Abstract
Objective
Little is known about neighborhood retail food environments and cardiometabolic health among Hispanic/Latino adults, who experience significant health inequities. This study aimed to examine associations of the neighborhood retail food environment with cardiometabolic health of Hispanic/Latino adults over a 6-year period.
Methods
Participants from the Hispanic Community Health Study/Study of Latinos San Diego Field Center (n = 3851) underwent assessments of cardiometabolic risk factors at baseline (2008–2011) and approximately 6 years later (2014–2017). Weight (body mass index [BMI] and overweight/obesity), hypertension (normotensive, prehypertensive, or hypertensive), and diabetes (normoglycemia, prediabetes, or diabetes) status were considered cardiometabolic risk factors. Participants’ home addresses at baseline and 6 years later were geocoded. The neighborhood retail environment was quantified using the Modified Retail Food Environment Index (mRFEI) within 800-m circular buffers around these two addresses. Complex survey regression analyses quantified associations between both baseline and changes in neighborhood retail food environment and cardiometabolic health of Hispanic/Latino adults over 6 years.
Results
A one standard deviation higher baseline standardized mRFEI score was associated with approximately 14.3% lower odds of transitioning from a healthy weight to overweight or from overweight to obesity (OR = 0.857, 95%CI [0.738, 0.994], p < 0.05) at year 6. No significant associations were found between the mRFEI and hypertension, diabetes status, or BMI.
Conclusions
A healthier retail food environment may be associated with healthier weight profiles among Hispanic/Latino adults over time, suggesting that improving healthy neighborhood food outlet availability may contribute to combating obesity among the Hispanic/Latino population.
Keywords: Food establishment, Social environment, Weight status, Diabetes, Hypertension
Introduction
Hispanic/Latino adults in the United States face significant health inequities [1]. For example, nearly half (45.6%) of Hispanic/Latino adults are classified with obesity [2]. Obesity contributes to a greater risk of developing type 2 diabetes, which affects Hispanic/Latino adults at more than double the rate of non-Hispanic White adults [3]. Obesity also contributes to a greater risk of developing hypertension, further exacerbating the risk of cardiovascular diseases within this population [4, 5].
The neighborhood retail food environment has been associated with adults’ weight status in previous studies, although few studies in this area reported results for Hispanic/Latino populations [6, 7]. In general, studies showed the neighborhood food environment plays an important role in individuals’ eating behaviors and energy intake [8]. For example, living near supermarkets and large grocery stores that provide high-quality healthy foods (e.g., fresh produce, whole grains, lean proteins, dairy) at relatively lower prices was associated with healthier diet quality and lower rates of obesity [9]. Better access to convenience stores that mainly provided unhealthy food (e.g., high-calorie, low nutrient density chips, candies, processed foods) has been positively related to overweight/obesity [10–12]. The presence of fast-food restaurants serving high-calorie/saturated-fat foods within residential neighborhoods has been associated with greater obesity prevalence [8]. Results are mixed on the relation between neighborhood food environment and type 2 diabetes mellitus, with some recent studies reporting no association [13, 14].
Research examining the relation between the neighborhood retail food environment and cardiometabolic health has shown that neighborhoods with a higher proportion of residents from marginalized racial and ethnic groups or with low incomes have less access to healthier food retail stores (i.e., grocery stores/supermarkets) and more access to relatively less-healthy food stores (i.e., convenience stores, fast food restaurants). Such disparities in the retail food environment may contribute to the disproportionate burden of obesity and related cardiometabolic conditions observed across racial and ethnic groups [7, 8, 15, 16]. However, the majority of the existing evidence on the relation of retail food environment to cardiometabolic health is from cross-sectional studies, which limits understanding of temporal precedence [7]. To strengthen causal inference, prospective studies are critically needed to clarify how neighborhood food environments influence cardiometabolic health over time.
The purpose of the current study was to assess associations of baseline and changes in neighborhood retail food environment with multiple cardiometabolic health indicators (i.e., weight status, hypertension status, diabetes status) in Hispanic/Latino adults over 6 years. We hypothesized that participants with healthier neighborhood retail food environments based on the availability of grocery stores/supermarkets at baseline or who experienced greater improvement in their retail food environments would have healthier weight changes and be more likely to maintain healthier levels of blood pressure and glycemic regulation over the 6-year period.
Methods
Study Population
The Hispanic Community Health Study/Study of Latinos (HCHS/SOL) is a longitudinal study involving 16,415 Hispanic/Latino adults aged 18 to 74 years at the time of screening. The present analysis utilized data from the San Diego field center. The target population in San Diego County was from the South Bay area, which is bordered by the Pacific Ocean and San Diego Bay to the west, and the U.S.-Mexico border to the south. This region comprises a blend of residential neighborhoods, commercial zones, businesses, shipyards, and recreational areas.
The methods and sampling for the HCHS/SOL have been previously detailed [17]. Participants had a baseline examination (visit 1) between 2008 and 2011, which included assessment of cardiometabolic health. Participants were contacted annually by telephone to identify clinical events and invited to attend a second examination roughly 6 years after the baseline (2014–2017, visit 2). The neighborhood retail food environment measures were collected as part of the Study of Latinos Community and Surrounding Areas Study (SOL CASAS) ancillary study (2015–2020) [18]. The current analysis used HCHS/SOL data from visit 1 and visit 2 that are tied to the SOL CASAS ancillary study with geocoded addresses for San Diego participants [18]. The current sample was 3851 San Diego HCHS/SOL participants for whom neighborhood (at visit 1 and visit 2) and cardiometabolic health outcomes (i.e., weight status, diabetes status, and hypertension status) (at visits 1 and 2) variables were collected. IRB approval was obtained from participating institutions, and all participants provided written informed consent. The analyses were conducted in 2024.
Measures
Three cardiometabolic health constructs were examined using data from visits 1 and 2: weight status, diabetes status, and hypertension status. Weight status was measured by BMI, calculated using objectively measured height and weight from clinical examinations, and categorized based on CDC criteria as healthy weight (18.5–24.9 kg/m2), overweight (25–29.9 kg/m2), or obesity (≥ 30 kg/m2). Diabetes status was determined through self-reported diagnosis, fasting plasma glucose (FPG), HbA1c, and a 2-h oral glucose tolerance test [19]. Hypertension status was determined by averaging three blood pressure measurements and self-reported diagnosis, with hypertension defined as SBP ≥ 140 mm Hg, DBP ≥ 90 mm Hg, or use of antihypertensive medication. Pre-hypertension was categorized as SBP 120–139 mm Hg or DBP 80–89 mm Hg, while normotension was classified as SBP < 120 mm Hg and DBP < 80 mm Hg [20]. We assessed the status of these variables at Visit 1 and examined changes at Visit 2 (e.g., progression from normotension to pre-hypertension; progression from overweight to obesity).
Participant home addresses at visits 1 and 2 were geocoded using SAS/GRAPH v9.3 (SAS Institute Inc., Cary, NC) geoprocessing procedures and U.S. Census Bureau TIGER/Line Shapefiles. Participants’ geocoded home locations were buffered using 800-m circular buffers. This buffer size was selected to model realistic walking distances from the home and to maximize environmental variability between participants by minimizing buffer overlap as the sample is concentrated in the southeastern part of San Diego County. The variables were computed in ArcGIS 10.5 (ESRI 2017, Redlands CA) or Google Earth Engine (Google 2017, Mountain View CA). Neighborhood retail food environment data were sourced from the California Health and Human Services (CHHS) for the years 2008–2009 and 2016–2017.
The key variable, the modified Retail Food Environment Index (mRFEI) [21], measures the number of retailers that provide more healthy food options (i.e., supermarkets and large grocery stores, fruit and vegetable markets) and retailers that have fewer healthy food options (i.e., convenience stores, small grocery stores, fast food restaurants). Out of the total number of food retailers considered healthy or less healthy, the mRFEI represents the percentage that are healthy and ranges from 0 to 100. Lower scores indicate there are more nonhealthy food locations than healthy food locations, and zero indicates no healthy food options.
Covariates included participant age, sex, educational attainment (less than high school/GED, high school/GED, or greater than high school/GED), household income (< $10,000, $10,000—$20,000, $20,000—$40,000, $40,000—$75,000, > $75,000, or did not respond), place of birth (Born in the U.S. or not), whether the participant moved personal residences between visits, and time between the Visit 1 and Visit 2 assessments, and neighborhood socioeconomic deprivation at visit 1, represented as a composite score calculated using principal components analysis (PCA) in SPSS (version 27.0). This composite was calculated for each participant’s 800-m buffer and included percentages of adults without a high-school diploma, unemployed adults, rented and crowded households, households in poverty, low-income households (≤ $30,000/year), female-headed households with children, households receiving public assistance, and those with public health insurance, as determined using census data.
Descriptive analyses were calculated using the complex sampling features in STATA to account for the HCHS/SOL sampling design and SOL CASAS sampling weights. To address missing data, inverse probability weights were calculated based on sociodemographic and individual characteristics detailed in the measures section, with the final weight being the product of the inverse probability and sampling weights [22]. Additionally, imputation was applied to account for missing data on covariates. Survey-weighted generalized linear models (GLMs) were used, including linear regression for continuous outcomes (BMI) and ordinal logistic regression for ordered categorical outcomes (3-level weight, diabetes, and hypertension status). Those models were used to investigate associations of the visit 1 neighborhood retail food environment and changes in the neighborhood retail food environment with visit 2 cardiometabolic health variables, adjusted for Visit 1 cardiometabolic health variables as additional covariates. For the ordinal models, the proportional odds assumption was tested and met, supporting the use of this approach. To assess robustness, we conducted sensitivity analyses using multinomial logistic regression, and results were consistent in direction and magnitude with those from the ordinal models. Changes in the retail food environment were calculated using a residualized change score from regressing the visit 2 neighborhood retail food environment values on the visit 1 neighborhood retail food environment values. Standardized regression coefficients were presented in addition to unstandardized coefficients to facilitate comparison of effect sizes across linear models, and odds ratios were reported for the ordinal models.
Results
Tables 1 and 2 show the descriptive statistics for the study population. About half of the population was female and did not have health insurance, with approximately two-thirds earning less than $40,000 per year. At baseline, the average BMI was 29.1 kg/m2 and ~ 38% of the population had obesity. Additionally, at baseline, 19% of the population had hypertension and 15% had diabetes. By visit 2, the average BMI had increased to 29.6 kg/m2 with ~ 40% of the population having obesity. At visit 2, 26% of the population met criteria for hypertension and 28% for diabetes. Approximately 13% of the population moved up one weight category (i.e., normal weight to overweight or overweight to obesity), and 0.5% moved up two (i.e., normal weight to obesity). For hypertension, 18% increased by one category (i.e., normotension to prehypertension or prehypertension to hypertension), and 4% increased by two (i.e., normotensive to hypertensive). For diabetes, 27% moved up one category (i.e., normoglycemia to prediabetes or prediabetes to diabetes), and 4% increased by two (i.e., normoglycemia to diabetes).
Table 1.
Baseline sociodemographic characteristics, neighborhood environment characteristics, and cardiometabolic health among Hispanic/Latino adults in HCHS/SOL and SOL CASAS
| Characteristics | n | Weighted % or weighted M (SD) |
|---|---|---|
| Visit 1 sociodemographic characteristics | ||
| Age, years | 3851 | 39.4 (0.5) |
| Female, % | 3851 | 53.3 |
| Born in the U.S. 50 states or DC % | 892 | 31.8 |
| Less than high school educational attainment, % | 1348 | 28.3 |
| Annual household income, $ | 3653 | |
| < 10,000 | 422 | 9.9 |
| 10,001–20,000 | 971 | 24.1 |
| 20,001–40,000 | 1352 | 35.2 |
| 40,001–75,000 | 676 | 21.2 |
| > 75,000 | 232 | 9.7 |
| Has health insurance, % | 1861 | 47.1 |
| Movers between Visit 1 and Visit 2 | 1123 | 27.9 |
| Visit 1 neighborhood environment characteristics | ||
| Modified retail food environment index (800 m) | 3851 | 17.3 (0.6) |
| Socioeconomic deprivation index (800 m) | 3851 | .8 (0.1) |
| Visit 1 cardiometabolic health | ||
| Weight status (BMI, kg/m2) | 3842 | 29.1 (0.2) |
| Normal weight category | 768 | 23.5 |
| Overweight category | 1457 | 38.9 |
| Obesity category | 1617 | 37.6 |
| Hypertension status | 3850 | |
| Normotension | 1872 | 54.9 |
| Prehypertension | 997 | 26.7 |
| Hypertension | 981 | 18.5 |
| Diabetes status | 3850 | |
| Normoglycemia | 1647 | 51.0 |
| Prediabetes | 1424 | 34.2 |
| Diabetes | 779 | 14.8 |
Note: HCHS/SOL, Hispanic Community Health Study/Study of Latinos; SOL CASAS, Study of Latinos Community and Surrounding Areas Study; BMI, body mass index
Table 2.
Visit 2 neighborhood environment characteristics, and cardiometabolic health among Hispanic/Latino adults in HCHS/SOL and SOL CASAS
| Characteristics | n | Weighted % Or weighted M (SD) |
|---|---|---|
| Visit 2 neighborhood environment characteristics | ||
| Modified retail food environment index (800 m) | 3758 | 23.1 (0.7) |
| Socioeconomic deprivation index (800 m) | 3758 | .8 (0.1) |
| Visit 2 cardiometabolic health | ||
| Weight status (BMI) | 2794 | 29.6 (0.3) |
| Normal weight category | 479 | 19.7 |
| Overweight category | 1111 | 39.9 |
| Obese category | 1204 | 40.4 |
| Hypertension status | 2859 | |
| Normotension | 1183 | 50.0 |
| Prehypertension | 668 | 23.6 |
| Hypertension | 1008 | 26.4 |
| Diabetes status | 2858 | |
| Normoglycemia | 810 | 35.4 |
| Prediabetes | 1034 | 36.1 |
| Diabetes | 1014 | 28.4 |
Note: HCHS/SOL, Hispanic Community Health Study/Study of Latinos; SOL CASAS, Study of Latinos Community and Surrounding Areas Study; BMI, body mass index
Complex survey regression analyses (see Table 3) indicated no cross-sectional associations between the baseline retail food environment and baseline cardiometabolic health measures. When examining the data prospectively, a healthier retail food environment at baseline, as measured by a higher mRFEI score, was associated with lower odds of worsening weight status 6 years later (OR = 0.857, p < 0.05), indicating that for each unit increase in the mRFEI score, the odds of moving to a higher weight category decreased by 14.3%. No significant associations were found between the baseline mRFEI and hypertension or diabetes status 6 years later.
Table 3.
Cross-sectional and prospective associations between neighborhood food environment characteristics and cardiometabolic health among Hispanic/Latino adults in HCHS/SOL and SOL CASAS
| Neighborhood retail food environment | BMI | Weight status | Hypertension status | Diabetes status | ||||
|---|---|---|---|---|---|---|---|---|
| B (95% CI) | P | OR(95% CI) | P | OR (95% CI) | P | B (95% CI) | P | |
| Cross sectional associationsa | .096 (− 0.255, 0.447) | 0.590 | 1.033 (0.933, 1.143) | 0.538 | 1.10 (0.985, 1.228) | 0.093 | 1.038 (0.943, 1.142) | 0.449 |
| Prospective associations healtha,b | ||||||||
| Baseline environment with changes in cardiometabolic health | − 0.134 (− 0.333, 0.066) | 0.188 | 0.857 (0.738, 0.994)* | 0.044 | 0.862 (0.730, 1.019) | 0.084 | 1.010 (0.865, 1.180) | 0.896 |
| Changes in environment with changes in cardiometabolic health | 0.126 (− 0.045, 0.296) | 0.147 | 1.053 (0.911, 1.218) | 0.484 | 1.055 (0.892, 1.249) | 0.532 | 1.038 (0.918, 1.175) | 0.554 |
BMI, body mass index (measured on a continuous scale using linear regression, coefficients, B, are reported). Weight status, hypertension status, and diabetes status were modeled as ordinal outcomes using ordinal logistic regression; OR represents the association of neighborhood retail food environment with the odds of increasing a category in weight status (i.e., worsening status from healthy weight to overweight or from overweight to obesity, hypertension status (i.e., worsening status from normotensive to prediabetes or from prediabetes to diabetes), or diabetes status (i.e., worsening status from normoglycemia to prediabetes or from prediabetes to diabetes), the variable mRFEI was standardized to a z-score prior to analysis
Models were adjusted for age, sex, Hispanic/Latino background, education, income, place of birth/duration of US residence, and neighborhood socioeconomic deprivation
Models were additionally adjusted for time between visit 1 and visit 2, whether the participant moved between visits, and baseline level of the respective outcome variable (to examine residualized change)
Discussions
The present study is among the first to examine changes in neighborhood retail food environment in relation to changes in cardiometabolic outcomes including weight, hypertension, and diabetes status in Hispanic/Latino communities [23]. The findings indicated that a healthier retail food environment, based on the availability of supermarkets/large grocery stores and/or fruit and vegetable markets relative to few convenience stores, small grocery stores, and fast food restaurants, had a moderate association with having a more favorable weight status over time. However, no significant associations were found between the neighborhood retail food environment and hypertension or diabetes.
Our finding that the availability of relatively healthy food retail outlets at baseline was associated with healthier weight status adds to inconsistent findings from longitudinal research on retail food outlet availability and body weight. For example, the Framingham Heart Study Offspring Cohort Study found a small negative (i.e., beneficial) association between proximity to a grocery store and BMI over time [24]. In contrast, longitudinal data from the Panel Study of Income Dynamics showed mixed results, with a small negative association between grocery store density and BMI for men with lower income, but a slight positive association for certain groups of women, such as those with higher income or without children [25]. Gibson (2011) specifically examined small grocery stores using data from the National Longitudinal Survey of Youth and found a negative (beneficial) association between increased density of small grocery stores and BMI among rural-to-urban movers, but a positive cross-sectional association between small grocery store density and BMI in urban settings, likely due to the prevalence of unhealthy food options [16]. In another study focusing on predominantly Black and Hispanic populations living in low- to moderate-income neighborhoods, small grocery store availability was associated with 1.22-unit higher BMI [26]. However, large grocery store and convenience store availability were not significantly associated with BMI [26]. These differing findings may be due to variations in study designs, population characteristics, and measures of food environments. Unlike previous studies, the current study used the mRFEI, which quantifies the proportion of healthy food outlets relative to all food outlets, offering a measure of the food environment’s relation to weight status.
In contrast to some previous studies, the current study did not find significant associations of baseline or changes in the retail food environment with hypertension or diabetes status. For example, a systematic review found that a greater presence of unhealthy food outlets, such as fast-food restaurants, was associated with higher odds of developing hypertension and diabetes [27]. Another study specifically identified a relation between the density of food outlets in Mexico and the risk of developing diabetes, noting that exposure to a higher density of supermarkets and healthier food outlets was inversely associated with diabetes risk, while exposure to convenience stores and fast-food outlets was positively associated [28]. It is possible that we were not able to detect an association for changes in the retail food environment due to little variation observed at the neighborhood level (e.g., the mean mRFEI score changed only slightly from 17.3 at time 1 to 23.1 at time 2). Another limitation of the food environment measures is we only assessed mRFEI within 800-m buffers. This complicates the interpretation of mRFEI results since some participants may travel to food outlets outside this buffer. More refined measures of food environment characteristics or the use of expanded classifications of healthy and less healthy food outlets might be needed to capture the impact of the food environment on cardiometabolic health outcomes more effectively.
Opportunities exist for urban planning and public health interventions focused on improving access to healthier foods to help support a healthier weight status, particularly in Hispanic/Latino communities [12, 15]. Thus, enhancing access to healthy food outlets such as supermarkets, grocery stores, and farmers’ markets may be important for healthier eating habits. Efforts to improve the healthfulness of food environments may be especially important in underserved areas, where access to healthy, affordable food is often limited, and where the prevalence of fast-food outlets and convenience stores may contribute to higher obesity rates [9, 25]. However, more research is needed, including in Hispanic/Latino communities. Experimental research is particularly important for understanding how such efforts may impact health over time [29].
This study had several strengths, including the use of mREFI measures for a detailed assessment of the neighborhood food environment and evaluation of multiple cardiometabolic variables. The large sample size and 6-year follow-up provide robust data to assess long-term prospective associations. Adjusting for economic deprivation is a key strength, as it ensures the associations detected are independent of factors like neighborhood income.
There were also some limitations to this study. First, the target population was drawn from a geographically focused area in San Diego, which limits the variability in environmental exposures. Specifically, standard deviations of the mRFEI were less than one on a scale of 100. The low scores indicated very few participants had what could be considered healthy food environments near their homes. This geographic concentration reduced the ability to detect significant associations and may reduce the generalizability of our findings to other regions, particularly those with more diverse food environments. Although our study followed participants over a 6-year period, the degree of change in BMI was relatively small, with BMI levels already high on average at baseline. This limited variation in BMI change may have reduced power to detect more robust associations between the neighborhood retail food environment and weight status. Another limitation was that the mRFEI groups small grocery stores as unhealthy, even though they some offer healthy options. We could not separate food categories to assess the relative impact of healthy versus unhealthy sources, which may affect the findings. We also could not fully account for the impact of residential mobility, as 28% of participants moved between visits. The study did not address potential endogeneity, as healthier and more affluent individuals may self-select into better-resourced neighborhoods, which could bias our findings despite controlling for socioeconomic factors.
Conclusion
Study findings indicated healthier retail food environments were linked to lower risk of becoming overweight or obese among Hispanic/Latino adults over 6 years. Given the very limited variability in neighborhood food environments in this sample, this association with an important health outcome is notable and worthy of further study. However, neighborhood food environments were not related to hypertension and diabetes. Addressing these health concerns requires a multi-faceted approach that considers environmental, behavioral, and biological factors. These results suggest that improving access to healthier food outlets in neighborhoods may serve as a critical entry point for reducing obesity disparities, while also providing the foundation for benefits to broader cardiometabolic health. Continued longitudinal research in diverse populations is needed to clarify these pathways and inform multi-level policy and intervention strategies aimed at advancing health equity.
Funding
The Hispanic Community Health Study/Study of Latinos is a collaborative study supported by contracts from the National Heart, Lung, and Blood Institute (NHLBI) to the University of North Carolina (HHSN268201300001I/N01-HC-65233), University of Miami (HHSN268201300004I/N01-HC-65234), Albert Einstein College of Medicine (HHSN268201300002I/N01-HC6-5235), University of Illinois at Chicago (HHSN268201300003I/N01-HC-65236 Northwestern University), and San Diego State University (HHSN268201300005I/N01- HC-65237). SOL CASAS is supported by grant DK106209 from the National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases.
Footnotes
Ethics Approval This study was approved by the Institutional Review Board (IRB) of the participating institutions, as detailed in the Funding Information section.
Consent to Participate All participants provided written informed consent prior to participation in the study.
Consent for Publication All authors have reviewed and approved the manuscript for submission and publication. Additionally, all authors consent to the publication of this work in its current form.
Conflict of interest The authors declare no competing interests.
Data Availability
The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.
Code Availability
The code used for data processing and analysis is available upon request from the corresponding author.
References
- 1.Koyama AK, McKeever Bullard K, Xu F, O nufrak S, Jackson SL, Saelee R, et al. Prevalence of cardiometabolic diseases among racial and ethnic subgroups in adults - behavioral risk factor surveillance system, United States, 2013–2021. MMWR Morb Mortal Wkly Rep. 2024;73(3):51–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Bryan S, Afful J, Carroll M, Te-Ching C, Orlando D, Fink S, et al. National health and nutrition examination survey 2017–March 2020 prepandemic data files -- development of files and prevalence estimates for selected health outcomes. 2021. [DOI] [PMC free article] [PubMed]
- 3.Bluher M Obesity: global epidemiology and pathogenesis. Nat Rev Endocrinol. 2019;15(5):288–98. [DOI] [PubMed] [Google Scholar]
- 4.Guzmán R, Escamilla, Evelio, Castellanos, Daniel, Chacón, Guillermo. Diabetes among U.S. Hispanics: trends and inequities.: Hispanic Equity Briefs.; 2024.
- 5.Elfassy T, Zeki Al Hazzouri A, Cai J, Baldoni PL, Llabre MM, Rundek T, et al. Incidence of hypertension among US Hispanics/Latinos: the Hispanic community health study/study of Latinos, 2008 to 2017. J Am Heart Assoc. 2020;9(12):e015031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Lovasi GS, Hutson MA, Guerra M, Neckerman KM. Built environments and obesity in disadvantaged populations. Epidemiol Rev. 2009;31(1):7–20. [DOI] [PubMed] [Google Scholar]
- 7.Lam TM, Vaartjes I, Grobbee DE, Karssenberg D, Lakerveld J. Associations between the built environment and obesity: an umbrella review. Int J Health Geogr. 2021;20(1):7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Feng J, Glass TA, Curriero FC, Stewart WF, Schwartz BS. The built environment and obesity: a systematic review of the epidemiologic evidence. Health Place. 2010;16(2):175–90. [DOI] [PubMed] [Google Scholar]
- 9.Larson NI, Story MT, Nelson MC. Neighborhood environments: disparities in access to healthy foods in the U.S. Am J Prev Med. 2009;36(1):74–81. [DOI] [PubMed] [Google Scholar]
- 10.Yan R, Bastian ND, Griffin PM. Association of food environment and food retailers with obesity in US adults. Health Place. 2015;33:19–24. [DOI] [PubMed] [Google Scholar]
- 11.Morland KB, Evenson KR. Obesity prevalence and the local food environment. Health Place. 2009;15(2):491–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Anderson Steeves E, Martins PA, Gittelsohn J. Changing the food environment for obesity prevention: key gaps and future directions. Curr Obes Rep. 2014;3(4):451–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Feldman JM, Lee DC, Lopez P, Rummo PE, Hirsch AG, Carson AP, et al. Assessing county-level determinants of diabetes in the United States (2003–2012). Health Place. 2020;63:102324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Thorpe LE, Adhikari S, Lopez P, Kanchi R, McClure LA, Hirsch AG, et al. Neighborhood socioeconomic environment and risk of type 2 diabetes: associations and mediation through food environment pathways in three independent study samples. Diabetes Care. 2022;45(4):798–810. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.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. [DOI] [PubMed] [Google Scholar]
- 16.Gibson DM. The neighborhood food environment and adult weight status: estimates from longitudinal data. Am J Public Health. 2011;101(1):71–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lavange LM, Kalsbeek WD, Sorlie PD, Aviles-Santa LM, Kaplan RC, Barnhart J, et al. Sample design and cohort selection in the Hispanic community health study/study of Latinos. Ann Epidemiol. 2010;20(8):642–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Gallo LC, Carlson JA, Sotres-Alvarez D, Sallis JF, Jankowska MM, Roesch SC, et al. The Hispanic community health study/study of latinos community and surrounding areas study: sample, design, and procedures. Ann Epidemiol. 2019;30:57–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.American Diabetes A. Diagnosis and classification of diabetes mellitus. Diabetes Care. 2010;33 Suppl 1(Suppl 1):S62–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chobanian AV, Bakris GL, Black HR, Cushman WC, Green LA, Izzo JL Jr, et al. Seventh report of the joint national committee on prevention, detection, evaluation, and treatment of high blood pressure. Hypertension. 2003;42(6):1206–52. [DOI] [PubMed] [Google Scholar]
- 21.Project HCDaI. Modified retail food environment index 2017. [Available from: https://data.chhs.ca.gov/dataset/modified-retail-food-environment-index.
- 22.SR S, IR W. Review of inverse probability weighting for dealing with missing data - PubMed. Statistical methods in medical research. 2013. Jun;22(3). [DOI] [PubMed] [Google Scholar]
- 23.Pérez-Ferrer C, Auchincloss AH, Barrientos-Gutierrez T, Colchero MA, Cardoso LdO, Menezes MCd, et al. Longitudinal changes in the retail food environment in Mexico and their association with diabetes. Health & Place. 2020/11/01;66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Block JP, Christakis NA, O’Malley AJ, Subramanian SV. Proximity to food establishments and body mass index in the Framingham Heart Study offspring cohort over 30 years. Am J Epidemiol. 2011;174(10):1108–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Powell LM, Han E. Adult obesity and the price and availability of food in the United States. Am J Agric Econ. 2011;93(2):378–84. [Google Scholar]
- 26.Zenk SN, Mentz G, Schulz AJ, Johnson-Lawrence V, Gaines CR. Longitudinal associations between observed and perceived neighborhood food availability and body mass index in a multiethnic urban sample. Health Educ Behav. 2017;44(1):41–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.de Albuquerque FM, Pessoa MC, De Santis Filgueiras M, Gardone DS, de Novaes JF. Retail food outlets and metabolic syndrome: a systematic review of longitudinal studies. Nutr Rev. 2022;80(6):1599–618. [DOI] [PubMed] [Google Scholar]
- 28.Perez-Ferrer C, Auchincloss AH, Barrientos-Gutierrez T, Colchero MA, de Oliveira Cardoso L, Carvalho de Menezes M, et al. Longitudinal changes in the retail food environment in Mexico and their association with diabetes. Health Place. 2020;66:102461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kraft AN, Thatcher EJ, Zenk SN. Neighborhood food environment and health outcomes in U.S. low-socioeconomic status, racial/ethnic minority, and rural populations: a systematic review. J Health Care Poor Underserved. 2020. 10.1353/hpu.2020.0083. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analyzed during the current study are available from the corresponding author upon reasonable request.
The code used for data processing and analysis is available upon request from the corresponding author.