39:54,220 --> 01:39:57,293
DR. GUADALUPE AYALA: A lot, yes. Maybe it’s
a reflection question too. 

970
01:39:57,293 --> 01:40:03,790
DR. ALICE AMMERMAN: I'll mention one thing that actually just came up with one of my
classes and students, I'm teaching a nutrition

971
01:40:03,790 --> 01:40:09,060
policy class now where students were asking
about living wage, and in North Carolina,

972
01:40:09,060 --> 01:40:13,190
we've been looking at school lunch staff
and the fact that they are not necessarily

973
01:40:13,190 --> 01:40:19,460
paid at a rate of other state employees, and
trying to actually working on legislation

974
01:40:19,460 --> 01:40:24,239
to try to change that. And then also, some
of you may know that school food services

975
01:40:24,239 --> 01:40:29,349
often charge an indirect rate by the school
so they actually have to pay for the use of

976
01:40:29,349 --> 01:40:33,940
it which cuts down on the amount of money
they have to pay their staff and to produce

977
01:40:33,940 --> 01:40:40,710
healthy meals. So those are a couple of kind
of small but important policies that could be addressed.

978
01:40:40,710 --> 01:40:46,580
DR. CHRISTINA ECONOMOS: Yeah, I think it's a really
important question. Certainly the restaurant

979
01:40:46,580 --> 01:40:52,469
industry has been working on minimum wage
and living wage. With the question about are

980
01:40:52,469 --> 01:40:57,010
nutrition scientists working on that, I think
it's a really important leverage point.

981
01:40:57,010 --> 01:41:05,409
If we want the entire food system to change,
then making sure that people are making enough

982
01:41:05,409 --> 01:41:12,060
money is really important for a variety of
reasons. The second question around health

983
01:41:12,060 --> 01:41:19,270
equity, and I think there's a lot of work going
on there. There needs to be more. But I think

984
01:41:19,270 --> 01:41:24,840
both of us presented data stratified as best
as possible from national surveillance data

985
01:41:24,840 --> 01:41:30,620
so we can start to look deeper into what's
actually going on. And I would say the more

986
01:41:30,620 --> 01:41:35,620
we can do that, the better because if you
look at the average in any particular area,

987
01:41:35,620 --> 01:41:41,540
or for a variable, it does not tell the story
at all. So I guess just encouraging people

988
01:41:41,540 --> 01:41:48,160
to pull the data apart as much as possible
and really look more deeply at who's impacted

989
01:41:48,160 --> 01:41:57,780
and then trying to go deeper to understand
what the drivers are, and why that impacts

990
01:41:57,780 --> 01:41:58,780
the current.

991
01:41:58,780 --> 01:42:00,250
DR. GUADALUPE AYALA: Right, that's great food
for thought to leave us with. Thank you both

992
01:42:00,250 --> 01:42:05,739
for sharing your time. Today was a really
interesting discussion, great presentations.

993
01:42:05,739 --> 01:42:11,270
So thank you for your leadership in the field
overall. To all of the participants, thank

994
01:42:11,270 --> 01:42:16,830
you for your very provocative questions, and
really helping us think about this field and how to

995
01:42:16,830 --> 01:42:24,210
move forward. So, we are now at a break time,
we have about 15 minutes. Starting the networking

996
01:42:24,210 --> 01:42:30,349
chat which you can engage with fellow attendees
on the topics just discussed or any other

997
01:42:30,349 --> 01:42:38,050
topics. We invite you to visit the networking
lounge and enjoy the conversation until about

998
01:42:38,050 --> 01:42:44,770
2:25 Eastern Time. Until then…and then after
that you can join us for the third panel, which

999
01:42:44,770 --> 01:42:50,679
is on measurement of food access and neighborhood
food environment. So thank you everyone. Thank

1000
01:42:50,679 --> 01:42:54,159
you panelists, thank you audience and we will
see you all shortly.

1001
01:42:54,159 --> 01:42:55,940
DR. ALICE AMMERMAN: Thank you.

1002
01:42:55,940 --> 01:43:01,929
DR. KAREN GLANZ: Good afternoon, everyone,
I hope you all had a good short break and

1003
01:43:01,929 --> 01:43:07,721
are ready for the next session. In this panel,
we're going to review measures of community

1004
01:43:07,721 --> 01:43:13,450
and consumer neighborhood and food environments,
including geographic information systems,

1005
01:43:13,450 --> 01:43:19,800
technologies, and technology-aided measurements.
Speakers will address what's known about the

1006
01:43:19,800 --> 01:43:24,281
validity and reliability of these measures.
Provide examples, and I think will provide

1007
01:43:24,281 --> 01:43:33,920
some provocative thoughts about where we're
going in the future. We are expecting Leslie

1008
01:43:33,920 --> 01:43:40,780
Lytle to join us to moderate the panel, but
as at the moment, she's having some technical

1009
01:43:40,780 --> 01:43:47,550
challenges. So I'm going to go right ahead
and introduce our panelists. We're covering

1010
01:43:47,550 --> 01:43:54,170
really three different areas. The first being
what I described earlier as community food

1011
01:43:54,170 --> 01:44:01,469
environment measurement. Next, measuring the
consumer food environment and then technology-aided

1012
01:44:01,469 --> 01:44:10,280
measurement of the food environment. Our first
panelist is Dr. Alana Rhone of the USDA's

1013
01:44:10,280 --> 01:44:15,739
Economic Research Service. She'll discuss
measuring the community food environment.

1014
01:44:15,739 --> 01:44:22,500
Next, Dr. Alison Gustafson of the University
of Kentucky will discuss measuring the consumer

1015
01:44:22,500 --> 01:44:28,170
food environment, and our third speaker will
be Dr. Marta Jankowska of the Beckman Research

1016
01:44:28,170 --> 01:44:34,099
Institute at the City of Hope, who will focus
on some innovative technology approaches to

1017
01:44:34,099 --> 01:44:38,849
measuring the food environment. And before
we get started, I want to encourage people,

1018
01:44:38,849 --> 01:44:45,820
as in the previous sessions, to ask questions
in the chat box, and we will get to as many

1019
01:44:45,820 --> 01:44:51,486
of those as we can during the Q&A. Alright,
let's jump right in, Alana.

1020
01:44:55,261 --> 01:44:57,080
DR. ALANA RHONE: ...I have the great pleasure

1021
01:44:57,080 --> 01:45:04,380
in presenting the Food Environment, Food Store
Access, and How to Measure the Food Environment

1022
01:45:04,380 --> 01:45:13,400
Using Geographical Information Systems. I
must note before I begin, the findings and

1023
01:45:13,400 --> 01:45:18,630
conclusions in this presentation are those
of myself and should not be construed to represent

1024
01:45:18,630 --> 01:45:29,119
any official USDA or U.S. government determination
or policy. The food environments and food

1025
01:45:29,119 --> 01:45:35,980
access is important because easier access
to retailers that sell healthy and affordable

1026
01:45:35,980 --> 01:45:44,900
food may allow consumers to be less reliant
on easily accessible sources of less healthy

1027
01:45:44,900 --> 01:45:54,160
food and improve their diet quality. Also,
access to a supermarket or large grocery store

1028
01:45:54,160 --> 01:46:01,650
is a problem for a small percentage of households.
Some consumers are constrained in their ability

1029
01:46:01,650 --> 01:46:06,440
to access healthy and affordable foods. About
14 percent of households live in low-income

1030
01:46:06,440 --> 01:46:13,660
neighborhoods, with a significant number of
households that do not have access to a vehicle

1031
01:46:13,660 --> 01:46:19,560
and also live far from the nearest source
of healthy food.

1032
01:46:19,560 --> 01:46:28,590
Easier access to retailers that sell healthy
and affordable food may also make it easier

1033
01:46:28,590 --> 01:46:37,730
for some Americans with access barriers to
achieve food security. In 2019, 10.5 percent

1034
01:46:37,730 --> 01:46:43,070
of American households were food insecure
or lacked enough food for an active, healthy

1035
01:46:43,070 --> 01:46:50,580
life for all household members at some time
in the year, and about 4.1 percent had very

1036
01:46:50,580 --> 01:46:57,670
low food security. Where at times food intake
of at least one household member was reduced

1037
01:46:57,670 --> 01:47:07,389
or eating patterns were disrupted. The economic
research service has many data products. There

1038
01:47:07,389 --> 01:47:15,000
are two mapping tools that are among those
products, and they look at the food environment

1039
01:47:15,000 --> 01:47:23,349
and food access specifically. Both mapping
tools utilize GIS technology to assess the

1040
01:47:23,349 --> 01:47:32,770
food environment. The first mapping tool is
the Food Environment Atlas. The Food Environment

1041
01:47:32,770 --> 01:47:40,909
Atlas assembles statistics on food environment
indicators and also provides a spatial overview

1042
01:47:40,909 --> 01:47:49,960
of a community's ability to access healthy
food. The Atlas assembles statistics on three

1043
01:47:49,960 --> 01:47:58,330
broad categories of food environment factors.
The first is food choices. Indicators of a

1044
01:47:58,330 --> 01:48:05,600
community's access to and acquisition of healthy,
affordable foods such as the number of food

1045
01:48:05,600 --> 01:48:12,800
stores and restaurants in a county; food and nutrition assistance program participation,

1046
01:48:12,800 --> 01:48:20,080
like the National School Lunch Program; and
local foods, such as information on food hubs,

1047
01:48:20,080 --> 01:48:29,679
the farm-to-school program, and the number
of farms in a county. The second category

1048
01:48:29,679 --> 01:48:39,800
is health and wellbeing. Indicators of a community's
success in maintaining healthy diets such

1049
01:48:39,800 --> 01:48:45,909
as food insecurity and diabetes and obesity
rates. The third category is community characteristics,

1050
01:48:45,909 --> 01:48:52,670
indicators of a community's characteristics
that might influence the food environment,

1051
01:48:52,670 --> 01:49:02,010
such as demographic composition and income
and poverty, among other things. This map

1052
01:49:02,010 --> 01:49:09,300
is from the Food Environment Atlas and shows
the number of farmers markets in a county

1053
01:49:09,300 --> 01:49:16,010
in 2018. The dark maroon-colored areas are
counties that have more than three farmers'

1054
01:49:16,010 --> 01:49:22,619
markets, and the yellow areas are counties
that have zero farmers' markets.

1055
01:49:22,619 --> 01:49:29,500
Using GIS software, we can map all our indicators
to give a better picture of the food environment

1056
01:49:29,500 --> 01:49:36,330
in a county and also see if there is any kind
of correlation between any of the indicators.

1057
01:49:36,330 --> 01:49:48,119
The next mapping tool is the Food Access Research
Atlas. The Food Access Research Atlas offers

1058
01:49:48,119 --> 01:49:56,840
census tract-level data on food access that
can be viewed, downloaded, or printed for government

1059
01:49:56,840 --> 01:50:05,510
community planning, or research purposes. The
Food Access Resources Atlas provides a spatial

1060
01:50:05,510 --> 01:50:13,119
overview of food access indicators for low-income
and other census tracts using different measures

1061
01:50:13,119 --> 01:50:21,250
of supermarket accessibility. It also estimates
food access data for the overall population

1062
01:50:21,250 --> 01:50:29,790
and subgroups within census tracts, such as
households without vehicles and SNAP participants.

1063
01:50:29,790 --> 01:50:36,710
The Atlas includes four measures of low
income and low access census tracts. The first

1064
01:50:36,710 --> 01:50:44,260
three measures are defined by low income status
of a census tract and proximity to the nearest

1065
01:50:44,260 --> 01:50:52,260
grocery store for a significant number or
share of people using different demarcations.

1066
01:50:52,260 --> 01:50:59,920
These measures are low income and low access
at 1 and 10 miles. Low income and low access

1067
01:50:59,920 --> 01:51:06,780
at half and 10 miles, and low income and low
access at 1 and 20 miles. The half-mile

1068
01:51:06,780 --> 01:51:15,360
and 1-mile demarcations are for urban areas,
and the 10 and 20 miles are for rural areas.

1069
01:51:15,360 --> 01:51:24,239
For example, the low income and low access
measure at 1 and 10 miles are low-income

1070
01:51:24,239 --> 01:51:31,210
census tracts, where a significant number
or share of the population is greater than

1071
01:51:31,210 --> 01:51:37,250
1 mile from the nearest supermarket, supercenter, or large grocery store for an urban

1072
01:51:37,250 --> 01:51:47,170
area, or greater than 10 miles for a rural
area. The fourth measure, which is the low

1073
01:51:47,170 --> 01:51:54,710
income and low access using vehicle access
and 20 miles are low-income census tracts,

1074
01:51:54,710 --> 01:52:01,750
where a significant number of households did
not have access to a vehicle and are more

1075
01:52:01,750 --> 01:52:08,599
than half a mile from the nearest supermarket.
Or a significant number or share of residents

1076
01:52:08,599 --> 01:52:17,869
are more than 20 miles from the nearest supermarket.
As mentioned, there are four measures of low

1077
01:52:17,869 --> 01:52:22,210
income and low access in the Food Access Research
Atlas.

1078
01:52:22,210 --> 01:52:29,580
However, in this slide, I will only be going
over two; the low income and low access at

1079
01:52:29,580 --> 01:52:37,920
one and 10 miles measure and the low income
and low access using vehicle access and 20

1080
01:52:37,920 --> 01:52:45,889
miles measure. These estimates show the change
in low-income and low supermarket access census

1081
01:52:45,889 --> 01:52:57,119
tracts from 2015 to 2019. The number of census
tracts classified as low income decreased

1082
01:52:57,119 --> 01:53:06,040
slightly from 2015 to 2019. This reflects
improvements in household income, whereas

1083
01:53:06,040 --> 01:53:18,291
from 2010 to 2015, the number of low-income
census tracts increased. The number of low

1084
01:53:18,291 --> 01:53:26,849
access, one and 10 miles census tracks increased
slightly from 27,527 census tracts in 2015

1085
01:53:26,849 --> 01:53:37,060
to 27,548 in 2019. When vehicle availability
and proximity to a supermarket are considered

1086
01:53:37,060 --> 01:53:51,699
together, such as in the low income and low
access using vehicle access and 20 miles,

1087
01:53:51,699 --> 01:54:01,920
estimates show a slight decrease of a little
over 1 percent point in this share of tracts

1088
01:54:01,920 --> 01:54:11,139
in 2019. When low income and low access are
compiled together, the number of low income

1089
01:54:11,139 --> 01:54:17,170
and low access tracts at the 1 and 10 miles
measure increased slightly.

1090
01:54:17,170 --> 01:54:22,849
And the low income and low access using vehicle
access and 20 miles measure decreased slightly.

1091
01:54:22,849 --> 01:54:39,170
A neat feature of the Atlas is by utilizing
GIS software, we can create maps showing food

1092
01:54:39,170 --> 01:54:47,119
access indicators by census tract using different
measures and indicators of supermarket accessibility.

1093
01:54:47,119 --> 01:54:55,659
Here is an example of how the Food Access
Research Atlas can be used. This map shows

1094
01:54:55,659 --> 01:55:03,369
low income and low access areas in the DMV,
DC, Maryland, and Virginia areas in 2019.

1095
01:55:03,369 --> 01:55:14,020
The yellow areas are census tracts that meet
the low income and low access using vehicle

1096
01:55:14,020 --> 01:55:22,940
access and 20 miles definition. The kelly
transparent green areas are areas that meet

1097
01:55:22,940 --> 01:55:29,989
the low income and low access definition at
the one and 10 miles measure. And the bright

1098
01:55:29,989 --> 01:55:37,730
lime green color tracts are tracts that meet
both definitions. These tracts may be the

1099
01:55:37,730 --> 01:55:41,780
worst off because they have many people without
vehicles and many who are quite far from the

1100
01:55:41,780 --> 01:55:47,430
nearest supermarket. Now I have overlaid our
map with a map of the wards in D.C. to get a

1101
01:55:47,430 --> 01:55:52,050
richer picture of where exactly these areas
are in the D.C.

1102
01:55:52,050 --> 01:56:02,699
As you can see, most of these areas are in
wards 7 and 8. These are Northeast

1103
01:56:02,699 --> 01:56:11,810
and Southeast DC. Now, let's take a look at
other ways using GIS technology can be used

1104
01:56:11,810 --> 01:56:23,050
to assess the Food Environment. As mentioned
before, for both mapping tools, the Food Environment

1105
01:56:23,050 --> 01:56:31,980
Atlas and the Food Access Research Atlas we
use ESRI GIS mapping software. In the Food

1106
01:56:31,980 --> 01:56:40,000
Access Research Atlas GIS tools are used to
assess distance to the nearest food store

1107
01:56:40,000 --> 01:56:49,840
for populations. Downscaling methods are used
to disaggregate to half kilometer square grids.

1108
01:56:49,840 --> 01:56:57,660
So why is downscaling using GIS tools important?
We downscale because we want to measure populations

1109
01:56:57,660 --> 01:57:03,880
at various distances to stores. For example.
And the first picture, which shows an image

1110
01:57:03,880 --> 01:57:13,330
of low-income census tracts in Little Rock,
Arkansas in 2019. Let's focus only on the

1111
01:57:13,330 --> 01:57:19,530
census tract boundaries, which are outlined
in gray. People live in these census tracts.

1112
01:57:19,530 --> 01:57:27,870
However, just by looking at the map, we don't
know if the census tract has areas that are

1113
01:57:27,870 --> 01:57:35,119
densely populated and are sparsely populated.
Also, food stores are sometimes distributed

1114
01:57:35,119 --> 01:57:47,010
unevenly around an area. Downscaling, which
is used in the second image, which is a pixilated

1115
01:57:47,010 --> 01:57:53,460
version of population distribution in Little
Rock, Arkansas in 2019, gives us a more granular

1116
01:57:53,460 --> 01:57:57,389
measure of population distribution within
the tract. The yellow areas have less than

1117
01:57:57,389 --> 01:58:06,360
10,000 people. The navy blue areas have less
than 100 people, and the gray areas have less

1118
01:58:06,360 --> 01:58:17,760
than 50 people on the map. The red lines are
interstate freeways. So downscaling is like

1119
01:58:17,760 --> 01:58:25,270
a distribution and density measure, all in
one. If we can downscale populations using

1120
01:58:25,270 --> 01:58:31,250
GIS technology, to those smaller areas, such
as half kilometer square grids just like

1121
01:58:31,250 --> 01:58:38,210
used in the second image, we can describe
access in populations a lot better. The data

1122
01:58:38,210 --> 01:58:46,040
products, and estimates, and this presentation
can be found on the ERS webpages. And this

1123
01:58:46,040 --> 01:58:52,310
concludes my presentation. Thank you.

1124
01:58:52,310 --> 01:59:00,790
DR. ALISON GUSTAFSON: Hello. I'm Dr. Alison Gustafson,
an associate professor here at the University

1125
01:59:00,790 --> 01:59:05,420
of Kentucky. And I'll be presenting today
on the consumer food environment measurement

1126
01:59:05,420 --> 01:59:13,000
approaches, challenges, and opportunities.
The goal of our talk today is first to understand

1127
01:59:13,000 --> 01:59:18,770
what the consumer food environment is in the
United States. Second, examine current measures

1128
01:59:18,770 --> 01:59:25,010
available to quantify the helpfulness of the
consumer food environment. Third, I'll then

1129
01:59:25,010 --> 01:59:30,349
discuss the gaps in measures related to reliability
and validity of the tools we currently have.

1130
01:59:30,349 --> 01:59:35,119
And lastly, I discuss emerging consumer food
environment changes and the need for enhanced

1131
01:59:35,119 --> 01:59:44,020
measures. First, I'd like to define the consumer
food environment. The definition today is

1132
01:59:44,020 --> 01:59:49,409
availability, price, quality, and variety
of foods for consumers coupled with advertising,

1133
01:59:49,409 --> 01:59:57,579
promotional messages witnessed by consumers.
Why is the consumer food environment so critical?

1134
01:59:57,579 --> 02:00:02,040
Not all stores are created equal. There is
a great variety of pricing, promotion, availability

1135
02:00:02,040 --> 02:00:07,344
of food items within and between different
neighborhoods.

1136
02:00:07,344 --> 02:00:13,909
We also know that in order to provide a sustainable,
healthy, and equitable food system, what is

1137
02:00:13,909 --> 02:00:20,320
inside the store where consumers shop needs
to be measured and addressed, as well as neighborhood

1138
02:00:20,320 --> 02:00:28,980
level factors. Where do we measure the consumer
food environment? We measure food venues where

1139
02:00:28,980 --> 02:00:35,180
consumers, customers, individuals can purchase
food. This includes, but is definitely not

1140
02:00:35,180 --> 02:00:43,710
only limited to these venues, grocery stores,
supercenters, gas stations, convenience stores,

1141
02:00:43,710 --> 02:00:50,620
corner stores, dollar stores, restaurants,
including fast food and fast-casual, vending

1142
02:00:50,620 --> 02:00:57,810
machines, farmers' markets, national parks
and rec locations. And now, most recently,

1143
02:00:57,810 --> 02:01:06,300
online grocery shopping. Dr. Ammerman will
go over this section in far more detail than

1144
02:01:06,300 --> 02:01:11,670
I will, but briefly, I'd like to just note
that in the consumer food environment today,

1145
02:01:11,670 --> 02:01:18,250
65 percent of food is purchased at traditional
stores, such as grocery stores and supercenters.

1146
02:01:18,250 --> 02:01:24,270
So that's majority of my talk today will be
centered around tools or measures that measure

1147
02:01:24,270 --> 02:01:29,530
the helpfulness of grocery stores, given that
most people spend a majority of their food

1148
02:01:29,530 --> 02:01:33,929
dollars at this type of location.
However, we also know that at convenience

1149
02:01:33,929 --> 02:01:39,670
stores such as corners stores and gas stations,
15 to 20 percent of people's food dollars

1150
02:01:39,670 --> 02:01:45,480
used these types of venues. Third, we have
non-traditional or mixed-use, such as dollar

1151
02:01:45,480 --> 02:01:50,520
stores or cooperatives. Ten percent of food
dollars are spent in dollar stores and thus

1152
02:01:50,520 --> 02:01:56,719
this emerging venue is critical to be measured
within the consumer food environment to understand

1153
02:01:56,719 --> 02:02:02,909
the helpfulness of the food that is available
where people shop. Lastly is small-scale locations

1154
02:02:02,909 --> 02:02:09,340
such as farmer's markets or community-supported
agriculture, such as a CSA. People spend less

1155
02:02:09,340 --> 02:02:14,699
than 1 percent of their food dollars are
spent here. That's for today. I will not delve

1156
02:02:14,699 --> 02:02:21,280
deep into the different audit tools that are
available for measuring the helpfulness of

1157
02:02:21,280 --> 02:02:25,380
a farmer's market. However, there are studies
that have utilized different tools to address

1158
02:02:25,380 --> 02:02:33,831
farmer's market helpfulness, and I can point
you in that direction. And now want to move

1159
02:02:33,831 --> 02:02:40,440
on to measuring the consumer food environment,
the tools that we currently have to date.

1160
02:02:40,440 --> 02:02:46,590
The Nutrition Environment Measurement Survey,
developed and tested by Dr. Karen Glanz and

1161
02:02:46,590 --> 02:02:52,180
her study team. NEMS is an observational tool
within the retail space to assess the availability

1162
02:02:52,180 --> 02:02:59,310
of healthy options, price, and quality. This
tool has inter-rater reliability and test-retest

1163
02:02:59,310 --> 02:03:07,360
reliability of approximately 0.73 to 1.0.
This tool is used in stores, convenience stores,

1164
02:03:07,360 --> 02:03:12,930
restaurants, including fast, casual, and fast
food; college campuses, vending, and has been

1165
02:03:12,930 --> 02:03:21,560
adapted in several locations in Texas, Brazil,
and with the Rudd Center. As we mentioned

1166
02:03:21,560 --> 02:03:25,780
in this slide, two sides previously, the consumer
food environment comprises significantly of

1167
02:03:25,780 --> 02:03:30,139
different venues, with a majority of people
spending their dollars at grocery stores and

1168
02:03:30,139 --> 02:03:36,860
convenient type locations. This tool has been
adapted to be used at these different locations,

1169
02:03:36,860 --> 02:03:43,630
which makes it a widely reputable and robust
tool. Next, there's the Thrifty Food Plan

1170
02:03:43,630 --> 02:03:51,040
tool that people use. The USDA provides a
representative helpful and minimal cost meal

1171
02:03:51,040 --> 02:03:53,409
plan.
People then have use, study authors have

1172
02:03:53,409 --> 02:03:59,850
then used this tool in 11 studies to date
to identify food and beverages that could

1173
02:03:59,850 --> 02:04:07,829
be purchased to meet the minimum requirement
for a healthful diet. Thus, as you know, in

1174
02:04:07,829 --> 02:04:12,670
the recent press and also through our recent
administration, there's been an expansion

1175
02:04:12,670 --> 02:04:18,309
of the SNAP benefits based on the Thrifty
Food Plan. What's the minimum requirement

1176
02:04:18,309 --> 02:04:25,119
to meet a healthful diet? So this tool will
then be used in stores to say what is the

1177
02:04:25,119 --> 02:04:32,079
minimum amount of food that's available for
customers and their stores to meet the TFP?

1178
02:04:32,079 --> 02:04:37,480
These are not obviously the only tools available
for measuring the consumer food environment.

1179
02:04:37,480 --> 02:04:45,940
There are a variety of tools that I am categorizing
as healthy baskets. One is the Healthy Eating

1180
02:04:45,940 --> 02:04:52,739
Indicator, or HEI, Shopping Basket. Many of
you might be familiar with that we use HEI

1181
02:04:52,739 --> 02:05:00,639
in dietary surveys. This tool or basket is
used to measure commonly consumed food and

1182
02:05:00,639 --> 02:05:07,489
culturally acceptable food are identified.
Practical considerations included meals, convenience,

1183
02:05:07,489 --> 02:05:14,909
and price.
This tool used 35 items from 17 different

1184
02:05:14,909 --> 02:05:20,770
categories. Another basket is the Healthy
Food Access Basket. This considered demographic

1185
02:05:20,770 --> 02:05:27,420
and food purchasing data were used to define
food in the healthy basket. This basket utilized,

1186
02:05:27,420 --> 02:05:34,739
this tool, excuse me, utilized an approach
of saying "we want to look at the foods most

1187
02:05:34,739 --> 02:05:39,210
commonly consumed." So it's one thing for a
store to have healthy food but if those foods

1188
02:05:39,210 --> 02:05:47,520
are not as commonly purchased, then it might
be irrelevant. And that...thus this group

1189
02:05:47,520 --> 02:05:53,630
utilized purchasing data to define their healthy
basket. They used 44 items were used in their

1190
02:05:53,630 --> 02:05:59,920
basket to meet the nutritional needs of most
families. So this may not look at families

1191
02:05:59,920 --> 02:06:08,150
who have disabled children. This may not include
families who are larger than most. However,

1192
02:06:08,150 --> 02:06:16,130
their approach was to look at purchasing data,
to define the measures. The last one I'll

1193
02:06:16,130 --> 02:06:21,119
talk about is the Market Basket Assessment
Tool. This looked at availability of fruits,

1194
02:06:21,119 --> 02:06:27,309
vegetables, low-fat dairy, eggs, lean meats,
and whole grain among our SNAP retailers.

1195
02:06:27,309 --> 02:06:35,270
They are still waiting on reliability and
validity reporting on the inter-rater reliability

1196
02:06:35,270 --> 02:06:42,020
measures. However, this tool focused more
on SNAP retailers rather than all types of

1197
02:06:42,020 --> 02:06:49,930
consumer food environment. And so again, there's
different tools for different purposes. So,

1198
02:06:49,930 --> 02:06:54,060
what are the most common items measured? The
two most common tools, as we've mentioned,

1199
02:06:54,060 --> 02:06:59,800
are NEMS and TFP. They measure a wide variety
of foods and beverages, including fruits,

1200
02:06:59,800 --> 02:07:04,790
vegetables, dairy, meat, frozen items, canned
items, milk, eggs, whole grain, beverages,

1201
02:07:04,790 --> 02:07:13,829
and a variety of snacks. Other less commonly
cited measures focus on simply fruits and

1202
02:07:13,829 --> 02:07:19,730
vegetables. So many times in market baskets,
these baskets are just focused on fruits and

1203
02:07:19,730 --> 02:07:23,590
vegetables. And I'll get into the good reasons
why they may only want to focus on one type

1204
02:07:23,590 --> 02:07:29,730
of food item. They also may only look at prices
of fruits and vegetables across different

1205
02:07:29,730 --> 02:07:36,810
venues. Different tools have different purposes,
and there is no one perfect measure. However,

1206
02:07:36,810 --> 02:07:42,250
we know that to measure the consumer food
environment, we need to think of the consumer,

1207
02:07:42,250 --> 02:07:51,810
and I think there needs to be more addressing
of the individual within the development of

1208
02:07:51,810 --> 02:07:54,429
these tools.
What is commonly measured among studies? Availability

1209
02:07:54,429 --> 02:08:01,639
is the most common factor that we measure.
Most studies utilizing an audit tool as the

1210
02:08:01,639 --> 02:08:05,760
ones I've previously talked about report on
availability of healthy relative to unhealthy

1211
02:08:05,760 --> 02:08:11,369
food within food venues. So, for example,
people will measure "is there availability

1212
02:08:11,369 --> 02:08:17,269
of fruits and vegetables." They will also then
measure availability of low-fat dairy relative

1213
02:08:17,269 --> 02:08:25,150
to higher fat dairy items as one example.
Price. About half of the studies will report

1214
02:08:25,150 --> 02:08:30,440
on price differences of healthy food relative
to the less healthy foods such as whole milk

1215
02:08:30,440 --> 02:08:37,579
price relative to low fat or skim milk. NEMS
will report on the price of higher fat meat

1216
02:08:37,579 --> 02:08:44,750
relative to lower fat meat, such as ground
beef, at 80 percent versus 90 percent. Quality

1217
02:08:44,750 --> 02:08:50,390
is the third factor. Only a third of the studies
report on quality of food within the stores,

1218
02:08:50,390 --> 02:08:55,600
such as bruises, bumps, or dents in cans.
Quality is a very important factor because

1219
02:08:55,600 --> 02:09:02,489
although the food may be available across
different venues such as apples, if those

1220
02:09:02,489 --> 02:09:08,030
apples are bruised in one venue versus another
venue, the consumer is less likely to buy

1221
02:09:08,030 --> 02:09:13,079
bruised apples, and thus quality is an important
factor that needs to be considered when measuring

1222
02:09:13,079 --> 02:09:18,920
the consumer food environment.
I'd now like to go over four different review

1223
02:09:18,920 --> 02:09:25,659
studies or systematic reviews of tools and
their recommendations. Our first author was

1224
02:09:25,659 --> 02:09:31,750
in 2008 and reported that 40 percent used
reliability or validity testing. Therefore,

1225
02:09:31,750 --> 02:09:36,679
less than half of the studies to date are
using our conducting reliability and validity

1226
02:09:36,679 --> 02:09:43,770
testing. And thus we're not sure if the measures
we're using to conduct our audits is actually

1227
02:09:43,770 --> 02:09:51,130
reliable or valid. Most are observational,
and many were cross-sectional in nature, meaning

1228
02:09:51,130 --> 02:09:58,630
at one-time point. And lastly, this author
addresses how these tools are best suited

1229
02:09:58,630 --> 02:10:04,849
for proposed outcomes needs to be addressed.
And I will get into that in a little bit.

1230
02:10:04,849 --> 02:10:11,179
The next systematic review was done by Dr.
Lytle, who is in this session, and she said

1231
02:10:11,179 --> 02:10:16,820
we need to pay attention to psychometric properties,
including reliability and validity, and address

1232
02:10:16,820 --> 02:10:23,150
utility of the measures across populations.
Thus, Native American populations are highly

1233
02:10:23,150 --> 02:10:29,790
different than white suburban location, and
that we need to address how these measures

1234
02:10:29,790 --> 02:10:33,540
vary across the population.
Tools need to describe their data reduction

1235
02:10:33,540 --> 02:10:38,231
methods and how they can relate to obesogenic
indices. And we need to improve the rigor

1236
02:10:38,231 --> 02:10:49,230
of the study design. I had conducted with
my co-author systematic review in 2011 and

1237
02:10:49,230 --> 02:10:53,920
indicated that neighborhood or place matters.
Store availability is different in geographic

1238
02:10:53,920 --> 02:10:59,719
areas, and we need to provide a multi-level
perspective when measuring the consumer food

1239
02:10:59,719 --> 02:11:05,210
environment. Not one single measure, such
as availability or price, or quality at one

1240
02:11:05,210 --> 02:11:11,020
time point is sufficient to adequately describe
the consumer food environment. Studies need

1241
02:11:11,020 --> 02:11:17,630
to measure availability, price, and quality
over time, and we need to move away from a

1242
02:11:17,630 --> 02:11:23,320
cross-sectional observational manner. Dr.
Karen Glanz recently conducted a systematic

1243
02:11:23,320 --> 02:11:29,650
review and found that the two most common
tools are NEMS and TFP. Again, most studies

1244
02:11:29,650 --> 02:11:36,630
were cross-sectional. Again, standardization
of measures is needed, as well as measurement

1245
02:11:36,630 --> 02:11:42,480
of quality, reliability, and validity. Thus,
all four study authors across all systematic

1246
02:11:42,480 --> 02:11:50,090
reviews across time points in the past 11,
over 11 years, we've all indicated there needs

1247
02:11:50,090 --> 02:11:58,010
to be validity and reliability measures.
There needs to be longitudinal measurement

1248
02:11:58,010 --> 02:12:03,880
of these, and we've all also indicated that
there needs to be a measurement of the geographic

1249
02:12:03,880 --> 02:12:08,909
location around these stores. And lastly,
there needs to be more focus on the development

1250
02:12:08,909 --> 02:12:16,690
of tools related to the outcome of interest.
Are gaps in measurement? I cannot stress enough

1251
02:12:16,690 --> 02:12:21,790
we need to have the validity and reliability
between and within measures. There needs to

1252
02:12:21,790 --> 02:12:26,850
be more sensitivity to detect change over
time regarding price, promotion, and availability.

1253
02:12:26,850 --> 02:12:32,690
For example, there's a new emerging non-meat
products and plant-based meat substitutes,

1254
02:12:32,690 --> 02:12:39,570
as noted by Dr. Karpyn, such as the Impossible
Burger. These tools need to be sensitive to

1255
02:12:39,570 --> 02:12:44,540
detect the change in the helpfulness or
the unhelpfulness for that matter of different

1256
02:12:44,540 --> 02:12:50,820
food items available for our consumers, such
as vegan items. Also changes in beverages

1257
02:12:50,820 --> 02:12:57,130
such as seltzer alcohols. Previous tools may
not be interested in measuring the availability

1258
02:12:57,130 --> 02:13:02,400
of alcohol, however, other tools may want
to make note of the changing landscape of

1259
02:13:02,400 --> 02:13:06,869
alcohol beverages available for consumption,
and those tools need to be sensitive to detect

1260
02:13:06,869 --> 02:13:11,159
these changes.
I will also note that many authors have noted

1261
02:13:11,159 --> 02:13:18,030
that we need to think about additions of culturally
relevant and appropriate foods. Not all cultures

1262
02:13:18,030 --> 02:13:23,239
look at food, fruits, vegetables the same
way. Or would they consume them in the same

1263
02:13:23,239 --> 02:13:29,520
way. We need to get out of a monolithic white
perspective and think of different cultures

1264
02:13:29,520 --> 02:13:36,579
and how they shop for their food. Adaptable
tools are needed, which are dynamic for individual

1265
02:13:36,579 --> 02:13:42,180
and neighborhood changes. Food venues selling
food continue to change, and audits need to

1266
02:13:42,180 --> 02:13:49,160
be readily adaptable. No one audit can be
static over time. We need to have sensitive,

1267
02:13:49,160 --> 02:13:54,679
adaptable tools. And this is holding true
regardless of the venue, whether it's grocery

1268
02:13:54,679 --> 02:14:00,610
stores, fast food or corner stores. Audits
need to be developed and tested from a diverse

1269
02:14:00,610 --> 02:14:06,179
perspective to reflect the cultures and neighborhoods
of our collective consumers. Our consumer

1270
02:14:06,179 --> 02:14:12,170
is not a white middle-aged woman. Our consumer
is anyone from a Latinx Hispanic household

1271
02:14:12,170 --> 02:14:18,690
to an LGBTQ. And there needs to be a wider
lens with how consumers shop within their

1272
02:14:18,690 --> 02:14:23,940
neighborhood and within their food venues.
Tools also need to be relevant for disease

1273
02:14:23,940 --> 02:14:29,460
prevention and obesity prevention, as Dr.
Lytle mentioned, tools need to be used with

1274
02:14:29,460 --> 02:14:35,849
the health outcome in mind. Perhaps we want
to look at a tool used for prevention of cardiovascular

1275
02:14:35,849 --> 02:14:42,840
disease and how does the consumer food environment
meet the prevention of those risk factors?

1276
02:14:42,840 --> 02:14:50,719
How does the tool meet the risk factors for
obesity? These tools need to think of the

1277
02:14:50,719 --> 02:14:56,489
health outcome in mind when being used within
intervention studies. I'd now like to briefly

1278
02:14:56,489 --> 02:14:59,750
move on to the new consumer food environment.

1279
02:14:59,750 --> 02:15:05,190
To make my point about how food venues
are constantly changing and we need tools

1280
02:15:05,190 --> 02:15:11,159
that are adaptable and sensitive to these
changes. We now know that there is a

1281
02:15:11,159 --> 02:15:18,460
gross growth with online shopping. Growth
of online shopping, especially with COVID-

1282
02:15:18,460 --> 02:15:23,940
19 and the emerging rising COVID Delta variant.
Once COVID-19 was confirmed in the United

1283
02:15:23,940 --> 02:15:26,929
States, there was a surge in online shopping
for various food and beverages, coinciding

1284
02:15:26,929 --> 02:15:32,770
with an increase of 48 percent in online sales.
Some business forecast models predict that

1285
02:15:32,770 --> 02:15:39,309
30 percent of all food and beverages will
be purchased online by 2025. There is a large

1286
02:15:39,309 --> 02:15:44,670
variance in availability of online shopping,
whether delivery or pickup between rural and

1287
02:15:44,670 --> 02:15:51,480
urban locations and across different types
of stores. Think Amazon to Walmart to Kroger.

1288
02:15:51,480 --> 02:15:56,920
There are no measures currently which address
price, availability and promotion of online

1289
02:15:56,920 --> 02:16:02,869
shopping. We need tools to be dynamic to be
adapted for new venues, selling new foods

1290
02:16:02,869 --> 02:16:10,159
for all cultures. In summary, all studies
need to engage and report on quality measures

1291
02:16:10,159 --> 02:16:13,829
of reliability and validity.
Longitudinal studies are needed to assess

1292
02:16:13,829 --> 02:16:19,540
changes over time within an intervention.
We need to provide a multi-level perspective

1293
02:16:19,540 --> 02:16:24,540
neighborhood, house, school, individual within
the consumer food environment. This can help

1294
02:16:24,540 --> 02:16:31,300
shed light on where resources are best utilized
to improve health outcomes among those most

1295
02:16:31,300 --> 02:16:36,770
vulnerable. I'd now like to stop and thank
you all for your time and attention and open

1296
02:16:36,770 --> 02:16:40,340
it for questions and comments.

1297
02:16:40,340 --> 02:16:52,960
DR. MARTA JANKOSWKA: Hi, my name is Marta Jankowska I'm
an Associate Professor at the Beckman Research

1298
02:16:52,960 --> 02:16:58,559
Center at the City of Hope Cancer Center in
Duarte, California. Please do feel free to

1299
02:16:58,559 --> 02:17:03,990
follow our lab on Twitter. We're @HDSCALECOLLAB.
Today, I'm going to be talking about two projects

1300
02:17:03,990 --> 02:17:09,309
that are utilizing different types of technologies
to enhance our measurement of the food environment.

1301
02:17:09,309 --> 02:17:13,561
The first uses crowdsourced health data to
come up with some unique food environment

1302
02:17:13,561 --> 02:17:18,421
measures. The second is focused more on measuring
exposure and interaction with the food environment

1303
02:17:18,421 --> 02:17:24,429
using Global Positioning System or G.P.S.
and geographic information systems, or GIS

1304
02:17:24,429 --> 02:17:29,370
data sources. So, measuring the food environment
is notoriously difficult. Government and industry

1305
02:17:29,370 --> 02:17:34,950
data sources are most commonly used, but they
have a lot of well documented issues. So,

1306
02:17:34,950 --> 02:17:40,240
to ensure that these data sets are showing
useful data, often ground truth validation

1307
02:17:40,240 --> 02:17:45,599
needs to be performed, which can be incredibly
costly and time consuming. So, several researchers

1308
02:17:45,599 --> 02:17:49,500
have been looking into crowdsourced data as
an alternative, which can possibly help fill

1309
02:17:49,500 --> 02:17:51,769
some of the content gaps in government and
industry data.

1310
02:17:51,769 --> 02:17:57,510
It can track these data sources for existing
or newly opened closed outlets, and possibly

1311
02:17:57,510 --> 02:18:03,389
reduce the burden of ground truth validation.
We're going to talk about a study we've been

1312
02:18:03,389 --> 02:18:09,939
working on in San Diego and some of the interesting
measures we've been able to come up with using

1313
02:18:09,939 --> 02:18:14,120
this data. Yelp data comes with several attributes
like the dollar sign price or average star

1314
02:18:14,120 --> 02:18:18,960
rating, but it also comes with the set of
reviews that people have written. We can mine

1315
02:18:18,960 --> 02:18:23,630
those reviews for several attributes about
the food outlets. So, we screened the business

1316
02:18:23,630 --> 02:18:29,500
reviews for Yelp from the years 2004 to 2016,
getting a total just over one million reviews

1317
02:18:29,500 --> 02:18:36,070
from almost 300,000 users for about 8,000
food establishments. And then we apply topic

1318
02:18:36,070 --> 02:18:41,550
modeling and sentiment analysis to the reviewed
text. So, topic modeling is a type of natural

1319
02:18:41,550 --> 02:18:45,750
language processing method for discovering
the abstract or hidden topics that might appear

1320
02:18:45,750 --> 02:18:51,670
in a collection of documents. LDA is one type
of topic modeling, and it classifies text

1321
02:18:51,670 --> 02:18:53,450
to that particular topic using probability
distributions.

1322
02:18:53,450 --> 02:18:59,390
The model describes the patterns of words
that are repeating together and occurring

1323
02:18:59,390 --> 02:19:04,479
frequently. So, what you get are several topics,
and under each topic, a set of corresponding

1324
02:19:04,479 --> 02:19:09,570
words that occur together frequently within
that topic. So, in this image, for example,

1325
02:19:09,570 --> 02:19:17,300
the yellow topic sees gene, DNA, and genetic
occurring together. So, we ran Yelp data through

1326
02:19:17,300 --> 02:19:22,980
LDA. We settled on 20 topics, and those are
shown in this table on the left, and named

1327
02:19:22,980 --> 02:19:27,349
those topics based on the types of dominant
words included in each topic. We then assigned

1328
02:19:27,349 --> 02:19:35,400
the most closely aligned topics to each review.
So, every review got coded as a topic. We

1329
02:19:35,400 --> 02:19:38,750
looked at the average type of distribution
by neighborhood throughout San Diego, and

1330
02:19:38,750 --> 02:19:42,580
you can quickly see some patterns emerged
of the types of restaurants that dominate

1331
02:19:42,580 --> 02:19:48,330
certain neighborhoods of the city. So, first
of all, everybody's unhappy on Yelp. So, several

1332
02:19:48,330 --> 02:19:53,750
neighborhoods see a dominance of the displeasure
category or displeasure topic shown here in

1333
02:19:53,750 --> 02:19:57,470
green, especially Carlsbad.
For some reason, they're very unhappy there.

1334
02:19:57,470 --> 02:20:04,250
But other topics definitely emerged. So, for
example, in central San Diego, there is a

1335
02:20:04,250 --> 02:20:09,140
huge popularity of breakfast and happy hour
spots in Coronado, which is a bit of a higher

1336
02:20:09,140 --> 02:20:11,520
end neighborhood. We see the topic of atmosphere
high end being dominant. For Alpine, which

1337
02:20:11,520 --> 02:20:19,410
is a bit more of a rural area in the eastern
part of the county, people are talking the

1338
02:20:19,410 --> 02:20:23,640
most about pizza. We mapped some of the topics
to see the distribution throughout the county.

1339
02:20:23,640 --> 02:20:28,490
So, for example, we see the neighborhoods
with higher dominance of Mexican restaurants

1340
02:20:28,490 --> 02:20:32,490
as classified by LDA analysis, and they are
concentrated in the south of the city, as

1341
02:20:32,490 --> 02:20:38,651
well as a short sort of corridor up through
the mid to Eastern areas. Here, we're looking

1342
02:20:38,651 --> 02:20:43,350
at aggregates, that these aggregations are
coming out of classification of each individual

1343
02:20:43,350 --> 02:20:47,609
food outlet. So, this can be particularly
useful in filling in some of the gaps that

1344
02:20:47,609 --> 02:20:51,280
we don't get from government or commercial
data where we don't know much about the types

1345
02:20:51,280 --> 02:20:54,830
of foods that are being sold.
We are now actually working on figuring out

1346
02:20:54,830 --> 02:20:59,790
how we might be able to classify words into
healthy versus unhealthy topics, to get a

1347
02:20:59,790 --> 02:21:06,570
ranking of the healthiest restaurant or market.
With the Yelp data, we can also look at pricing

1348
02:21:06,570 --> 02:21:11,460
and that sentiment or customer satisfaction.
And this is an...actually could be a really

1349
02:21:11,460 --> 02:21:16,750
important factor to consider in terms of assessing
not only the presence or absence of specific

1350
02:21:16,750 --> 02:21:22,520
food types, but their quality and their ability
to satisfy basic customer needs. So, in the

1351
02:21:22,520 --> 02:21:26,790
data analysis, the algorithm is looking for
specific sets of words that would imply positive

1352
02:21:26,790 --> 02:21:32,340
or negative sentiments or opinions. These
are then aggregated into a scale from positive

1353
02:21:32,340 --> 02:21:37,399
to negative sentiments. In this figure, sentiments
are aggregated by neighborhood. But again,

1354
02:21:37,399 --> 02:21:43,080
we can also examine specific food outlets
to see the sentiment of that particular location.

1355
02:21:43,080 --> 02:21:49,240
As seen here, we can also see sentiment changed
over time. So, in some neighborhoods like

1356
02:21:49,240 --> 02:21:53,180
central San Diego, we don't really see much
change in sentiment at all.

1357
02:21:53,180 --> 02:21:56,180
In Chula Vista, we see a decrease in overall
sentiment outcomes, so, people are getting

1358
02:21:56,180 --> 02:21:59,950
less and less happy with their reviews. But
in Harbison Crest, we see a lot of variation

1359
02:21:59,950 --> 02:22:06,290
throughout the years, although this may be
due to a small amount of restaurants in this

1360
02:22:06,290 --> 02:22:11,620
area. So, the opening or closing of even one
restaurant will likely have an impact on average,

1361
02:22:11,620 --> 02:22:20,900
as seen for the neighborhood. Another area
we're looking into with the Yelp data is ethnic

1362
02:22:20,900 --> 02:22:27,210
preferences. So, again, while a grocery store
or a restaurant may be present in a neighborhood,

1363
02:22:27,210 --> 02:22:34,750
does it actually have the types of foods that
the dominant racial or ethnic groups in the

1364
02:22:34,750 --> 02:22:37,479
area prefer? So, you can use natural language
processing to pick out other languages, for

1365
02:22:37,479 --> 02:22:40,220
example, Spanish, and then match restaurants
with the predominance of Spanish words. We

1366
02:22:40,220 --> 02:22:43,850
can compare them to neighborhoods with high
prevalence of Hispanic ethnic density. So,

1367
02:22:43,850 --> 02:22:48,930
on the left here, we're seeing what the Yelp
reviews with more of the Spanish language

1368
02:22:48,930 --> 02:22:53,940
are. Whereas on the right, we're looking at
where Hispanic ethnic density is from the

1369
02:22:53,940 --> 02:22:58,970
budget census related variables.
And we can overlap these and see if the restaurants

1370
02:22:58,970 --> 02:23:04,460
are actually meeting the needs of the population
that they are attempting to serve. We are

1371
02:23:04,460 --> 02:23:09,530
similarly looking at restaurants and food
stores classified under Asian type categories

1372
02:23:09,530 --> 02:23:14,780
in Yelp to find census tracts that can be
considered Asian food deserts. And we're comparing

1373
02:23:14,780 --> 02:23:18,750
them with Asian food insecurity as measured with
SNAP data from the American Community Survey.

1374
02:23:18,750 --> 02:23:24,740
I'm not going to go into depth into this analysis
because we actually have a poster being presented

1375
02:23:24,740 --> 02:23:33,149
during the conference from our undergraduate
interns. So, please do go check that out.

1376
02:23:33,149 --> 02:23:34,880
So, some conclusions from this work so far.
There definitely are issues with crowdsourced

1377
02:23:34,880 --> 02:23:40,360
data, such as representation, or who is contributing
to the data, the coverage of the data. Although

1378
02:23:40,360 --> 02:23:45,670
over time, this is getting better and better,
and there can be marketing bias. So, for example,

1379
02:23:45,670 --> 02:23:52,830
Yelp has been known to try to convince businesses
to spend ad dollars to get better reviews.

1380
02:23:52,830 --> 02:23:57,790
But Yelp and other customers data can work
in tandem with more traditional data sources

1381
02:23:57,790 --> 02:24:02,710
to fill in missing gaps.
It can aid in ground truth efforts by reducing

1382
02:24:02,710 --> 02:24:06,319
the number of outlets that need to be ground
truth, and it can add variables of interest.

1383
02:24:06,319 --> 02:24:11,600
It can fill content gaps, and it can be used
to classify food stores by ethnic types. So,

1384
02:24:11,600 --> 02:24:18,550
I'm going to pivot now and talk about some
of the ways that GPS and GIS can be used to

1385
02:24:18,550 --> 02:24:22,990
obtain better measures of exposure to the
food environment. We've definitely seen a

1386
02:24:22,990 --> 02:24:26,069
huge increase in interest of not just the
food environments where people live, but being

1387
02:24:26,069 --> 02:24:30,550
able to assess the types of environments people
encounter during their daily travel patterns,

1388
02:24:30,550 --> 02:24:35,910
or even more specifically, the types of food
environments that somebody actually enters

1389
02:24:35,910 --> 02:24:41,450
and interacts with. In this image, which is
courtesy of Lukar Thornton, we see that people

1390
02:24:41,450 --> 02:24:45,670
purchase foods in a variety of locations beyond
their neighborhood. So, we want to be able

1391
02:24:45,670 --> 02:24:51,410
to accurately measure this type of behavior,
ideally passively without passing on a lot

1392
02:24:51,410 --> 02:24:57,149
of burden to participants. We have a few studies
where we have put outward facing cameras on

1393
02:24:57,149 --> 02:25:00,880
participants to be able to objectively classify
when they enter a food establishment.

1394
02:25:00,880 --> 02:25:07,040
And we paired that with a GPS tracker and then
underlined GIS measures of food environments.

1395
02:25:07,040 --> 02:25:14,120
And what we get as a result is this picture
on the right, where we see somebody passing

1396
02:25:14,120 --> 02:25:19,260
through a bunch of food environments and ultimately
an environment they stop and engage with.

1397
02:25:19,260 --> 02:25:24,439
So, here they're passing through several fast
food or taqueria outlet possibilities, but

1398
02:25:24,439 --> 02:25:29,630
they actually end up stopping at lunch just
south of the U.S.-Mexican border, and we can

1399
02:25:29,630 --> 02:25:34,899
actually validate that with the imagery that
we've collected here as well. This type of

1400
02:25:34,899 --> 02:25:41,690
data is important to gather because as ground
truth data to then test developed algorithms

1401
02:25:41,690 --> 02:25:47,609
that don't have image data that are only relying
on the GPS and GIS data. We wanted to know how

1402
02:25:47,609 --> 02:25:51,600
accurate this type of measurement can be so
we can use this type of ground truth data

1403
02:25:51,600 --> 02:25:58,820
to create rules for when we classify somebody
as in or out of a restaurant. An example is

1404
02:25:58,820 --> 02:26:02,520
in the images above, where we've set a five
minute rule of being within the buffer of

1405
02:26:02,520 --> 02:26:06,970
a food outlet location as the rule for actually
interacting with that food outlet rather than

1406
02:26:06,970 --> 02:26:11,729
just passing through or passing next to it.
In the lower images, we can see how this plays

1407
02:26:11,729 --> 02:26:15,300
out in real life with the GPS data passing
through the drive thru of the food outlet,

1408
02:26:15,300 --> 02:26:21,700
and then we can actually validate that food
purchase with the imagery. So, in combining

1409
02:26:21,700 --> 02:26:29,960
GPS and GIS data, eventually, ideally, through
the smartphone, we can be incredibly, it can

1410
02:26:29,960 --> 02:26:37,330
be incredibly useful in terms of moving towards
passive measures of the food environment exposure

1411
02:26:37,330 --> 02:26:41,260
and engagement, rather than burdening users
with surveys or constant reporting. And I

1412
02:26:41,260 --> 02:26:45,930
think eventually these types of methods are
going to be key in implementing just in time

1413
02:26:45,930 --> 02:26:50,640
adaptive intervention strategies related to
food environment engagement. Right now, we

1414
02:26:50,640 --> 02:26:55,279
are relying heavily on distance based notifications.
In these types of applications, we're just

1415
02:26:55,279 --> 02:27:01,500
using a geofence. But as we collect more data
about behaviors, such as time spent in a location,

1416
02:27:01,500 --> 02:27:05,770
trip mode, such as car or walking, which
we can actually get from the accelerometer

1417
02:27:05,770 --> 02:27:10,890
data on the phone and other contextual factors
like stoplights, we can really hone in on

1418
02:27:10,890 --> 02:27:15,510
the likelihood that someone is actually interacting
with a food outlet rather than just passing

1419
02:27:15,510 --> 02:27:20,380
through.
As we're increasingly getting more interested

1420
02:27:20,380 --> 02:27:22,931
in mobility and the food environment, we're
going to want to reduce participant burden

1421
02:27:22,931 --> 02:27:29,189
in these types of studies and move towards
passive measurement methods. We need more

1422
02:27:29,189 --> 02:27:34,250
ground truth studies to understand the types
of errors that passive sensing methods can

1423
02:27:34,250 --> 02:27:38,900
produce and hopefully, with enough research,
we can start applying these methods in actual

1424
02:27:38,900 --> 02:27:46,470
interventions. Just want to say thank you
to the people in our labs that helped with

1425
02:27:46,470 --> 02:27:52,610
all of these studies, and also thank you to
the NIH and NSF for funding this research.

1426
02:27:52,610 --> 02:28:20,561
Please feel free to reach out to me to keep
the conversation going, and I'm happy to take

1427
02:28:20,561 --> 02:28:21,700
any questions.

1428
02:28:21,700 --> 02:28:28,990
DR. KAREN GLANZ: Alright, thank you. Thank you, Alana,
Alison, and Marta, for those interesting and

1429
02:28:28,990 --> 02:28:35,980
provocative talks and expanding our scope
in our ways of thinking of measurement, especially

1430
02:28:35,980 --> 02:28:43,760
Marta's talk, which kind of introduces us
to some new and emerging technologies. We're

1431
02:28:43,760 --> 02:28:50,550
going to go to a question and answer session.
And for that, I'd like to introduce our moderator,

1432
02:28:50,550 --> 02:28:58,050
Leslie Lytle. Leslie will not be seen live
on camera, but you have a thumbnail picture

1433
02:28:58,050 --> 02:29:06,550
of her, so, she can be seen that way. Dr.
Lytle is a professor from The University of

1434
02:29:06,550 --> 02:29:11,650
North Carolina at Chapel Hill, has done a
tremendous amount of work in this area, and

1435
02:29:11,650 --> 02:29:14,250
so, I'll turn it over to you, Leslie.

1436
02:29:14,250 --> 02:29:20,120
DR. LESLIE LYTLE: Yeah, thank you so much again.
I'm really sorry. I had trouble with the audio

1437
02:29:20,120 --> 02:29:26,200
and visual earlier, but I was really interested
in the presentations. Thank you panelists

1438
02:29:26,200 --> 02:29:30,569
for your great presentations, and we have
quite a few good questions, so, I'm going

1439
02:29:30,569 --> 02:29:38,359
to jump right in. There are a couple of questions
for Alana that I think are related. They are

1440
02:29:38,359 --> 02:29:44,500
specifically about the Environmental Atlas
and Food Access Research. One question is

1441
02:29:44,500 --> 02:29:51,280
about any plans that you have for updating
that atlas with more current data. And then

1442
02:29:51,280 --> 02:30:01,340
there was also a question regarding the accuracy
of the databases used for the food atlas identification

1443
02:30:01,340 --> 02:30:06,430
of food locations. Those of us who work in
this area know there's quite a few studies

1444
02:30:06,430 --> 02:30:13,730
about ground truthing, with various findings
about accuracy of the list services and databases

1445
02:30:13,730 --> 02:30:20,090
that are available commercially. So, if
you could, just mention briefly plans for updating

1446
02:30:20,090 --> 02:30:25,930
the databases, as well as what you know about
ground truthing and accuracy.

1447
02:30:25,930 --> 02:30:34,760
ALANA RHONE: So, we do have plans to update both
atlases. The Food Environment Atlas and the

1448
02:30:34,760 --> 02:30:43,210
Food Access Research Atlas with more current
data. We are currently in the process of updating

1449
02:30:43,210 --> 02:30:49,750
the Food Environment Atlas and our information
on the location of our store directories.

1450
02:30:49,750 --> 02:30:59,320
We use two different datasets: one being stores
that are authorized to accept SNAP, and another

1451
02:30:59,320 --> 02:31:06,960
is...that's called STARS and then another data set, which
is called TDLinx, which is a Nielsen directory,

1452
02:31:06,960 --> 02:31:14,569
and it gives a store list, an annual snapshot
of stores. And we merge both of these directories

1453
02:31:14,569 --> 02:31:21,130
together so that we may accurately capture
stores and the right location and that they

1454
02:31:21,130 --> 02:31:27,370
are really open. So, that's how we kind of,
you know, make sure that we have accurate

1455
02:31:27,370 --> 02:31:30,660
information using our stored data set.

1456
02:31:30,660 --> 02:31:38,450
DR. LESLIE LYTLE: OK, thanks. Also, a follow up,
Alana, someone asked about how rural...how

1457
02:31:38,450 --> 02:31:41,810
you defined rural for the distance to a supermarket.

1458
02:31:41,810 --> 02:31:53,130
ALANA RHONE: Yes. I would love to go over that. So,
rural status and urban status comes from...directly

1459
02:31:53,130 --> 02:32:00,040
from the Bureau of Defense's urbanized area
definitions, where rural areas are sparsely

1460
02:32:00,040 --> 02:32:09,080
populated areas with fewer than 2,500 people
and urban areas are areas with more than 2,500

1461
02:32:09,080 --> 02:32:10,080
people.

1462
02:32:10,080 --> 02:32:15,440
DR. LESLIE LYTLE: OK, thank you very much. I
actually have a question for Alison. Alison,

1463
02:32:15,440 --> 02:32:21,960
I thought that your data on the proportion
of online shoppers was just really fascinating

1464
02:32:21,960 --> 02:32:27,770
and to think that 30 percent of all food will
be purchased online by 2025 kind of blows

1465
02:32:27,770 --> 02:32:34,500
my mind. I'm hearing there's this discussion
out in the field about the potential for marketing

1466
02:32:34,500 --> 02:32:43,560
through online shoppers or providing incentives
when someone is shopping online. It seems

1467
02:32:43,560 --> 02:32:47,930
like that could open up a whole new can of
worms.

1468
02:32:47,930 --> 02:32:54,580
DR. ALISON GUSTAFSON: Yes. So, Leslie, we have a study right
now funded through Share Our Strength...No Kid Hungry,

1469
02:32:54,580 --> 02:33:00,569
people might be familiar with that, looking
at how to navigate healthy marketing within

1470
02:33:00,569 --> 02:33:07,229
the online platform. And we also have a couple
of grants going in around that online healthy

1471
02:33:07,229 --> 02:33:14,990
marketing, especially for SNAP now that SNAP
online has been expanded to so many different

1472
02:33:14,990 --> 02:33:21,080
storefronts. So, that's something that we
need to help the public-private partnership,

1473
02:33:21,080 --> 02:33:24,819
that we need to be helping those retailers
with their healthy marketing. And I don't

1474
02:33:24,819 --> 02:33:31,240
know if you saw last week maybe, Instacart
will be taking quite a bit of money from advertisers

1475
02:33:31,240 --> 02:33:38,690
to be doing marketing in their online space.
So, that's something we're trying to work

1476
02:33:38,690 --> 02:33:45,779
on as far as intervention work and also policy
change with Share Our Strength and with USDA,

1477
02:33:45,779 --> 02:33:51,550
especially since so many SNAP dollars are
about to be redeemed online. So, that's something

1478
02:33:51,550 --> 02:33:58,370
that we all need to kind of put our ear to
the ground for and start to have conversations

1479
02:33:58,370 --> 02:34:00,110
around that.

1480
02:34:00,110 --> 02:34:07,290
DR. MARTA JANKOWSKA: (CROSSTALK) that there's a lot of
mobile based apps like Waze, for example,

1481
02:34:07,290 --> 02:34:13,479
is already taking marketing dollars from fast
food outlets. So, as you're on your route

1482
02:34:13,479 --> 02:34:17,950
and Waze is telling you where to go, you'll
get a pop up that says, "Hey, you're half

1483
02:34:17,950 --> 02:34:24,290
a mile from this McDonald's, and you can get
a dollar off your burger. Check it out." So,

1484
02:34:24,290 --> 02:34:30,390
there's a lot of innovation going on in the
marketing space for online and mobility based

1485
02:34:30,390 --> 02:34:34,500
data that, as researchers, we are incredibly
behind on.

1486
02:34:34,500 --> 02:34:41,061
DR. LESLIE LYTLE: Yeah. Yeah, I think that's going
to be a fascinating area to keep our eyes

1487
02:34:41,061 --> 02:34:50,140
on. Marta, a related tech question for you:
How has GPS and GIS research design controlled

1488
02:34:50,140 --> 02:34:57,290
for neighborhood demographic and resource
access, such as internet and smartphones?

1489
02:34:57,290 --> 02:35:03,270
And also, is there more? Are there more data
collected and available in areas with higher

1490
02:35:03,270 --> 02:35:04,270
income?

1491
02:35:04,270 --> 02:35:11,740
DR. MARTA JANKOWSKA: So, a lot of the research that comes
out of that Pew Research Center is showing

1492
02:35:11,740 --> 02:35:17,530
that the disparities in access to at least
smartphones are really decreasing over time.

1493
02:35:17,530 --> 02:35:23,740
So, at this point, we're seeing penetration
of smartphone access quite extensively throughout

1494
02:35:23,740 --> 02:35:28,470
different types of populations in the United
States. Internet access is a little different,

1495
02:35:28,470 --> 02:35:39,000
and there, especially, just like with our industry
provided data sets, rural data is a little

1496
02:35:39,000 --> 02:35:46,660
less accessible and less prevalent than urban
data. That being said, in the GPS related

1497
02:35:46,660 --> 02:35:53,120
studies, a lot of those are being collected
with sampling frameworks in mind to try to

1498
02:35:53,120 --> 02:35:58,700
get a wider range of access. So, it really
depends. Is it researcher enabled, in which

1499
02:35:58,700 --> 02:36:04,151
case, usually, hopefully, the researcher has
thought about getting a wide range of people

1500
02:36:04,151 --> 02:36:11,400
involved? Or is it collected from commercial
datasets where commercial sources are providing

1501
02:36:11,400 --> 02:36:15,470
smartphone data? And in that case, again,
there's less concern, I think, on their end

1502
02:36:15,470 --> 02:36:22,520
about who they're representing.
But smartphone penetration has really gone

1503
02:36:22,520 --> 02:36:25,250
far in the past 10 years.

1504
02:36:25,250 --> 02:36:33,760
DR. LESLIE LYTLE: Yeah, I think this is a question
probably for Marta or Alison. There's a question

1505
02:36:33,760 --> 02:36:40,110
that says,: what are your thoughts about shopping
behavior in the research narrative? This appears

1506
02:36:40,110 --> 02:36:45,720
to be an oversight in the literature. I think
that shopping behavior can be tracked via

1507
02:36:45,720 --> 02:36:51,779
technology, but also I think from Allison's
talk on consumer behavior, you might have

1508
02:36:51,779 --> 02:36:53,490
some insights.

1509
02:36:53,490 --> 02:37:03,530
DR. ALISON GUSTAFSON: I can go first, Marta. Yeah, so, one
thing. Yeah. And so, and it's a little bit

1510
02:37:03,530 --> 02:37:10,649
about what Marta was saying of in the moment
decision making. And so, we, myself and Jared McGuirt, 

1511
02:37:10,649 --> 02:37:18,360
who's at UNC Greensboro, are working
with some different state providers. So WIC

1512
02:37:18,360 --> 02:37:26,410
and SNAP at the state level and using technology
that we put inside the store called Beacon

1513
02:37:26,410 --> 02:37:31,790
to then give alerts on their phone while they're
shopping to help the consumer make different

1514
02:37:31,790 --> 02:37:37,431
choices while they're in the moment of their
shopping behavior. And then we're also looking

1515
02:37:37,431 --> 02:37:45,630
at that in the online space. So, while you're
shopping online, is there a way to help push

1516
02:37:45,630 --> 02:37:51,200
or behaviorally nudge to make healthier
purchases while you're actually putting items

1517
02:37:51,200 --> 02:37:53,580
in your cart in the online space?

1518
02:37:53,580 --> 02:37:59,771
DR. LESLIE LYTLE: Thanks. Marta, anything else to
add?

1519
02:37:59,771 --> 02:38:07,970
DR. MARTA JANKOWSKA: Yeah, and I would add to that that
all of these new resources that we're seeing

1520
02:38:07,970 --> 02:38:13,510
like Instacart and other online based options,
I mean, they're collecting all that data.

1521
02:38:13,510 --> 02:38:22,050
So, whether or not as researchers can get
access to it, I haven't seen much being done

1522
02:38:22,050 --> 02:38:27,460
in that space yet. I haven't heard of people
trying to get that data, but I think there's

1523
02:38:27,460 --> 02:38:31,260
a lot of potential. Yeah. Alison, go ahead.

1524
02:38:31,260 --> 02:38:37,590
DR. ALISON GUSTAFSON: Yeah. So, we actually are in the works
with Instacart, seeing if they are open to

1525
02:38:37,590 --> 02:38:45,670
share their SNAP purchasing data with us.
So, we are hoping that we can keep that bridge

1526
02:38:45,670 --> 02:38:52,550
and be able to have access to all of their
SNAP purchasing data. But I want to note that

1527
02:38:52,550 --> 02:38:57,430
in rural communities, it's obviously not as
robust as it is in urban sites.

1528
02:38:57,430 --> 02:39:06,040
DR. LESLIE LYTLE: Yeah, that makes sense. A related
question, is sort of about using that technology

1529
02:39:06,040 --> 02:39:12,100
for intervention use. So, it's talking about
how might one use ecological momentary assessment

1530
02:39:12,100 --> 02:39:19,399
with smartphones or mobile devices in order
to characterize individual environment interactions

1531
02:39:19,399 --> 02:39:28,710
and something like a just in time adaptable
intervention? Are we seeing any work in that

1532
02:39:28,710 --> 02:39:33,190
space where we're trying to use something
like an ecological momentary assessment with

1533
02:39:33,190 --> 02:39:36,530
a smartphone to intervene and prompt behavior?

1534
02:39:36,530 --> 02:39:41,420
DR. MARTA JANKOWSKA: Yeah, I think there's actually quite
a bit of research on that, especially because

1535
02:39:41,420 --> 02:39:52,810
the reliability of just using the GIS data
with the GPS data is quite low still. So,

1536
02:39:52,810 --> 02:39:56,250
we're trying to get better at coming up with
these rule sets of when someone's actually

1537
02:39:56,250 --> 02:40:00,811
in a restaurant, or going to the restaurant
trying to predict that, but that is still

1538
02:40:00,811 --> 02:40:06,601
pretty recent. And so, a lot of people, a
lot of researchers that are using just in time

1539
02:40:06,601 --> 02:40:13,210
interventions with food related outcomes,
they have to use EMA because there's really

1540
02:40:13,210 --> 02:40:18,870
no other way of knowing did someone eat? Is
someone about to eat? What kind of choices

1541
02:40:18,870 --> 02:40:23,860
they're about to make. So there is quite a
bit of research using EMA, and I think that

1542
02:40:23,860 --> 02:40:31,069
right now we're just going to have to keep
doing that. Eventually, I mean, this is a bit

1543
02:40:31,069 --> 02:40:34,210
high in the sky, but I think eventually we
can get to a place where passive sensors on

1544
02:40:34,210 --> 02:40:41,460
the smartphone might be able to do pretty
good at predicting when someone's about to

1545
02:40:41,460 --> 02:40:47,859
purchase food. Interesting. I want to bring
the conversation back to Alana.

1546
02:40:47,859 --> 02:40:55,790
We have a question here about: Has GPS/GIS
food data been integrated with other social

1547
02:40:55,790 --> 02:41:02,830
determinants of health data, such as transportation
access, education, housing status, and employment

1548
02:41:02,830 --> 02:41:07,250
status? You talked a little bit about transportation.
Can you say anything more about the other

1549
02:41:07,250 --> 02:41:11,250
social determinants of health and those other
characteristics?

1550
02:41:11,250 --> 02:41:24,890
ALANA RHONE: Yes, I sure can. In the Food Environment
Research Atlas, we do include social determinants

1551
02:41:24,890 --> 02:41:35,150
of health. We do have things such as people's
access to recreational centers, poverty and

1552
02:41:35,150 --> 02:41:44,530
income information also in the Food Environment
Atlas. But in the Food Access Research Atlas,

1553
02:41:44,530 --> 02:41:52,060
we really basically only look at income and
vehicle access. We also have information in

1554
02:41:52,060 --> 02:42:01,819
there about SNAP participants, and we also
look at race as well. So, you can find the

1555
02:42:01,819 --> 02:42:07,040
number of people, for example, who have a
certain race or ethnicity that live far from

1556
02:42:07,040 --> 02:42:08,899
a grocery store.

1557
02:42:08,899 --> 02:42:15,750
DR. LESLIE LYTLE: Alright, great. Thank you for clarifying
that. Another question for Marta: Can you

1558
02:42:15,750 --> 02:42:22,910
talk about any available, actually tried and
tested smartphone apps that could help us

1559
02:42:22,910 --> 02:42:26,590
measure the food environment and track our
spending habits?

1560
02:42:26,590 --> 02:42:36,910
DR. MARTA JANKOWSKA: Yeah, that's a good question. I don't
know any off the top of my head. I haven't

1561
02:42:36,910 --> 02:42:41,500
done a lot of research into what's out there.
I know there's a lot out there, and a lot of

1562
02:42:41,500 --> 02:42:48,359
it is not tested, so, people are just trying
to make quick money off of these types of

1563
02:42:48,359 --> 02:42:52,840
apps. Yeah, that's the most I can say about
it.

1564
02:42:52,840 --> 02:42:59,470
DR. LESLIE LYTLE: Yeah, I'm sure apps are coming.
We probably just don't have them tried and

1565
02:42:59,470 --> 02:43:07,940
true yet. I think we have time for one more
question. And I think that this probably could

1566
02:43:07,940 --> 02:43:14,540
go to any of our three panelists. In your
opinion, will we someday be able to better

1567
02:43:14,540 --> 02:43:21,960
predict individual consumption using food
environment and sales data? And what would

1568
02:43:21,960 --> 02:43:28,380
that look like? You know, I think this question
is getting to the point of we all believe

1569
02:43:28,380 --> 02:43:33,149
that the food environment affects people's
eating behavior, their nutritional status,

1570
02:43:33,149 --> 02:43:39,569
and their health outcomes. But we still have
some real challenge in actually seeing those

1571
02:43:39,569 --> 02:43:45,689
connections in the data that we produce. Do
you believe that we're going to be getting

1572
02:43:45,689 --> 02:43:51,240
better at that, on using food environment
and sales data in the future?

1573
02:43:51,240 --> 02:43:58,980
DR. ALISON GUSTAFSON: I'll go first, just because I do a
lot of work with receipt data, and I think

1574
02:43:58,980 --> 02:44:09,160
we have a ways to go. So, I've done some work
with Simon French and their expertise with

1575
02:44:09,160 --> 02:44:14,781
receipt data. And I think, right, they're
just capturing what's being purchased and

1576
02:44:14,781 --> 02:44:21,260
then who's actually consuming it
from that. We are still not even sure yet.

1577
02:44:21,260 --> 02:44:28,060
So, I personally would love to see if someone
knows a developer that...where, as a researcher,

1578
02:44:28,060 --> 02:44:34,300
someone could hand the receipt and they could,
in a sense, annotate but digitally what they

1579
02:44:34,300 --> 02:44:39,210
were thinking, who they were with, what their
geographic food environment was around them

1580
02:44:39,210 --> 02:44:43,040
and then what they intend to do with that
food that they just purchased.

1581
02:44:43,040 --> 02:44:52,470
DR. LESLIE LYTLE: Yeah, but then we run into unconventional
food sources I think farmers' markets, where

1582
02:44:52,470 --> 02:44:56,410
we would really like to see people moving
to farmers' markets, but there's no receipts

1583
02:44:56,410 --> 02:44:57,910
given at farmers' markets.

1584
02:44:57,910 --> 02:45:02,430
DR. ALISON RHONE: And even that. And one of your question,
which is always a good point, and I do work

1585
02:45:02,430 --> 02:45:06,620
with food pantries, so, there's no receipt
for food pantries. There's no...you know, so that kind

1586
02:45:06,620 --> 02:45:11,430
of what they're using, what they get from
the WIC office. So, there's always those

1587
02:45:11,430 --> 02:45:17,140
all those other sources of food. But for the
places where you could get a receipt or a

1588
02:45:17,140 --> 02:45:22,510
digital sale, I think, but I still think there's
a lot of gap even in that. So, Leslie, you

1589
02:45:22,510 --> 02:45:27,590
know, so, I'd like to see us maybe just even
work on that. But...

1590
02:45:27,590 --> 02:45:31,180
DR. MARTA JANKOWSKA: Yeah, I agree, just because food inherently
is such a social phenomenon. And so, you know,

1591
02:45:31,180 --> 02:45:39,022
whether you're going to a party or potluck
and you're ingesting a bunch of food that

1592
02:45:39,022 --> 02:45:45,229
other people brought, whether you're buying
snacks for your kid's soccer team, I mean,

1593
02:45:45,229 --> 02:45:50,279
in order to get that narrow of a pipeline,
you know, you have to essentially be a singular

1594
02:45:50,279 --> 02:45:55,410
person, only purchasing food for yourself
and without any other interaction. But I don't

1595
02:45:55,410 --> 02:46:01,190
think is possible. I mean, there's a very
small section of the population that that

1596
02:46:01,190 --> 02:46:08,640
could be possible. But I think in my mind,
it's less of like the overarching goal,

1597
02:46:08,640 --> 02:46:16,660
and it's more of a where can, we without too
much user burden, collect information that

1598
02:46:16,660 --> 02:46:25,430
might help us provide better choices to a
person. So, it's less about knowing every

1599
02:46:25,430 --> 02:46:30,520
little thing that they put into their body
and more about what are the general habits

1600
02:46:30,520 --> 02:46:35,500
that we can help shape while they are in the
middle of food purchasing, which I think is

1601
02:46:35,500 --> 02:46:37,290
a really exciting area.

1602
02:46:37,290 --> 02:46:45,750
DR. LESLIE LYTLE: Yeah, yeah, I I agree. We are out
of time. I want to thank the panelists for

1603
02:46:45,750 --> 02:46:52,290
their great responses to these very interesting
questions. Before we move on to the next break

1604
02:46:52,290 --> 02:46:59,290
again, I want to thank you for your questions
and the panelists for sharing their presentations.

1605
02:46:59,290 --> 02:47:06,729
We are going to take a quick five minute break
from about 3:30 to 3:35. Session six, which

1606
02:47:06,729 --> 02:47:13,290
is Innovative Interventions to Address Neighborhood Food Environment will begin at 3:35 in the

1607
02:47:13,290 --> 02:47:19,870
auditorium. So, thanks again for this wonderful
conference and we'll look forward to learning

1608
02:47:19,870 --> 02:47:23,800
a lot more interesting stuff in the next day
and a half ahead.

1609
02:47:23,800 --> 02:47:26,870
DR. KAREN GLANZ: Welcome back from your break, everyone. We're

1610
02:47:26,870 --> 02:47:32,819
ready to move into session 6. We have a
pretty packed afternoon here. As we move into

1611
02:47:32,819 --> 02:47:38,110
the session, I want to remind everyone to
add your questions to the chatbox during the

1612
02:47:38,110 --> 02:47:44,851
presentation so we can address them in the
Q and A session that follows. We've had robust

1613
02:47:44,851 --> 02:47:50,710
input of questions, we haven't been able to
get to all of them, but many of them have

1614
02:47:50,710 --> 02:47:59,529
been very insightful and added to the presentations.
We're going to start the next panel that highlights

1615
02:47:59,529 --> 02:48:04,340
interventions and evaluation strategies intended
to make neighborhood food environments more

1616
02:48:04,340 --> 02:48:11,590
health enhancing. Speakers will describe various
intervention studies and their findings and

1617
02:48:11,590 --> 02:48:18,120
identify and discuss some research gaps and
ways to move this science forward. I'm especially

1618
02:48:18,120 --> 02:48:23,930
pleased to introduce Dr. Mary Story from
Duke University and the Healthy Eating Research

1619
02:48:23,930 --> 02:48:30,370
national program. She is going to lead
the moderator panel. So, Mary, over to you.

1620
02:48:30,370 --> 02:48:36,910
DR. MARY STORY: Thank you, Karen, and welcome
everyone to session 6. This has been a terrific

1621
02:48:36,910 --> 02:48:44,070
panel today. And Karen, you gave a wonderful
overview of the state of the science. And

1622
02:48:44,070 --> 02:48:52,870
we've had panels on health impacts and measurements,
and now we turn to food retail interventions,

1623
02:48:52,870 --> 02:48:59,850
and we'll be dealing with three specific areas
and communities grocery stores and supermarkets,

1624
02:48:59,850 --> 02:49:06,920
small food stores and sources, and food pantries.
So, we're fortunate to have presentations

1625
02:49:06,920 --> 02:49:14,160
from three leading scholars who have conducted
extensive innovative research in this area.

1626
02:49:14,160 --> 02:49:22,300
First, let me introduce Dr.Tamara Dubowitz,
who's a senior policy researcher at the RAND

1627
02:49:22,300 --> 02:49:29,210
Corporation. She's a social epidemiologist
and has done some of the most innovative and

1628
02:49:29,210 --> 02:49:37,050
rigorous work on neighborhood food environments,
including SES and health disparities, and

1629
02:49:37,050 --> 02:49:43,660
the introduction of supermarkets in communities
where they lack large food stores. And she'll

1630
02:49:43,660 --> 02:49:50,420
be taking on this latter topic today. Our
second speaker will be Dr. Joel Gittelsohn,

1631
02:49:50,420 --> 02:49:57,050
a professor in international health at the
Bloomberg School of Public Health at Johns

1632
02:49:57,050 --> 02:50:01,010
Hopkins University.
Joel's training is in medical anthropology

1633
02:50:01,010 --> 02:50:08,550
and public health nutrition. He's led multiple
nutrition and food-based intervention trials

1634
02:50:08,550 --> 02:50:15,110
in Native American communities and has done
landmark work on interventions in small food

1635
02:50:15,110 --> 02:50:23,710
stores in Baltimore, which is his topic today,
and the end of session we'll have Dr. Christina

1636
02:50:23,710 --> 02:50:28,240
Roberto, Associate Professor of Health Policy
in the School of Medicine at the University

1637
02:50:28,240 --> 02:50:36,029
of Pennsylvania. She's a clinical psychologist
and epidemiologist, and her research is expansive

1638
02:50:36,029 --> 02:50:42,811
and innovative in the areas of policies and
interventions to promote healthy eating. And

1639
02:50:42,811 --> 02:50:49,670
today she'll speak on community interventions
and food pantries and within food retail environments.

1640
02:50:49,670 --> 02:50:58,750
Unfortunately, Christina cannot join us live
for the Q&A after the presentation, so

1641
02:50:58,750 --> 02:51:06,670
she will not be able to answer any questions.
Now during any of...during the presentations.

1642
02:51:06,670 --> 02:51:12,500
Please enter your questions into the question
and answer box during the talks. And also

1643
02:51:12,500 --> 02:51:18,120
if you could, just indicate if it's for Tamara
or Joel or both.

1644
02:51:18,120 --> 02:51:28,290
So, with that, I'm really pleased to have
the presentations on now.

1645
02:51:28,290 --> 02:51:38,450
DR. TAMARA DUBOWITZ: Good afternoon, and thank
you so much for coming to hear about what

1646
02:51:38,450 --> 02:51:46,250
we've learned about placing supermarkets in
food deserts. First, I think it's really important

1647
02:51:46,250 --> 02:51:52,410
to understand what we are talking about when
we refer to food deserts. Technically, food

1648
02:51:52,410 --> 02:51:58,800
deserts have been defined as geographic areas
or places with limited access to healthful,

1649
02:51:58,800 --> 02:52:05,750
varied foods. Now, the United States Department
of Agriculture, or the USDA, definition has

1650
02:52:05,750 --> 02:52:13,700
looked at low-income and low access areas.
A low-income area has been defined by a poverty

1651
02:52:13,700 --> 02:52:23,580
rate of 20 percent or greater. A low access
urban area is where 33 percent of the population

1652
02:52:23,580 --> 02:52:30,870
lives ten or more miles from the nearest large
grocery store. A low access rural area is

1653
02:52:30,870 --> 02:52:40,529
where 33 percent of the population lives ten
or more miles from the nearest grocery store.

1654
02:52:40,529 --> 02:52:47,770
Yet low access to a supermarket is more nuanced
than it may seem. Well, 23.5 million people

1655
02:52:47,770 --> 02:52:54,069
live in low-income areas that are more than
a mile from a supermarket. About 11.5 million

1656
02:52:54,069 --> 02:52:59,990
are low-income and live in such areas. There
are different numbers of folks who fit into

1657
02:52:59,990 --> 02:53:04,490
these categories who have access to vehicles
or not.

1658
02:53:04,490 --> 02:53:10,560
So, I want to stop here for just a second
because this is really important when we're

1659
02:53:10,560 --> 02:53:16,890
trying to gauge whether these policies that
intend to incentivize supermarkets to locate

1660
02:53:16,890 --> 02:53:27,050
in food or supermarket deserts will ultimately
change food shopping behaviors or diet. We

1661
02:53:27,050 --> 02:53:32,680
know that places with limited food access
are characterized by higher levels of racial

1662
02:53:32,680 --> 02:53:40,500
segregation and income inequality. But it
is unclear whether food access or income

1663
02:53:40,500 --> 02:53:46,250
constraints our greater barriers. Supermarkets
and large grocery stores have lower prices

1664
02:53:46,250 --> 02:53:52,520
than smaller stores, and we know that when
they can, low-income households shop where

1665
02:53:52,520 --> 02:54:00,120
food prices are lower. We also know that easy
access to all food, and not just lack of access

1666
02:54:00,120 --> 02:54:07,200
to specific healthy foods may be an important
factor in explaining increases in obesity.

1667
02:54:07,200 --> 02:54:12,300
And we know that understanding market conditions
that contribute to differences in food access

1668
02:54:12,300 --> 02:54:21,500
is critical to policy intervention and design.
The Healthy Food Financing Initiative or HFFI

1669
02:54:21,500 --> 02:54:26,520
is the largest national policy that has sought
to address food deserts.

1670
02:54:26,520 --> 02:54:31,620
This is a public-private partnership that
provides grants and loans to finance the construction

1671
02:54:31,620 --> 02:54:38,540
and development of grocery stores and other
healthy food retailers in underserved areas.

1672
02:54:38,540 --> 02:54:45,880
It's jointly administered at the federal level
by USDA, HHS, and the Treasury Department.

1673
02:54:45,880 --> 02:54:53,710
And since 2011, this model has leveraged $220
million in federal funding into an estimated

1674
02:54:53,710 --> 02:55:01,729
$1 billion in additional private investments,
resources, loans, and tax incentives. The

1675
02:55:01,729 --> 02:55:10,020
ultimate goal is to attract fresh food retailers
to invest in underserved communities. So,

1676
02:55:10,020 --> 02:55:15,350
the big question is, has it worked? For now,
I'm focusing on studies that have been able

1677
02:55:15,350 --> 02:55:21,470
to design and implement natural experiments.
There are really only a small handful of studies

1678
02:55:21,470 --> 02:55:27,551
that have examined supermarkets before and
after. In Leeds and Glasgow in the U.K., there's

1679
02:55:27,551 --> 02:55:35,250
a set of studies. One in Leeds that had a pre-post-intervention
design. The one in Glasgow used a pre-post

1680
02:55:35,250 --> 02:55:43,280
design with a comparison area. Now, both showed
that the new stores created switchers or individuals

1681
02:55:43,280 --> 02:55:50,200
who switched to now shop at the new store.
And while these studies did not report changes

1682
02:55:50,200 --> 02:55:56,851
in diet, they did report that the switchers
had better psychological health or had changes

1683
02:55:56,851 --> 02:56:03,670
in their psychological health. Another study
in Philadelphia, Pennsylvania, and in Trenton,

1684
02:56:03,670 --> 02:56:12,790
as the comparison, has two sets of pre-store
opening data, and so far they found correlates

1685
02:56:12,790 --> 02:56:22,660
of diet to be perceived nutrition, environment,
perceived food availability, and income. The

1686
02:56:22,660 --> 02:56:28,910
study that I am most equipped to talk about
is PHRESH or the Pittsburgh Hill Homewood Research

1687
02:56:28,910 --> 02:56:35,920
on Neighborhood Change and Health, which at
its inception was called the Pittsburgh Hill

1688
02:56:35,920 --> 02:56:42,090
Home Research on eating, shopping, and health.
The original study started to examine a new

1689
02:56:42,090 --> 02:56:47,790
supermarket in the Hill District neighborhood
in Pittsburgh, Pennsylvania. Homewood was

1690
02:56:47,790 --> 02:56:55,290
the control neighborhood. Now, we collected
baseline data in 2011, a full-service supermarket

1691
02:56:55,290 --> 02:57:04,140
opened in 2013, and we had follow-up dietary
and food purchasing data collected in 2014,

1692
02:57:04,140 --> 02:57:14,130
2018, and currently. Some of our main results
are published in a 2015 Health Affairs Paper

1693
02:57:14,130 --> 02:57:21,010
titled, "Diet and Perceptions Change With Supermarket
Introduction in a Food Desert, But Not Because

1694
02:57:21,010 --> 02:57:27,960
of Supermarket Use." So, we found that this
new full-service supermarket led to some improvements

1695
02:57:27,960 --> 02:57:34,050
in local resident's diet, as well as to increase
satisfaction with their neighborhood as a

1696
02:57:34,050 --> 02:57:39,130
place to live.
However, the study found no significant improvements

1697
02:57:39,130 --> 02:57:44,311
in resident's fruit and vegetable intake,
whole grain intake, weight, or rates of obesity.

1698
02:57:44,311 --> 02:57:51,350
And importantly, the improvements in diet
and neighborhood satisfaction were not significantly

1699
02:57:51,350 --> 02:57:57,680
linked with frequency of shopping at the new
supermarket. Both frequent and infrequent

1700
02:57:57,680 --> 02:58:05,750
shoppers experienced the same improvements.
And just some quick illustrations to show

1701
02:58:05,750 --> 02:58:12,000
that there were indeed differences. Here we
see that our Hill District neighborhood residents

1702
02:58:12,000 --> 02:58:19,319
or the intervention residents did find a net
decrease in caloric intake compared with our

1703
02:58:19,319 --> 02:58:28,880
control group or our Homewood neighborhood
residents. And here you can see a decrease

1704
02:58:28,880 --> 02:58:34,620
among our Hill District residents in sugar
consumption, compared with a small increase

1705
02:58:34,620 --> 02:58:43,979
among our Homewood residents in sugar consumption
during the same period. Next, similarly, we

1706
02:58:43,979 --> 02:58:50,740
found a net decrease in empty calories among
our Hill District or intervention neighborhood

1707
02:58:50,740 --> 02:58:55,930
residents compared with our Homewood or comparison
neighborhood residents.

1708
02:58:55,930 --> 02:59:03,279
While this illustrates that both our intervention
and comparison, neighborhood residents decreased

1709
02:59:03,279 --> 02:59:09,910
their fruit and vegetable intake between the
baseline or prior to the grocery store opening

1710
02:59:09,910 --> 02:59:18,210
and after the grocery store opened. Before
the supermarket opening, residents in both

1711
02:59:18,210 --> 02:59:24,591
neighborhoods thought that healthy foods were
not easily accessible, with just 16 percent

1712
02:59:24,591 --> 02:59:29,979
of respondents perceiving fruits and vegetables
as easily accessible in the Hill District

1713
02:59:29,979 --> 02:59:39,920
and just 22 percent reporting the same thing
in Homewood in 2011. And here you can see

1714
02:59:39,920 --> 02:59:45,189
quite obvious changes that after the supermarket
opened, and we asked the exact same question

1715
02:59:45,189 --> 02:59:53,000
to the exact same people in 2014. Seventy two percent
of Hill District residents reported that

1716
02:59:53,000 --> 02:59:58,739
they perceived fruits and vegetables to be
easily accessible, while just 27 percent of

1717
02:59:58,739 --> 03:00:06,370
our Homewood participants reported the same
thing. Importantly, we found pretty large

1718
03:00:06,370 --> 03:00:13,790
changes in satisfaction with one's neighborhood
as a place to live, with Hill District residents

1719
03:00:13,790 --> 03:00:21,000
reporting a pretty large increase between
2011 and 2014 and Homewood residents staying

1720
03:00:21,000 --> 03:00:26,760
the same.
Now, when we looked at all of these outcomes

1721
03:00:26,760 --> 03:00:34,989
and whether they were related to regular supermarket
use, the only outcome that was related to

1722
03:00:34,989 --> 03:00:45,080
regular use of the supermarket was better-perceived
access to healthy foods. It is important to

1723
03:00:45,080 --> 03:00:52,630
note that when we looked specifically at SNAP
participants, we did see benefits in both

1724
03:00:52,630 --> 03:00:59,020
food security and diet after the opening of
the full-service supermarket. And these results

1725
03:00:59,020 --> 03:01:06,780
are published in a separate analysis in a
more recent paper in Health Affairs from August

1726
03:01:06,780 --> 03:01:16,420
of 2020. Now, because so many of the changes
were not associated with use of the supermarket,

1727
03:01:16,420 --> 03:01:22,280
we went back to our data to try and understand
whether it could have been improvements in

1728
03:01:22,280 --> 03:01:28,380
neighborhood socioeconomic conditions, perhaps
related to the new supermarket that may in

1729
03:01:28,380 --> 03:01:35,160
fact be responsible for these improvements
in resident diet. What we found was that,

1730
03:01:35,160 --> 03:01:43,569
in fact, neighborhood socioeconomic conditions
may have contributed to these dietary shifts.

1731
03:01:43,569 --> 03:01:50,790
However, it was not a simple story. We saw
through our pathway analyses that these shifts

1732
03:01:50,790 --> 03:01:56,600
seemed to operate differently for renters
versus homeowners.

1733
03:01:56,600 --> 03:02:07,670
I also want to note here that we collected
data again in 2018, which was four years after

1734
03:02:07,670 --> 03:02:15,980
our initial follow-up data collection to understand
whether these shifts held. Now our preliminary

1735
03:02:15,980 --> 03:02:22,660
and unpublished results suggest that, in
fact, the dietary shifts did not hold. However,

1736
03:02:22,660 --> 03:02:30,680
the neighborhood satisfaction and the perceived
access to healthy foods did seem to hold.

1737
03:02:30,680 --> 03:02:38,430
Please stay tuned for more results as we are
getting ready to submit them for publication.

1738
03:02:38,430 --> 03:02:45,250
I do want to note here that there's been a
lot of other work, not necessarily based on

1739
03:02:45,250 --> 03:02:50,470
natural experiments. But based on secondary
data analyses that have shown that household

1740
03:02:50,470 --> 03:02:55,820
resources are, in fact, a stronger predictor
of dietary quality than geographic access

1741
03:02:55,820 --> 03:03:02,080
to healthy food options. Some of this work
has been done by Markovsky and Snyder, who

1742
03:03:02,080 --> 03:03:07,620
used a national sample of proprietary data
on food shopping behavior, and they found

1743
03:03:07,620 --> 03:03:13,460
that households in low income and low supermarket
access census tracts may travel farther to

1744
03:03:13,460 --> 03:03:19,560
shop for food, but don't necessarily change
their purchasing habits because of distance.

1745
03:03:19,560 --> 03:03:26,870
Similarly, Ver Ploeg, Wilde, and colleagues
have analyzed food apps data and have found

1746
03:03:26,870 --> 03:03:34,890
similar results. So, while there's been a
good amount of research centered around food

1747
03:03:34,890 --> 03:03:41,970
deserts, supermarkets, and diet, only a few
studies have been able to capitalize on rigorous,

1748
03:03:41,970 --> 03:03:48,840
scientific design to examine the impact of
eliminating a food desert to investigate

1749
03:03:48,840 --> 03:03:55,029
whether the opening of a supermarket changes
food purchasing behaviors or diet of residents,

1750
03:03:55,029 --> 03:04:02,090
and there have been mixed results
on whether diet changes. We have also found

1751
03:04:02,090 --> 03:04:08,580
that people don't necessarily go to the nearest
supermarket. That is, factors other than proximity

1752
03:04:08,580 --> 03:04:17,060
matter. We've also found that other neighborhood
characteristics matter. We've found that the

1753
03:04:17,060 --> 03:04:28,380
pathways are potentially very complex. So
given all of this, what are our conclusions

1754
03:04:28,380 --> 03:04:37,050
and next steps? Well, first I ask whether
we're too focused on diet as an outcome, that

1755
03:04:37,050 --> 03:04:46,069
is perhaps there are other outcomes that are
just as and if not more important. Secondly,

1756
03:04:46,069 --> 03:04:52,250
supermarkets do not open in isolation. There
are other neighborhood conditions that are

1757
03:04:52,250 --> 03:04:58,330
likely changing, including housing, transportation,
general neighborhood aesthetics, and socioeconomic

1758
03:04:58,330 --> 03:05:04,340
conditions. There are likely other neighborhood
factors and characteristics that are tied

1759
03:05:04,340 --> 03:05:12,510
with supermarkets' openings. There are a spectrum
of possibilities regarding supermarket quality,

1760
03:05:12,510 --> 03:05:20,500
and this is actually very important and also
very hard to measure and to quantify. What

1761
03:05:20,500 --> 03:05:25,380
I've learned from all of this might be something
we already know, and that is really there

1762
03:05:25,380 --> 03:05:31,340
don't seem to be magic bullet solutions and
relationships, and pathways are part of the

1763
03:05:31,340 --> 03:05:36,979
processes.
I think it's really important to try and consider

1764
03:05:36,979 --> 03:05:46,760
all of this and look at it within all of these
different domains of complexity. Thank you so much.

1765
03:05:46,760 --> 03:05:55,250
DR. JOEL GITTELSOHN: I'm really pleased to be here today
to speak on this topic, Improving the Small

1766
03:05:55,250 --> 03:06:02,080
Food Source Environment, Overview and Future Directions. In this talk, I'm going to provide

1767
03:06:02,080 --> 03:06:07,330
a little bit of background, talk a bit about
what we've learned in terms of interventions

1768
03:06:07,330 --> 03:06:12,340
and small food sources, address some of the
key challenges and gaps and talk a bit about

1769
03:06:12,340 --> 03:06:18,460
future directions. First of all, some introductions. When I refer to small food sources, I'm talking

1770
03:06:18,460 --> 03:06:25,080
about small retail food sources, such as corner
stores, bodegas, convenience stores, and the

1771
03:06:25,080 --> 03:06:30,060
like. I'm also speaking about prepared food
sources, such as carryout restaurants and other

1772
03:06:30,060 --> 03:06:34,790
kinds of independently owned restaurants.
What I'm not talking about are chain fast

1773
03:06:34,790 --> 03:06:46,310
food restaurants. I'm not speaking about supermarkets
or other large grocery stores. So let's start

1774
03:06:46,310 --> 03:06:51,790
off by taking a picture of a particular context.
This is a map of the city of Baltimore, and

1775
03:06:51,790 --> 03:06:58,120
you can see the city is highlighted in several
different ways. One of the ways that we focused

1776
03:06:58,120 --> 03:07:03,330
on is in terms of identification of healthy
food priority areas, formerly known as food

1777
03:07:03,330 --> 03:07:09,160
deserts. About 25 percent of the residents
of Baltimore City live in healthy food priority

1778
03:07:09,160 --> 03:07:16,910
areas where there is low healthy food availability,
a low median household income, limited personal

1779
03:07:16,910 --> 03:07:24,819
vehicle access, and a greater distance to a
supermarket. And in this particular setting

1780
03:07:24,819 --> 03:07:30,890
of Baltimore City, despite the fact of low
access to larger food sources such as supermarkets,

1781
03:07:30,890 --> 03:07:36,860
there is a heavy access and use of small corner
stores, small grocery stores and carryout

1782
03:07:36,860 --> 03:07:41,279
restaurants. As depicted in this picture,
you can see all the different colored dots

1783
03:07:41,279 --> 03:07:48,210
represent all the small, different types of
small food sources. And in this particular

1784
03:07:48,210 --> 03:07:53,350
setting, and in many other settings, there
appear to be associations between the presence

1785
03:07:53,350 --> 03:07:58,910
of these types of food sources, and food insecurity
and obesity.

1786
03:07:58,910 --> 03:08:05,750
And you can see this map shows essentially
an overlap between areas of high food insecurity

1787
03:08:05,750 --> 03:08:13,040
in obesity and the areas that you saw before
that were healthy food priority areas. So

1788
03:08:13,040 --> 03:08:19,149
then why do we think working in small food
source interventions is a good idea? For several

1789
03:08:19,149 --> 03:08:24,610
different reasons. First of all, they're ubiquitous.
These small food sources are already in low income

1790
03:08:24,610 --> 03:08:29,330
disadvantaged communities, and so therefore
there's no need to build entirely new stores

1791
03:08:29,330 --> 03:08:34,630
and supermarkets. They're already there. They're
also very regularly used. They tend to be

1792
03:08:34,630 --> 03:08:41,390
open five to seven days a week, 12 or more
hours a day, and many residents report using

1793
03:08:41,390 --> 03:08:47,580
these small food sources on a daily basis.
Some two or three times a day. Also, much

1794
03:08:47,580 --> 03:08:52,670
of the high fat, high sugar, high sodium foods
that people consume come from these sources.

1795
03:08:52,670 --> 03:08:59,160
So this represents an opportunity to address
the the root of some of the dietary issues

1796
03:08:59,160 --> 03:09:06,270
that we are considering for these settings.
Prepared food sources also account as for

1797
03:09:06,270 --> 03:09:09,640
more than 50 percent of calories in some of
these settings.

1798
03:09:09,640 --> 03:09:15,000
And so therefore, that's part of the reason
we selected prepared food sources, as well

1799
03:09:15,000 --> 03:09:22,540
as retail food sources as part of the topic
of this particular presentation. Another reason

1800
03:09:22,540 --> 03:09:27,950
to work with small food sources may not be
so apparent immediately, but the owners of

1801
03:09:27,950 --> 03:09:33,490
these food sources need to have good relationships
with community members. They need to try to

1802
03:09:33,490 --> 03:09:40,029
respond to the needs of community members
in order to stay in business. Well, what have

1803
03:09:40,029 --> 03:09:44,729
some of the interventions looked like to improve
small food sources? There's essentially been

1804
03:09:44,729 --> 03:09:51,050
four different approaches that have been used,
demand type interventions, supply type interventions,

1805
03:09:51,050 --> 03:09:56,000
infrastructure, and training. Demand interventions
are ones that focus on communication with

1806
03:09:56,000 --> 03:10:00,930
the consumers or the customers of these small
food sources. So point of purchase promotions

1807
03:10:00,930 --> 03:10:05,950
such as shelf labeling, posters, flyers, menu
labeling, and prepared food sources and so

1808
03:10:05,950 --> 03:10:11,720
forth, as well as interactive activities,
including taste tests. Then, we have supply

1809
03:10:11,720 --> 03:10:17,620
types of interventions which act to increase
or decrease the availability of healthy or

1810
03:10:17,620 --> 03:10:23,960
less healthy foods, and perhaps making them
a little bit easier to locate within the store.

1811
03:10:23,960 --> 03:10:28,710
We can talk about infrastructural changes,
such as the provision of produce refrigerators,

1812
03:10:28,710 --> 03:10:34,960
and we can talk about training of small store
owners and managers introducing methods of

1813
03:10:34,960 --> 03:10:39,899
how to make healthier food selections, how
to source those foods, how to store them safely,

1814
03:10:39,899 --> 03:10:47,090
how to promote these foods to their customers.
And most of the work that has taken place

1815
03:10:47,090 --> 03:10:52,080
in the past in these small food sources has
focused on combining two or more of these

1816
03:10:52,080 --> 03:10:59,291
approaches. About a decade ago, we did a couple
of systematic reviews. The first you can see

1817
03:10:59,291 --> 03:11:05,000
is listed is for small food stores, so for
retail food stores. And the second is for

1818
03:11:05,000 --> 03:11:10,029
prepared food sources. And I'm going to be
providing a little bit of summary of some

1819
03:11:10,029 --> 03:11:15,680
of the main findings of those studies. I've
also have updated the information a bit by

1820
03:11:15,680 --> 03:11:22,300
some of the more recent reviews, such as the
one by Pinard listed at the bottom. And if

1821
03:11:22,300 --> 03:11:27,380
you look at the more recent intervention approaches
in small food stores, in addition to what

1822
03:11:27,380 --> 03:11:32,040
I've described before, you can see that a
number of the interventions more recently

1823
03:11:32,040 --> 03:11:37,350
have focused on strategies such as incorporating
social marketing approaches.

1824
03:11:37,350 --> 03:11:42,641
So audience segmentation and targeting approaches,
behavioral, economic or nudging strategies.

1825
03:11:42,641 --> 03:11:46,250
There's been increased interest and use
of different types of pricing strategies to

1826
03:11:46,250 --> 03:11:51,930
encourage the selection of healthier foods,
increased interest in emphasis on store conversions

1827
03:11:51,930 --> 03:11:58,410
or larger types of infrastructure changes
beyond just introducing a refrigerator, and

1828
03:11:58,410 --> 03:12:04,170
an emphasis on community partnerships on strengthening
the relationships between store owners and

1829
03:12:04,170 --> 03:12:12,010
community members. What have these different
studies found over the years? Well, I think

1830
03:12:12,010 --> 03:12:17,700
there's been a lot of promising findings.
First of all, in general, almost all of these

1831
03:12:17,700 --> 03:12:24,130
interventions of any type have shown increases
in the stocking of the promoted foods or beverages.

1832
03:12:24,130 --> 03:12:32,060
When it's been measured, there's been a impact
on sales of these foods typically found, and

1833
03:12:32,060 --> 03:12:38,760
there's been an impact seen in customer purchasing
of these items. When it's been measured, we

1834
03:12:38,760 --> 03:12:43,729
see some dietary changes, and in the very
few times it's been assessed, we do see some

1835
03:12:43,729 --> 03:12:46,569
positive health outcomes occasionally with
these interventions.

1836
03:12:46,569 --> 03:12:54,090
But I have to emphasize that measurement of
health outcomes in small food source interventions

1837
03:12:54,090 --> 03:13:00,550
is actually quite rare. So quite a few challenges
remain for small food source interventions,

1838
03:13:00,550 --> 03:13:06,090
and I'd like to list them out and talk a little
bit about how we and others have addressed

1839
03:13:06,090 --> 03:13:11,910
them. So I think one of the major challenges
I hear about is how to scale up and sustain

1840
03:13:11,910 --> 03:13:18,330
small food source interventions. Related to
this is a lack of attention to suppliers of

1841
03:13:18,330 --> 03:13:26,100
these small food sources. A challenge that
remains is that level of community engagement.

1842
03:13:26,100 --> 03:13:31,090
Although there are reasons for this engagement
to take place, it can vary dramatically. There

1843
03:13:31,090 --> 03:13:36,730
can be a lack of technology in stores, for
example, for inventory management to track

1844
03:13:36,730 --> 03:13:44,399
sales and so forth and lack of rigorous evaluations,
taking it all the way to health outcomes or

1845
03:13:44,399 --> 03:13:50,460
even sometimes to dietary outcomes. It's just
not there in a lot of the studies, even nowadays,

1846
03:13:50,460 --> 03:13:55,739
that are being conducted. So these are some
of the challenges. I wanted to address a few

1847
03:13:55,739 --> 03:14:04,780
of these a little more closely and talk about,
first of all, this challenge of scaling up.

1848
03:14:04,780 --> 03:14:11,280
In order to scale up the the sort of numerous
small food source interventions that have

1849
03:14:11,280 --> 03:14:15,420
been conducted, we need to address what I
call the supplier or the distribution problem

1850
03:14:15,420 --> 03:14:23,270
that exists for small, independent food sources.
And in Baltimore City, what we found is that

1851
03:14:23,270 --> 03:14:27,729
it's a very different picture in terms of
the distribution modality if you are a bag

1852
03:14:27,729 --> 03:14:33,511
of potato chips than if you are a bag of groceries.
If you are a bag of potato chips, there's

1853
03:14:33,511 --> 03:14:42,260
a very strong supply demand loop between wholesalers
or distributors and the corner stores. Those

1854
03:14:42,260 --> 03:14:48,479
foods are delivered to the corner stores.
This creates a strong sales of those foods.

1855
03:14:48,479 --> 03:14:54,830
There's a perceived strong demand, a very,
very robust feedback loop. This is strengthened

1856
03:14:54,830 --> 03:15:00,010
by formal and informal agreements between
wholesalers and distributors, and the corner

1857
03:15:00,010 --> 03:15:05,620
stores, such as free delivery, incentives,
reduced cost for items, provision of freezers

1858
03:15:05,620 --> 03:15:10,489
or refrigerators or display racks, that sort
of thing. The situation is almost entirely

1859
03:15:10,489 --> 03:15:17,550
the reverse. If you are a bag of groceries
or produce in a corner store setting, the

1860
03:15:17,550 --> 03:15:21,880
supply demand feedback loop is very weak,
and in fact, for the most part, corner store

1861
03:15:21,880 --> 03:15:26,500
owners have to go get it themselves.
They have to drive to the wholesaler or distributor

1862
03:15:26,500 --> 03:15:32,250
and pick up these foods. There is costly or
no delivery available, and where there is

1863
03:15:32,250 --> 03:15:38,160
delivery, it often requires a high level of
minimum purchase to make it worthwhile. So

1864
03:15:38,160 --> 03:15:45,851
very much disadvantaged in terms of access
to healthier foods for small retail food sources,

1865
03:15:45,851 --> 03:15:51,520
at least in terms of corner stores. We've
actually been trying to address this challenge

1866
03:15:51,520 --> 03:15:58,370
in Baltimore City through the development
of a mobile app we call BUD, which stands

1867
03:15:58,370 --> 03:16:03,760
for Baltimore Urban Food Distribution app.
And the goal of BUD is to develop an affordable

1868
03:16:03,760 --> 03:16:08,529
solution for corner stores to access and have
delivered healthier foods and beverages, ultimately

1869
03:16:08,529 --> 03:16:13,080
leading to improved access to locally sourced
healthier foods and a more resilient food

1870
03:16:13,080 --> 03:16:19,990
system. And essentially, what the app does
is it connects corner stores with local producers

1871
03:16:19,990 --> 03:16:28,060
and wholesalers. Wholesalers and producers
can set up reduced cost product, healthier

1872
03:16:28,060 --> 03:16:32,061
products that are reduced costs through collective
purchasing.

1873
03:16:32,061 --> 03:16:37,720
Corner owners can sign up for and collectively
purchase a BuddyUp! deal so they can get

1874
03:16:37,720 --> 03:16:43,970
even a small amount of produce or other healthier
foods for a relatively inexpensive price and

1875
03:16:43,970 --> 03:16:51,120
have it delivered to their store. So this
is in process, and this is a type of intervention

1876
03:16:51,120 --> 03:16:58,570
strategy to address the supply, the distribution
and supplier challenge. Some follow up questions

1877
03:16:58,570 --> 03:17:09,199
that we are faced with are things like how
to handle distribution costs effectively and

1878
03:17:09,199 --> 03:17:14,199
essentially in terms of keeping the costs
down low. The fact that in fact most corner

1879
03:17:14,199 --> 03:17:19,229
stores are not interested in just stocking
healthy foods, that in order to make the use

1880
03:17:19,229 --> 03:17:26,460
of the app available and desirable to them,
we probably have to make the app stock other

1881
03:17:26,460 --> 03:17:34,590
foods, possibly unhealthy foods. Another challenge
or follow up that we need to consider is the

1882
03:17:34,590 --> 03:17:39,690
fact that we're talking about prepared food
sources as well as retail food sources. So

1883
03:17:39,690 --> 03:17:45,040
we need to see if these supplier related challenges
are similar for prepared food sources. So

1884
03:17:45,040 --> 03:17:48,069
there's some further work that needs to be
done in this area.

1885
03:17:48,069 --> 03:17:55,520
A second gap relates to this whole issue of
sustainability, which in some ways is overlapping

1886
03:17:55,520 --> 03:18:00,640
with the concept of profitability from the
store perspective, as well as public health

1887
03:18:00,640 --> 03:18:07,489
benefits. And any solution or program or policy
for sustainable change in small food sources

1888
03:18:07,489 --> 03:18:14,630
has to be tied to profitability in my view.
Much intervention work overlooks this issue

1889
03:18:14,630 --> 03:18:20,600
of sustainability, and an approach that we've
used or one strategy is to take a system science

1890
03:18:20,600 --> 03:18:27,239
view of things. An example of this is at one
time Baltimore was considering implementation

1891
03:18:27,239 --> 03:18:32,340
of the Staple Foods Ordinance. And so, we worked
with them to develop a systems dynamic model

1892
03:18:32,340 --> 03:18:37,540
to simulate the inclusion of different foods
and beverages in different amounts in a proposed

1893
03:18:37,540 --> 03:18:43,410
ordinance, and use the systems dynamic model
to recommend modifications to the ordinance.

1894
03:18:43,410 --> 03:18:47,300
This is a visual that was created through
a group model building process of the flow

1895
03:18:47,300 --> 03:18:53,311
of different things. You can see that there's
a ordinance characteristics, there's supply

1896
03:18:53,311 --> 03:18:59,220
characteristics, there's consumer characteristics.
This is a interface that shows some of the

1897
03:18:59,220 --> 03:19:04,250
slider bars that have been done to modify some
of the different parameters. And we tested

1898
03:19:04,250 --> 03:19:10,960
four different simulation: a basic SNAP simulation,
a SNAP depth of stock simulation, the Minneapolis

1899
03:19:10,960 --> 03:19:15,560
Staple Foods Ordinance requirements, and WIC
requirements. And these are some of the details

1900
03:19:15,560 --> 03:19:21,930
of the simulation in terms of required minimum
stock. Comparatively, the level of enforcement

1901
03:19:21,930 --> 03:19:30,910
that we perceive to be present or required.
And what we're able to output in this case,

1902
03:19:30,910 --> 03:19:36,149
in this particular image is what would be
the profits associated with each of these

1903
03:19:36,149 --> 03:19:41,500
different simulations of a different type
of a staple foods ordinance. And you can see

1904
03:19:41,500 --> 03:19:51,149
that certain of these simulations or revealed
more profitable types of outcomes in terms

1905
03:19:51,149 --> 03:19:56,520
of the store perspective than others. So, the
snap depth of stock simulation showed that,

1906
03:19:56,520 --> 03:20:09,660
in fact, it was very unprofitable to implement
that particular policy versus the Minneapolis

1907
03:20:09,660 --> 03:20:13,319
Staple Foods Ordinance would actually provide
a small profit.

1908
03:20:13,319 --> 03:20:18,930
There are a lot of different outputs that
are possible with this type of simulation,

1909
03:20:18,930 --> 03:20:22,750
and we're currently writing and publishing
on this work.

1910
03:20:22,750 --> 03:20:27,790
DR. JOEL GITTTELSOHN: Some follow up questions,
do system science methods like group model

1911
03:20:27,790 --> 03:20:33,500
building replace more traditional formative
research and community engagement approaches?

1912
03:20:33,500 --> 03:20:39,840
And, of course, how best to combine system
science approaches with other types of strategies

1913
03:20:39,840 --> 03:20:47,330
for planning. Finally, I want to just bring
up this issue that of multilevel, multicomponent

1914
03:20:47,330 --> 03:20:53,729
interventions as a particular gap or area
of work. When I speak of multilevel, multicomponent

1915
03:20:53,729 --> 03:21:00,489
interventions, I talk about combining different
intervention strategies in a thoughtful way.

1916
03:21:00,489 --> 03:21:05,580
And so we need to think about complementarity
of the different interventions, the interventions

1917
03:21:05,580 --> 03:21:10,540
being referential to each other. So, for example,
a school intervention refers to or builds

1918
03:21:10,540 --> 03:21:15,010
on a retail food store intervention. They
need to be mutually supportive with one building

1919
03:21:15,010 --> 03:21:20,270
on another and supporting another. So, if
you're going to work with small food sources,

1920
03:21:20,270 --> 03:21:25,779
you should need to work with suppliers. You
need to work with and effect demand by customers

1921
03:21:25,779 --> 03:21:34,910
as an example, and this is just a visual image
of the B’More Healthy Communities for Kids

1922
03:21:34,910 --> 03:21:40,370
intervention conceptual framework showing
the different components of this multilevel

1923
03:21:40,370 --> 03:21:44,470
intervention, which was implemented successfully
in Baltimore City.

1924
03:21:44,470 --> 03:21:49,660
And giving some idea about how the different
levels or components of the intervention were

1925
03:21:49,660 --> 03:21:56,790
reinforced and complemented each other through
the use of agent based modeling through social

1926
03:21:56,790 --> 03:22:04,830
media, through the stocking of healthier foods,
and so forth. So, some of the follow up questions

1927
03:22:04,830 --> 03:22:09,500
and there are many beyond these, what are
the community level effects of multilevel

1928
03:22:09,500 --> 03:22:14,970
interventions, not just the effects that would
be measured at an individual level? How best

1929
03:22:14,970 --> 03:22:20,370
to measure these effects? How to sustain these
kinds of complex programs and initiatives?

1930
03:22:20,370 --> 03:22:27,050
These are all things that we need to address
in future work. So, in summary, small food

1931
03:22:27,050 --> 03:22:33,310
sources, both retail and prepared, are ubiquitous
in low income communities. They tend to be

1932
03:22:33,310 --> 03:22:39,130
regularly used and are partially responsible,
at least for the unhealthy food access, diet

1933
03:22:39,130 --> 03:22:44,899
patterns, and diet related chronic disease.
Interventions in these settings typically

1934
03:22:44,899 --> 03:22:51,260
target supply, demand, infrastructure, training
of staff or some combination of these strategies,

1935
03:22:51,260 --> 03:22:57,720
and they've had positive impacts in the literature
on stocking, sales, revenues, purchasing, and

1936
03:22:57,720 --> 03:23:02,340
consumption of healthier foods.
Small food sources are, of course, are part

1937
03:23:02,340 --> 03:23:08,439
of larger food systems which will influence
supply and distribution issues, and figuring

1938
03:23:08,439 --> 03:23:14,230
out interventions which can address these
multiple levels are are important. System

1939
03:23:14,230 --> 03:23:19,240
science offers one set of tools that can help
us to better understand and plan for cost

1940
03:23:19,240 --> 03:23:25,689
effective, sustainable interventions. And
related to all of this, multilevel interventions,

1941
03:23:25,689 --> 03:23:32,390
which incorporate small food sources, show
great promise in achieving some of these objectives.

1942
03:23:32,390 --> 03:23:40,060
I want to thank you for your attention today,
and please feel free to contact me for future

1943
03:23:40,060 --> 03:23:43,180
questions or requests for information. Thank
you.

1944
03:23:43,180 --> 03:23:49,270
DR. CHRISTINA ROBERTO: Hi, everyone. My name is
Christina Roberto, and I'm delighted to be

1945
03:23:49,270 --> 03:23:53,489
able to participate in this workshop today.
The title of my talk is promoting healthy

1946
03:23:53,489 --> 03:23:59,910
choices in food pantries. And in the short
time we have together, what I'm going to try

1947
03:23:59,910 --> 03:24:04,939
to do is cover what we know about interventions
and food pantries to promote healthy eating.

1948
03:24:04,939 --> 03:24:08,620
I'll then share some insights from psychology
and behavioral economics that can be used

1949
03:24:08,620 --> 03:24:12,770
to inform the design of food pantry interventions.
And then I'll tell you a little bit about

1950
03:24:12,770 --> 03:24:17,399
some work we're doing in my lab, the PEACH
lab, on a behavioral economics informed randomized

1951
03:24:17,399 --> 03:24:24,490
controlled trial to promote healthy choices
in food pantries. To address what is known

1952
03:24:24,490 --> 03:24:28,800
on food pantry based interventions to promote
healthy choices, there's a recent literature

1953
03:24:28,800 --> 03:24:35,859
review by on colleagues and Public Health
Nutrition from 2019. And I went through that

1954
03:24:35,859 --> 03:24:39,650
review carefully to try to identify studies
to present today. They had identified seven

1955
03:24:39,650 --> 03:24:44,489
randomized controlled trials in seven pre-post
studies, and I thought in our short time together,

1956
03:24:44,489 --> 03:24:48,220
we would focus on the randomized controlled
trials because those tell us the most in terms

1957
03:24:48,220 --> 03:24:52,600
of causal impacts of these interventions.
Fortunately, in this workshop, we have folks

1958
03:24:52,600 --> 03:24:55,569
like Hilary Seligman and Caitlin Caspi, who
are really expert in this area and doing very

1959
03:24:55,569 --> 03:24:59,750
innovative work, and I'm sure they'll share
some of their research as well. So, if we

1960
03:24:59,750 --> 03:25:05,470
drill down into the RCTs, you actually find
that some of them are multiple papers about

1961
03:25:05,470 --> 03:25:12,029
one intervention. One of them, it wasn't clear
to me that the outcomes related to a different

1962
03:25:12,029 --> 03:25:15,890
health or behavioral outcomes that weren't meaningful
for our conversation around food insecurity

1963
03:25:15,890 --> 03:25:18,990
and healthy choices. So, I won't be discussing
that one. So, I'm just going to talk about

1964
03:25:18,990 --> 03:25:23,840
the first three on your screen there. And
I do want to put an asterisk on the Wilson

1965
03:25:23,840 --> 03:25:27,590
one. Brian one is the last author on that,
and a lot of his work has been discredited

1966
03:25:27,590 --> 03:25:32,180
in the last several years, but he is not the
first author, so I thought I would share it

1967
03:25:32,180 --> 03:25:42,340
with the asterisk that it's a study that would
need to be replicated. So, this first study

1968
03:25:42,340 --> 03:25:47,790
that I'll share is a test of the Food Stamp Nutrition Education Program that was offered

1969
03:25:47,790 --> 03:25:52,330
at community sites, including food pantries,
was also available to clients in their home.

1970
03:25:52,330 --> 03:25:57,670
And this was a basic education program that
covered five different topics. They included,

1971
03:25:57,670 --> 03:26:04,020
MyPyramid discussion around food groups,
food safety, shopping behaviors, and research

1972
03:26:04,020 --> 03:26:07,470
management involvement. So, talking about
things like shopping on a budget. And in the

1973
03:26:07,470 --> 03:26:14,500
program, 30 to 60 minute lessons were provided
and the lessons were tailored. So, in this

1974
03:26:14,500 --> 03:26:19,660
study, they randomized women to either that
food study nutrition education program or

1975
03:26:19,660 --> 03:26:24,570
to a control group. And I just want to highlight
a couple of the methods, concerns, and the

1976
03:26:24,570 --> 03:26:28,160
randomization. So, clients that were arriving
together were randomized to the same group,

1977
03:26:28,160 --> 03:26:34,190
but that wasn't stratified, and so it likely
led to these group sizing balances. And as

1978
03:26:34,190 --> 03:26:37,990
far as I can tell, I don't think they were
clustering standard errors to account for

1979
03:26:37,990 --> 03:26:45,160
people in the same group. So, these are just
some limitations to this RCT. In terms of

1980
03:26:45,160 --> 03:26:51,229
the actual study design, one lesson was provided
to both the Food Stamp Nutrition Education

1981
03:26:51,229 --> 03:26:54,590
group and the control group. And then that
Food Stamp Nutrition Education group received

1982
03:26:54,590 --> 03:27:00,050
four additional lessons, and the control group
received lessons after the study was over.

1983
03:27:00,050 --> 03:27:06,330
What they found is that the intervention significantly
increased food security by point 0.37 points

1984
03:27:06,330 --> 03:27:12,460
on a 6.0 scale compared to the control group.
I'll mention this suggests some promising

1985
03:27:12,460 --> 03:27:21,040
data on the basic nutrition education to improve
food security and a food pantry and other

1986
03:27:21,040 --> 03:27:26,710
settings. So, although nutrition education
can be very useful and is, you know, an important

1987
03:27:26,710 --> 03:27:30,600
component of many interventions, I think the
field is recognized over many decades that

1988
03:27:30,600 --> 03:27:34,890
education is not going to be sufficient. And
so I'll talk about this intervention, the

1989
03:27:34,890 --> 03:27:39,790
Freshplace food pantry intervention, which
really tries to go beyond education. And this

1990
03:27:39,790 --> 03:27:43,340
intervention is designed to address the root
causes of poverty, and it was developed by

1991
03:27:43,340 --> 03:27:48,060
three community agencies and in collaboration
with the university to do the evaluation.

1992
03:27:48,060 --> 03:27:53,310
Now, a choice pantry is one where you can
go and select the food that you want, as opposed

1993
03:27:53,310 --> 03:27:58,109
to a lot of other pantries where food might
be prepackaged in bags, and you don't have

1994
03:27:58,109 --> 03:28:00,630
much choice.
The goal of this intervention was to work

1995
03:28:00,630 --> 03:28:05,649
with clients to help them set small, achievable
behavior change goals, and I'll describe a

1996
03:28:05,649 --> 03:28:09,700
little bit more about what it involved. So,
it included monthly meetings with a project

1997
03:28:09,700 --> 03:28:14,250
manager to receive motivational interviewing
using goal setting and provide social support,

1998
03:28:14,250 --> 03:28:20,410
and then it had targeted referrals to community
services like SNAP and Energy Assistance.

1999
03:28:20,410 --> 03:28:24,121
And finally, there were tailored offerings.
So, a six week cooking classes, nutrition

2000
03:28:24,121 --> 03:28:29,140
education, but also access to computers for
job searches, depending on different needs

2001
03:28:29,140 --> 03:28:35,130
of different clients. In this randomized trial,
they randomized adults to either the Freshplace

2002
03:28:35,130 --> 03:28:38,990
intervention I just described or a control
which were food pantries where bags of food

2003
03:28:38,990 --> 03:28:45,229
were received. And you can see the ns' on
your screen. And then they did quarterly follow

2004
03:28:45,229 --> 03:28:48,130
ups for 12 months. At baseline, about half
the sample reported very low food security.

2005
03:28:48,130 --> 03:28:56,569
The results from this study found that the
Freshplace group was half as likely as the

2006
03:28:56,569 --> 03:28:57,569
control group to experience very low food
security.

2007
03:28:57,569 --> 03:29:01,939
In addition, those in the Freshplace group
gained 4.1 points in self-sufficiency across

2008
03:29:01,939 --> 03:29:11,739
all the different time points when compared
to control, and the households with the lower

2009
03:29:11,739 --> 03:29:14,110
monthly incomes benefited more from it. And
finally, the Freshplace group significantly

2010
03:29:14,110 --> 03:29:17,850
increased their fruit and vegetable intake
compared to the control, and that amounted

2011
03:29:17,850 --> 03:29:26,380
to two additional points on the measure they
were using, or one additional serving per

2012
03:29:26,380 --> 03:29:31,930
day. So, they are really some promising results
from this intervention. This next study took

2013
03:29:31,930 --> 03:29:34,140
place at a food pantry in New York. And when
the researchers were visiting the pantry,

2014
03:29:34,140 --> 03:29:37,540
they noticed there was a dessert table that
had desserts like frosted sheet cakes, brownies,

2015
03:29:37,540 --> 03:29:42,229
cookies, pies, assorted pastries. But they
also noticed that one of the dessert items

2016
03:29:42,229 --> 03:29:47,790
was a protein bar, and relative to the other
items that seemed healthier. And so they wanted

2017
03:29:47,790 --> 03:29:52,160
to encourage people to select the protein
bar. So, the intervention manipulated two

2018
03:29:52,160 --> 03:29:56,330
things, they manipulated whether the protein
bar was displayed at the front of the table

2019
03:29:56,330 --> 03:30:01,810
or at the back of the table, and then they
also manipulated the packaging it was in.

2020
03:30:01,810 --> 03:30:09,030
It was either in its original packaging in
a box or in a clear bag without the original

2021
03:30:09,030 --> 03:30:15,830
packaging. OK. So, this study took place over
five successive Tuesdays between October and

2022
03:30:15,830 --> 03:30:22,040
November, and they alternated treatment by
days. There were 443 people who participated

2023
03:30:22,040 --> 03:30:27,120
because that was the number of the clients
attending the food pantry on those days and

2024
03:30:27,120 --> 03:30:32,569
they would stop data collection if any of
the dessert products ran out. So, let's take

2025
03:30:32,569 --> 03:30:37,860
a look at the first set of results from this
study. On the y axis are the proportion of

2026
03:30:37,860 --> 03:30:40,920
people that actually chose the protein bar,
and you can see there's an advantage for putting

2027
03:30:40,920 --> 03:30:47,000
it in the front, though it is not statistically
significant, and you can see there's an advantage

2028
03:30:47,000 --> 03:30:53,040
for keeping it in its original boxed packaging,
though again, it's not statistically significant.

2029
03:30:53,040 --> 03:30:55,850
When they break down by the four different
cells, the only one that emerges as statistically

2030
03:30:55,850 --> 03:30:59,700
significant with a higher proportion of people
taking the bar is when it's placed in the

2031
03:30:59,700 --> 03:31:08,540
front and in the original boxed packaging.
OK, so, I started off with this nice systematic

2032
03:31:08,540 --> 03:31:10,930
review that looked like it had a ton of studies
in it. But once you really dug into the ones

2033
03:31:10,930 --> 03:31:14,550
where we have, you know, good randomized,
at least the ones where we have randomized

2034
03:31:14,550 --> 03:31:21,100
controlled data at this point where we can
feel more certain about what we know, you

2035
03:31:21,100 --> 03:31:25,200
realize that it's really limited and a lot
more research needs to be done. There are

2036
03:31:25,200 --> 03:31:29,680
tons of gaps, and I feel that, you know, the
field is really wide open. The pre-post studies

2037
03:31:29,680 --> 03:31:33,130
that I didn't get into, I think, have some
really innovative ideas, and I think the next

2038
03:31:33,130 --> 03:31:38,310
phase for those are going to be moving into
randomized controlled trials. And my last

2039
03:31:38,310 --> 03:31:44,910
part of the talk I'm going to share a food
pantry experiment that we're doing in my lab

2040
03:31:44,910 --> 03:31:50,510
that's funded by NIH NCI and I don't have
results for you, but I will share sort of

2041
03:31:50,510 --> 03:31:57,290
how we design the intervention and what the
research design looks like. So, I just want

2042
03:31:57,290 --> 03:32:02,000
to acknowledge my wonderful research team
and community partners. These are the collaborators

2043
03:32:02,000 --> 03:32:07,380
on the study, and we're working with a community
based organization called the Jewish Federation

2044
03:32:07,380 --> 03:32:10,260
of Greater Philadelphia, and Brian Gralnick
and Deidre Mulligan are two points of contact

2045
03:32:10,260 --> 03:32:15,780
there, and they've just been absolutely amazing
partners and we call this study the Mitzvah Market study.

2046
03:32:15,780 --> 03:32:20,310
We designed the study really thinking about

2047
03:32:20,310 --> 03:32:24,290
principles from psychology and behavioral
economics. So, there's a fundamental insight

2048
03:32:24,290 --> 03:32:29,580
in psychology that's now, you know, really
well known where we think that there are kind

2049
03:32:29,580 --> 03:32:32,547
of two systems of thought in how people make
judgments and decisions. So, a good way to

2050
03:32:32,547 --> 03:32:38,410
think about this is there's a kind of system
one that's fast and automatic, effortless

2051
03:32:38,410 --> 03:32:45,420
we make a lot of judgments based on associations
and emotions when we're in a system one kind

2052
03:32:45,420 --> 03:32:49,040
of mindset. Sometimes though, we're engaging
system two, and that's when we're doing the

2053
03:32:49,040 --> 03:32:54,870
kind of thinking that's slow, controlled and
effectual. And so most of our food decisions

2054
03:32:54,870 --> 03:32:59,040
and choices that we're making, we're often
in system one at the grocery store, you know,

2055
03:32:59,040 --> 03:33:03,189
at a restaurant just kind of making these
decisions quickly as opposed to, you know,

2056
03:33:03,189 --> 03:33:06,620
picking up a cereal box, turning around and
carefully studying the nutrition facts label

2057
03:33:06,620 --> 03:33:14,060
that'd be engaging kind of system through
reasoning. So, a lot of what I'll present

2058
03:33:14,060 --> 03:33:18,680
has been, of course, widely popularized by
the book Nudge, which makes the case that

2059
03:33:18,680 --> 03:33:23,171
choice context really matters, and there's,
you know, been decades of psychological research

2060
03:33:23,171 --> 03:33:28,220
on this, and the idea of "Nudge" is that they
don't abridge freedom.

2061
03:33:28,220 --> 03:33:31,560
So, they're not, you know, they're not bands,
they're not taxes, they don't change price

2062
03:33:31,560 --> 03:33:37,000
incentives, they're trying to redesign the
choice experience to encourage healthier choices

2063
03:33:37,000 --> 03:33:41,880
and choices that are in people's self-interest
and without incurring that sort of additional

2064
03:33:41,880 --> 03:33:47,460
penalties. So, for the study that I'm going
to describe to you, we had a really unique

2065
03:33:47,460 --> 03:33:51,670
opportunity here to change Mitzvah Market's
online ordering system. So, this is a choice

2066
03:33:51,670 --> 03:33:56,010
pantry that has an online ordering system
where people place orders about once a month,

2067
03:33:56,010 --> 03:34:03,240
and you can see a picture of it on your screen
there. So, for this study, we took this opportunity

2068
03:34:03,240 --> 03:34:07,680
to redesign the online interface using insights
from psychology and behavioral economics.

2069
03:34:07,680 --> 03:34:12,600
So, for example, one foundational insight
is we're influenced by the placement of things

2070
03:34:12,600 --> 03:34:18,409
and cues that we see. So, based on some work
of that Anne Thorndyke has done demonstrating

2071
03:34:18,409 --> 03:34:23,610
that traffic light food labeling increased
the purchase of healthy green items in a hospital

2072
03:34:23,610 --> 03:34:27,260
cafeteria and decreased the purchase of less
healthy red items.

2073
03:34:27,260 --> 03:34:32,560
We've also adopted a traffic light labeling
system. In that same study, that Anna and colleagues

2074
03:34:32,560 --> 03:34:38,210
did show that making water more available
in the cafeteria also boosted sales. So, we're

2075
03:34:38,210 --> 03:34:44,150
also using a choice architecture intervention
where we display the healthiest option at

2076
03:34:44,150 --> 03:34:47,021
the top of a web page within a certain food
isle, and I'll show you what some of this looks

2077
03:34:47,021 --> 03:34:51,580
like in a moment. The next insight we applied
was the idea that people are highly prone

2078
03:34:51,580 --> 03:34:56,500
to stick with default options. There's a very
famous organ donation study that I think most

2079
03:34:56,500 --> 03:35:00,891
people are familiar with at this point that
shows that when you have to opt out, so you're

2080
03:35:00,891 --> 03:35:04,720
defaulted to be an organ donor in those
countries where you're defaulted to be an

2081
03:35:04,720 --> 03:35:09,790
organ donor, they have much higher rates of
organ donation. And so we've applied this

2082
03:35:09,790 --> 03:35:14,550
to our online store by pre-populating shopping
baskets with healthy, preferred items. So,

2083
03:35:14,550 --> 03:35:19,790
for example, when you log in, brown rice might
already be defaulted into your shopping cart,

2084
03:35:19,790 --> 03:35:22,400
and you can kick it out if you don't want
it.

2085
03:35:22,400 --> 03:35:26,689
But it's a nudge to try to encourage you to
order the brown rice. And Jamie Coutinho has

2086
03:35:26,689 --> 03:35:31,979
some really nice work using this kind of pre-populated
shopping basket, which is what we drew from

2087
03:35:31,979 --> 03:35:37,460
to create that intervention. The other insight
is the notion that other people signal what

2088
03:35:37,460 --> 03:35:42,270
to do and what is appropriate. So, we tend
to be very influenced by other people's actions

2089
03:35:42,270 --> 03:35:47,000
and social norms. And so what we ended up
doing in our store is displaying messages

2090
03:35:47,000 --> 03:35:53,830
highlighting commonly selected healthy items.
So, for example, you know, we might say that

2091
03:35:53,830 --> 03:35:58,130
in any given month, bananas are the most popular
fruit being selected in the pantry, and that's

2092
03:35:58,130 --> 03:36:04,290
data driven, that's based on the actual data
from the pantry. And finally, our last intervention

2093
03:36:04,290 --> 03:36:09,430
is called healthy swaps, where if you click
on an item, let's say white rice, for example,

2094
03:36:09,430 --> 03:36:15,010
a pop up like this would come up, nudging
you to switch to brown rice. And the pantry

2095
03:36:15,010 --> 03:36:18,210
before we even start working with them has
a point system where they were discounting

2096
03:36:18,210 --> 03:36:19,660
some of the healthier items.
So, this is also highlighting the fact that

2097
03:36:19,660 --> 03:36:27,170
it's fewer points, it's more economical to
pick the brown rice relative to the white

2098
03:36:27,170 --> 03:36:37,729
rice. So, here's a picture from our online
store when you first log into the store, you

2099
03:36:37,729 --> 03:36:39,320
can see that social norms messaging at the
top. This example says "Flavorful, diced canned

2100
03:36:39,320 --> 03:36:46,410
tomatoes are very popular with Mitzvah Market
clients in July. Try some today," to try to

2101
03:36:46,410 --> 03:36:48,279
really highlight and normalize what most people
are doing in terms of their fruit vegetable

2102
03:36:48,279 --> 03:36:52,550
purchases. And then there's a little description
for people for how they're going to interpret

2103
03:36:52,550 --> 03:36:56,640
those different traffic light symbols that
we're using, right. So, you can see that green,

2104
03:36:56,640 --> 03:37:01,989
yellow, red there. You'll note that we have
the whole store done in both Russian and English

2105
03:37:01,989 --> 03:37:08,239
because there's a very large Russian immigrant
population that goes to this pantry.

2106
03:37:08,239 --> 03:37:12,100
Then you can see on this next screen, here's
an example of what it looks like as people

2107
03:37:12,100 --> 03:37:16,330
are shopping. You can see the picture of one
of the aisles. This is a green aisle with

2108
03:37:16,330 --> 03:37:20,770
the brown rice labeled green at the top.
And then towards the bottom, you get items

2109
03:37:20,770 --> 03:37:24,279
like white rice, which are labeled yellow
towards the bottom and working in conjunction

2110
03:37:24,279 --> 03:37:30,630
with the traffic lights. And I just want to
highlight the SWAP system, Supporting Wellness

2111
03:37:30,630 --> 03:37:35,710
at Pantries, developed by Katie Martin and
others, which is again, using this kind of

2112
03:37:35,710 --> 03:37:39,470
traffic light technique, but what's really
great about it is it's actually a way that

2113
03:37:39,470 --> 03:37:44,979
food pantries can get a sense of the stock
that they're, what they're offering in their

2114
03:37:44,979 --> 03:37:51,410
pantries and how they're stocking their pantries.
So, regular green traffic lights are used

2115
03:37:51,410 --> 03:38:00,040
as they're ordering food from food banks
to help make sure they have a good balance

2116
03:38:00,040 --> 03:38:05,850
of different options for clients. So I want
to end just by sharing a little bit about

2117
03:38:05,850 --> 03:38:07,160
the research design we're using for our NIH R01 funded by NCI.

2118
03:38:07,160 --> 03:38:11,120
We're recruiting 500 participants at the pantry
who are able to use the touchscreen ordering

2119
03:38:11,120 --> 03:38:16,900
system, which is the majority of people that
go there. And we're examining how our intervention

2120
03:38:16,900 --> 03:38:21,780
in the suite of behavioral economic tools
is going to influence what people select using

2121
03:38:21,780 --> 03:38:26,220
that data that comes from the online ordering
system. We're looking at how it impacts their

2122
03:38:26,220 --> 03:38:30,630
fruit and vegetable intake, both by self-report
and something called the Veggie meter, which

2123
03:38:30,630 --> 03:38:35,699
is able to detect levels of beta carotene,
for example, in the skin. And we're measuring

2124
03:38:35,699 --> 03:38:40,220
weight and blood pressure. Our initial study
design was to collect this information at

2125
03:38:40,220 --> 03:38:44,060
baseline three months after the intervention
and then 12 months after the intervention

2126
03:38:44,060 --> 03:38:50,380
had started. Now we like everyone who've been
really impacted by COVID-19, and so in the

2127
03:38:50,380 --> 03:38:54,620
new version of the study, we're only going
to be able to look at three month outcomes.

2128
03:38:54,620 --> 03:38:58,479
But I think we're still really excited to
see how this impacts shorter term behavior

2129
03:38:58,479 --> 03:39:04,359
and health. And the hope of all this work
is that we can inform low cost food pantry

2130
03:39:04,359 --> 03:39:06,649
interventions.
But I also think there are some interesting

2131
03:39:06,649 --> 03:39:10,010
applications for online ordering platforms
in general, which are becoming more and more

2132
03:39:10,010 --> 03:39:17,990
popular, particularly in light of the COVID-19
pandemic. So I want to thank you all very

2133
03:39:17,990 --> 03:39:23,600
much for your time. Please come visit us online
at the Peach Lab and my contact information

2134
03:39:23,600 --> 03:39:37,370
is on the screen. And thanks for the opportunity
to be able to share this work at this exciting

2135
03:39:37,370 --> 03:39:38,820
workshop.

2136
03:39:38,820 --> 03:39:50,570
DR. MARY STORY: So, welcome back. Thank
you both Tamara and Joel for such engaging

2137
03:39:50,570 --> 03:40:01,040
presentations. And we have several questions.
And Joel was having some computer connection

2138
03:40:01,040 --> 03:40:04,760
problems, so Tamara you get to answer all the
questions.

2139
03:40:04,760 --> 03:40:08,290
DR. TAMARA DUBOWITZ: Great I'll answer all the questions.

2140
03:40:08,290 --> 03:40:16,450
DR. MARY STORY: So Joel will come back on, but
in the meantime, we have several questions

2141
03:40:16,450 --> 03:40:26,210
and we have about 10 minutes left for questions.
And the first question for you Tamara was:

2142
03:40:26,210 --> 03:40:33,150
can you summarize your understanding from
your work and others of the differences in

2143
03:40:33,150 --> 03:40:42,500
effects of opening new supermarkets and food
deserts on different subgroups? You mentioned

2144
03:40:42,500 --> 03:40:46,840
SNAP participants, but also socioeconomic
groups.

2145
03:40:46,840 --> 03:40:57,370
DR. TAMARA DUBOWITZ: Got it, so that's a pretty large question
in terms of just summarizing the effect of

2146
03:40:57,370 --> 03:41:02,520
new supermarkets on all different subgroups.
I think I could probably speak best to most

2147
03:41:02,520 --> 03:41:11,020
of the research that's been done has looked
at opening supermarkets in food deserts that

2148
03:41:11,020 --> 03:41:19,569
are normally or by definition, lower income
neighborhoods with lower income residents.

2149
03:41:19,569 --> 03:41:29,600
So it in terms of subgroups, we can look at
different racial/ethnic groups. I don't think

2150
03:41:29,600 --> 03:41:36,090
that very much work has looked at all of the
different subgroups and how this might impact

2151
03:41:36,090 --> 03:41:45,130
groups differentially. Our own work has been
able to look at, in particular at SNAP participants.

2152
03:41:45,130 --> 03:41:53,670
Because in the residents we've been following,
and I'll say that the way that we were, recruited

2153
03:41:53,670 --> 03:42:01,030
our cohort was done by recruiting a geographically
random sample from our intervention neighborhood

2154
03:42:01,030 --> 03:42:07,290
and from a controlled neighborhood. Both of
these neighborhoods are low income and predominantly

2155
03:42:07,290 --> 03:42:18,510
Black. So, our study has really focused on
Black neighborhoods and the subgroup that

2156
03:42:18,510 --> 03:42:26,930
we're most familiar with are SNAP participants.
It does seem that snap participants in particular

2157
03:42:26,930 --> 03:42:36,160
did see more positive results than in our
intervention neighborhoods than snap participants

2158
03:42:36,160 --> 03:42:42,229
in our comparison neighborhood. So that's
probably the subgroup that I can speak to

2159
03:42:42,229 --> 03:42:47,790
most easily. I'll stop there so we can answer
some other questions.

2160
03:42:47,790 --> 03:42:54,620
DR. MARY STORY: OK, so on your last slide, Tamara,
you asked, is research too focused on diet

2161
03:42:54,620 --> 03:43:02,210
as an outcome. What, could you expand on that,
like, what other outcomes do you think should

2162
03:43:02,210 --> 03:43:03,210
be answered?

2163
03:43:03,210 --> 03:43:11,960
DR. TAMARA DUBOWITZ: Sure, and you know, I kind of wish that
Joel was here also, but Mary, I think that

2164
03:43:11,960 --> 03:43:19,910
you're so well equipped to talk to this as
well. So even though I think that one of the

2165
03:43:19,910 --> 03:43:28,210
huge questions here is how do we create shifts
in diet? And I think that diet has been one

2166
03:43:28,210 --> 03:43:35,569
of the main outcomes that my own research
has focused on. One of the things that I think

2167
03:43:35,569 --> 03:43:42,479
we've learned is that diet is part of a much
more complicated picture, and I think most

2168
03:43:42,479 --> 03:43:50,970
people probably know this of many other (a),
health outcomes. But (b), when we're talking

2169
03:43:50,970 --> 03:43:59,080
about things like supermarkets, that is one
really important neighborhood change in a

2170
03:43:59,080 --> 03:44:04,500
food, it's really changing that quote unquote,
"food environment" is really important in

2171
03:44:04,500 --> 03:44:09,729
these neighborhoods. But concurrently, there
are other developments that are happening.

2172
03:44:09,729 --> 03:44:17,400
Housing improvement, transportation improvements,
green space improvements, all kinds of other

2173
03:44:17,400 --> 03:44:24,500
sort of aesthetic things that might be happening
that yes, they might be impacting diet, but

2174
03:44:24,500 --> 03:44:31,500
they might be impacting other psychological
distress, perceived stress, hope, satisfaction

2175
03:44:31,500 --> 03:44:39,790
with one's neighborhood as a place to live.
And I think it's very difficult, but also

2176
03:44:39,790 --> 03:44:44,920
really important to try and isolate these
different pathways and to try and understand

2177
03:44:44,920 --> 03:44:54,760
how and that they all actually do relate to
one another. So when I, although a lot of

2178
03:44:54,760 --> 03:45:03,000
our research has focused on diet as a main
outcome, it's not, it's become much less obvious

2179
03:45:03,000 --> 03:45:12,720
to us that diet is the most important outcome
that we should be looking at when all of these

2180
03:45:12,720 --> 03:45:17,130
other health outcomes are related to diet
and all of these other neighborhood changes

2181
03:45:17,130 --> 03:45:20,340
are related to the food environment.

2182
03:45:20,340 --> 03:45:27,000
DR. MARY STORY: Yeah, I was wondering, you know,
Healthy Food Finance Initiative is wonderful,

2183
03:45:27,000 --> 03:45:33,939
and it's brought so many grocery stores into
communities that really need them. But a supermarket

2184
03:45:33,939 --> 03:45:38,840
and grocery store, they have well, they have
healthy food, they also bring in, they also

2185
03:45:38,840 --> 03:45:45,850
have a lot of unhealthy foods as well. And
so, I know the Healthy Food Financing Initiative

2186
03:45:45,850 --> 03:45:53,180
doesn't put any parameters on, you know, they
just want to bring the grocery store or supermarket

2187
03:45:53,180 --> 03:46:01,760
in. They're not required to have pricing strategies
or other in-store strategies. So I was wondering

2188
03:46:01,760 --> 03:46:08,939
like, you didn't find any significant differences
in fruits and vegetables or whole grains.

2189
03:46:08,939 --> 03:46:16,970
And I'm wondering if you think like there's
other, the factors that you mentioned, I think

2190
03:46:16,970 --> 03:46:27,010
are really important. But what about requiring
or having more pricing strategies, which we

2191
03:46:27,010 --> 03:46:33,229
know I think would really work if you could
incentivize and have healthier or having incentivizing

2192
03:46:33,229 --> 03:46:40,199
the healthy foods or else the promotional
practices and marketing practices and then

2193
03:46:40,199 --> 03:46:48,159
even looking at like what Joel was talking
about with the distribution issues that changing

2194
03:46:48,159 --> 03:46:54,820
the trade promotion practices that are
used by manufacturers and distributors.

2195
03:46:54,820 --> 03:47:02,529
If those other factors, interventions within
the grocery store, along with bringing in

2196
03:47:02,529 --> 03:47:10,550
a small food store or a supermarket would
really be kind of help shift them the diet

2197
03:47:10,550 --> 03:47:11,550
quality.

2198
03:47:11,550 --> 03:47:14,520
DR. TAMARA DUBOWITZ: Yeah. Welcome back Joel.

2199
03:47:14,520 --> 03:47:17,140
DR. JOEL GITTELSOHN: Thank you. I hope I'm really back.

2200
03:47:17,140 --> 03:47:22,800
DR. TAMARA DUBOWITZ: Yeah, you're back just in time to
answer all the rest of the questions.

2201
03:47:22,800 --> 03:47:24,460
DR. JOEL GITTELSOHL: OK, hit me.

2202
03:47:24,460 --> 03:47:30,010
DR. TAMARA DUBOWITZ: I'll answer what your question quickly,
Mary, before I know that there's probably

2203
03:47:30,010 --> 03:47:38,210
a lot of questions for Joel as well. But in
terms of parameters for HSFI or the Healthy

2204
03:47:38,210 --> 03:47:45,120
Food Financing Initiative and what kinds of
grocery stores end up locating in neighborhoods,

2205
03:47:45,120 --> 03:47:50,609
it's really difficult. And one of the issues
is that this is supposed to be an incentive

2206
03:47:50,609 --> 03:47:55,189
to get grocers to locate in places where typically
they don't want to locate. They don't want

2207
03:47:55,189 --> 03:48:01,020
to locate in these areas in supermarket deserts
because they don't think that they can make

2208
03:48:01,020 --> 03:48:08,090
money. So if HSFI started saying, "Well,
you know, you can only locate if you're willing

2209
03:48:08,090 --> 03:48:16,409
to only make sure that your healthy foods
are displayed in the most prominent areas

2210
03:48:16,409 --> 03:48:22,949
and you limit your junk food, which we know
you make more money on." The practicality of

2211
03:48:22,949 --> 03:48:28,960
this probably wouldn't work. But then on the
other hand, we're seeing supermarkets located

2212
03:48:28,960 --> 03:48:34,319
in food deserts that are providing just as
many unhealthy foods as they are healthy foods.

2213
03:48:34,319 --> 03:48:43,699
I'll say one thing about this, and that is,
although I think that this is a question of

2214
03:48:43,699 --> 03:48:49,819
opening supermarkets and allowing for the
access of healthy foods, it's much more than

2215
03:48:49,819 --> 03:48:53,439
that.
It's opening up a business that is creating

2216
03:48:53,439 --> 03:49:02,100
life in a community and is an economic anchor.
And I think if I were to make parameters, which

2217
03:49:02,100 --> 03:49:07,540
I don't think that they can do, I would make
parameters that said, this needs to be a grocery

2218
03:49:07,540 --> 03:49:12,989
store that works with the community and is
responsive to community needs and involves

2219
03:49:12,989 --> 03:49:20,239
the community in as many ways, shapes, and
forms that it possibly can. And I would actually

2220
03:49:20,239 --> 03:49:32,700
focus less on food and more on creating a
space that residents feel hopeful and happy

2221
03:49:32,700 --> 03:49:37,939
about. And that's a little bit radical when
we're talking about diet and dietary shifts,

2222
03:49:37,939 --> 03:49:43,340
I think. Or maybe not radical, but maybe doesn't
make total sense. But that's what I've learned

2223
03:49:43,340 --> 03:49:45,950
from the past decade of our research.

2224
03:49:45,950 --> 03:49:53,300
DR. MARY STORY: And some of the other stories
that I've heard, it's like when there's a

2225
03:49:53,300 --> 03:49:59,760
food co-op that's owned by members of that
community, it really does become a community

2226
03:49:59,760 --> 03:50:08,040
place as well as just a place they can
shop. OK, well, I'm going to jump into other

2227
03:50:08,040 --> 03:50:14,729
questions because we only have a few minutes
left. Joel, a couple of people asked that your

2228
03:50:14,729 --> 03:50:22,409
BUD app sounds absolutely amazing. Can you
please share more about how you decided to

2229
03:50:22,409 --> 03:50:29,080
use this specific app, and where can they
find information about BUDDY and the eligibility

2230
03:50:29,080 --> 03:50:30,080
requirements?

2231
03:50:30,080 --> 03:50:38,290
DR. JOEL GITTELSOHN: So the whole BUD experience and the
reason for the BUDexperience and app is that

2232
03:50:38,290 --> 03:50:43,479
what we found was that suppliers, well, it
was really easy for a bag of potato chips

2233
03:50:43,479 --> 03:50:48,360
to make it to a corner store, but it was very
difficult for an apple to make it to a corner

2234
03:50:48,360 --> 03:50:56,659
store. And this related to all sorts of incentivization
and really robust feedback loop for junk

2235
03:50:56,659 --> 03:51:04,189
food and deliveries of junk food, free delivery,
and so forth, but not anything like that for

2236
03:51:04,189 --> 03:51:09,670
healthier foods, such as fresh produce. So
we basically, we found that most of the corner

2237
03:51:09,670 --> 03:51:13,370
store owners in Baltimore at least have to
go out and go get anything healthy they want

2238
03:51:13,370 --> 03:51:17,149
to put in their stores as opposed to having
it delivered to them. And it's not worth it

2239
03:51:17,149 --> 03:51:23,270
for them. And I think this is a really unexplored
area. So the Bud app was, the idea of the

2240
03:51:23,270 --> 03:51:28,440
BUD app was developed to address this distribution
and supply problem for corner stores. It's

2241
03:51:28,440 --> 03:51:34,090
sort of a lot of the work that I've done previously
has been multilevel, multi-component, but

2242
03:51:34,090 --> 03:51:39,830
this is really addressing one particular piece
of this issue.

2243
03:51:39,830 --> 03:51:45,560
We received an R34 to develop the app, so
we're in the process of developing the app.

2244
03:51:45,560 --> 03:51:50,560
We're just about to begin piloting it in just
a few stores before we do a larger pilot in

2245
03:51:50,560 --> 03:51:59,939
about 20 stores. And yeah, I mean, I think
it's a really interesting idea. It's my first

2246
03:51:59,939 --> 03:52:05,149
time developing software. I'm not directly
writing of the software myself obviously,

2247
03:52:05,149 --> 03:52:12,090
but it's a huge challenge and involves lots
of usability testing and so forth. I'm very

2248
03:52:12,090 --> 03:52:16,210
interested in collaborating with other people
who are interested in expanding this model

2249
03:52:16,210 --> 03:52:21,250
in the future. Right now, our funding is just
to do this work in Baltimore as more of a

2250
03:52:21,250 --> 03:52:22,250
feasibility trial.

2251
03:52:22,250 --> 03:52:29,449
DR. MARY STORY: You know what, I just got a message
saying, I need to wrap this up. I'm really

2252
03:52:29,449 --> 03:52:37,250
sorry, we had several questions and I'm sorry
we couldn't get to them all. But I really

2253
03:52:37,250 --> 03:52:43,890
want to thank both of you again for these
amazing presentations, and we appreciate the

2254
03:52:43,890 --> 03:52:51,340
questions from the audience. Next, we're going
to head into our Networking Break for you

2255
03:52:51,340 --> 03:53:01,400
to engage with your fellow attendees on the
topics just discussed. You can visit the Networking

2256
03:53:01,400 --> 03:53:06,280
Lounge and join that conversation. And
at 4:55, you'll need to return to the auditorium

2257
03:53:06,280 --> 03:53:12,370
for the final presentation of the day which
highlights gaps and opportunities to address

2258
03:53:12,370 --> 03:53:18,700
the neighborhood food environment. Thank you
very much, and it was a pleasure to moderate

2259
03:53:18,700 --> 03:53:19,850
this session.

2260
03:53:19,850 --> 03:53:22,720
DR. TAMARA DUBOWITZ: Thank you so much.

2261
03:53:22,720 --> 03:53:27,900
DR. KAREN GLANZ: Welcome back, everyone. Hopefully,
you've had a chance to have some discussions

2262
03:53:27,900 --> 03:53:33,890
during the networking break, or at least you
had a chance to maybe get up off your seat

2263
03:53:33,890 --> 03:53:39,909
and stand up. We like to take activity breaks
in some meetings, but everybody's on their

2264
03:53:39,909 --> 03:53:45,710
own with this virtual workshop. I'm delighted
to introduce to you our final speaker for

2265
03:53:45,710 --> 03:53:51,530
the day, Dr. Kristen Cooksey-Stowers from the
University of Connecticut. She's going to

2266
03:53:51,530 --> 03:53:57,279
speak on gaps and opportunities to address
the neighborhood food environment, including

2267
03:53:57,279 --> 03:54:02,850
emerging work from the field and some really
interesting work that she herself has been

2268
03:54:02,850 --> 03:54:08,529
leading. So I'll turn it over to Kristen and
again encourage you to put your questions

2269
03:54:08,529 --> 03:54:09,790
in the chat box.

2270
03:54:09,790 --> 03:54:21,960
DR. KRISTEN COOKSEY: Good afternoon, everyone.
I hope you have enjoyed day two of this workshop

2271
03:54:21,960 --> 03:54:26,739
on the State of the Science on the Neighborhood Food Environment, Health Impacts, and Interventions. 

2272
03:54:26,739 --> 03:54:31,270
Thank you to NIH and the organizers
for the opportunity to join today's session.

2273
03:54:31,270 --> 03:54:35,870
Today, I will be discussing the topic of Gaps and Opportunities in the Neighborhood Food 

2274
03:54:35,870 --> 03:54:42,569
Environment: Strategies to Advance Health Equity. So, what does health equity actually

2275
03:54:42,569 --> 03:54:47,239
mean? Braveman and colleagues define it as
the principle underlying a commitment to reduce

2276
03:54:47,239 --> 03:54:53,949
and ultimately eliminate disparities in health
and its determinants, including social determinants.

2277
03:54:53,949 --> 03:54:58,620
And we have this call to action. If an effort
does not address poverty, discrimination or

2278
03:54:58,620 --> 03:55:02,770
their health-damaging consequences for groups
of people who have historically been excluded

2279
03:55:02,770 --> 03:55:08,850
or marginalized, it's probably not a health
equity effort. We also have the critical first

2280
03:55:08,850 --> 03:55:14,291
do no harm principle, which I think this quote
nicely captures. That without careful design

2281
03:55:14,291 --> 03:55:20,040
and implementation, policy systems and environmental
strategies may inadvertently widen health

2282
03:55:20,040 --> 03:55:25,220
disparities.
This idea was vitally important prior to the

2283
03:55:25,220 --> 03:55:29,880
COVID-19 and structural racism pandemics,
what experts are referring to as the twin

2284
03:55:29,880 --> 03:55:43,850
pandemics, and it is even more critical moving
forward. Here is an adaptation of the World

2285
03:55:43,850 --> 03:55:48,620
Health Organization's framework for how to
alleviate health inequities. As you see, it

2286
03:55:48,620 --> 03:55:52,630
highlights policy and governance as key structural
determinants to leverage and keep in mind.

2287
03:55:52,630 --> 03:55:59,649
It also points out how other determinants
of equity in society, like gender, racism,

2288
03:55:59,649 --> 03:56:06,240
income and even citizenship status shape intermediary
determinants of health like, but not limited

2289
03:56:06,240 --> 03:56:13,760
to the focus of today's workshop, neighborhood
food environments, and ultimately impact health

2290
03:56:13,760 --> 03:56:19,710
inequities downstream. In other words, health
and well-being is determined by social equity.

2291
03:56:19,710 --> 03:56:25,739
Notably, this is a simplified version of the
model and doesn't show all of the potential

2292
03:56:25,739 --> 03:56:30,320
interrelationships. Instead, we're focusing
on basic pathways and food availability here.

2293
03:56:30,320 --> 03:56:35,630
Reflecting on this framework, the design implementation
of equity focused neighborhood food environment

2294
03:56:35,630 --> 03:56:40,430
interventions should account for upstream
structural determinants of health and thus

2295
03:56:40,430 --> 03:56:48,250
be more effective in addressing major long
term impacts on health inequities downstream.

2296
03:56:48,250 --> 03:56:52,920
The first point that I want to leave you with
today is that intersecting social determinants

2297
03:56:52,920 --> 03:56:58,460
of health matter. There have been cumulative
impacts of racism in food and policing amid

2298
03:56:58,460 --> 03:57:08,310
COVID-19. Much of my research has focused
on how upstream factors like policy, place

2299
03:57:08,310 --> 03:57:13,740
and race shape inequitable neighborhood food
environments, particularly food swamps, where

2300
03:57:13,740 --> 03:57:17,830
unhealthy food retailers inundate healthier
alternatives. So this research prior to COVID-19,

2301
03:57:17,830 --> 03:57:23,770
I was able to engage in national discourse
around evidence based approaches to identifying

2302
03:57:23,770 --> 03:57:30,190
food areas. For people of color, including
but not limited to Black Americans, structural

2303
03:57:30,190 --> 03:57:35,050
racism across systems has posed cumulative
barriers to optimal health and well-being

2304
03:57:35,050 --> 03:57:40,430
in light of COVID-19. The same communities
are faced with too many unhealthy food retailers,

2305
03:57:40,430 --> 03:57:46,352
over policing, too few job opportunities and
medical resources to cope with the pandemic.

2306
03:57:46,352 --> 03:57:51,640
And as we know, individuals with pre-existing
conditions like diabetes and hypertension

2307
03:57:51,640 --> 03:57:56,040
are among those with heightened COVID-19 risk.
People of color were more likely to have conditions

2308
03:57:56,040 --> 03:58:00,290
going into the pandemic. We also know that
people of color are more likely to be food

2309
03:58:00,290 --> 03:58:04,560
insecure and live in inequitable neighborhood
food swamp and food desert environments prior

2310
03:58:04,560 --> 03:58:10,970
to COVID-19. Living in a food swamp area where
fast food and junk food are the most successful

2311
03:58:10,970 --> 03:58:17,370
food options is a serious structural challenge,
especially in the middle of a global pandemic.

2312
03:58:17,370 --> 03:58:21,229
One of our recent studies using a national
sample of adults in the US showed that Black

2313
03:58:21,229 --> 03:58:26,550
Americans were most likely to report that
they reside in a food swamp. Earlier in the

2314
03:58:26,550 --> 03:58:30,350
pandemic, grocery stores remain open while
smaller food retailers were closed, but we

2315
03:58:30,350 --> 03:58:37,050
know there were inequities in access to supermarket
and grocery store access prior to the pandemic.

2316
03:58:37,050 --> 03:58:42,870
Recent CDC data show that ethnic disparities
in COVID-19 hospitalizations, deaths and in

2317
03:58:42,870 --> 03:58:47,609
vaccination rates persist. Judging from pre-pandemic
data on disparities in traffic stops, arrests

2318
03:58:47,609 --> 03:58:53,040
and police use of force, we know there is
racism in the U.S. policing system.

2319
03:58:53,040 --> 03:58:58,020
The civil unrest across the world in response
to the murders of George Floyd and Breonna

2320
03:58:58,020 --> 03:59:04,180
Taylor led to devastated food stores in many
neighborhoods. Future interventions targeting

2321
03:59:04,180 --> 03:59:08,870
food environment should account for how intersecting
policies impact the lived experiences of Black

2322
03:59:08,870 --> 03:59:15,149
Americans and other marginalized groups. This
2020 article by Leon and colleagues outlines

2323
03:59:15,149 --> 03:59:19,600
how healthy food retail has changed during
the COVID-19 pandemic. The authors write, "The

2324
03:59:19,600 --> 03:59:23,990
pandemic has placed unprecedented strain on
the U.S. food system and changed the way food

2325
03:59:23,990 --> 03:59:28,630
is distributed, sold, obtained and prepared
and consumed." And though not the primary focus

2326
03:59:28,630 --> 03:59:33,160
of the publication, they also comment on the
relevance of structural racism to neighborhood

2327
03:59:33,160 --> 03:59:38,640
food environment research. They write, "Many
communities have also been impacted by uprisings

2328
03:59:38,640 --> 03:59:43,620
against police brutality and structural racism
that may have damaged, disrupted or destroyed

2329
03:59:43,620 --> 03:59:48,030
food retail outlets and other infrastructure,
creating even more food access issues."

2330
03:59:48,030 --> 03:59:55,020
For the next part of my presentation, I will
go through the following domains to identify

2331
03:59:55,020 --> 04:00:00,040
some research gaps and opportunities for neighborhood
food environment research to advance health

2332
04:00:00,040 --> 04:00:07,370
equity amid the twin pandemics, including
new grocery stores, existing food retailers,

2333
04:00:07,370 --> 04:00:11,950
online grocery shopping, food banks and food
pantries, and community kitchens. Research

2334
04:00:11,950 --> 04:00:18,359
on the grocery stores can generally be characterized
by majority pre/post quasi experiments, limited

2335
04:00:18,359 --> 04:00:23,410
randomized controlled trials, limited use
of community based and qualitative research

2336
04:00:23,410 --> 04:00:28,620
methods. Still, grocery stores have been linked
with improvements in perceived food access,

2337
04:00:28,620 --> 04:00:35,370
neighborhood satisfaction, psychological health
outcomes and food security among SNAP participants.

2338
04:00:35,370 --> 04:00:41,529
Studies show no impact on overall diet. And
studies show mixed findings regarding fruit

2339
04:00:41,529 --> 04:00:48,580
and vegetable consumption and in BMI. To advance
health equity, there are opportunities for

2340
04:00:48,580 --> 04:00:56,060
this area of research to involve more rural
settings, qualitative research methods, community

2341
04:00:56,060 --> 04:01:01,670
based research methods, integration of health
care services and multilevel interventions

2342
04:01:01,670 --> 04:01:07,021
integrating within- store components.
Regarding scope, future work should consider

2343
04:01:07,021 --> 04:01:13,740
providing more information about food insecurity,
unhealthy food and beverage intake, social

2344
04:01:13,740 --> 04:01:19,670
support for healthy eating, community support
and social cohesion, employment outcomes and

2345
04:01:19,670 --> 04:01:25,490
the concurrence of COVID related closures
among small food stores and restaurants nearby.

2346
04:01:25,490 --> 04:01:30,840
And last but not least, future work should
further explore the fear of displacement and

2347
04:01:30,840 --> 04:01:35,470
gentrification among residents. For the last
five years, I've worked very closely with

2348
04:01:35,470 --> 04:01:40,280
the Invest Health Hartford team and local residents groups
to pursue a new grocery store in the North

2349
04:01:40,280 --> 04:01:44,890
Hartford Promise Zone and gentrification.
And displacement associated with a new grocery

2350
04:01:44,890 --> 04:01:52,449
store is a real concern for the community.
Next, I'll talk about research on modifying

2351
04:01:52,449 --> 04:01:57,580
environmental changes within existing food
retail. I'll focus particularly on corner

2352
04:01:57,580 --> 04:02:03,290
stores and grocery stores. In terms of methods,
this research area reflects a mix of exploratory

2353
04:02:03,290 --> 04:02:07,909
studies, store audits, natural experiments
and a strong use of community based research

2354
04:02:07,909 --> 04:02:11,630
methods, mixed methods and customer intercept
surveys.

2355
04:02:11,630 --> 04:02:16,250
Generally, findings from these studies point
to place based disparities in the quality

2356
04:02:16,250 --> 04:02:21,120
of food environments within these store types
where less healthy foods and drinks and more

2357
04:02:21,120 --> 04:02:26,750
salient and prominently placed. Results show
that corner stores and small grocers are more

2358
04:02:26,750 --> 04:02:31,680
prevalent in low income communities, and that
food insecure populations shop at them more

2359
04:02:31,680 --> 04:02:39,930
frequently. In terms of intervention outcomes,
minimum stocking standards have had no impact

2360
04:02:39,930 --> 04:02:45,300
on healthy food availability or the availability,
of culturally preferred produce and whole

2361
04:02:45,300 --> 04:02:50,540
grains. Healthy corner store interventions
targeting equipment and technical assistance

2362
04:02:50,540 --> 04:02:54,330
show mixed findings regarding improvements
in healthy food and beverage availability

2363
04:02:54,330 --> 04:02:58,920
and customer food purchases and consumption.
For example, a recent publication on a four

2364
04:02:58,920 --> 04:03:04,140
year evaluation of the North Carolina Healthy
Food Small Retailer Program found that there

2365
04:03:04,140 --> 04:03:08,380
were increases in the supply of healthy food
and beverages, but no change in Veggie

2366
04:03:08,380 --> 04:03:11,850
Meter, or objectively assessed food vegetable
intake.

2367
04:03:11,850 --> 04:03:18,159
Corner store interventions targeting both food
availability and affordability more consistently

2368
04:03:18,159 --> 04:03:25,540
show increased produce purchases and sales
across healthy food categories. Another recent

2369
04:03:25,540 --> 04:03:29,880
systematic review looked at the efficiency
of interventions within grocery stores and

2370
04:03:29,880 --> 04:03:35,110
assessed 36 qualitative studies and included
30 quantitative studies in a meta analysis.

2371
04:03:35,110 --> 04:03:39,430
Their results show that interventions with
pricing combined with promotion and prompts

2372
04:03:39,430 --> 04:03:44,100
have the largest effects. Regarding sales,
results show that interventions within existing

2373
04:03:44,100 --> 04:03:48,380
grocery stores have significant effects on
the sales of healthy foods and beverages,

2374
04:03:48,380 --> 04:03:56,970
including increased sales of fruits and vegetables.
Research in this area will be strengthened

2375
04:03:56,970 --> 04:04:03,330
by larger sample sizes, more diverse settings,
subgroup analysis, standardized reporting

2376
04:04:03,330 --> 04:04:10,790
of sample sizes, objective diet and food purchase
data, seasonality adjustments. We also need

2377
04:04:10,790 --> 04:04:15,470
to learn more about linkages to home food
environments, the impacts of store closures

2378
04:04:15,470 --> 04:04:23,500
due to COVID-19 and 2020 civil unrest, and
a broader use of equity focused post-intervention

2379
04:04:23,500 --> 04:04:28,460
outcomes, including disparities in the quality
of food purchases and diet, but also issues

2380
04:04:28,460 --> 04:04:34,000
around trust, social cohesion, well-being
and psychological health outcomes.

2381
04:04:34,000 --> 04:04:39,710
Last, research on modifying food environments
within existing food stores can further advance

2382
04:04:39,710 --> 04:04:44,330
health equity by leveraging community based
research approaches to developing interventions

2383
04:04:44,330 --> 04:04:49,220
from the design stage, targeting unhealthy
foods and beverages, aiming to increase the

2384
04:04:49,220 --> 04:04:54,700
availability of culturally preferred nutrient-rich
items tailored to rural settings, and aiming

2385
04:04:54,700 --> 04:04:59,729
to overcome the many challenges reflected
in qualitative research with small food retailers.

2386
04:04:59,729 --> 04:05:06,310
And some of these new challenges may be related
to COVID-19. These interventions should be

2387
04:05:06,310 --> 04:05:11,199
inclusive, sustainable and accessible to diverse
populations in terms of cost feasibility and

2388
04:05:11,199 --> 04:05:16,720
translated in a variety of languages, and
integrate other systems and social determinants

2389
04:05:16,720 --> 04:05:22,949
of health like transportation and housing,
health care and faith-based groups. Next,

2390
04:05:22,949 --> 04:05:27,069
online grocery shopping has become more popular
and necessary since the start of the COVID-19

2391
04:05:27,069 --> 04:05:33,900
pandemic. Also in 2020, the USDA rapidly
expanded its SNAP online purchasing pilot.

2392
04:05:33,900 --> 04:05:39,090
In terms of methods, experiments, simulations,
qualitative research and mixed methods are particularly

2393
04:05:39,090 --> 04:05:45,290
common. To date, studies point to increased
accessibility when multiple forms of payment

2394
04:05:45,290 --> 04:05:50,029
are accepted. And recent research on points
of decision prompts and pre-refilled grocery

2395
04:05:50,029 --> 04:05:55,260
shopping carts find increases in healthy food
selection in terms of healthy eating Index

2396
04:05:55,260 --> 04:06:01,149
2015 scores, high fiber items, fruits and
vegetables, and items with lower sodium, fat

2397
04:06:01,149 --> 04:06:06,371
and cholesterol. Recent research in this area
also suggests there is lower access to online

2398
04:06:06,371 --> 04:06:11,790
grocery shopping, particularly with curbside
pickup for low income, WIC and SNAP participants,

2399
04:06:11,790 --> 04:06:17,630
and overall, lower access to nutrition information
across populations. And compared to in-store

2400
04:06:17,630 --> 04:06:22,580
shopping, results show lower impulse spending
on candy, frozen desserts and grain based

2401
04:06:22,580 --> 04:06:28,590
desserts. A 2019 study manipulating product
positioning and food swaps showed a reduction

2402
04:06:28,590 --> 04:06:34,290
in the saturated fat content of selected food
items. And some qualitative research, recent

2403
04:06:34,290 --> 04:06:39,330
research found that individuals with overweight
or obesity particularly prefer online grocery

2404
04:06:39,330 --> 04:06:44,149
shopping options with nutrition rating information,
interactive healthy eating aisles and healthy

2405
04:06:44,149 --> 04:06:49,050
shopping preference settings.
Opportunities for online grocery shopping

2406
04:06:49,050 --> 04:06:54,649
research to advance health equity may include
learning more about the effects of COVID-19

2407
04:06:54,649 --> 04:07:00,490
across diverse settings, relative spending
on unhealthy versus healthy options, disparities

2408
04:07:00,490 --> 04:07:06,180
in affordability and pricing, disparities
in access by geography among non-English speaking

2409
04:07:06,180 --> 04:07:13,300
participants, low income, program use across
sociodemographic groups and delving more into

2410
04:07:13,300 --> 04:07:19,470
who is using delivery versus pickup options,
including curbside options. Barriers to use,

2411
04:07:19,470 --> 04:07:24,779
including perceived theft and crime, food
quality, availability of cultural food, comfortability

2412
04:07:24,779 --> 04:07:32,960
with technology, and also to examine linkages
to job creation. There are also opportunities

2413
04:07:32,960 --> 04:07:38,159
for more formative research for targeted interventions
with diverse samples like racial ethnic minorities

2414
04:07:38,159 --> 04:07:43,390
and people with lower educational attainment.
Some targeted formative work has already been

2415
04:07:43,390 --> 04:07:48,659
done with SNAP and WIC participants, mothers
and rural families. More work can also be

2416
04:07:48,659 --> 04:07:52,069
done exploring the efficacy of bundling online
grocery shopping with other social determinants

2417
04:07:52,069 --> 04:07:55,989
of health interventions.
And there is a need for standardized approaches

2418
04:07:55,989 --> 04:08:02,850
to online grocery shopping experiments. As
an example of thinking across social determinants,

2419
04:08:02,850 --> 04:08:08,359
this study points to online grocery shopping
delivery services as a viable option for addressing

2420
04:08:08,359 --> 04:08:15,409
food disparities in areas with limited transportation.
And Rummo and colleagues published this

2421
04:08:15,409 --> 04:08:20,040
standardized guide to developing online grocery
stores for testing nutrition related policies

2422
04:08:20,040 --> 04:08:25,529
and interventions. The authors point out that
standardized approaches will support reproducibility

2423
04:08:25,529 --> 04:08:31,080
and meta analysis of intervention effects.
We conducted a systematic review involving

2424
04:08:31,080 --> 04:08:35,739
community kitchens and found that they are
understudied, particularly in the US. Although

2425
04:08:35,739 --> 04:08:41,649
studies that have been conducted on this area,
there are diverse settings with diverse populations.

2426
04:08:41,649 --> 04:08:46,189
For example, the Healthy Kitchen Healthy Children
program hires women to deliver subsidized

2427
04:08:46,189 --> 04:08:51,479
healthy snacks to schoolchildren in Palestinian
refugee camps in Lebanon. The majority of

2428
04:08:51,479 --> 04:08:56,880
studies, though, utilize qualitative methods
for formative research or process evaluations.

2429
04:08:56,880 --> 04:09:01,270
The qualitative research in this area points
to strong program reach for populations with

2430
04:09:01,270 --> 04:09:08,130
high food insecurity and high health inequities.
The results show that people participate in

2431
04:09:08,130 --> 04:09:12,450
community kitchen programs for a variety of
reasons, including overall positive perceptions,

2432
04:09:12,450 --> 04:09:19,069
increased dignity, social cohesion, social
skills, community participation and cooking

2433
04:09:19,069 --> 04:09:27,250
skills. Research on factors impacting sustainability
increased financial and entrepreneurial skills.

2434
04:09:27,250 --> 04:09:32,800
As supported by about county health rankings,
community kitchens have great potential for

2435
04:09:32,800 --> 04:09:38,430
advancing health equity by promoting food
security, nutrition education and as a community

2436
04:09:38,430 --> 04:09:43,399
asset more broadly. Future work on community
kitchens should move beyond process, evaluations,

2437
04:09:43,399 --> 04:09:48,810
to explore impact on health inequities,
expand formative research to design interventions,

2438
04:09:48,810 --> 04:09:54,250
and have greater geographic representation,
including but not limited to the United States.

2439
04:09:54,250 --> 04:09:59,290
Regarding scope, research on community kitchens
can explore whether they have had any physical

2440
04:09:59,290 --> 04:10:05,080
changes in access due to COVID-19 and prioritize
equity oriented outcomes related to employment,

2441
04:10:05,080 --> 04:10:08,660
financial and entrepreneurial skills, mental
health and social cohesion.

2442
04:10:08,660 --> 04:10:14,210
The last domain I'll discuss today is
food pantries. You've heard a lot about some

2443
04:10:14,210 --> 04:10:17,760
innovative research in this area already and
you'll hear more tomorrow, so I'll just briefly

2444
04:10:17,760 --> 04:10:23,720
highlight a few additional equity considerations.
We published this study on structural characteristics

2445
04:10:23,720 --> 04:10:30,369
of the food banking system and obesity inequities
among food insecure pantry clients. Our results

2446
04:10:30,369 --> 04:10:35,409
from key stakeholder interviews highlight
opportunities to better leverage federal policies

2447
04:10:35,409 --> 04:10:40,580
like TEFAP or The Emergency Food Assistance
Program to address disparities in access to

2448
04:10:40,580 --> 04:10:45,370
quality information, to dismantle structural
racism, stereotypes and other biases and mistrust in communities

2449
04:10:45,370 --> 04:10:52,350
of color, to enhance food pantries as inclusive
safe spaces for immigrants and other historically

2450
04:10:52,350 --> 04:10:59,750
marginalized groups to equitably access food.
Future work in this area should focus on understudied

2451
04:10:59,750 --> 04:11:04,870
health disparities in the food pantry context
including heart disease, mental health and

2452
04:11:04,870 --> 04:11:11,190
wellbeing. Understudied populations and food
pantry research include ethnic minorities,

2453
04:11:11,190 --> 04:11:16,949
immigrant populations, rural communities and
LGBTQ youth, for example.

2454
04:11:16,949 --> 04:11:22,020
We need to learn more about the effects of
COVID-19 on food distribution models in the

2455
04:11:22,020 --> 04:11:26,580
food banking system. For example, pre-pandemic,
there was a shift to more dignified client

2456
04:11:26,580 --> 04:11:31,260
choice food pantry models, but the urgent
nature of the pandemic caused pantries to

2457
04:11:31,260 --> 04:11:36,150
resort back to pre-packed boxes. It'll be
helpful to explore whether these changes will

2458
04:11:36,150 --> 04:11:41,739
have long term impact on the system. We also
need to learn more about the availability

2459
04:11:41,739 --> 04:11:46,060
of culturally preferred produce, grains, etc. So, learn more about social connectedness

2460
04:11:46,060 --> 04:11:51,369
in pantries, disparities in access to health
promoting resources, structural racism and

2461
04:11:51,369 --> 04:11:56,380
any linkages between discrimination, stress
and food choices and consumption in pantry

2462
04:11:56,380 --> 04:12:01,880
setting. Regarding interventions targeting
the food banking system, future work should

2463
04:12:01,880 --> 04:12:07,860
integrate other systems linked to social determinants
of health like housing and education, expand

2464
04:12:07,860 --> 04:12:14,200
the use of community based research approaches
to developing and evaluating PSE strategies

2465
04:12:14,200 --> 04:12:18,159
targeting inequities.
More formative research to enhance the cultural

2466
04:12:18,159 --> 04:12:23,790
relevance of nutrition related interventions.
The expanded use of technology and digital

2467
04:12:23,790 --> 04:12:29,551
interventions to disseminate information across
the food banking system more equitably. Leveraging

2468
04:12:29,551 --> 04:12:35,340
natural experiments linked to COVID related
changes in the TEFAP program, including community

2469
04:12:35,340 --> 04:12:40,080
eligibility changes and national experiments
linked to changes in written nutrition policies.

2470
04:12:40,080 --> 04:12:45,369
The twin pandemics have affected many aspects
of the neighborhood food environment, people's

2471
04:12:45,369 --> 04:12:50,790
lives including pre-existing social inequities,
and people's food behaviors, and their connection

2472
04:12:50,790 --> 04:12:55,370
to the neighborhood food environment. Social
context matters for advancing health equity

2473
04:12:55,370 --> 04:12:59,990
via neighborhood food environment research.
The twin pandemics will likely impact the

2474
04:12:59,990 --> 04:13:06,770
scope of and approach to this research for
the foreseeable future. Strategies for advancing

2475
04:13:06,770 --> 04:13:12,760
equity oriented neighborhood food environment
research include understanding recent environmental

2476
04:13:12,760 --> 04:13:19,619
changes and potential disparities in access;
community based and citizen science research

2477
04:13:19,619 --> 04:13:26,409
approaches to intervention, design, implementation,
and evaluation; pursuing interventions that

2478
04:13:26,409 --> 04:13:31,720
work across systems and integrate other social
determinants of health like health care, education,

2479
04:13:31,720 --> 04:13:37,409
civil rights, and policing; and expanding our
use of equity oriented measures and outcomes

2480
04:13:37,409 --> 04:13:43,720
like dietary disparities, but also mental
health outcomes, well-being, employment, perceptions,

2481
04:13:43,720 --> 04:13:50,680
attitudes as well as experiences. We should
apply theoretical approaches that consider

2482
04:13:50,680 --> 04:13:56,920
community assets, strengths and resilience,
as well as consider the life course and intergenerational

2483
04:13:56,920 --> 04:14:06,369
perspectives. Thank you. Any questions?

2484
04:14:06,369 --> 04:14:19,650
DR. KAREN GLANZ: Thank you Kristen for that fantastic
talk and lots of food for thought. And that

2485
04:14:19,650 --> 04:14:24,710
fits into our issues of food environment,
health disparities and so on. We have a couple

2486
04:14:24,710 --> 04:14:34,119
of questions from people in the audience for
you. So, one of the first questions is simple

2487
04:14:34,119 --> 04:14:38,850
relating to some things you're talking about
near the end of your talk. Can you recommend

2488
04:14:38,850 --> 04:14:44,180
any measures to assess citizen science (INAUDIBLE)?

2489
04:14:44,180 --> 04:14:52,649
DR. KRISTEN COOKSEY-STOWERS: Yes. Yes. So, one of the
things that we've really been focusing on

2490
04:14:52,649 --> 04:15:00,159
in Hartford in terms of community based research
approaches are the perceptions. Perceived

2491
04:15:00,159 --> 04:15:07,439
food access, particularly perceived food swamp
exposure, which is an area that I'm most equipped

2492
04:15:07,439 --> 04:15:13,590
to speak with in terms of equity in the 
food environment, because it turns out that

2493
04:15:13,590 --> 04:15:17,590
food swamp measures are strong predictor of
diet related health inequities. And there's

2494
04:15:17,590 --> 04:15:23,739
a lot more work to be done in this area. But
so far we see correlations between obesity

2495
04:15:23,739 --> 04:15:29,239
disparities at the national level, we see
correlations with food swamp measures in terms

2496
04:15:29,239 --> 04:15:35,830
of what their reporting and dietary habits.
Other work by Colin Rommel and others shows

2497
04:15:35,830 --> 04:15:43,370
relationships between adolescent health disparities
and Joel Gittelsohn's team has found some

2498
04:15:43,370 --> 04:15:49,189
work looking at food swamps and housing vacancy.
So it's a strong predictor in terms of some

2499
04:15:49,189 --> 04:15:55,279
of the many social inequities that we've heard
about today, and we're doing some citizen

2500
04:15:55,279 --> 04:15:57,189
science work in Hartford and (INAUDIBLE) in
particular.

2501
04:15:57,189 --> 04:16:03,359
And it's really turning out to be a strong
framework to use.

2502
04:16:03,359 --> 04:16:12,600
DR. KAREN GLANZ: Right. OK. Another question we
have. Given the changing food prices and increasing

2503
04:16:12,600 --> 04:16:19,490
promotions seem to be among the promising
strategies to increase healthy purchases in

2504
04:16:19,490 --> 04:16:24,939
food retail settings, what would you recommend
as the top priorities for researchers and

2505
04:16:24,939 --> 04:16:27,520
funders to inform changes in those areas
of price and increased promotions?

2506
04:16:27,520 --> 04:16:36,220
DR. KRISTEN COOKSEY-STOWERS: Thank you for that question,
it's a really good question. I mean, I think

2507
04:16:36,220 --> 04:16:40,939
the first thing that's really important to
keep in mind is that we should be letting

2508
04:16:40,939 --> 04:16:48,080
the community answer that for us before we
start and to not go in with a preconceived

2509
04:16:48,080 --> 04:16:56,949
notion on which particular items we should
be promoting, both in terms of price and promotion.

2510
04:16:56,949 --> 04:17:02,470
We know that those strategies work overall,
but we still need to learn a lot more about,

2511
04:17:02,470 --> 04:17:09,062
you know, for whom and which items in particular.
And so I'm a strong proponent of community

2512
04:17:09,062 --> 04:17:16,408
based work because the answer to a lot of
our questions is that it depends. So, Dr. Schwartz

2513
04:17:16,408 --> 04:17:21,819
and I have done some work in food pantries
in Hartford, and we started out finding out

2514
04:17:21,819 --> 04:17:28,380
from the pantry directors what are the items
that they're looking to '"move", right? Because

2515
04:17:28,380 --> 04:17:36,080
there's a conception and concern that if they
invest in and work toward stocking healthier

2516
04:17:36,080 --> 04:17:41,100
produce options is that it wouldn't move,
right? So, we started out with asking what

2517
04:17:41,100 --> 04:17:47,409
are the items that you're concerned about
moving?

2518
04:17:47,409 --> 04:17:53,000
And marrying that with some client level data
in terms of what items they would prefer to

2519
04:17:53,000 --> 04:17:58,909
see their pantry stock more. And let some
of that formative of research guide the intervention

2520
04:17:58,909 --> 04:18:07,340
and which items we chose to promote in a ingredient
bundle, which did sort of, as we know, as

2521
04:18:07,340 --> 04:18:12,471
other speakers spoke about today, convenience.
And the bundling approach is a strategy that

2522
04:18:12,471 --> 04:18:18,149
helps us, but rather than go in and decide what
items we wanted to manipulate and promote,

2523
04:18:18,149 --> 04:18:23,189
we let the clients as well as the pantry directors
inform that discussion.

2524
04:18:23,189 --> 04:18:29,670
DR. KAREN GLANZ: So I'm going to throw you a question
of my own in relation to that.

2525
04:18:29,670 --> 04:18:30,670
DR. KRISTEN COOKSEY-STOWERS: Go ahead.

2526
04:18:30,670 --> 04:18:37,319
DR. KAREN GLANZ: Just to get your response. So
let's suppose you asked the food pantry customers

2527
04:18:37,319 --> 04:18:46,090
users what it is they would like to see more
of. And they say, we want more Ho Hos

2528
04:18:46,090 --> 04:18:50,220
and we want more cheaper hotdogs. What would
you, what would you do with that information?

2529
04:18:50,220 --> 04:19:00,069
DR. KRISTEN COOKSEY: I honestly think that's
a great question, Karen. And I think I wear

2530
04:19:00,069 --> 04:19:05,290
sort of two hats here in Hartford. I'm an
academic, but I'm also a board member of the

2531
04:19:05,290 --> 04:19:10,529
Connecticut Food Bank in our Food Share. And
that is a huge concern among donors and among

2532
04:19:10,529 --> 04:19:17,930
key stakeholders that does not pan out in
the data. The clients are like, it reminds me of

2533
04:19:17,930 --> 04:19:22,109
this sort of data on what participants showing
that they're very much in the know. They're

2534
04:19:22,109 --> 04:19:28,580
very kind, or they're not trying to game the
system. Very similar sort of consistent findings

2535
04:19:28,580 --> 04:19:34,380
in our work but also on Dr Katz's work showing
that clients want to see more nutrient-dense

2536
04:19:34,380 --> 04:19:42,181
foods. They are very sophisticated in how
they're navigating their food donors and their

2537
04:19:42,181 --> 04:19:46,850
environment throughout the month. I did some
cognitive interviews preparing for pantry

2538
04:19:46,850 --> 04:19:50,939
intervention in Hartford. And you can just
see the wheels turning, you know. On Wednesday.,

2539
04:19:50,939 --> 04:19:58,310
I can go to the pantry for my protein, and
then by the end of the month I can go to SNAP

2540
04:19:58,310 --> 04:20:02,310
and stock up, right? And so that is not what
we see in the data in terms of them wanting

2541
04:20:02,310 --> 04:20:07,271
to see more unhealthy food.
That is, you know, that those items are cheaper.

2542
04:20:07,271 --> 04:20:11,720
Those items are, particularly in an equitable
food around areas like food swamps, that it's

2543
04:20:11,720 --> 04:20:18,680
more than they can, then they want to see.
So I definitely hear the concern, including

2544
04:20:18,680 --> 04:20:25,489
among donors and stakeholders in this space.
But clients are very vocal about wanting to

2545
04:20:25,489 --> 04:20:31,010
see more healthy food that they can always
afford in grocery stores.

2546
04:20:31,010 --> 04:20:38,010
DR. KAREN GLANZ: OK. But I'm being, I'm getting
a signal that we have only time for one more

2547
04:20:38,010 --> 04:20:47,380
question. So, this is a question about the
community and health care settings and kind

2548
04:20:47,380 --> 04:20:54,199
of resource management and access interaction
with the food environment. Do you have any

2549
04:20:54,199 --> 04:21:02,840
good examples of either demonstration projects
or a resource where social workers have been

2550
04:21:02,840 --> 04:21:09,410
involved in nutrition counseling and financial
management of food access and so forth?

2551
04:21:09,410 --> 04:21:16,390
DR. KRISTEN COOKSEY-STOWERS: I have to be honest, I'm
not familiar with any particular sort of large

2552
04:21:16,390 --> 04:21:21,180
scale studies in this space, but we have been
doing some work looking at how the health

2553
04:21:21,180 --> 04:21:27,770
care system can better integrate with the
food banking system. We conducted a national

2554
04:21:27,770 --> 04:21:37,260
study with CEOs of food banks. And this idea
of better inclusion of social workers indeed

2555
04:21:37,260 --> 04:21:43,630
on staff is really percolating the system
that we need to be working across...across like sort of our

2556
04:21:43,630 --> 04:21:51,029
disciplines or sectors. And so I'm not aware
of any large scale studies but social workers

2557
04:21:51,029 --> 04:21:57,119
are definitely coming up a lot in terms of
how can we work in a more intersectoral way

2558
04:21:57,119 --> 04:21:58,580
in pantry setting?

2559
04:21:58,580 --> 04:22:04,560
DR. KAREN GLANZ: Alright, Thank you. Thank you
so much for a wonderful talk and for taking

2560
04:22:04,560 --> 04:22:07,330
time to answer these questions from the audience.

2561
04:22:07,330 --> 04:22:13,260
DR. KRISTEN COOKSEY-STOWERS: Thank you all. Thank you
for the opportunity. Have a nice night.

2562
04:22:13,260 --> 04:22:19,640
DR. KAREN GLANZ: Sure. We're almost to the end
of the second day of our workshop. And I see

2563
04:22:19,640 --> 04:22:27,140
over 300 of you are still with us. So that's
terrific. We've covered a ton today, and tomorrow

2564
04:22:27,140 --> 04:22:33,170
morning we're going to do a little bit of
reflection on much of what was covered today

2565
04:22:33,170 --> 04:22:40,900
as we go on and hear even more examples of
innovative research, reviews of research that's

2566
04:22:40,900 --> 04:22:48,140
being conducted, and also about policies and
programs that federal government is undertaking

2567
04:22:48,140 --> 04:22:54,640
to address both food insecurity and the neighborhood
food environment. I want to pick on three

2568
04:22:54,640 --> 04:23:00,779
things to just end the day that I think have
been somewhat crosscutting, but that particularly

2569
04:23:00,779 --> 04:23:08,620
stuck out to me in a super full day. The one
has to do with social determinants, and Kristen

2570
04:23:08,620 --> 04:23:16,050
has highlighted this quite a bit in her talk
as well. The idea that poor conditions coexist

2571
04:23:16,050 --> 04:23:24,210
with poor neighborhood food environments,
they often coexist with minority racial makeup

2572
04:23:24,210 --> 04:23:34,050
in neighborhoods, poverty, discrimination,
poor housing. But they also, as I showed in

2573
04:23:34,050 --> 04:23:40,830
some map, they also coexist with higher rates
of smoking, low access to health care and

2574
04:23:40,830 --> 04:23:46,689
a number of other things.
So, the issues are far beyond food. And they're

2575
04:23:46,689 --> 04:23:51,920
not just in terms of food and healthy eating
and determinants of healthy eating that we

2576
04:23:51,920 --> 04:23:57,210
need to consider in the big picture if we
really want to improve health and quality

2577
04:23:57,210 --> 04:24:06,909
of life overall. And another point along the
same lines is that we need to both look for

2578
04:24:06,909 --> 04:24:13,040
and encourage programs that cut across different
aspects of social determinants of health,

2579
04:24:13,040 --> 04:24:19,300
of quality, of life, of physical health risk
factors and diet. Again, Kristen mentioned

2580
04:24:19,300 --> 04:24:24,470
the Promise Zone model as one of the areas
that she's working in. I've also worked in

2581
04:24:24,470 --> 04:24:32,140
the Promise Zone. This was a model that President
Obama started to actually get large departments

2582
04:24:32,140 --> 04:24:38,270
in the federal government working together
that had not. Housing and urban development

2583
04:24:38,270 --> 04:24:46,760
working with Health and Human Services, working
with the U.S. Department of Agriculture. It

2584
04:24:46,760 --> 04:24:51,630
really is a tremendous idea that probably
needs to be revived and expanded, not just

2585
04:24:51,630 --> 04:24:58,310
on the federal level, but also on local levels.
And the last point that I just want to point

2586
04:24:58,310 --> 04:25:07,340
to as food for thought is that we're only
limited by our creativity, our innovation

2587
04:25:07,340 --> 04:25:12,510
and the novelty of ideas.
And I've heard about public private partnerships

2588
04:25:12,510 --> 04:25:19,319
for decades. And yet some of the most interesting
and promising interventions that we've heard

2589
04:25:19,319 --> 04:25:26,030
about today fall in that space. For example,
the Silver Diner restaurant intervention that

2590
04:25:26,030 --> 04:25:31,380
Chris Economos talked about where the restaurant
chain actually changed their menu and changed

2591
04:25:31,380 --> 04:25:38,960
their default working with health experts.
And the results were really impressive. Another

2592
04:25:38,960 --> 04:25:45,729
one that we've heard of late this afternoon
from Joel Gittelsohn is working with the owners

2593
04:25:45,729 --> 04:25:55,770
and managers of small stores using the BUD
app to help to try to correct errors and gaps

2594
04:25:55,770 --> 04:26:02,359
in the food supply chain as a way to get the
healthier foods into these stores, something

2595
04:26:02,359 --> 04:26:09,069
that was a real obstacle for the retail industry
in that space. So I think those are just a

2596
04:26:09,069 --> 04:26:14,489
couple of examples, there are many more. But
just to plant those ideas in everyone's head.

2597
04:26:14,489 --> 04:26:23,399
Thanks to everyone for being with us for a
great day of this workshop. And we will return

2598
04:26:23,399 --> 04:26:28,758
again tomorrow at 12:30 Eastern Time for the
third day of our workshop.

2599
04:26:28,758 --> 04:26:39,462
I hope as many of you as can will join us. We have, again, a wonderful line up of speakers and some time for discussion and synthesis of what we've talked about.

2600
04:26:39,462 --> 04:26:47,066
In regards to food insecurity and neighborhood food environments in this workshop. So, thanks everyone and have a good evening.