A farmers' market at a federally qualified health center improves fruit and vegetable
intake among low-income diabetics
Darcy A. Freedmana,b,⁎, Seul Ki Choib,c, Thomas Hurleyb, Edith Anadud, James R. Hébertb,e
aUniversity of South Carolina, College of Social Work, DeSaussure Hall, Columbia, SC 29208, USA
bUniversity of South Carolina, Cancer Prevention and Control Program, 915 Greene St., Columbia, SC 29208, USA
cUniversity of South Carolina, Department of Health Promotion, Education, and Behavior, 800 Sumter St., Columbia, SC 29208, USA
dFamily Health Centers, Inc., 3310 Magnolia St., Orangeburg, SC 29115, USA
eUniversity of South Carolina, Department of Epidemiology and Biostatistics, 800 Sumter St., Columbia, SC 29208, USA
a b s t r a c ta r t i c l e i n f o
Available online 4 February 2013
Prevention & control
Community health centers
Diabetes mellitus, type 2
Objective. A 22-week federally qualified health center (FQHC)-based farmers' market (FM) and personal
financial incentive intervention designed to improve access to and consumption of fruits and vegetables
(FVs) among low-income diabetics in rural South Carolina was evaluated.
Methods. A mixed methods, one-group, repeated-measures design was used. Data were collected in 2011
before (May/June), during (August), and after (November) the intervention with 41 diabetes patients from
the FQHC. FV consumption was assessed using a validated National Cancer Institute FV screener modified
to include FV sold at the FM. Sales receipts were recorded for all FM transactions. A mixed-model, repeated
measures analysis of variance was used to assess intervention effects on FV consumption. Predictors of
changes in FV consumption were examined using logistic regression.
Results. A marginally significant (p=0.07) average increase of 1.6 servings of total FV consumption per
day occurred. The odds of achieving significant improvements in FV consumption increased for diabetics
using financial incentives for payment at the FM (OR: 38.8, 95% CI: 3.4–449.6) and for those frequenting
the FM more often (OR: 2.1, 95% CI: 1.1–4.0).
Conclusions. Results reveal a dose–response relationship between the intervention and FV improvements
and emphasize the importance of addressing economic barriers to food access.
© 2013 Elsevier Inc. All rights reserved.
Obesity rates in the United States are rising (Finkelstein et al.,
2012; Flegal et al., 2010) resulting in increases in type 2 diabetes
(Hu et al., 2001). Interventions designed to improve fruit and vegeta-
ble (FV) consumption are important strategies for preventing and
treating obesity and diabetes (Ford et al., 2012; Gillies et al., 2007;
Montonen et al., 2004; Yamaoka and Tango, 2005). Improving FV in-
take among Americans, however, has proven to be difficult (Grimm
et al., 2010).
Recently, there has been a focus on increasing individual con-
sumption of FV by improving access, availability, and affordability to
FV in communities (Grimm et al., 2010). This focus emerged because
populations disproportionately burdened by diet-related health con-
ditions (Pan et al., 2009) are less likely to have healthy food retailers
in their community (Dutko et al., 2012). Moreover, behaviorally-
based interventions have not resulted in sustained improvements in
diet (Jetter and Cassady, 2006).
Farmers' markets (FMs) are targeted approaches for improving ac-
cess to FV (Centers for Disease Control and Prevention, 2011). Prima-
ry health care settings such as federally qualified health centers
(FQHC) provide an ideal context for establishing FMs because they
are situated in underserved communities (Health Resources and
Services Administration, 2012). Moreover, locating at a FQHC makes
an explicit connection between FMs and preventive medicine.
Only a few studies reported have examined the influence of FM in-
terventions on FV consumption; most have limitations related to
study design and measurement of FV intake, are focused on the
Women's Infant and Children (WIC) program participants, and are
based in urban areas (McCormack et al., 2010). A recent review of
the nutritional implications of FMs concluded “…there is limited re-
search assessing the specific health benefits of farmers' markets”
(McCormack et al., 2010).
Four FM interventions used repeated, validated measures to ex-
amine changes in FV intake (Abusabha et al., 2011; Anderson et al.,
2001; Evans et al., 2012; Herman et al., 2008). Not one of these is
Preventive Medicine 56 (2013) 288–292
⁎ Corresponding author at: University of South Carolina, DeSaussure Hall, Columbia,
SC 29208, USA. Fax: +1 803 777 3498.
E-mail addresses: firstname.lastname@example.org (D.A. Freedman), email@example.com
(S.K. Choi), THURLEY@mailbox.sc.edu (T. Hurley), firstname.lastname@example.org (E. Anadu),
JHEBERT@mailbox.sc.edu (J.R. Hébert).
0091-7435/$ – see front matter © 2013 Elsevier Inc. All rights reserved.
Contents lists available at SciVerse ScienceDirect
journal homepage: www.elsevier.com/locate/ypmed
focused on high-risk sub-populations like people with diabetes,
conducted in a rural context, or conducted in a setting that serves med-
ically underserved populations. In this analysis we addressed this gap
by evaluating the influence of a FQHC-based FM and personal financial
incentive intervention on FV consumption among low-income, dia-
betics in a rural context using validated measures of FV intake and ob-
jective measures of FM usage.
Design, setting, and participants
We used a community-based participatory research (CBPR) approach
that involved a partnership between the University of South Carolina
and Family Health Centers, Inc., an FQHC located in a majority minority
(63% African American) rural county in South Carolina (U.S. Census Bureau,
2010). A one-group repeated-measures design was used. The study was ap-
proved by the university Institutional Review Board.
Adult participants were eligible if they were patients at the FQHC with a
diabetes diagnosis as of March 1, 2011 (N=2306 patients). Health center
staff randomly selected 345 diabetics; each received a mailing that described
the purpose of the study including information about the chance to receive
$50 in vouchers to shop at the FM. Due to HIPAA, patient names and contact
information were not revealed to the research staff; we are unable to track
mailings that were undeliverable. It was impossible to determine the total
sample that actually received an invitation to participate in the study. Inter-
ested potential participants were required to contact the research staff to ex-
press interest and determine eligibility. A total of 63 patients expressed
interest; 9 could not be reached to schedule a survey, 9 were ineligible
ple includes 45 diabetics. Informed consent was obtained from all participants;
this involvedverbalreviewofthe writtenconsentform to describe the purpose
and process of the research prior to voluntary agreement to participate.
The intervention was designed using community feedback, which is de-
tailed in a documentary about the market (Murphy and Jacobs, 2011). We
conducted a visioning exercise with 50 community members to understand
their hopes for a new FM using a modified version of a nominal group process
(Johnson et al., 2011). A 10-member Community Advisory Council was
formed to guide the development of the FM including representatives from
the FQHC, local schools/universities, agricultural extension, faith-based insti-
tutions, and community volunteers. Pre-market interest surveys were
conducted with patients at the FQHC and local farmers.
Community feedback informed the two components of the FM interven-
tion. First, onsite produce-only FMs operated at the FQHC once per week
(10 am–2 pm) for 22 weeks from June to October 2011. The market was man-
aged by a community member hired through grant funding. Supplemental Nutri-
tion Assistance Program (SNAP) vouchers were accepted at the FM through a
central point-of-purchase electronic benefit transfer system. Most of the vendors
were certified to accept Senior and WIC Farmers' Market Nutrition Program
Vouchers. Second, study participants were enrolled in a personal financial incen-
tive program that provided up to $50 in vouchers to purchase FV at the FM.
Vouchers were provided after participants completed two study-related surveys
(described below) and were paid for through grant resources. Voucher usage at
the $50 maximum due to a recording error.
We collected data with the diabetic patient sample at three time points: be-
fore the FM intervention (T1, May/June 2011), midway through the intervention
(T2, August 2011), and immediately after the intervention (T3, November 2011).
Data werecollectedvia structured surveyseitherin-personoroverthetelephone
by trained research assistants. Vouchers to shop at the FM were provided to par-
ticipants after completing surveys at T1 and T2 ($25 each time). A stipend of $20
was provided after the first and second surveys and $40 after the third. A total of
The main outcome, assessed at each time point, was FV intake measured
using a modified version of the 10-item NCI FV screener (Greene et al., 2008;
Peterson et al., 2008; Thompson et al., 2002). The modified 19-item version
included nine additional FV available at the FM (e.g., peach, apple, orange,
cantaloupe, cabbage, broccoli, squash/zucchini, sweet potato, tomato). De-
mographics, social context, and health status information were collected at
T1 using close-ended questions from the Behavioral Risk Factor Surveillance
System survey (Table 1). We developed one question to assess self-reports of
seven diet-related health conditions.
Characteristics of diabetics enrolled in the farmers' market intervention in rural South
Carolina, June–October, 2011 (N=41).
Number of people living in householda
Body mass index at baseline (T1)b
Shopping days at farmers' market
Number of sales transactions at farmers' market
Amount of money spent at farmers' market, $
Divorced or separated
Less than high school
High school graduate or GED
Some college or technical school
College graduate or more
Annual household income (last year)c
Less than $10,000
$10,000 to $19,999
$20,000 to $29,999
Household food assistance (e.g., SNAP, WIC,
and/or free or reduced price lunches)
Household financial assistance (e.g., TANF,
Medicaid, Disability, SSI)
Primary form of transportationc
Ride with friend or family
Bus or taxi
Unable to work
Employed for wages
Out of work for 1 year or more
Not employed for wages (e.g., homemaker, student)
Worried about having enough money to buy
nutritious meals in past year
Always or usually
Rarely or never
Self-reported health statusc
Self-reported disease status
High Blood Pressure
aIncluding the respondent.
bBased on self-reports of height and weight at T1.
cTotals do not add up to 100% due to missing data.
D.A. Freedman et al. / Preventive Medicine 56 (2013) 288–292
A receipt of each sales transaction (N=3747) at the FQHC-based FM was
recorded on an optically scannable form; 438 of the receipts were related to
purchases made by the study participants. Receipts were recorded by trained
research assistants and included the following information: date, participant
ID, volume of produce purchased measured in units (e.g., 1 unit=1 peach or
1 basket of okra), cost of produce purchased, and form of payment.
Descriptive statistics were computed to describe sociodemographic charac-
teristics of participants, FM use, and food items purchased. FV consumption esti-
mation followed NCI'sguidelines(2010) Participants' reports of the frequency of
intake and portion size of each food item were converted to average daily fre-
quency and MyPyramid servings. Missing values for frequency of intake were
reported portion size of other fruit for fruit items. Daily frequency and servings
were multiplied to generate average daily servings of single food items. Total
FV consumption was calculated by summing all food items. Participants without
surveys at T2 or T3 (n=1) and those reporting greater than three standard de-
viations above the mean (approximately 12.5 servings/day) at any time point
(n=3) were excluded resulting in an analytic sample of 41 participants.
A mixed-model, repeated measures analysis of variance (ANOVA) was
used to assess the effects of the FM and personal financial incentive interven-
tion on FV consumption over time. The sample was dichotomized to explore
factors associated with increases in FV consumption. “Increasers” were 56.1%
of the sample whose average FV consumption at T2 and T3 was at least 0.5
servings greater than their consumption at T1 whereas the “Non-increaser”
group constituted 43.9% of the sample that did not increase FV consumption
over time. This level of change was selected because effective behavioral
interventions typically improve diet by about 0.5 servings per day
(Ammerman et al., 2002). Students' t-tests were used to compare FV con-
sumption between the two groups at each time point. Logistic regression
was conducted to examine potential predictors of changes in FV consump-
tion. Self-reported height and weight were used to calculate body mass
index ([BMI=weight (kg)/height (m2)]; continuous), payment type (study
voucher only versus voucher+other form of payment), number of FM visits
(continuous), total amount of money spent at the FM (continuous), and receipt
of food assistance in the past year (yes/no) were included in the regression
model. The goodness of fit of the regression model was adequate (χ2=11.30,
p=0.19), as assessed by the Hosmer and Lemeshow statistic. All statistical
tests were performed using the SAS 9.2 (SAS Institute Inc., Cary, NC). The level
of significance was set at pb0.05.
The diabetic sample was majority African American, female, and
older (Table 1). Most participants were obese (BMI≥30 kg/m2)
(World Health Organization, 2000) with an average BMI of 34.9±
6.8 kg/m2calculated from self-reports of height and weight at T1.
Participants had high rates of economic and food insecurity: most
earned b$10,000 per year and 51.2% reported being at least somewhat
worried about having enough money to buy nutritious meals during
the past year. At T1, 75.6% had not shopped at a FM in the month be-
fore the FQHC-based FM opened.
All participants came to the FQHC-based FM on at least 2 dates
throughout the 22-week season (average, 4.5 days; range, 2–15). On
average, participants made 10.7 (range, 5–28) sales transactions at the
FM with an average of 2.5 (range, 1–6) transactions per day indicating
that participants frequented multiple vendors at the market during
their visit. In total, the 41 study participants made 438 sales transac-
of $53.30 (range, $29–126) throughout the market season and an aver-
age of $5.49 during each sales transaction. Participants purchased food
using multiple forms of payment often in the same sales transaction.
Most participants (70.7%) paid for purchases with the study vouchers
and at least one other form of payment (e.g., cash, SNAP) whereas
29.3% only used the study vouchers for payment.
Fruit and vegetable consumption before (T1), during (T2), and after (T3) the federally qualified health center-based farmers' market intervention in rural South Carolina, June–October,
Mean difference95% confidence interval
of the difference
Total fruit & vegetable
aTime points include T1 (May/June 2011), T2 (August 2011), and T3 (November 2011).
bFruit includes 100% juice, peach, apple, orange, cantaloupe, and other fruits.
cVegetable includes lettuce salad, cabbage, broccoli, squash or zucchini, sweet potatoes, French fries or fried potatoes, white potatoes, cooked dried beans, tomatoes, tomato
sauce, vegetable soup, and other vegetables.
Fruit and vegetable consumption among diabetics participating in a federally qualified health center-based farmers' market intervention in rural South Carolina, June–October, 2011
who did and did not increase consumption of fruits and vegetables over time.
Non-increasera(n=18)Increasera(n=23) Mean difference95% confidence interval
of the difference
Mean fruit and vegetable consumptionc
Mean fruit and vegetable consumptionc
−5.16 to −0.85
aIncreaser is defined as having average fruit and vegetable consumption at T2 and T3 that was at least 0.5 servings per day greater than consumption at T1. Non-increaser is
defined as having average fruit and vegetable consumption at T2 and T3 less than 0.5 servings per day greater than consumption at T1.
bTime points include T1 (May/June 2011), T2 (August 2011), and T3 (November 2011).
cExpressed as servings per day.
D.A. Freedman et al. / Preventive Medicine 56 (2013) 288–292
Total FV intake increased from 5.9 servings per day at T1 to 7.5 and
6.5 servings per day at T2 and T3, respectively (Table 2). The increase
of 1.6 servings per day in total FV consumption from T1 to T2 was
marginally significant (p=0.07). FV intake at T3 was less than at
T2, but was not statistically significant.
In the diabetic cohort at T1, we found that Increasers consumed
fewer servings of FV per day compared to the Non-increasers
(T1: 4.9 vs. 7.3, p=0.02), whereas at T3 Increasers consumed signif-
icantly more FV servings per day compared to the Non-increasers
(T3: 7.8 vs. 4.8, p=0.01) (Table 3).
The odds of being an Increaser in FV consumption were higher
for diabetics who only used vouchers for payment at the market
(OR: 38.8, 95% CI: 3.35–445.0) and for those who visited the FM more
often (OR: 2.07, 95% CI: 1.09–3.95) (Table 4). Increasing FV intake was
not associated with BMI at baseline, receipt of food assistance, or total
amount of money spent at the FM.
While FMs are beginning to open at a variety of health care delivery
sites (Estabrook et al., 2012; George et al., 2011), this is the first
FQHC-based FM intervention in the scientific literature. Findings high-
light the benefit of a FQHC-based FM and personal financial incentive
intervention designed to improve diet among diabetics. FV consump-
tion increased by 1.6 servings per day from baseline to the mid-point
(August) of the intervention and remained about half a serving higher
thanbaseline after themarketended (November). HigherFVconsump-
tion patterns at the end of the FM intervention compared to baseline is
noteworthy because FV consumption patterns tend to be higher in the
summer compared to fall/winter months (Locke et al., 2009; Ziegler et
al., 1987). Only a few studies have used repeated, validated measures
to examine the influence of FMs on FV consumption (McCormack et
al., 2010) finding improvements between 0.4 and 2.4 servings per day
(Abusabha et al., 2011; Evans et al., 2012; Herman et al., 2008). The
FQHC-based FM model contributed to increases in FV consumption at
levels equivalent to or better than behaviorally-based interventions
(Ammerman et al., 2002).
Results illuminate a dose–response relationship between the FM
intervention and increases in FV consumption among the diabetic
cohort. More frequent usage of the FM was associated with higher
odds of increasing FV consumption. Findings also emphasize the im-
portance of the personal financial incentive program. The relatively
small financial incentive ($50) was quite beneficial to the diabetics.
Those who only used the financial vouchers for payment at the FM
were significantly more likely to increase FV consumption compared
to those who used the voucher and at least one other form of pay-
ment. Findings suggest that the FM and personal financial incentive
intervention was particularly beneficial for those consuming the low-
est levels of FVs at baseline.
Strengths of this research include the use of a random sample, val-
idated tools for measuring FV intake, and a repeated measures design.
The context of the research is another strength; most FM research is
focused on urban settings (McCormack et al., 2010) whereas this
study occurred in a rural context. The sample is both a strength and
limitation. The cohort of diabetics represents a population disparately
affected by disease and hard-to-reach. The sample, however, may not
be representative given its small size. The study design was enhanced
by using a CBPR approach to engage community members in study
development and implementation and by using mixed methods.
Lack of a control group is a limitation. Future research is warranted
that includes a larger sample and a more robust study design. Finally,
there are limitations related to the sales transaction data collection
process. There is a chance that some sales transactions made by the
diabetic cohort were not recorded due to the busyness of the market;
thus, findings may underrepresent FM utilization and benefit.
Findings offer evidence for developing FMs at health centers as a
strategy for improving patient health. FQHC-based FMs may be in-
strumental to providing preventive healthcare services to patient
populations, particularly in contexts with high rates of diet-related
health conditions or limited access to healthy food retailers (i.e.,
food deserts) or both. FQHC-based FMs have the potential to serve
as a “farmacy” for patients to access nutrients fundamental to good
health, especially if patients shop at the market on a regular basis.
Personal financial incentives to improve economic access to FMs, sim-
ilar to co-payment programs that facilitate patient access to pharma-
ceuticals, may further enhance the benefit of a FM intervention.
community-based participatory research
federally qualified health center
fruit and vegetable
Health Insurance Portability and Accountability Act of 1996
National Cancer Institute
Conflict of interest statement
The authors declare that there are no conflicts of interest.
This publication was supported by the South Carolina Cancer Pre-
vention and Control Research Network funded under Cooperative
Agreement Number 3U48DP001936-01W1 from the Centers for
Disease Control and Prevention and the National Cancer Institute.
J.R. Hébert and T. Hurley were supported by an Established Investiga-
tor Award in Cancer Prevention and Control from the Cancer Training
Branch of the National Cancer Institute (K05 CA136975; JR Hébert,
P.I.).We are thankful for our partners at Family Health Centers, Inc.
and the South Carolina Primary Healthcare Association and for re-
search assistance from Jason Greene, Kassy Alia, Natalia Carvahlo,
Shanna Hastie, Amy Mattison Faye, and Samira Khan. We thank the
anonymous reviewers for their guidance in refining this manuscript.
Abusabha, R., Namjoshi, D., Klein, A., 2011. Increased access and affordability of pro-
duce improves perceived consumption of vegetables in low-income seniors.
J. Am. Diet. Assoc. 111, 1548–1555.
Ammerman, A.S., Lindquist, C.H., Lohr, K.N., Hersey, J., 2002. The efficacy of behavioral
interventions to modify dietary fat and fruit and vegetable intake: a review of the
evidence. Prev. Med. 35, 25–41.
Odds ratios and 95% confidence intervals of increases in fruit and vegetable consump-
tion over time among diabetics frequenting a FQHC-based farmers' market in rural
South Carolina, June–October, 2011 (N=41).
Study voucher only
Study voucher+other form of payment
Number of farmers' market visits
Total amount of money spent at the farmers' market
Receipt of food assistance in the past year
D.A. Freedman et al. / Preventive Medicine 56 (2013) 288–292
Anderson, J.V., Bybee, D.I., Brown, R.M., et al., 2001. 5 a day fruit and vegetable inter- Download full-text
vention improves consumption in a low income population. J. Am. Diet. Assoc.
Centers for Disease Control and Prevention, 2011. Strategies to Prevent Obesity and
Other Chronic Diseases: The CDC Guide to Strategies to Increase the Consumption
of Fruits and Vegetables. U.S. Department of Health and Human Services, Atlanta, GA.
Dutko, P., Ver Ploeg, M., Farrigan, T., 2012. Characteristics and influential factors of food
deserts. In: Economic Research Service (Ed.), U.S. Department of Agriculture,
Estabrook, B., Zapka, J., Lemon, S.C., 2012. Evaluating the implementation of a hospital
work-site obesity prevention intervention: applying the RE-AIM framework.
Health Promot. Pract. 13, 190–197.
Evans, A.E., Jennings, R., Smiley, A.W., et al., 2012. Introduction of farm stands in low-
income communities increases fruit and vegetable among community residents.
Health Place 18, 1137–1143.
Finkelstein, E.A., Khavjou, O.A., Thompson, H., et al., 2012. Obesity and severe obesity
forecasts through 2030. Am. J. Prev. Med. 42, 563–570.
Flegal, K.M., Carroll, M.D., Ogden, C.L., Curtin, L.R., 2010. Prevalence and trends in obe-
sity among US adults, 1999–2008. JAMA 303, 235–241.
Ford, E.S., Bergmann, M.M., Boeing, H., Li, C., Capewell, S., 2012. Healthy lifestyle behav-
iors and all-cause mortality among adults in the United States. Prev. Med. 55,
George, D.R., Kraschnewski, J.L., Rovniak, L.S., 2011. Public health potential of farmers'
markets on medical center campuses: a case study. Am. J. Public Health 101,
Gillies, C.L., Abrams, K.R., Lambert, P.C., et al., 2007. Pharmacological and lifestyle inter-
ventions to prevent or delay type 2 diabetes in people with impaired glucose tol-
erance: systematic review and meta-analysis. BMJ 334, 299–302.
Greene, G.W., Resnicow, K., Thompson, F.E., et al., 2008. Correspondence of the NCI
Fruit and Vegetable Screener to repeat 24-h recalls and serum carotenoids in be-
havioral intervention trials. J. Nutr. 138, 200S–204S.
Grimm, K.A., Blanck, H.M., Scanlon, D.S., et al., 2010. State-specific trends in fruit and
vegetable consumption among adults—United States, 2000–2009. MMWR 59,
Health Resources and Services Administration, 2012. About Health Centers. U.S.
Department of Health and Human Services, Washington, DC (Available: http://
bphc.hrsa.gov/about/index.html. Accessed December 15, 2012).
Herman, D.R., Harrison, G.G., Afifi, A.A., Jenks, E., 2008. Effect of a targeted subsidy on
intake of fruits and vegetables among low-income women in the Special Supple-
mental Nutrition Program for Women, Infants, and Children. Am. J. Public Health
Hu, F.B., Manson, J.E., Stampfer, M.J., et al., 2001. Diet, lifestyle, and the risk of type 2
diabetes mellitus in women. N. Engl. J. Med. 345, 790–797.
Jetter, K.M., Cassady, D.L., 2006. The availability and cost of healthier food alternatives.
Am. J. Prev. Med. 30, 38–44.
Johnson, C., Freedman, D.A., Joosten, Y., Duke, M., 2011. Cultivating an agenda for
change: a dynamic model for community food assessments. Arête 32, 61–89.
Locke, E., Coronado, G.D., Thompson, B., Kuniyuki, A., 2009. Seasonal variation in fruit and
McCormack, L.A., Laska, M.N., Larson, N.I., Story, M., 2010. Review of the nutritional im-
plications of farmers' markets and community gardens: a call for evaluation and
research efforts. J. Am. Diet. Assoc. 110, 399–408.
Montonen, J., Knekt, P., Jarvinen, R., Reunanen, A., 2004. Dietary antioxidant intake and
risk of type 2 diabetes. Diabetes Care 27, 362–366.
Murphy, T., Jacobs, T., 2011. Planting healthy roots: a look at the right choice fresh start
farmers' market. Produced by University of South Carolina, Columbia, SC. Available:
http://www.youtube.com/watch?v=viPtYFwzD6I (Accessed December 21, 2012).
National Cancer Institute, 2010. Fruit and vegetable screeners in the Eating at America's
Table Study (EATS): scoring the all-day screener. National Institutes of Health
Accessed June 15, 2012).
Pan, L., Galuska, D.A., Sherry, B., et al., 2009. Differences in prevalence of obesity among
black, white, and Hispanic adults—United States, 2006–2008. MMWR 58, 740–744.
Peterson, K.E., Hebert, J.R., Hurley, T.G., et al., 2008. Accuracy and precision of two short
screeners to assess change in fruit and vegetable consumption among diverse
populations participating in health promotion intervention trials. J. Nutr. 138,
Thompson, F.E., Subar, A.F., Midthune, D., et al., 2002. Fruit and vegetable assessment:
performance of 2 new short instruments and a food frequency questionnaire.
J. Am. Diet. Assoc. 102, 1764–1772.
U.S. Census Bureau, 2010. Decennial Census. U.S. Census Bureau, Washington, DC.
WorldHealth Organization,2000.Obesity:preventing andmanagingtheglobalepidemic.
WHO Technical Report Series.World Health Organization, Geneva, Switzerland.
Yamaoka, K., Tango, T., 2005. Efficacy of lifestyle education to prevent type 2 diabetes.
Diabetes Care 28, 2780–2786.
Ziegler, R.G., Wilcox III, H.B., Mason, T.J., Bill, J.S., Virgo, P.W., 1987. Season variation in
intake of carotenoids and vegetables and fruits among white men in New Jersey.
Am. J. Clin. Nutr. 45, 107–114.
D.A. Freedman et al. / Preventive Medicine 56 (2013) 288–292