Abstract
Accessibility of healthy food is an important predictor for several health outcomes, but its association with life expectancy is unclear. We evaluated the association between U.S. Department of Agriculture’s Food Research Atlas measures of healthy food accessibility and life expectancy at birth across contiguous U.S. census tracts using spatial modeling analysis. Both income and healthy food accessibility were associated with life expectancy at birth, as indicated by shorter life expectancy in low-income census tracts when comparing tracts with similar healthy food accessibility level, and in low-access tracts when comparing tracts with similar income level. Compared to high-income/high-access census tracts, life expectancy at birth was lower in high-income/low-access (− 0.33 years; 95% confidence interval − 0.42, − 0.28), low-income/high-access (− 1.45 years; − 1.52, − 1.38), and low-income/low-access (− 2.29 years; − 2.38, − 2.21) tracts after adjusting for socio-demographic characteristics and incorporating vehicle availability. Effective interventions to increase healthy food accessibility may improve life expectancy.
Supplementary Information
The online version contains supplementary material available at 10.1007/s11524-023-00742-x.
Introduction
Life expectancy at birth varies substantially in the U.S., ranging from 56.3 years to 97.5 years across census tracts in 2010–2015 [1], due to numerous factors, including differences in living environment, access to appropriate health care, and exposure to risk factors, such as poor diet [2]. Healthy diet at any age is important for living a healthier and longer life [3, 4]. In early ages, moreover, type of diet may shape lifelong eating behavior [5]. Limited healthy food accessibility is a potential contributor to poor diet and has been associated with some adverse health outcomes, including heart disease, cancer, and diabetes [6, 7]. Poor diet has also been linked to obesity, a major risk and comorbidity factor to all these adverse outcomes [8, 9]. However, it remains unknown to what extent healthy food accessibility is associated with life expectancy. In this study we evaluate the association between healthy food accessibility and life expectancy at birth in census tracts across contiguous U.S. (District of Columbia and all states excluding Alaska and Hawaii).
Methods
Most recent data on life expectancy at birth by census tract (2012–2015) were obtained from the Centers for Disease Control and Prevention [10]. Sociodemographic characteristics (education level and racial/ethnic composition) by census tract (2011–2015) were based on the American Community Survey data [11]. To measure healthy food accessibility, we used the U.S. Department of Agriculture’s (USDA’s) Food Research Atlas (2015) indices on income/poverty level and distance-based geographic access to healthy food for each census tract. Low-income census tracts were defined as tracts with a poverty rate ≥ 20% or a median household income ≤ 80% of the state-wide/metropolitan area-wide median household income [12]. Low-access census tracts were defined as tracts with ≥ 500 people or a share ≥ 33% of the total population living > 1 mile from the nearest supermarket, supercenter, or large grocery store for an urban area or > 10 miles for a rural area [12]. We then defined 4 categories of census tracts with (1) high income and high healthy food accessibility (high-income/high-access), (2) high income and low healthy food accessibility (high-income/low-access), (3) low income and high healthy food accessibility (low-income/high-access), and (4) low income and low healthy food accessibility (low-income/low-access).
Previous studies have reported that 20% of U.S. households experience major transportation barriers to accessing healthy foods [13]. Thus, we also repeated the analyses using a stricter definition for low healthy food accessibility from the USDA’s Food Research Atlas as an alternative measure, in which low-access census tracts (both urban and rural) were defined as tracts where ≥ 100 households were > 0.5 mile away from the nearest supermarket and had no access to a vehicle; or ≥ 500 people or ≥ 33% of the population lived > 20 miles from the nearest supermarket, regardless of vehicle availability [12].
Spatial Bayesian regression models (R package BayesX) [14] were used to calculate regression coefficients and corresponding 95% confidence intervals (CIs) for associations between healthy food accessibility and life expectancy at birth, with adjustments for percentage of individuals with a high school education or less (as a proxy for lower socioeconomic status), percentage of non-White population (as a proxy for racial residential segregation), and population density. Spatial effect was incorporated based on an adjacency matrix of geographic neighbors (weights based on rook's case) for each census tract. Maps of average life expectancy and healthy food accessibility were generated using ArcGIS Pro version 3.1.0.
Results
After excluding 6,884 census tracts (9.3%) with missing data, 67,138 census tracts in contiguous U.S. were included in this analysis (Table 1). The average life expectancy at birth was higher in high-income/high-access (80.0 years) and high-income/low-access (80.2) census tracts than in low-income/high-access (76.8) and low-income/low-access (76.1) census tracts. Census tracts with low healthy food accessibility were predominantly concentrated in the Southeast and West, where an excessive number of tracts with lower life expectancy were located as well (Supplemental Figure A1).
Table 1.
Association between healthy food accessibility and life expectancy at birth in the United States
Healthy food accessibility | No. (%) of census tracts |
Average life expectancy at birth, year (SD)a | Regression coefficient, baseline (95% CI)b | Regression coefficient, fully adjusted (95% CI)c |
---|---|---|---|---|
Overalld,e | ||||
High income/high access | 21,850 (32.5) | 80.0 (0.01) | Reference | Reference |
High income/low access | 16,943 (25.2) | 80.2 (0.01) | 0.52 (0.46, 0.57) | 0.28 (0.21, 0.34) |
Low income/high access | 19,878 (29.6) | 76.8 (0.01) | − 3.65 (− 3.71, − 3.58) | − 1.56 (− 1.62, − 1.49) |
Low income/low access | 8,467 (12.6) | 76.1 (0.01) | − 3.40 (− 3.50, − 3.32) | − 1.58 (− 1.66, − 1.50) |
Stricter definition for low healthy food accessibilityf,g | ||||
High income/high access | 33,263 (49.5) | 80.2 (0.01) | Reference | Reference |
High income/low access | 5,530 (8.2) | 79.4 (0.01) | − 0.71 (− 0.81, − 0.62) | − 0.33 (− 0.42, − 0.24) |
Low income/high access | 18,133 (27.0) | 77.2 (0.01) | − 3.49 (− 3.55, − 3.44) | − 1.45 (− 1.52, − 1.38) |
Low income/low access | 10,212 (15.2) | 75.5 (0.01) | − 4.61 (− 4.67, − 4.54) | − 2.29 (− 2.37, − 2.22) |
SD, standard deviation; CI, confidence interval
aBased on average life expectancy at birth values for each census tract weighted by the census tract population
bModels adjusted for spatial effect, population density
cModels adjusted for spatial effect, population density, % population aged ≥ 18 years with high school education or less (as a proxy of lower socioeconomic status), and % non-White population (as a proxy for racial residential segregation)
dLow-income census tracts are defined as tracts with a poverty rate ≥ 20% or a median household income ≤ 80% of the state-wide/metropolitan area-wide median family income
eLow-access census tracts excluding vehicle access are defined as tracts with ≥ 500 people or a share ≥ 33% of the total population living > 1 mile from the nearest supermarket, supercenter, or large grocery store for an urban area or > 10 miles for a rural area
fLow-income census tracts in the alternate measure of food accessibility are defined the same as the original measure excluding vehicle access, consistent with (d)
gLow-access census tracts including vehicle access were defined as a tract where ≥ 100 households are > 0.5 mile from the nearest supermarket and have no access to a vehicle; or ≥ 500 people or ≥ 33% of the population live > 20 miles from the nearest supermarket, regardless of vehicle access
In models adjusted for all sociodemographic covariates, compared to high-income/high-access census tracts, life expectancy at birth was lower in low-income/low-access tracts (− 1.58 years; 95%CI: − 1.66, − 1.50) and in low-income/high-access tracts (− 1.56 years; 95%CI: − 1.62, − 1.49), but it was slightly higher in high-income/low-access tracts, likely affluent suburban areas (0.28 years; 95%CI: 0.21, 0.34) (Table 1). However, when using the stricter definition for low healthy food accessibility, which also included vehicle availability or living > 20 miles away from a healthy food store, life expectancy at birth was lower in other 3 groups compared to the reference group. With the stricter definition, life expectancy at birth was lowest in low-income/low-access tracts (− 2.29 years; 95%CI: − 2.37, − 2.22), followed by low-income/high-access (− 1.44 years; 95%CI: − 1.52, − 1.38), and high-income/low-access (− 0.33 years; 95%CI: − 0.42, − 0.24) tracts.
Discussion
In this nationwide study, we found that life expectancy at birth was shorter in census tracts with a lower median household income or higher poverty level, consistent with previous literature [15]. However, shorter life expectancy at birth was also associated with low healthy food accessibility when considering the stricter measure that included vehicle availability and longer distances to a healthy food store.
Previous studies have reported an association between poor accessibility to healthy food and poorer health or adverse health outcomes. For example, in a cross-sectional study of young adults aged 24–34 years from the National Longitudinal Study of Adolescent to Adult Health, living in a food desert was associated with poor cardiovascular health as defined by the American Heart Association, which included body mass index, smoking, physical activity, diet, blood pressure, blood glucose and cholesterol [7]. A cohort study from the California Cancer Registry, comprised of individuals diagnosed with stage II to III breast or colorectal cancer, found that cancer-specific mortality among individuals with breast cancer and colorectal cancer was higher in food desert census tracts as compared to non-food desert tracts after controlling for treatment, health risks, socioeconomic status, race, and health insurance [16].
Conversely, some studies have shown that an association between living in areas with poor healthy food accessibility and poorer health outcomes disappeared after adjusting for income or price of food [7, 17]. Some other studies have not found any associations between healthy food accessibility and health outcomes [18–20]. In our study, however, healthy food accessibility was associated with life expectancy as indicated by shorter life expectancy in low-access tracts, even when comparing tracts with similar income level, when vehicle availability was considered. Moreover, this association persisted even with the inclusion of census tract education level (a proxy for socioeconomic status), percent non-White population, and population density (a proxy for greater availability of resources and services) in the model. The differences in associations between healthy food accessibility and health outcomes across studies could be due to the variation in measures or methodologies between studies, as well as possible true differences in the association between healthy food accessibility and various health outcomes.
Several initiatives have been introduced to increase healthy food accessibility in low-income/low-access areas. These initiatives encompass establishing new healthy food retailers (e.g., farmers markets, grocery stores, bodegas, and mobile retailers), as well as improving the quality, diversity, and quantity of healthier foods at current stores [21]. Supporting public transportation systems may also improve healthy food accessibility, particularly in rural areas and communities with small populations [13], and among people with lower incomes and racial/ethnic minority groups who may have limited access to personal vehicles [22]. Changes to public policies to increase the affordability of healthy foods may also be beneficial to individuals with limited income, especially people using food assistance programs; e.g. improvements to Supplemental Nutrition Assistance Program (SNAP) to incentivize purchasing of healthy foods [23].
To our knowledge, this is the first nationwide study estimating associations between healthy food accessibility and most recent life expectancy at birth estimates at the census tract level – the most granular data aggregation level for life expectancy in the U.S. – while accounting for census tract income level. A major limitation of this study is that we were not able to examine the association of individual-level income and healthy food accessibility with life expectancy, indicating the need for further research. Another limitation is the assumption that healthy food would be available at the same level in all healthy food stores, while the availability may be lower in stores in predominantly Black and lower-income neighborhoods than in other neighborhoods [24]. Moreover, we were not able to take into account some other factors that my influence accessibility to healthy foods, such as shopping behavior (e.g., willingness to travel longer distances to purchase healthy foods) and availability of reliable and efficient public transportation, due to the lack of data.
Results of this study suggest an association between poor healthy food accessibility and shorter life expectancy at birth at the census tract level in the U.S. Effective interventions to increase healthy food accessibility may improve diet quality and population health.
Supplementary Information
Below is the link to the electronic supplementary material.
Funding
There was no external sponsor.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Boing AF, Boing AC, Cordes J, Kim R, Subramanian SV. Quantifying and explaining variation in life expectancy at census tract, county, and state levels in the United States. Proc Natl Acad Sci. 2020;117:17688–17694. doi: 10.1073/pnas.2003719117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ekmekcioglu C. Nutrition and longevity - From mechanisms to uncertainties. Crit Rev Food Sci Nutr. 2020;60:3063–3082. doi: 10.1080/10408398.2019.1676698. [DOI] [PubMed] [Google Scholar]
- 3.Schlenker ED, Feurig JS, Stone LH, Ohlson MA, Mickelsen O. Nutrition and health of older people. Am J Clin Nutr. 1973;26:1111–1119. doi: 10.1093/ajcn/26.10.1111. [DOI] [PubMed] [Google Scholar]
- 4.Black M, Bowman M. Nutrition and healthy aging. Clin Geriatr Med. 2020;36:655–669. doi: 10.1016/j.cger.2020.06.008. [DOI] [PubMed] [Google Scholar]
- 5.Scaglioni S, De Cosmi V, Ciappolino V, Parazzini F, Brambilla P, Agostoni C. Factors influencing children’s eating behaviours. Nutrients. 2018;10:706. doi: 10.3390/nu10060706. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.McCullough M, Chantaprasopsuk S, Islami F, Um C, Rees-Punia E, Wang Y, Leach C, Sullivan L, Patel A. Socioeconomic and geographic predictors of poor diet quality in a large U.S. cohort of adult men and women. Cancer Epidemiol Biomarkers Prev. 2022;31:1512–1513. doi: 10.1158/1055-9965.EPI-22-0479. [DOI] [Google Scholar]
- 7.Testa A, Jackson DB, Semenza DC, Vaughn MG. Food deserts and cardiovascular health among young adults. Public Health Nutr. 2021;24:117–124. doi: 10.1017/S1368980020001536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Cooksey-Stowers K, Schwartz MB, Brownell KD. Food swamps predict obesity rates better than food deserts in the United States. Int J Environ Res Public Health. 2017;14(11):1366. 10.3390/ijerph14111366. [DOI] [PMC free article] [PubMed]
- 9.Centers for Disease Control and Prevention. Fast stats: leading causes of death. 2022. https://www.cdc.gov/nchs/fastats/leading-causes-of-death.htm. Accessed 20 Sept 2022.
- 10.Arias E, Escobedo LA, Kennedy J, Fu C, Cisewki J. U.S. Small-area life expectancy estimates project: methodology and results summary. Vital Health Stat 2. 2018;(181):1–40. [PubMed]
- 11.U.S. Census Bureau. American Community Survey (ACS) 2011–2015. 5-Year Estimates. https://www.socialexplorer.com/tables/ACS2015_5yr. Obtained from Social Explorere. Accessed 20 Sept 2022.
- 12.U.S. Department of Agriculture. Food Access Research Atlas. https://www.ers.usda.gov/data-products/food-access-research-atlas/. Accessed 20 Sept 2022.
- 13.Dumas BL, Harris DM, McMahon JM, Daymude TJ, Warnock AL, Moore LV, Onufrak SJ. Prevalence of municipal-level policies dedicated to transportation that consider food access. Prev Chronic Dis. 2021;18:E97. doi: 10.5888/pcd18.210193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Belitz C, Brezger A, Kneib T, Lang S. BayesX - Software for bayesian inference in structured additive regression models. Version 2.0.1. 2011. https://www.uni-goettingen.de/de/bayesx/550513.html.
- 15.Chetty R, Stepner M, Abraham S, Lin S, Scuderi B, Turner N, Bergeron A, Cutler D. The association between income and life expectancy in the United States, 2001–2014. JAMA. 2016;315:16. doi: 10.1001/jama.2016.4226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Fong AJ, Lafaro K, Ituarte PHG, Fong YA-O. Association of living in urban food deserts with mortality from breast and colorectal cancer. Ann Surg Oncol. 2021;28:8. doi: 10.1245/s10434-020-09049-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ghosh-Dastidar B, Cohen D, Hunter G, et al. Distance to store, food prices, and obesity in urban food deserts. Am J Prev Med. 2014;47(5):587–595. 10.1016/j.amepre.2014.07.005. [DOI] [PMC free article] [PubMed]
- 18.Zhen C. Food Deserts: myth or reality? Annu Rev Resour Economics. 2021;13:109–129. doi: 10.1146/annurev-resource-101620-080307. [DOI] [Google Scholar]
- 19.Izeogu C, Idigo AJ, Howard AD, Lyerly MJ. Abstract TP195: geographic proximity to healthy food not associated with increased prevalence of cardiovascular risk factors in stroke patients. Stroke 2019;50:ATP195-ATP195.
- 20.Kelli HM, Kim JH, Samman Tahhan A, et al. Living in food deserts and adverse cardiovascular outcomes in patients with cardiovascular disease. J Am Heart Assoc. 2019;8(4):e010694. 10.1161/JAHA.118.010694. [DOI] [PMC free article] [PubMed]
- 21.Centers for Disease Control and Prevenstion (CDC). Increase Access to Healthy Foods and Beverages. https://www.cdc.gov/nccdphp/dnpao/state-local-programs/nutrition.html. Accessed 20 Sept 2022.
- 22.Klein NJ, Smart MJ. Car today, gone tomorrow: the ephemeral car in low-income, immigrant and minority families. Transportation. 2017;44:495–510. doi: 10.1007/s11116-015-9664-4. [DOI] [Google Scholar]
- 23.Centers for Disease Control and Prevenstion (CDC). State Initiatives Supporting Healthier Food Retail: An Overview of the National Landscape. https://www.cdc.gov/obesity/downloads/healthier_food_retail.pdf. Accessed 20 Sept 2022.
- 24.Franco M, Diez Roux AV, Glass TA, Caballero B, Brancati FL. Neighborhood characteristics and availability of healthy foods in Baltimore. Am J Prev Med. 2008;35(6):561–567. 10.1016/j.amepre.2008.07.003. [DOI] [PMC free article] [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.