Investigating Environmental and Socioeconomic Contributors to Adult Obesity in the Rio Grande Valley

Abstract

Obesity in the Rio Grande Valley (RGV) of Texas remains a critical public health concern, with rates (46.9%) significantly exceeding those of Texas (36%) and the U.S. (34%) (p < 0.001). This study used 2024 County Health Rankings data to analyze environmental and socioeconomic contributors to obesity across 240 Texas counties, with a population-weighted focus on Hidalgo, Cameron, Starr, and Willacy counties. The RGV exhibited markedly poorer access to exercise, higher rates of physical inactivity, lower Food Environment Index scores (FEI = 4.3 vs. 5.7 in Texas), higher rates of uninsurance, worse patient-to–primary care physician (PCP) ratios (2152:1 vs. Texas 1660:1), and increased preventable hospitalizations. Multiple linear regression identified physical inactivity (β = 0.6, p = 0.01) and access to exercise (β = −0.02, p = 0.02) as significant predictors of obesity. Notably, higher uninsured rates were associated with lower reported obesity, likely due to underdiagnosis in the absence of routine care. These findings emphasize the need for targeted interventions addressing food access, environmental and recreational infrastructure, along with healthcare infrastructure in the RGV, where socioeconomic disadvantage and structural barriers magnify the impact of national obesity trends on the regional level.

1. Introduction

1.1. Obesity

Obesity is a complex, multifactorial disease whose prevalence continues to rise across the United States, where the proportion of adults affected rose from 30.5% in 1999 to 41.9% between 2017 and March 2020 [1]. It is a key modifiable risk factor for many comorbidities and disabilities; any efforts towards understanding its etiology are critical to reducing the medical burden at the population level.

While often framed in terms of individual behavior, obesity is shaped by a dynamic interplay of biological, behavioral, and modifiable environmental factors. Social determinants can further compound these risks, with decades of findings suggesting an income-dependent association between socioeconomic status (SES) and obesity [2]. In low-income countries, a higher SES correlates with obesity, while in high-income countries, a lower SES is more strongly associated [2,3]. Despite these findings, however, studies of county-level, low-SES populations remain limited. The aim of this study is to examine such social and environmental contributors on the Rio Grance Valley (RGV), a majority-Hispanic region in South Texas with some of the highest obesity and poverty rates in the country. Therefore, its investigation presents a compelling case study between the relationship between obesity risk and various environmental and socioeconomic determinants.

1.2. Rio Grande Valley

Composed of the fourth southernmost Texas counties (Hidalgo, Cameron, Starr, and Willacy), the RGV is a geographic region that runs along the United States (US)–Mexico border. With such proximity, this region’s significant Hispanic population has a heavy influence on cultural diversity, cuisine, and community life [4]. However, the RGV is also an economically disadvantaged and medically underserved area of Texas. Recent epidemiological studies indicate that the prevalence of obesity in the RGV substantially exceeds both state and national averages, with profound implications for the region’s overall health burden. For instance, large-scale health surveys estimate that between 45% and 50% of Mexican–American adults in the RGV are classified as obese [5].

The multifactorial etiology of obesity in the RGV includes a complex interplay of genetic, environmental, and socioeconomic factors. Although genetic predisposition is estimated to account for approximately 50% of the variance in body mass index (BMI), environmental and lifestyle factors such as diet, physical inactivity, and poverty are recognized as major contributors to the region’s obesity epidemic [6]. Notably, dietary patterns characterized by calorie-dense, nutrient-poor foods are prevalent, significantly contributing to android obesity, a form closely linked to type 2 diabetes and metabolic syndrome [1]. Additionally, physical inactivity remains pervasive in the RGV, compounded by limited access to recreational facilities and safe spaces for exercise.

The socioeconomic landscape further exacerbates the obesity crisis in RGV. Lower educational attainment, language barriers, and economic instability are consistently associated with higher obesity rates in the region, particularly among Hispanic populations, creating a complex landscape where genetic predispositions intersect with socio-environmental factors [7,8].

Genetic studies underscore the interplay between genetic susceptibility and environmental influences. A recent study focusing on young Mexican–Americans in the RGV has identified significant genetic associations related to obesity and metabolic traits [9]. Specifically, polymorphisms such as rs1800497 ankyrin repeat and kinase domain containing 1 (ANKK1), rs846910 hydroxysteroid 11-beta dehydrogenase 1 (HSD11B1), and rs1205 c-reactive protein (CRP) have demonstrated significant associations with obesity-related phenotypes, including waist circumference, insulin resistance, and fasting glucose levels in Mexican–American populations. For instance, rs1800497 (ANKK1) has been significantly correlated with central obesity (p = 0.009) and insulin resistance (p = 0.03 after permutation testing), rs846910 (HSD11B1) with elevated triglycerides (p = 0.03) and Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) (p = 0.04), and rs1205 (CRP) with fasting glucose (p = 0.007), although some associations did not maintain statistical significance after correction [9].

Despite the persistent obesity burden, recent data suggest a modest decline in obesity rates in the RGV, with a 1.2% reduction observed since 2011, accounting for approximately 5500 fewer obese individuals [10]. This decline contrasts with the continued upward trend in obesity prevalence across Texas overall. Consequently, targeted public health interventions that address the underlying socioeconomic and environmental determinants of obesity remain critical to mitigating the health disparities faced by the RGV’s predominantly Hispanic population.

2. Materials and Methods

2.1. Data Source

Data for this study were obtained from the 2024 County Health Rankings & Roadmaps (CHR&R) database, a comprehensive public health resource which compiles health-related metrics from national and state-level sources, including the Behavioral Risk Factor Surveillance System (BRFSS), the United States Diabetes Surveillance System, and the American Community Survey, to assess health outcomes and determinants across U.S. counties [10]. The CHR&R database calculates each new measure using a k-means clustering analysis defined in the ’Methods’ section of their website and reevaluates these measures annually for validity with source and analytical considerations in mind [10,11]. For this analysis, data were extracted on Adult Obesity, Access to Exercise Opportunities, Food Environment Index, Limited Access to Healthy Foods, Physical Inactivity, Poor or Fair Health, Median Household Income, Income Inequality, Uninsured Adults, and Population. These variables were selected based on their relevance in investigating environmental and socioeconomic influences surrounding the RGV, and were examined at three geographical levels: the national average, Texas state average, and at the specific county level. Variables such as sexually transmitted infection rates, teen birth rates, and drug overdose mortality were excluded based on demonstrating a limited or indirect theoretical relationship to obesity, or because they were inconsistent with the study’s focus on modifiable environmental and socioeconomic determinants.

2.2. Statistical Analyses

Of the obtained county data, an additional RGV row was created using an average of its four constituent counties’ data weighted by population. Descriptive statistics were generated for the United States, Texas, and the aggregated county-level data, as well as for each individual county within the RGV. Counties with missing values of interest (n = 14) were excluded, leaving 240 remaining observations. To compare obesity prevalence rates between the RGV counties, the state of Texas, and the United States overall, a chi-squared test was conducted. The test indicated a statistically significant difference in obesity percentages across the three regions (χ²(2) = 1.5 × 10⁵, p < 0.001). Then, a series of pairwise proportion z-tests, conducted using the prop.test() in R without Yates’ continuity correction, were conducted to assess specific differences in pairs. It is notable that all the z-tests, despite accounting for the very large population numbers, concluded as statistically significant. To both take advantage of these large n values and more practically express these differences in health outcomes, effect sizing through odds ratios (ORs) is also included. Visualizations are made in R using package ggplot2 to map adult obesity vs. Food Environment Index (FEI) and Income Inequality vs. median income for the groups of interest. To explore the relationship between the dependent variable of adult obesity and the several explanatory variables, a multiple linear regression model was made in base R using the 240 (14 omitted from initial 254 due to missing data) Texas counties. Assumptions of linearity, normality, and homoscedasticity were checked using diagnostic plots. All coefficient βs were standardized to be on the same scale (mean = 0, SD = 1). Finally, adult obesity and FEI heatmaps were made using 2024 TIGER/Line^® Shapefiles: Counties (and equivalent) .cpg file into R and applying the pheatmap package [12].

3. Results

3.1. RGV Compared to State and Country Levels

The RGV obesity prevalence was calculated by summing the estimated number of obese individuals in each county, based on county-level obesity rates and population sizes (total population = 661,313). Here, we found that obesity prevalence within the RGV (46.9%) was significantly higher than that of Texas (36.0%), χ²(1) = 6.7 × 10⁴, p < 0.001. and the United States (34.0%), χ²(1) = 1.0 × 10⁵, p < 0.001 (

Table 1. Obesity prevalence % by RGV counties, weighted average % of RGV area, Texas, and U.S.

Indicator	Willacy	Hidalgo	Starr	Cameron	RGV	Texas	United States
% Obesity	45	48	44	45	46.9	36	34

). The state of Texas itself also had a significantly higher obesity rate than the national average, χ²(1) = 5.0 × 10⁴, p < 0.001. Thus, we observe a stepwise gradient increasing from national to state to regional levels, illustrating how a national health crisis is felt most acutely at the regional level. This analysis is conducted for the other categorical variables of interest and is summarized in

Table 2. Obesity-related health indicators and access to care in RGV Counties vs. Texas and U.S. [10].

Indicator	Willacy	Hidalgo	Starr	Cameron	RGV	Texas	United States
Avg. Physically Unhealthy Days	5.8	5.7	6.0	5.3	5.6	3.8	3.9
% Fair/Poor Health	37	37	38	37	37	20	17
% Access to Exercise	62	64	62	74	65	82	84
Food Environment Index	3.6	4.4	3.0	4.8	4.3	5.7	7.4
Primary Care Physician Ratio	2540:1	2158:1	3885:1	2034:1	2152:1	1660:1	1330:1
Preventable Hospitalizations (Days)	3725	3145	4460	3242	3150	2968	2666

.

3.2. Indicators Related to Obesity

Among obesity-related health indicators in RGV counties, several notable findings highlight disparities in access and resources compared to state and national averages (Table 2). The RGV column is the weighted average of the four counties. Variable definitions may be found in the addendum section below. The FEI (Food Environment Index) is notably low, with three counties ranking within the bottom 19 statewide. This pattern is also visually apparent in Figure 1, where the RGV stands out with both a higher obesity rate and a lower FEI compared to the rest of Texas, reinforcing the region’s disproportionate burden. All RGV counties rank in the lower half of recorded counties for exercise access. It is also worth noting that RGV counties have a primary care access ratio of 2152:1. This is an additional 500 patients per primary care provider compared to other counties in Texas. Figure 2 illustrates the inverse relationship between adult obesity rates and the FEI across the four RGV counties and their average. These RGV counties show higher obesity rates corresponding with lower FEI values, indicating more limited access to healthy food options. In contrast, both Texas and the US exhibit relatively lower obesity rates alongside higher FEI scores, highlighting a potential link between the food environment and obesity prevalence.

Table 3. Odds ratios (ORs) and 99% confidence intervals (CIs) for obesity and related health/environmental factors in the Rio Grande Valley (RGV) compared to Texas and the United States [10].

	Proportion z-Test	OR	CI (99%)	Chi-Square
Obesity				6.9 × 104 (p < 0.05)
RGV-TX	X2 = 5.7 × 103	1.57	1.56–1.57
RGV-US	X2 = 9.5 × 103	1.09	1.09–1.09
TX-US	X2 = 4.5 × 103	1.71	1.70–1.72
Uninsured				3.0 × 106 (p < 0.05)
RGV-TX	X2 = 1.6 × 105	2.07	2.06–2.08
RGV-US	X2 = 8.0 × 105	4.38	4.36–4.40
TX-US	X2 = 2.4 × 106	2.11	2.11–2.11
Poor Fair health				4.6 × 105 (p < 0.05)
RGV-TX	X2 = 1.7 × 105	2.10	2.09–2.11
RGV-US	X2 = 3.0 × 105	2.56	2.55–2.57
TX-US	X2 = 1.8 × 105	1.22	1.22–1.22
Physical inactivity				2.4 × 105 (p < 0.05)
RGV-TX	X2 = 2.1 × 105	1.84	1.83–1.85
RGV-US	X2 = 3.2 × 105	2.05	2.04–2.06
TX-US	X2 = 8.2 × 104	1.12	1.11–1.12
Access to exercise				3.9 × 105 (p < 0.05)
RGV-TX	X2 = 1.2 × 105	0.44	0.43–0.44
RGV-US	X2 = 1.8 × 105	0.38	0.38–0.38
TX-US	X2 = 7.7 × 104	0.87	0.87–0.87
Limited heathy food				6.3 × 105 (p < 0.05)
RGV-TX	X2 = 2.4 × 105	2.82	2.8–2.83
RGV-US	X2 = 4.6 × 105	3.84	3.81–3.86
TX-US	X2 = 1.9 × 105	1.36	1.36–1.36

highlights that the Rio Grande Valley (RGV) exhibits substantial disparities in health and environmental factors associated with obesity when compared to both Texas and the United States. The prevalence of obesity in the RGV is approximately 1.58 times higher than in Texas (OR = 1.576) and significantly higher than the national average (OR = 1.091) [10]. The most pronounced disparity is seen in the limited access to healthy food, where RGV residents are nearly 3.84 times more likely to face food access barriers than the national population (OR = 3.84, 99% CI: 3.81–3.86), making it the strongest positive association observed. Additionally, physical inactivity is significantly more common in the RGV, with residents over twice as likely to be physically inactive compared to the U.S. average (OR = 2.051). The region also demonstrates markedly lower access to exercise opportunities, with an odds ratio of 0.384 compared to the U.S., indicating substantial environmental constraints on physical activity. Compounding these issues, the RGV has far higher rates of uninsured individuals, with an OR of 4.384 relative to the national average, and a poorer self-rated health status, with an OR of 2.566. Collectively, these findings underscore the multiple and intersecting barriers that contribute to elevated obesity risk in the RGV.

While these bivariate comparisons highlight stark disparities, they do not account for how these factors interact or overlap. To identify which variables most strongly and independently contribute to obesity when considered together, a multivariable linear regression analysis was conducted.

The impact of various health and environmental factors on obesity through a multivariable linear regression model is shown in

Table 4. The table presents the impact of various health and environmental factors on obesity through a multivariable linear regression model [10].

	Standardized Coefficient Estimate	Standard Error	T Value	Probability
Poor Fair Health	4.7 × 10−2	2.98 × 10−1	0.16	0.87
Food Environment Index	−6.8 × 10−2	1.03 × 10−1	−0.66	0.51
Physical Inactivity	6.9 × 10−1	2.86 × 10−1	2.41	0.02
Access to Exercise	−1.0 × 10−1	4.92 × 10−2	−2.10	0.04
Limited Healthy Food	−4.6 × 10−2	8.95 × 10−2	−0.52	0.61
Uninsured	−1.7 × 10−1	6.14 × 10−2	−2.76	0.01

. The strongest positive association with obesity is Physical Inactivity (0.7), suggesting that higher inactivity rates significantly increase obesity risk, while Access to Exercise (−0.1) and Uninsured (−0.2) show notable negative associations [10].

As shown in

Table 5. Summary of the model fit statistics for a multivariable linear regression assessing factors impacting obesity [10].

Residual Standard Error	0.7 on 233 degrees of freedom
Multiple R Squared	0.48
Adjusted R Squared	0.46
F-statistic	35.6 on 6 and 233 DF
p-value	<2.2 × 10−16

, the multiple linear regression model to examine the association between obesity and various county-level health and socioeconomic indicators was statistically significant (F(8, 231) = 27.3, p < 0.001), explaining approximately 48% of the variance in adult obesity (R² = 0.48, Adjusted R² = 0.47). Controlling for other factors, physical inactivity was positively associated with obesity (β = 0.6, p = 0.01), indicating that for each 1% increase in the physically inactive population, adult obesity rates increased by approximately 0.56 percentage points. Conversely, improving access to exercise was significantly associated with lower obesity rates (β = −0.02, p = 0.02). Possibly explained by less health maintenance visits and explicit medical awareness of obesity status, counties that had higher percentages of uninsured citizens were associated with lower obesity (β = −0.4, p = 0.009). Other predictors included in the study were not significantly associated with obesity in this model (p > 0.05).

4. Discussion

The purpose of this study was to characterize the environmental and socioeconomic factors which contribute to elevated obesity rates in the RGV. The findings from this study reveal that many of the known social determinants of health which impact obesity prevalence elsewhere are amplified within the Valley. Significant differences in physical activity, availability of exercise, access to nutritious foods, self-reported health, and insurance status were identified between the RGV when compared to other Texas counties and the United States more broadly. While the County Health Rankings database has been widely used to assess health disparities across regions, such as between rural and non-rural Texas counties or to describe disease within smaller states, studies conducted towards to adjacent multi-county regions that analyze similar contributors to obesity could not be found [13,14,15,16].

Physical inactivity is a central contributor to the global obesity epidemic and a major independent risk factor for chronic disease and premature mortality. Residents of the RGV are twice as likely to be physically inactive when compared to the rest of the United States (odds ratio [OR]: 2.1 [99% confidence interval: 2.04–2.06]). This is an unsurprising finding, as access to exercise is significantly limited when compared to the rest of Texas (odds ratio [OR]: 0.4 [99% confidence interval: 0.38–0.38]), while Texas has reached near parity with the United States. The existence of this large disparity implies that investment in recreational and sporting infrastructure, along with the promotion of physical activity within the RGV, would greatly contribute to a reduction in obesity in the region. Previous research has shown that even small increases in activity among previously sedentary individuals yield substantial health gains, particularly in reducing cardiometabolic risk [17]. However, one underrecognized barrier to physical activity and a potential confounder in the RGV is its climate. Historical weather data have found an apparent temperature of the summer days in McAllen to start from 108 °F, having an eleven-day streak above 116 °F in June 2023 [18]. High heat can not only discourage residents from exercising outside but pose a heat risk for those unprotected by climate-resilient infrastructure. Despite limited academic attention, the lived experience of hot weather on daily activity in the region is clear and should be considered when examining the physical inactivity patterns within the RGV.

Uninsured status becomes more prevalent within the valley in comparison to the rest of Texas (odds ratio [OR]: 2.1 [99% confidence interval: 2.06–2.08]), and even more so on the national level (odds ratio [OR]: 4.4 [99% confidence interval: 4.36–4.40]). Data from the 2009 Medical Expenditure Panel Survey estimated the rate of obesity among the American adult uninsured, publicly insured, and privately insured populations at 28.3%, 40.2%, and 29.1%, respectively [19]. Despite this finding in adults, data have shown that uninsured adolescents aged 10–17 are more likely to be obese than their insured counterparts (odds ratio [OR]: 1.1 [95% confidence interval: 1.0–1.2]) [20]. In line with these previous findings, our multi-variate linear analysis across Texas counties revealed a negative correlation between the percentage of uninsured in a county and prevalence of obesity (β = −0.4, p = 0.009). With a calculated Variance Inflation Factor (VIF) = 1.7, the possibility of any multicollinearity is unlikely. This finding may reflect underdiagnosis of obesity among uninsured individuals due to reduced healthcare access and fewer routine clinical encounters. Additionally, obesity may be less likely to be formally recorded in populations with limited engagement in preventive care. Other potential confounders include socioeconomic variables not fully accounted for in the model, such as employment type, healthcare-seeking behavior, or undocumented status, which may influence both insurance coverage and reported health outcomes. Further investigation is needed to clarify whether this association reflects true epidemiologic patterns or limitations in measurement and reporting.

While low insurance coverage, inadequate housing, and access to exercise are also significantly lower in the RGV compared to both state and national averages, they are not as directly impactful or as readily modifiable as food insecurity, which remains the most pressing and preventable determinant of obesity in the region. Before discussing strategies, it is essential to briefly define the Food Environment Index (FEI) and explain how it is measured. The FEI is a metric developed by the U.S. Department of Agriculture to assess access to healthy food and food security, ranging from 0 (worst) to 10 (best) [10]. In the RGV, the FEI is notably low, with food insecurity rates of approximately 20%, higher than both the Texas and national averages [10]. By focusing on improving the FEI through targeted interventions, there is potential to address one of the most direct and preventable factors contributing to the region’s high obesity rates.

Figure 3 illustrates food access disparities in the RGV, as defined by the U.S. Department of Agriculture Economic Research Service, using data from 2019. The map categorizes census tracts based on proximity to supermarkets and vehicle access into three distinct indicators: 1 and 10 miles (urban/rural thresholds for a third of residents), 1/2 and 10 Miles, 1 and 20 Miles, and Low Vehicle Access, defined as over 100 households lacking vehicle access while being over 1/2 mile from a supermarket or having many residents live over 20 miles away [21].

The abundance of colored areas illustrates the transportation challenges, and the long distances residents must travel to access supermarkets and fresh produce. Many areas in the RGV qualify as food deserts, where the nearest grocery store is beyond 1 mile in urban areas or 10 miles away in rural areas [10]. Addressing these geographic and mobility challenges—through improving physical access, expanding healthy food supply, and improving transportation infrastructure—could substantially improve FEI scores across the RGV.

Transportation initiatives such as mobile grocery markets or shuttle services to grocery stores can bridge the gap for residents without vehicle access. This has precedent in the region, as the MolinaCares Accord (charitable organization) in collaboration with the Food Bank of the Rio Grande Valley and H-E-B (grocery chain) deployed a mobile food-bank trunk which has operated out of Hidalgo County since 2022 [22]. The purpose of this initiative is to expand access to fresh nutritious foods for those most in need. The Food Bank of the Rio Grande Valley currently delivers 48 million meals annually and feeds 76,000 people weekly (5.5% of the total RGV population) [22,23].

Additionally, policy can be shaped by local and state officials to incentivize the elimination of these food desert regions. For example, reducing the gross income tax or granting property tax exemptions to select entrepreneurs which aim to establish grocery store locations in underserved areas can increase food access, while simultaneously stimulating local economic development. Proof of concept and public demand for such stores is already established, as several independent grocers have carved out a niche providing locally grown, fresh, whole, raw, natural, and organic foods [22,24]. Food supply, too, can alter available nutrition profiles. Similar incentives which aim to support farmers’ markets and the stocking of healthy foods in stores can improve nutritional health. Federal and state-level programs that aim to improve the accessibility of these same healthy foods such as the Supplemental Nutrition Assistance Program (SNAP) and Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) can complement increased fresh food supply by improving access and encouraging healthier choices in underserved areas.

To our knowledge, however, few studies have examined how these structural factors relate to obesity within underserved U.S. regions using County Health Rankings (CHR&R) data [25]. The only peer-reviewed example remains a 2022 study focused on Shelby County, Tennessee. Our study builds on this gap by applying updated 2024 CHR&R data to the Rio Grande Valley, providing a timely comparison and expanding understanding of how environmental and socioeconomic barriers influence obesity at the regional levels.

It must be noted that there are inherent limitations when using a public database for an ecological study, in that associations are drawn from aggregate group-level data, which may not fully reveal individual-level relationships. Additionally, public databases often lack detailed and granular data for investigating, and controlling for, potential confounders. The nature of these large datasets limits the ability to adjust for factors that may bias the observed associations, and the temporal sequence between exposure and outcome is frequently unclear.

5. Conclusions

This analysis highlights the disproportionately high burden of obesity in the Rio Grande Valley (RGV), where the obesity prevalence of 46.9% significantly exceeds not just the state but the national level. Compelling environmental and socioeconomic conditions that underscore this disparity include the low Food Environment Index (FEI), levels of physical inactivity, and poor access to exercise. Counterintuitively, we discovered an inverse relationship between uninsured status and reported obesity. This aligns with previous national level estimates which predict a lower prevalence of obesity within uninsured populations. We postulate this may reflect underdiagnosis and underreporting due to true prevalence, although further investigation is needed. We encourage future researchers to investigate the strength and validity of the inverse correlation between uninsured status and obesity to better characterize this often-overlooked population. Regardless, these modifiable structural inequities should be considered by governments and public health bodies aiming to reduce obesity, whether in the RGV, or in other underserved regions facing similar challenges.