Identifying rural food deserts: methodological considerations for food environment interventions.

Lebel, Alexandre ; Noreau, David ; Tremblay, Lucie 等

Adequate nutrition is fundamental from infancy until adult life and is among the most important determinants of health. (1-3) Healthy dietary choices are associated with a higher life expectancy; unhealthy choices are associated with substantial comorbidities, including obesity, diabetes and other chronic diseases. (4,5) While individual dietary behaviours remain important, (6) there is an increasing interest in ecological approaches to improving nutrition, including access to healthy foods. (7,8) Promotional interventions for healthy eating are essential, but it is widely recognized that such interventions are more efficient when healthy choices are facilitated by enhanced accessibility to a variety of nutritious foods at a reasonable price. (9) If such accessibility is unevenly distributed among regions or between demographic subgroups, health determinant disparities may emerge, (10) including food insecurity. (11) Food insecurity occurs when all members of a household do not have reliable access to food in sufficient quantity and quality to maintain an active and healthy lifestyle. (12)

In 2012, 13% of Canadian households experienced food insecurity, (5) which is similar to the 14.5% observed in the US. (4) In 2012, all the 10 counties in the Health Region of Chaudiere-Appalaches (HR-CA), a mainly rural administrative region in the province of Quebec with approximately 400,000 people, perceived food insecurity as a priority area for policy action. (13) According to regional HR-CA stakeholders, most production of fresh food is oriented to the market outside the region. In order to develop context-specific interventions, the Regional Public Health Authority (RPHA-CA) planned to characterize the food supply offered and its accessibility in the Authority's Action Plan 2013-2018. (14)

Various methods to measure accessibility to the food environment have been developed during the last decade. The use of geographic information systems (GIS) as a tool for spatial analysis is one approach. (15) It involves mapping food establishments across a region, measuring spatial access to these facilities, (16) and determining the association with individual dietary and other health behaviours. (15) The popularity of GIS in public health studies has generated a great number of food access measures and methods. In some cases, differing methods of characterizing food access could provide different results and associations. Some studies have focused on a specific food source, such as fast food restaurants or convenience stores, others aimed to describe and compare accessibility between neighbourhoods or regions. A review of these studies has demonstrated evidence for disparities in food access, income and race (17,18) and has shown an impressive heterogeneity in measurements, (19) which are not always comparable. Findings from other high-income countries have been sparse and equivocal. (20) The most common measurements rely on either density of food sources per square/area or physical proximity in road distance network. (21) These studies have been particularly useful in identifying problematic areas and have been closely linked to the notion of food deserts. (22) Food deserts are typically defined as "poor urban areas, where residents cannot buy affordable, healthy food". (23,24)

The majority of studies aiming to measure the food environment focus on higher density urban environments, using measurements that are perhaps unsuitable for less dense and more scarcely populated regions. (25,26) The concept of a food desert is contested regarding rural environments since it is expected that rural dwellers always have access to motorized transport and that they maximize shopping efficiency by making large-volume shopping trips. (27) Yet, low-income or elderly rural residents may not have access to safe and reliable transportation, which may lead to low shopping frequency. (28) In effect, rural households with fewer resources may be constrained, the long distances and less frequent trips to their primary food store reducing the regularity of healthy food availability at home. Aside from the issue of physical distance between individuals and resources, this situation makes the food desert phenomenon similar to that for low-income urban residents, i.e., healthy food is difficult to reach and is often more expensive. (29) However, the measurement of the food environment remains challenging in rural settings, since both population and food supply sources are unevenly distributed in these territories. Few studies address this problem and may be responsible for mischaracterizing food deserts. (30) For example, according to desk top information, the US Department of Agriculture (USDA) defines a food desert as a "low-income tract in which at least 500 people or 33 percent of the population live more than 1 mile (in urban areas) or more than 10 miles or 16 km (in rural areas) from the nearest supermarket, supercenter, or large grocery store". (31) Such a definition may limit the identification of rural food deserts because of three important shortcomings: it does not consider the relative geographic distribution between households and food supply sources; it only considers large food sources that are often nonexistent in rural communities; and it overlooks the quality of what is offered within food sources. With the increasing proportion of low-income and elderly people in rural areas, the access to fresh, diverse and affordable food products becomes a critical issue, and more precise measurements are needed to identify potential food deserts in rural areas. (10,30)

The objective of this study was to build a methodology to describe the food environment in a rural area. A specific description of the food environment is more likely to identify potential food deserts and may provide an opportunity for stakeholders to develop intersectoral interventions and policies to address the current food insecurity problem (32) currently perceived in the HR-CA.

METHODS

We developed objective ecologic measurements to assess the regional food environment using a mixed methodology approach, grounded in the closely linked concepts of food security and food deserts. Food security rests on four pillars: food availability, access, utilization and stability. (11) The food desert concept has been typically operationalized in three ways: problems with food accessibility, problems with food affordability and problems with food availability. (33) Food utilization and stability have typically been the domain of individuals and are not considered to be ecologic measurements. Accessibility describes whether an individual has physical access to food retailers selling healthy items. Affordability relates to the cost of food within these retailers. Availability indicates the relative diversity of food choices available in the food retail economy of a particular neighborhood.

We constructed ecologic measures of the food environment in three steps: 1) modelling the ecumene, a wide concept essentially referring to the area inhabited by human society; (34) 2) assessing food store quality; and 3) identifying potential food deserts. Geographic information was validated directly in the field. Inter-rater statistics were computed for in-store observations. Final results were compared with the food desert parameters proposed by the USDA, as well as with the perceptions of a group of regional stakeholders involved in the Regional Food Security Committee.

Modelling the ecumene

A major limitation for the identification of food deserts in rural environments is the need to take into account the enormous variation of the relative geographic distribution between people and resources. In effect, the proximity of food supply sources is disproportionately distributed between households in rural settings, where some people live relatively close to a wide range of food sources and others may have access to only a single store with limited produce, such as a convenience store or a gas station, within 10 or 20 km. In this study, the ecumene was assessed by computing the road network of each household to food establishments in all communities. Communities were defined as municipalities, subsections of municipalities or groups of municipalities in which individuals share common needs and activities. A group of regional stakeholders designated these communities as such while considering attachment or belonging to a community that may influence individuals' travel behaviour. (35) The deprivation level of the communities was assessed using the regional weighted deprivation index provided by the Quebec National Public Health Institute using the Canadian census 2006. This index is widely used in Canada and was computed from the communities' mean income, proportion of people without a high school diploma and unemployment rate, and was divided into quintiles. (36) A topologic road network database was created using DMTI CanMap Streetfiles 2013 and a GIS (ArcGIS, 10.1).

The Quebec Ministry of Municipal Affairs, Regions and Land Occupancy (MAMROT) provided the municipal property assessment roll of 2010 as well as the civic address and the function of every listed property. We gathered information on all buildings categorized as "food retail": supermarkets, convenience stores, large general stores (e.g., Walmart) and even gas stations, which often provide a large variety of food products in rural municipalities. All food outlets were mapped and visited to confirm their existence and location. All roads where no food outlets were reported were also visited in order to locate outlets missing from the MAMROT database. Missing outlets (n = 30) were removed from the database, and those that were not reported (n = 43) were added, thus providing a complete and exhaustive list of 153 food outlets. Six places refused the assessment and were not considered. This validation indicated that the MAMROT data had a positive predictive value of 0.74, which means that it delivered accurate information for 74% of registered food stores three years after its publication in 2010, and a sensitivity of 0.80, which means that 20% of reported food stores in the original data were not observed in the field in 2014.

Assessment of food store quality

In-store evaluations were performed to acquire information on the food affordability and availability. For all visited stores, we categorized 25 food products present in the National Nutritious Food Basket (2008) and distributed among the four groups of Canada's Food Guide (vegetables and fruit 4, meat and alternatives 10, grain products 6 and milk and alternatives 5) according to availability, freshness and price. The information was used to construct four synthetic indexes: affordability, freshness, diversity and relative availability.

Affordability: The observed price of each food product was subtracted from the regional mean price of the same product. We categorized food as above or below the regional mean.

Freshness: For the vegetables and fruit, a simple qualitative assessment determined whether the product was fresh or not. For products with an expiration date, we considered the product to be fresh if the product had not yet "expired". A store was considered to be selling fresh food if all products were reported to be fresh.

Diversity: The diversity index was built using an entropy-based formula similar to the land-use mix index. (37) It synthesised the importance and availability of each of the four categories of food as shown in Equation 1, where [p.sub.i] is the proportion of the type of product in one food category and n is the number of food categories (n = 4). The denominator [ln.sub.n] allows for the calculation of an index bounded between 0 and 1, where 0 qualifies as a store selling a single food product category (no diversity) and 1 indicates a store that has a great number of products (high diversity). The 25% of food stores having the highest scores (fourth quartile) were considered as selling a high diversity of food products.

-(([[summation].sub.n][p.sub.i] * ln [p.sub.i])/[ln.sub.n]) (1)

Relative availability: As suggested by Pouliot and Hamelin, (38) the shelf space provided for healthy (vegetables and fruit) and unhealthy (sodas and chips) foods was estimated using step count. The ratio of the estimated shelf space for each type allowed for the identification of stores offering more shelf space for healthy than unhealthy products.

Food products quality assessment validation: Using two teams of two observers, we conducted this assessment in two regional counties from October to December of 2013 and another two in October to early December 2014. The observers had been previously trained to look for the right produce and to avoid outliers or inaccurate observations, such as weekly sales and different volumes of product. Inter-rater agreement (Kappa statistics) was computed for all the reported information in a subset of 20 food stores. Most Kappa statistics were above 0.90 (mean 0.92), indicating a very high concordance among observers. However, the freshness of whole wheat bread scored a moderate concordance with a coefficient of 0.57. This indicator was not considered reliable and was removed from the analyses.

Identification of potential food deserts

Food stores that qualified as affordable and selling fresh, diverse and high-quality food products were selected in the GIS and labeled as "high store". A road network distance was calculated between each residential building to the closest high store. Residences located above 16 km to a high store (31) and in most deprived quintile communities were identified as potential food deserts.

Comparison with USDA method and stakeholders' perceptions

We used the USDA method to define potential food deserts. (31) The methodology was applied by keeping only retailers identified as "supermarkets". The results were then compared to highlight the differences in the exposed population. All results were presented to the Regional Food Security Committee of Chaudiere-Appalaches and collaborators in March 2015. This interdisciplinary committee comprises 17 professionals working in the field of education, public health, agriculture, food banks and other non-government organizations (NGOs) and includes as many other collaborators from various institutions (for a total n of close to 40). We obtained their perceptions regarding this representation of the food environment and inquired whether the information matched their knowledge of the region.

RESULTS

All regions have extensive numbers of food establishments; few residences have no access within 16 km. When in-store food affordability, freshness, diversity, quality and relative availability are incorporated, the proportion of residences with poor accessibility varied from 2.8% to 40.4% among regions. When residences located in deprived communities only were considered, accessibility varied from 0.4% to 32.3% among regions (Table 1). The Etchemin region may be particularly vulnerable since the mean distance to reach a high store was 14.8 km; over 68% of the population live in a deprived community. The difference between the mean and the median distance to a high store indicates the importance of the variation within and among regions, and highlights the heterogeneity of local contexts of food environment.

Using the USDA criteria to define rural food deserts led to substantially altered results. For example, Table 1 shows that 5,115 households were living in a potential food desert distributed among all four regions, whereas using USDA criteria this dropped to 686 households located in only one region.

Mapping provided more precision regarding the location of residential clusters located in a deprived community and living further than 16 km from a high store (Figure 1). The map also reveals vulnerable communities that are materially deprived but where accessibility to a good food source relies on only one store.

The results of the food environment description were presented to the Regional Food Security Committee of Chaudiere-Appalaches (March 20, 2015). Many stakeholders recognized what they generally observed in the communities but were impressed by how many households were located in a potential food desert and by the variety of contexts in which potential food deserts were identified. The presentation inspired many alternative interventions, such as developing a cooperative businesses network, mobile food suppliers or enhancing in-store diversity. Nevertheless, it was clear to the Committee that the structure of the ecumene makes the development of cost-effective solutions challenging, that none of these alternatives would entirely address the perceived food insecurity problem and that intersectoral cooperation among regional stakeholders (e.g., producers, suppliers, NGOs) would be needed. An objective description of the food environment was seen as a valuable complementary tool and as a first essential step in building the case for decision-makers to develop adapted interventions for the region.

[FIGURE 1 OMITTED]

DISCUSSION

Knowledge of the food environment is a critical dimension to address when considering the food insecurity problem in the population. (39) Because it is particularly challenging to describe the food environment in a rural setting, this study proposes a more precise methodology to identify potential food deserts and has revealed an impressive variety of local contexts. Regional stakeholders in HR-CA recognized this objective measurement as a better complementary tool to their knowledge of the situation than what is usually used and as better contributing to robust arguments directed at decision-makers for facilitating interventions. Regional stakeholders understand the complexity of the food market dynamics and are not looking for a single and easy solution for all vulnerable areas. Although many proposed interventions are known to them, analyzing the food environment by considering the ecumene more closely strengthens the need for intersectoral interventions. Identifying the most vulnerable areas (potential food deserts) may also help to prioritize some areas for intervention or point to the presence of support for initiatives in those areas.

The ecumene model provides a more precise assessment of the food environment and allows the identification of a variety of local contexts, since accessibility to food sources is not constrained by census tract boundaries, as for the USDA criteria. Another benefit of the proposed methodology is its flexibility, as it could easily be transferred to study any rural food environment. Yet food items and thresholds chosen for building each indicator were based on relative measurements specifically in HR-CA and were defined with the help of regional public health authority professionals, including a nutritionist. Consequently, any measurement or threshold proposed in this study should not be considered as an absolute indicator of food-store quality and should be adapted to the area under investigation. Although four synthetic indices (affordability, freshness, diversity and relative availability) are directly associated with the notion of food insecurity, we further recommend validating the thresholds used to identify a good food source with the knowledge of regional stakeholders.

The main limitation of this methodology is that it requires field observation and many data validation procedures. Although these procedures are demanding, our experience has shown that the alternative of using exclusively desk information may be not only less reliable but also misleading for decision-making for at least two reasons: 1) a significant number of errors were found in administrative databases; 2) several nontraditional food sources, such as gas stations, are often the only reachable source of healthy food and must be considered.

Assessment of the food environment in rural areas is methodologically challenging. Nevertheless, in order to provide a relevant complementary tool to stakeholders, we recommend considering field observations and the relative distribution of people and resources (the ecumene) as a necessity, not an option.

REFERENCES

(1.) Mozaffarian D, Ludwig DS. The 2015 US dietary guidelines: Lifting the ban on total dietary fat. JAMA 2015; 313(24):2421-22. PMID: 26103023. doi: 10.1001/ jama.2015.5941.

(2.) Kiefte-de Jong JC, de Vries JH, Bleeker SE, Jaddoe VW, Hofman A, Raat H, et al. Socio-demographic and lifestyle determinants of 'Western-like' and 'Health conscious' dietary patterns in toddlers. Br J Nutr 2013; 109(1):137-47. PMID: 22475342. doi: 10.1017/S0007114512000682.

(3.) USDA, D.G.A.C. Scientific Report of the 2015 Dietary Guidelines Advisory Committee. Washington, DC: USDA and US Department of Health and Human Services, 2015.

(4.) Coleman-Jensen A, Rabbitt MP, Gregory C, Singh A. Household Food Security in the United States in 2014, 194. Washington, DC: US Department of Agriculture, Economic Research Service, 2015.

(5.) Tarasuk V, Mitchell A, Dachner N. Household Food Insecurity in Canada, 2012. Toronto, ON: Research to Identify Policy Options to Reduce Food Insecurity (PROOF), 2014.

(6.) Egger G, Swinburn B. An "ecological" approach to the obesity pandemic. Br Med J 1997; 315(7106):477-80. PMID: 9284671. doi: 10.1136/bmj.315. 7106.477.

(7.) Townsend N, Foster C. Developing and applying a socio-ecological model to the promotion of healthy eating in the school. Public Health Nutr 2013; 16(6):1101-8. PMID: 22014870. doi: 10.1017/S1368980011002655.

(8.) Golden SD, Earp JAL. Social ecological approaches to individuals and their contexts: Twenty years of Health Education & Behavior health promotion interventions. Health Educ Behav 2012; 39(3):364-72. PMID: 22267868. doi: 10.1177/1090198111418634.

(9.) Giskes K, Kamphuis CB, van Lenthe FJ, Kremers S, Droomers M, Brug J. A systematic review of associations between environmental factors, energy and fat intakes among adults: Is there evidence for environments that encourage obesogenic dietary intakes? Public Health Nutr 2007; 10(10):1005-17. PMID: 17381942. doi: 10.1017/S1368980007665525.

(10.) McEntee J, Agyeman J. Towards the development of a GIS method for identifying rural food deserts: Geographic access in Vermont, USA. Appl Geogr 2010; 30(1):165-76. doi: 10.1016/j.apgeog.2009.05.004.

(11.) Barrett CB. Measuring food insecurity. Science 2010; 327(5967):825-28. PMID: 20150491. doi: 10.1126/science.1182768.

(12.) Radimer K, Olson C. Understanding hunger and developing indicators to assess it in women and children. J Nutr Educ 1992; 24(1):36S-44S. doi: 10. 1016/S0022-3182(12)80137-3.

(13.) HR-CA. L'identification des deserts alimentaires en milieu rural. Direction de sante publique de Chaudiere-Appalaches, Sainte_Marie de Beauce, 2015; 10.

(14.) HR-CA. Plan d'action regional en securite alimentaire de Chaudiere-Appalaches 2013-2018. In: D.d.s. publique (Ed.): Gouvernement du Quebec, 2013; 15.

(15.) Charreire H, Casey R, Salze P, Simon C, Chaix B, Banos A, et al. Measuring the food environment using geographical information systems: A methodological review. Public Health Nutr 2010; 13(11):1773-85. PMID: 20409354. doi: 10. 1017/S1368980010000753.

(16.) Glanz K, Sallis JF, Saelens BE, Frank LD. Healthy nutrition environments: Concepts and measures. Am J Health Promot 2005; 19(5):330-33. PMID: 15895534. doi: 10.4278/0890-1171-19.5.330.

(17.) Caspi CE, Sorensen G, Subramanian SV, Kawachi I. The local food environment and diet: A systematic review. Health Place 2012; 18(5):117287. PMID: 22717379. doi: 10.1016/j.healthplace.2012.05.006.

(18.) Walker RE, Keane CR, Burke JG. Disparities and access to healthy food in the United States: A review of food deserts literature. Health Place 2010; 16(5):876-84. PMID: 20462784. doi: 10.1016/j.healthplace.2010.04.013.

(19.) Lytle LA. Measuring the food environment: State of the science. Am J Prev Med 2009; 36(4):S134-44. PMID: 19285204. doi: 10.1016/j.amepre.2009.01.018.

(20.) Beaulac J, Kristjansson E, Cummins S. Peer reviewed: A systematic review of food deserts, 1966-2007. Prev Chronic Dis 2009; 6(3):A105. PMID: 19527577.

(21.) Apparicio P, Abdelmajid M, Riva M, Shearmur R. Comparing alternative approaches to measuring the geographical accessibility of urban health services: Distance types and aggregation-error issues. Int J Health Geogr 2008; 7(1):7. PMID: 18282284. doi: 10.1186/1476-072X-7-7.

(22.) Morland KB. Local Food Environments: Food Access in America. xxiv, 323, 8 pages de planches non numerotees.

(23.) Cummins S, Macintyre S. A systematic study of an urban foodscape: The price and availability of food in greater Glasgow. Urban Stud 2002; 39(11):2115-30. doi: 10.1080/0042098022000011399.

(24.) Ver Ploeg M, Breneman V, Dutko P, Williams R, Snyder S, Dicken C, Kaufman P. Access to affordable and nutritious food: Updated estimates of distance to supermarkets using 2010 data, ERR-143. Washington, DC: U.S. Department of Agriculture, Economic Research Service, November 2012.

(25.) Bitler M, Haider SJ. An economic view of food deserts in the United States. J Policy Anal Manage 2011; 30(1):153-76. doi: 10.1002/pam.20550.

(26.) Smith C, Morton LW. Rural food deserts: Low-income perspectives on food access in Minnesota and Iowa. J Nutr Educ Behav 2009; 41(3):176-87. PMID: 19411051. doi: 10.1016/j.jneb.2008.06.008.

(27.) Hartley D, Anderson N, Fox K, Lenardson J. How does the rural food environment affect rural childhood obesity? Child Obes 2011; 7(6):450-61.

(28.) Lucan SC, Gustafson A, Jilcott Pitts SB. The concept of "rural food deserts" is still meaningful. Child Obes 2012; 8(5):484-85. PMID: 23061503. doi: 10. 1089/chi.2012.0063.

(29.) Black J. Local food environments outside of the United States--a look to the north: Examining food environments in Canada. In: Morland KB (Ed.), Local Food Environments: Food Access in America. Boca Raton, FL: Taylor and Francis Group, 2015; 231-69.

(30.) Sadler RC, GillilandJA, Arku G. An application of the edge effect in measuring accessibility to multiple food retailer types in Southwestern Ontario, Canada. IntJHealth Geogr 2011; 10:34. PMID: 21575162. doi: 10.1186/1476-072X-10-34.

(31.) USDA, E. Food Access Research Atlas, 2012.

(32.) Mader E, Busse H. Hungry in the heartland: Using community food systems as a strategy to reduce rural food deserts. J Hunger Environ Nutr 2011; 6(1):45-53. doi: 10.1080/19320248.2011.549377.

(33.) Cummins S. Food Deserts. The Wiley Blackwell Encyclopedia of Health, Illness, Behavior, and Society, 2014.

(34.) Berque A. Ecoumene: introduction a l'etude des milieux humains. Paris: Belin, 2000; 271.

(35.) Liu QA, Ryan V, Aurbach H, Besser T. The influence of local church participation on rural community attachment. Rural Sociol 1998; 63(3):432.

(36.) Pampalon R, Raymond G. A deprivation index for health and welfare planning in Quebec. Chronic Dis Can 2000; 21(3):104-13. PMID: 11082346.

(37.) Frank LD, Andresen MA, Schmid TL. Obesity relationships with community design, physical activity, and time spent in cars. Am J PrevMed 2004; 27(2):87-96. PMID: 15261894. doi: 10.1016/j.amepre.2004.04.011.

(38.) Pouliot N, Hamelin A-M. Disparities in fruit and vegetable supply: A potential health concern in the greater Quebec City area. Public Health Nutr 2009; 12(11):2051-59. PMID: 19344543. doi: 10.1017/S1368980009005369.

(39.) Buck-McFadyen EV. Rural food insecurity: When cooking skills, homegrown food, and perseverance aren't enough to feed a family. Can J Public Health 2015; 106(3):e140-46. PMID: 26125240. doi: 10.17269/cjph.106.4837.

Alexandre Lebel, PhD, [1,2] David Noreau, MSc, [1,3] Lucie Tremblay, RDN, MSc, [4] Celine Oberle, MSc, [1] Maurie Girard-Gadreau, MSc, [1] Mathieu Duguay, MSc, [1] Jason P. Block, MD, PhD [5]

Author Affiliations

[1.] Graduate School of Land Management and Regional Planning, Laval University, Quebec, QC

[2.] Evaluation Platform on Obesity Prevention, Quebec's Heart and Lung Institute, Laval University, Quebec, QC

[3.] Ministere des Affaires municipales et de l'Occupation du territoire, Quebec, QC

[4.] Regional Public Health Authority, Chaudiere-Appalaches, Ste-Marie de Beauce, QC

[5.] Obesity Prevention Program, Department of Population Medicine, Harvard Medical School, Harvard Pilgrim Health Care Institute, Boston, MA, USA

Correspondence: Alexandre Lebel, PhD, Graduate School of Land Management and Regional Planning, Laval University, Pavillon Felix-Antoine-Savard, Office 1616, 2325, allee des Bibliotheques, Quebec, QC G1V 0A6, Tel: 418-656-7685, E-mail: alexandre. lebel@esad.ulaval.ca

Acknowledgements: This research was partly funded by the Fonds de recherche du Quebec-Sante, the Graduate School of Land management and Regional Planning of the Laval University, the Evaluation Platform on Obesity Prevention of the Quebec's Heart and Lung Institute, and the Regional Public Health Authority, Chaudiere-Appalaches. The authors would like to acknowledge members of the Regional Food Security Committee of Chaudiere-Appalaches for sharing their knowledge.

Conflict of Interest: None to declare. Table 1. Regional characteristics to identify potential food deserts for the ecumene model and US Department of Agriculture (USDA) criteria Ecumene model Robert-Cliche n % Population 19,288 14.8% Number of residences 6500 14.3% Located in a deprived comunity 1898 29.2% Located beyond 16 km of a high store * 181 2.8% Located in a potential food desert 24 0.4% Mean distance to closest high store * 5.4 km Median distance to closest high store * 3.2 km USDA criteria n % Population in derived census tract 0 0.0% Population beyond 16 km supermarket 110 3.9% Population in potential food desert 0 0.0% Ecumene model Etchemins n % Population 17,245 13.2% Number of residences 6850 15.1% Located in a deprived comunity 4711 68.8% Located beyond 16 km of a high store * 2770 40.4% Located in a potential food desert 2211 32.3% Mean distance to closest high store * 14.8 km Median distance to closest high store * 14.4 km USDA criteria n % Population in derived census tract 3370 31.9% Population beyond 16 km supermarket 196 6.9% Population in potential food desert 0 0.0% Ecumene model Beauce-Sartigan n % Population 50,962 39.0% Number of residences 16,727 36.8% Located in a deprived comunity 5021 30.0% Located beyond 16 km of a high store * 3188 19.1% Located in a potential food desert 2339 14.0% Mean distance to closest high store * 7.2 km Median distance to closest high store * 2.2 km USDA criteria n % Population in derived census tract 2462 23.3% Population beyond 16 km supermarket 745 26.2% Population in potential food desert 0 0.0% Ecumene model Appalaches n % Population 43,120 33.0% Number of residences 15,323 33.8% Located in a deprived comunity 3661 23.9% Located beyond 16 km of a high store * 725 4.7% Located in a potential food desert 541 3.5% Mean distance to closest high store * 5.3 km Median distance to closest high store * 2.3 km USDA criteria n % Population in derived census tract 4720 44.7% Population beyond 16 km supermarket 1797 63.1% Population in potential food desert 686 100% Ecumene model Region Total % Population 130,615 100% Number of residences 45,400 34.8% Located in a deprived comunity 15,291 11.7% Located beyond 16 km of a high store * 6864 5.3% Located in a potential food desert 5115 3.9% Mean distance to closest high store * 7.4 km Median distance to closest high store * 3.5 km USDA criteria Total % Population in derived census tract 10,552 8.1% Population beyond 16 km supermarket 2848 2.2% Population in potential food desert 686 0.5% * Affordable and selling fresh, diverse and high-quality food products.