No. 4, October 2008
Assessing Retail Fruit and Vegetable Availability in Urban and Rural Underserved Communities
Akiko S. Hosler, PhD, Deepa T. Rajulu, MS, Bonnie L. Fredrick, MS, Adrienne E. Ronsani, MS
Suggested citation for this article: Hosler AS, Rajulu DT, Fredrick BL, Ronsani AE. Assessing retail fruit and vegetable availability in urban and rural underserved communities. Prev Chronic Dis 2008;5(4).
oct/07_0169.htm. Accessed [date].
Fruits and vegetables (F&Vs) are important parts of a healthy, balanced diet. Consumption of F&Vs is low among
residents of socioeconomically disadvantaged communities. We investigated and compared retail F&V availability in urban and rural underserved communities in New York
All food retail stores and farmers’ markets (N = 263) in downtown Albany and in Columbia and Greene counties in New York
State were visited and surveyed. Food stores were classified as F&V stores if they stocked more than the minimum varieties of fresh F&Vs defined by this study and as fruit-for-snack stores if they had ready-to-eat fruits only. Store density per 10,000 residents was calculated as a standardized measure of F&V availability. Adjustment weights
were created to incorporate store size and business hours into the analysis.
The weight-adjusted density (per 10,000 residents) of F&V stores was 4.6 in Albany’s minority neighborhood (reference category), 11.4 in Albany’s racially mixed neighborhood (P = .01), 7.8 in Columbia and Greene counties’ rural community (P = .10), and 9.8 in Columbia and Greene counties’ small-town community (P = .02). Significant differences were not found in fruit-for-snack stores, which ranged from 2.0 per 10,000 in the mixed neighborhood to 3.4 per 10,000 in the rural community.
The urban minority neighborhood had the most barriers to fresh F&Vs in
retail outlets, even when compared with the rural community. The low
availability of retail F&Vs in the minority neighborhood was attributed to the
lack of supermarkets and not the absolute lack of food stores. Public health intervention strategies to increase retail F&V availability in urban minority neighborhoods may aid in mitigating these disparities.
Back to top
Fruits and vegetables (F&Vs) are important parts of a healthy, balanced diet. Many studies have documented the benefits of F&V consumption for weight control (1,2) and prevention of chronic diseases, including cardiovascular disease (3,4), diabetes (5,6), and certain types of cancer (7). On the basis of recommendations by the Dietary Guidelines for Americans (8), Healthy People 2010 (HP 2010) has set national objectives for increasing proportions of Americans who consume at
least 2 daily servings of fruit and 3 daily servings of vegetables, with at least 1 serving being a dark-green or
orange-colored vegetable (9). Yet most Americans do not consume the daily recommended amount of F&Vs (10,11). Consumption of F&Vs is reportedly low among socioeconomically disadvantaged, rural, and poor residents (12).
Reasons that people do not consume the recommended daily servings of F&Vs are complex. Environmental barriers, notably the limited availability of fresh produce in local retail stores, can be a factor (12-14). Studies have found that individuals living in close proximity to supermarkets were more likely to consume F&Vs than were people living in neighborhoods without
supermarkets or located farther away from supermarkets (15-17). The inverse association between distance to the nearest supermarket and
F&V consumption appears to be pronounced among African Americans (15) and low-income individuals (16). One study found that urban
African American women who shopped at their neighborhood small-food stores consumed fewer F&Vs than did their higher-income peers who were able to shop in suburban supermarkets (18).
Studies suggest that consumption of F&Vs by disadvantaged individuals can be increased if they have access to a supermarket or food store that provides an adequate amount of affordable fresh produce (13,19,20). However, during the last few decades spatial disparities in access to retail fresh F&Vs have increased considerably (21). Supermarkets and grocery stores with produce departments are much less likely to be found in economically disadvantaged minority neighborhoods than
in middle-class or affluent white neighborhoods (21-27). Residents in remote, rural communities also face barriers to obtaining
retail fresh F&Vs (12).
We measured, quantified, and compared retail availability of fresh F&Vs in selected urban and rural communities in New York
State. We developed a field survey tool; a store classification scheme; and a standardized, weight-adjusted method to calculate store density as a measure for fresh F&V retail availability. Our goals were to identify communities with barriers to obtaining fresh produce and to suggest interventions for improving retail availability
of fresh F&Vs.
Back to top
The study communities were Columbia County, Greene County, and a downtown portion of the city of Albany, the designated underserved intervention communities for the Albany Prevention Research Center’s Diabetes Prevention and Control Project (28). We defined downtown Albany (“Albany” hereafter) as 4 geographically contiguous
zip code areas (12202, 12206, 12207, and 12210). An analysis of 2000 US census data showed that these areas formed a region
with a population of low socioeconomic status within the city of Albany (29). Our previous study showed that residential racial composition was the sole predictor for
availability of low-fat milk and high-fiber bread in these same communities (29). We divided Albany into a minority neighborhood, which was defined by
census block groups (CBGs) with at least 50% nonwhite and/or Hispanic populations, and a mixed neighborhood, which was defined by CBGs with
less than 50% nonwhite and/or Hispanic populations.
We divided Columbia and Greene counties, where the population is predominantly non-Hispanic white, into a rural community and a small-town community. The rural community was composed of CBGs with populations that are 100% rural, as defined by the US Census Bureau (30), and the small-town community was composed of the two counties’ remaining CBGs. This division contrasted sparsely populated rural regions with more populous small towns along the Hudson River.
We visited and surveyed all food stores in the study communities during July through September 2003. We defined a “food store” as a retail store that was open for business for at least 5 days per week during the time of the survey and that stocked at least 1 of the following food items: milk, bread, or fresh produce. Farmers’ markets were included regardless of the number of days they were open each week. We used 3 sources to obtain a list of food stores: 1) the New York State
Department of Agriculture and Markets’ (NYSDAM’s) current inspected food store database (accessed using a Freedom of Information Act request), 2) the NYSDAM’s online guide for farm-fresh products and farmers’ markets (31), and 3) online local yellow pages. We initially identified 426 potential food stores, including 7 farmers’ markets. We contacted all of the food stores by phone to verify whether they fit our food store definition and whether they were located in the study area.
After eliminating nonqualified, closed, and duplicate stores, 263 stores remained and were surveyed. A detailed process of identifying food stores is shown in Figure 1.
Figure 1. Flow Chart Depicting How Food Stores Were
Identified, Albany, Columbia County, and Greene County, New York, 2003. [A
text description of this figure is also
The 1-page survey assessed more than 10 types of food and nonfood items. Among other indicators, F&V presence and variety were assessed. A team of 2 or 3 trained
surveyors conducted the in-store surveys. If fewer than 10 types of fresh F&Vs were sold, surveyors wrote down each type of F&V; if more than 10 types of F&Vs were sold, surveyors noted that there was a large variety. Surveyors noted whether at least 1 dark-green or
was present, and they collected information about store hours and number of cash registers used. We used the number of cash registers as a measure for store size, as suggested by a previous study (25). To ensure interrater reliability, results of the assessment were recorded when all members of the survey team agreed on the assessment decisions. As a standardized protocol, surveyors asked permission to conduct a survey, and all stores granted permission. Written consent was obtained from an
employee of each store who was authorized to make customer-related decisions.
The location of each store was measured at the front door by coordinates using a hand-held global positioning system (GPS) device (Thales Navigation, Inc, Alexandria, Virginia). Survey data were analyzed using SPSS version 12.0 (SPSS, Inc, Chicago, Illinois), and GPS information was processed using MapInfo software (Pitney Bowes Software, Inc, Troy, New York).
We initially grouped surveyed food stores using a business-type classification system developed by Morland et al (22). However, we found that considerable variations
in fresh F&V presence and variety existed within the same business type. For instance, 46% of gas station stores had some fresh F&Vs, but 54% did not carry F&Vs at all. Among convenience stores
in the same chain, some locations had produce sections but others did not. Furthermore, “supermarket” did not
automatically mean that the store had a produce department; 1 large national chain’s local stores had no produce departments at all. Therefore, we devised a store classification system to properly categorize the stores by the availability and variety of fresh F&Vs sold (Figure 2).
Figure 2. Flow Chart Depicting How Food Stores Were
Classified, Albany, Columbia County, and Greene County, New York, 2003. [A
text description of this figure is also
We categorized the business as an F&V store if it stocked at least 2 types of fresh fruit, excluding lemons and limes, and at least 3 types of fresh vegetables, including at least 1 dark-green
or orange-colored vegetable. We further grouped the F&V stores into 3 types: super produce stores, or supermarkets with a produce department; year-round produce stores, which were nonsupermarket F&V stores that operate year-round, including grocery stores,
convenience stores, and specialty food stores; and seasonal produce stores. The last category included seasonal farm (ie, barn) stores, roadside huts and stands, and farmers’ markets.
Stores that did not meet the F&V measure but did carry at least 1 type of ready-to-eat fresh fruit, such as apples, oranges, and bananas, were designated as fruit-for-snack stores. We determined that recognition of this type of store would be useful for promoting fruit consumption (ie,
choosing a piece of fruit in lieu of candy or a snack food). The remaining stores were grouped into the non-F&V store category.
As a standardized proximal measure of fresh F&V availability, we calculated store density per 10,000 residents and compared the densities of the 4 communities. Data from the 2000 census were used to determine total number of residents. A simple store density calculation would treat all stores equally, including large supermarkets, small corner stores, and farmers’ markets that open only 4 to 7 hours per week. To adjust for store size and operating hours, we devised a simple
adjustment weight. The number of cash registers, a surrogate measure for store size (25),
was natural-log–transformed and added the constant of 1 to normalize the right-skewed distribution with the minimum value of 1. Store hours were measured by per-week operating hours. Store hours were normally distributed, with 98 hours as both the mean and the median values. A 98-hours-per-week operation is equivalent to opening from 7:00
a.m. to 9:00
p.m. daily, a time period we determined was adequate
for most residents to shop for F&Vs. We designated a weight of 1 for a store with 1 cash register that was open for 98 hours per week. The weight increased if the store had more than 1 cash register and decreased if the store operated fewer than 98 hours per week.
Therefore, the adjustment weight (wi) was obtained from 1 plus the natural-log–transformed number of cash registers multiplied by the store hours per week divided by 98. For stores open 98 hours per week or longer, a constant of 98 was used:
wi = (1+ nlog ci )(hi / 98) where wi = adjustment weight, ci = number of cash registers, and hi = business hours per week if hi <98, hi = 98 for all other stores.
In aggregate, the adjustment weight amplifies the effects of super produce stores almost 3 times but reduces the effects of seasonal produce stores by half. Other types of stores remain mostly unchanged.
We compared unadjusted and weight-adjusted store density per 10,000 residents of the study communities, using the minority neighborhood in Albany as the reference category. The z score was used for statistical analysis of significance. Significance was determined at an α level of .05. The institutional review board of
the New York State Department of Health reviewed and approved the study protocol.
Back to top
Characteristics of the study communities are presented in
Table 1. The minority neighborhood had the highest
percentage of racial and ethnic minorities (74.1%) of the 4 communities, and the
rural community had the lowest percentage (4.6%). One in 3 residents in the
minority neighborhood and 1 in 4 residents in the mixed neighborhood were living
below poverty. In both the mixed and minority neighborhoods, three-quarters of housing units were renter-occupied. At least one-third of households in
both the minority and mixed neighborhoods did not have a vehicle. The rural community, with a total population of nearly 62,000, had the lowest rates among the
4 study communities of racial/ethnic minorities (4.6%), individuals living in poverty (8.9%), renter-occupied housing (24.6%), and households without a vehicle (5.9%), and it had the highest rate of citizens
65 years or older (16.3%).
A total of 263 eligible food stores and farmers’ markets were categorized
with our classification system. Of stores designated as F&V stores, which
made up 36.5% of all food stores, 16 were super produce stores, 54 were year-round produce stores, and 26 were seasonal produce stores. Most F&V stores stocked far more than the minimum fresh F&Vs defined by this study. Additionally, 37 stores (14%) were fruit-for-snack stores, and 130 stores (49%) were non-F&V stores.
Seventeen stores classified as non-F&V stores carried some lemons, limes, and/or vegetables, but none were dark green or orange-colored. A summary of statistics for the numbers of cash registers and store hours, as well as adjustment weights, is shown in
The availability of fresh F&Vs was measured by store density (Table
3). Unadjusted population density of F&V stores was 4.2 per 10,000 residents for the minority neighborhood, 7.3 per 10,000 residents for the mixed neighborhood, 6.5 per 10,000 residents for the rural community, and 6.8 per 10,000 residents for the small-town community. No statistically significant differences were found among the communities.
Population densities of all stores were weight-adjusted to factor in store size and business hours.
For every 10,000 residents the communities had the following number of F&V stores: 4.6 for the minority neighborhood, 11.4 for the mixed neighborhood, 7.8 for the rural community, and 9.8 for the small-town community. The store densities were significantly higher for the mixed neighborhood (P = .01) and the small-town community (P = .02) than for the minority
neighborhood. This was largely because the adjustment weight upwardly corrected the contribution of super produce stores, which were present in the mixed neighborhood (7.3 per 10,000 residents), rural community (2.4 per 10,000 residents), and small-town community (4.0 per 10,000 residents) but not in the minority neighborhood. The contribution of seasonal stores was downwardly corrected by the weight, primarily because of their shorter business hours, but the small-town community still had
a significantly higher density of seasonal stores (1.6 per 10,000 residents, P = .04) than did the minority neighborhood, where there were no seasonal stores. No significant differences were found among year-round produce stores (ranged from 4.0 per 10,000
residents in the mixed neighborhood to 4.6 in the minority neighborhood), fruit-for-snack stores (ranged from 2.0 per 10,000
residents in the mixed neighborhood to 3.4 in the rural community), and all food stores combined (ranged from 17.7 per 10,000
the rural community to 25.4 per in the mixed community).
Back to top
Our study demonstrated that the urban minority neighborhood was the most disadvantaged in terms of retail F&V availability, as measured by the population density of F&V stores. The low
retail availability of fresh F&Vs in this community appears to be largely because of the lack of super produce stores and not the absolute lack of food stores. In fact, the urban minority community had a large number of food stores, and, of the 4 communities, the highest population density of
year-round produce stores and non-F&V stores and the second-highest population density of fruit-for-snack stores in the weight-adjusted model. This is a case of relative deficiency, in which high-impact super produce stores were conspicuously missing. In contrast, the urban mixed neighborhood had 3 large super produce stores. The higher density
by population of supermarkets and large grocery stores in racially mixed areas compared with predominantly minority areas in the urban setting has been
reported by other studies conducted elsewhere in the United States (22,26,27).
Although the results were not statistically significant, the rural community had higher overall fresh F&V availability compared with the urban minority neighborhood. The rural community had the lowest density
by population of total food stores but had the highest density of fruit-for-snack stores and the second highest density of year-round produce stores. The high availability of ready-to-eat fruits can be explained by the existence of many family-owned fruit orchards and berry
farms in this community. The small-town community had the second highest overall F&V availability, with a balanced representation of super produce, year-round produce, and seasonal produce stores. In the winter when seasonal produce stores are not in business, both communities in Columbia and Greene counties probably see some decline in F&V availability. However, the decline probably
does not reach the level of the urban minority neighborhood because these seasonal stores make
a small contribution to the overall F&V availability in the rural and small-town communities.
Our study has several limitations. We used a cross-sectional survey of food stores and did not adjust for the variability caused by harvest, delivery, or reshelving schedule. We did not directly assess quantity or quality of fresh F&Vs because of
a lack of standard measurements suited for field studies. The designation of neighborhoods and communities was made on the basis of CBG data and may differ from actual areas designated as communities. The rural community encompasses a large
tract of land, and most remote portions of the community may have more environmental barriers
to retail F&V availability. The store classification system and the adjustment weight were new tools and have not been tested in other communities. This study, by design, did not assess the availability of non-retail fresh F&Vs such as those grown in private or community gardens. Availability of canned, jarred, and frozen F&Vs was not assessed because of time constraints,
the lack of standard
measurements, and the presence of salt, sugar, and nutrient loss that can occur in the process of canning, jarring, and freezing F&Vs.
A growing need exists to develop public health policies and innovative intervention strategies to increase retail availability of fresh F&Vs in disadvantaged communities. To do so, scientific yet simple field
methods to measure retail food availability need to be established that can be applied to both urban and rural communities. The “food deserts” controversy in the United Kingdom revealed that a few small-scale exploratory studies of retail food availability in poor urban
neighborhoods were used repeatedly — and in some instances were misinterpreted — to form food and nutrition policies (32). This incident highlights the importance of applying sound science in community food assessment for policy recommendations.
Because very few existing food availability studies directly compared rural communities with urban communities, findings from this study will be useful for making policy recommendations and planning interventions from a global perspective. This study suggests that urban minority neighborhoods should be a priority for improving retail fresh F&V availability. Although bringing new retail
produce stores into an urban environment is not an easy task, 1 study explains that collaboration among
stakeholders, including community leaders, business owners, growers, media, and local government, can greatly improve the chances of successful introduction and retention of the stores (23). Working with the existing fruit-for-snack stores to add more F&Vs to their inventory is another intervention idea. In-store promotion and nutrition education, such as taste testing, giveaways, and cooking demonstrations to increase local consumer demand, should accompany this effort.
Back to top
This study was supported by the Centers for Disease Control and Prevention through the Albany Prevention Research Center Cooperative Agreement
No. U48/DP000028. We thank Maria Lopez, Trudi McGhee, Michelle Kizer, and Patricia Gutierrez for their assistance in data collection and data entry.
Back to top
Corresponding Author: Akiko S. Hosler, PhD, Assistant Professor, Department of Epidemiology and Biostatistics, University at Albany School of Public Health, One University Pl, Rensselaer, NY 12144-3456. Telephone: 518-402-1561. E-mail: firstname.lastname@example.org.
Author Affiliations: Deepa T. Rajulu, Adrienne E. Ronsani, Bureau of Chronic
Disease Epidemiology and Surveillance, New York State Department of Health,
Albany, New York; Bonnie L. Fredrick, Wadsworth Center Genetic Services, New
York State Department of Health, Albany, New York.
Back to top
- Newby PK, Muller D, Hallfrisch J, Qiao N, Andres R, Tucker KL.
Dietary patterns and changes in body mass index and waist circumference in adults. Am J Clin Nutr 2003;77(6):1417-25.
- Rolls BJ, Ello-Martin JA, Tohill BC.
What can intervention studies tell us about the relationship between fruit and vegetable consumption and weight management? Nutr Rev 2004;62(1):1-17.
- Joshipura KJ, Hu FB, Manson JE, Stampfer MJ, Rimm EB, Speizer FE, et al.
The effect of fruit and vegetable intake on risk for coronary heart disease. Ann Intern Med 2001;134(12):1106-14.
- Dauchet L, Amouyel P, Hercberg S, Dallongeville J.
Fruit and vegetable consumption and risks of coronary heart disease: a meta-analysis of cohort studies. J Nutr 2006;136(10):2588-93.
- van Dam RM, Rimm EB, Willett WC, Stampfer MJ, Hu FB.
Dietary patterns and the risk for type 2 diabetes mellitus in U.S. men. Ann Intern Med 2002;136:201-9.
- Ford ES, Mokdad AH.
Fruit and vegetable consumption and diabetes mellitus incidence among U.S.
adults. Prev Med 2001;32(1):33-9.
- Terry P, Terry JB, Wolk A.
Fruit and vegetable consumption in the prevention of cancer: an update. J Intern Med 2001;250(4):280-90.
- Dietary Guidelines for Americans, 2005. Washington (DC): U.S. Department
of Health and Human Services, U.S. Department of Agriculture. http://www.health.gov/dietaryguidelines/dga2005/document.
Accessed September 10, 2007.
- Healthy People 2010: understanding and improving health. Vol. 2.
Washington (DC): U.S. Department of Health and Human Services. http://www.healthypeople.gov.
Accessed September 10, 2007.
- Heimendinger J, Van Duyn MA, Chapelsky D, Foerster S, Stables G.
The national 5 A Day for Better Health Program: a large-scale nutrition
intervention. J Public Health Manag Pract 1996;2(2):27-35.
- National Cancer Institute. NCI 5 A Day attitude and behavioral tracking: national omnibus survey results, November 2003. Bethesda (MD): National Cancer Institute; 2003.
- Krebs-Smith SM, Kantor LS.
Choose a variety of fruits and vegetables daily: understanding the complexities. J Nutr 2001;131(2S-1):487S-501S.
- Glanz K, Yaroch AL.
Strategies for increasing fruit and vegetable intake in grocery stores and communities: policy, pricing, and environmental change. Prev Med 2004;39(Suppl
- Seymore JD, Yaroch AL, Serdula M, Blanck HM, Khan LK.
Impact of nutrition environmental interventions on point-of-purchase behavior in adults: a review. Prev Med 2004;39(Suppl
- Morland K, Wing S, Diez Roux A.
The contextual effect of the local food environment on residents’ diets: the
atherosclerosis risk in communities study. Am J Public Health 2002;92(11):1761-7.
- Rose D, Richards R.
Food store access and household fruit and vegetable use among participants in the US Food Stamp Program. Public Health Nutr 2004;7(8):1081-8.
- Laraia BA, Siega-Riz AM, Kaufman JS, Jones SJ.
Proximity of supermarkets is positively associated with diet quality index for pregnancy. Prev Med 2004;39(5):869-75.
- Zenk SN, Schulz AJ, Hollis-Neely T, Campbell RT, Holmes N, Watkins G, et al.
Fruit and vegetable intake in African Americans: income and store characteristics. Am J Prev Med 2005;29(1):1-9.
- Kristal AR, Goldenhar L, Muldoon J, Morton RF.
Evaluation of a supermarket intervention to increase consumption of fruits and vegetables. Am J Health Promot 1997;11(6):422-5.
- Sloane DC, Diamant AL, Lewis LB, Yancey AK, Flynn G, Nascimento LM, et al.
Improving the nutritional resource environment for healthy living through community-based participatory research. J Gen Intern Med 2003;18(7):568-75.
- Curtis KA, McClellan S. Falling through the safety net: poverty, food assistance and shopping constraints in an American City. Urban Anthropol 1995;24:93-135.
- Morland K, Wing S, Diez Roux A, Poole C.
Neighborhood characteristics associated with the location of food stores and
food service places. Am J Prev Med 2002;22(1):23-9.
- Shaffer A. The persistence of L.A.’s grocery gap: the need for a new food policy and approach to market development. Los Angeles (CA): Center for Food and Justice, Urban and Environmental Policy Institute, Occidental College; 2002.
- Zenk SN, Schulz AJ, Israel BA, James SA, Bao S, Wilson ML.
Neighborhood racial composition, neighborhood poverty, and the spatial accessibility of supermarkets in metropolitan Detroit. Am J Public Health 2005;95(4):660-7.
- Horowitz C, Colson KA, Hebert PL, Lancaster K.
Barriers to buying healthy foods for people with diabetes: evidence of environmental disparities. Am J Public Health 2004;94(9):1549-54.
- Baker EA, Schootman M, Barnidge E, Kelly C. The role of race and poverty
in access to foods that enable individuals to adhere to dietary guidelines.
Prev Chronic Dis 2006;3(3).
- Powell LM, Slater S, Mirtcheva D, Bao Y, Chaloupka FJ.
Food store availability and neighborhood characteristics in the United States. Prev Med 2007;44(3):189-95.
- Shaw BA, Gallant MP, Riley-Jacome M, Spokane LS.
Assessing sources of support for diabetes self-care in urban and rural underserved communities. J Community Health 2006;31(5):393-412.
- Hosler AS, Varadarajulu D, Ronsani AE, Fredrick BL, Fisher BD.
Low-fat milk and high-fiber bread availability in food stores in urban and rural communities. J Public Health Manage Pract 2006;12(6):556-62.
- Question and answer center. Washington (DC): U.S. Census Bureau. https://ask.census.gov/cgi-bin/askcensus.cfg/php/enduser/std_alp.php.
Accessed September 10, 2007.
- Farm and market search. Albany (NY): New York State Department of
Agriculture and Markets. http://www.agmkt.state.ny.us. Accessed September
- Cummins S, Macintyre S.
“Food deserts” — evidence and assumption in health policy making. BMJ 2002;325(7361):436-8.
Back to top