Epidemiologic studies have identified socioeconomic status as a potent social determinant of health. Poor access to healthcare, lack of education, inadequate financial resources, experience with violence, and constant vigilance associated with low socioeconomic status explain increased morbidity and mortality in populations that live in poverty. Genotype-by-environment interaction (GxE) studies detect if genetic factors interact with the environment to produce phenotypic variation. Our results provide evidence of interactions between SES and metabolic syndrome at a molecular level in Mexican Americans living in South Texas.

Socioeconomic status (SES) is a composite measure of economic and social status that incorporates education, income, and occupation elements. SES influences health through the socialization of positive health habits, education, and the ability to access healthcare resources (Marmot et al., 1991; Marmot, 2003a; Marmot, 2003b; Marmot, 2015; Pillay-van Wyk and Bradshaw, 2017; Rosengren et al., 2019; Averbuch et al., 2022). SES is inversely correlated to disease-specific morbidity/mortality risk for cardiovascular disease, metabolic syndrome, hypertension, diabetes, frailty, and obesity (Chandola et al., 2007; Bijlsma-Rutte et al., 2018; Conway et al., 2018; Allegrini et al., 2020; Manusov et al., 2021; Diego et al., 2023). The SES effects on health are even more prominent in Latino populations, especially Mexican Americans, where a higher prevalence of obesity, diabetes, hypertension, and hyperlipidemia are observed (Hazuda et al., 1991; Heidi Ullmann et al., 2011; Hostinar et al., 2017a; Conway et al., 2018; Diego et al., 2023). An estimated 35 percent of Mexican Americans meet the American Heart Association criteria for metabolic syndrome (MS defined as high triglycerides (>150 mg/dL or on cholesterol medicine, High-Density Lipoprotein (HDL) < 40 mg/dL men and 50 mg/dL women, blood pressure <135/85 or on medications, fasting blood glucose> 100 mg/dL) (Beltrán-Sánchez et al., 2016; Manusov et al., 2019). The prevalence of MS is higher in certain Mexican American groups (older adults, women, non-recent immigrants, US South) (Fernández-Rhodes et al., 2021).

There is an inverse statistical correlation between SES measures and metabolic syndrome (MS) and all-cause mortality in Mexican Americans living in south Texas (Hazuda et al., 1991; Bustamante-Villagómez et al., 2021; He et al., 2021). The effects of SES are traditionally explained by social advantage/deprivation, social mobility, chronic generational poverty, neighborhood deprivation, early-life education health habits (exercise, healthy eating, sleep, quality of life), immigration, vulnerability to environmental stressors, and acculturation (Hazuda et al., 1988; Wei et al., 1996; Stafford and Marmot, 2003; Riosmena and Dennis, 2012; Diego et al., 2015; Kelly and Ismail, 2015; Khambaty et al., 2020; McDoom et al., 2020).

Significant GxE interactions exist between modifiable environmental risk factors and genes that influence cardiovascular risk, (Diego et al., 2023), hyperlipidemia, obesity, and non-alcoholic fatty liver disease (sedentary lifestyle, smoking, and hormonal imbalance) (Hazuda et al., 1988; Heidi Ullmann et al., 2011; Arya et al., 2018; Dugravot et al., 2020; Muscatell et al., 2020; Ciciurkaite, 2021; Manusov et al., 2022).

Recent statistical advances in the evaluation of GxE interaction are shedding light on possible mechanisms by which genes respond to environmental stressors. Specific genes have been determined to play roles in lipid metabolism, inflammation, and insulin regulation, with pro-inflammatory pathways also influencing obesity development. Several studies have examined the effect of gene-environment interactions for loci regulating inflammatory pathways. Environmental factors may modify risk through epigenetic or transcriptional control mechanisms or by affecting downstream regulation of metabolic pathways (Muscatell et al., 2020).

We hypothesize that environmental factors (defined by Duncan’s Socioeconomic Index, educational level, and income) influence the expression of genes related to risk for metabolic syndrome. We investigate the interaction between the SES environment and MS risk factors using a genotype-by-environment interaction (GxE) approach (Blangero et al., 2013). GxE describes how the genetic architecture underlying a given trait (MS risk factor) responds to or depends on environmental changes (SES). Whereas epidemiological studies have demonstrated the inverse association of SES with MS-related morbidity and mortality, GxE analyses can provide information on how genes interact with the environment. To formally evaluate the hypothesis that the genes underlying MS risk factors exhibit GxE interaction effects, we examined potential GxE interaction using variance components models and likelihood-based statistical inference in the phenotypic expression of metabolic syndrome in data from the Mexican American Family Studies (MAFS) (Blangero et al., 2013; Diego et al., 2023). MAFS was initially designed to characterize the genetic determinants of risk for cardiovascular disease (CVD) in Mexican Americans of South Texas. The MAFS is a large, longitudinal pedigree-based study of genetic determinants of risk for complex disease in Mexican Americans, which was initiated in 1992 (Hazuda et al., 1991).

The Institutional Review Board at the University of Texas Health Science Center at San Antonio approved the MAFS study protocols. All study participants provided written informed consent.

The ages of individuals in our sample range from 18 to 94 years of age, with a mean age of 35 years. The descriptive statistics for the MS trait data are presented in Table 1. Missing data were not imputed. Based on an established definition of MS the prevalence of MS (Engin, 2017) in our total sample of 1,140 individuals is 27.21%.

In Table 2, we present the heritability estimates for each of the MS traits, along with the covariates included in the BMS most-supported model and the percent variance they explained. Only four traits (waist circumference, total serum cholesterol, systolic blood pressure, and diastolic blood pressure) showed evidence of a significant household variance component (Table 2).

Test results of the GxE model versus model 1 or model 2, corrected for multiple hypothesis testing, are presented in Table 3. Five traits showed the potential for GxE effects (fasting glucose, 2-h glucose, total serum cholesterol, systolic blood pressure, and diastolic blood pressure). We note that only systolic blood pressure has household variance and correlation functions added to the GxE model (Table 4).

Results of the specific GxE hypothesis tests and of the tests of homogeneous residual environmental variance, homogeneous household variance component, and ρ c = 1 are presented in Table 4. We found evidence of GxE interaction due to heterogeneity in additive genetic variance in fasting glucose and 2-h glucose (Table 4; Figure 1). We also found evidence of heterogeneity in residual environmental variance for fasting glucose and 2-h glucose (Table 4; Figure 1). Because additive genetic variance is decreasing while residual environmental variance is increasing with zSES for 2-h glucose and fasting glucose, total phenotypic variance exhibits a superficial quadratic-like behavior along the SES environment (Figure 1). We also found evidence of GxE due to genetic correlation being less than 1 for 2-h glucose (Table 4; Figure 1). Of the 3 traits (total serum cholesterol, systolic blood pressure, and diastolic blood pressure) that exhibited significant household effects and that progressed to this stage of the analyses, only systolic blood pressure showed significant heterogeneity in its household variance (Table 4; Figure 1). The household variance function for systolic blood pressure behaved in a similar manner to the additive genetic variance functions for fasting glucose and 2-h glucose. Given that the additive genetic and residual environmental variance functions for systolic blood pressure were constant (Table 4), the behavior of the phenotypic variance function was determined entirely by the household variance function, which explains their similarity (Figure 1).

In Figure 2, we express the joint behavior of the additive genetic variance and correlation functions in producing covariance heterogeneity for 2- hour glucose.

Our results are consistent with the epidemiological relationship between SES and MS risk factors in Latinos as well as other ethnic groups (Hostinar et al., 2017b; Engin, 2017; Blanquet et al., 2019). The clinical implications of the heritability of the 12 phenotypes studied imply that although the genetic components are moderate, the environmental influences are greater and complementary. Hispanic/Latinos exhibit an inverse association between SES and metabolic syndrome of varying magnitudes across SES variables. For the glucose variables, there was evidence of GxE via heterogeneity in additive genetic variance and we consistently observed that additive genetic variance was higher in the lower end of the SES spectrum (Figures 1, 2). The genetic effects underlying cardiovascular disease are similarly dynamically modulated along the SES spectrum (Schultz et al., 2018; Carter et al., 2019; Powell-Wiley et al., 2022; Diego et al., 2023).

To the general patterns regarding SES, mortality, and morbidity measures, we can add the inverse association between SES and additive genetic variance in 2-h glucose and fasting glucose. We found evidence of GxE for fasting glucose and 2-h glucose (Figure 2) that suggests that different sets of genes are active in different regions of the SES spectrum. These results indicate that the gene networks influencing glucose are upregulated in lower relative SES environments. Our results demonstrate that the SES environment modulates the underlying genetic architecture, and this interaction is related to glucose homeostasis (Puolakka et al., 2016; Corsi and Subramanian, 2019).

After accounting for sex effects before GxE analysis, the inverse association between SES and morbidity measures is still significantly more robust in women than in men among Mexican Americans from San Antonio (Stern et al., 1984). Accounting for behavioral variables, especially physical activity, attenuates the strength of the inverse relationship between SES and morbidity measures of MS (Walker et al., 2014; Tanaka et al., 2018).

There are important household effects for systolic blood pressure. The household variance function for systolic blood pressure was also inversely related to SES. In addition to the dynamic genetic effects discussed above, this is another novel twist to the general pattern of an inverse relationship between measures of MS traits and SES.

Our conclusions have real-world applicability in clinical practice and public health strategies. Addressing health disparities, especially in vulnerable populations like Mexican Americans in south Texas, should be a priority. Our findings highlight the importance of reducing health inequities by addressing both genetic and socioeconomic factors contributing to MS risk. By recognizing the interplay between genetics and SES in the development of metabolic syndrome, healthcare professionals and policymakers can work together to develop targeted interventions and policies that improve the health outcomes of communities, particularly those facing socioeconomic challenges, similar to the study population.

The identification of gene-environment interactions (GxE) related to metabolic syndrome (MS) risk factors suggests that healthcare interventions can be tailored based on an individual’s genetic profile and socioeconomic status (SES). Understanding that the genetic effects underlying MS traits are modulated by SES provides insights into targeting high-risk populations. Healthcare providers and public health initiatives can focus their efforts on communities with lower SES, as these individuals may have a higher genetic susceptibility to MS-related risk factors. Public health strategies should emphasize health education and promotion, especially in lower SES communities. This includes raising awareness about the importance of healthy lifestyle choices such as diet, especially in relationship to carbohydrates. Recognizing the inverse association between SES and additive genetic variance in glucose variables (fasting and 2-h glucose) suggests that individuals from lower SES backgrounds may benefit from earlier screening and detection of MS risk factors. This can help identify at-risk individuals and initiate preventive measures at an earlier stage.

Public health policies can be designed to address the social determinants of health that are linked to SES. Initiatives to improve access to education, reduce chronic generational poverty, enhance neighborhood environments, and promote healthy behaviors can have a positive impact on reducing MS risk. Healthcare providers should consider a patient’s SES as a vital factor in assessing their risk for metabolic syndrome. By incorporating SES into clinical practice, clinicians can provide more comprehensive and tailored care to individuals, considering both genetic and environmental factors.

Our results support the presence of GxE interaction between genetic factors and SES in determining the risk for metabolic syndrome in Mexican Americans.

We are now further exploring the GxE interaction using mRNA sequencing data to identify the contribution of specific genes and canonical pathways.

We present evidence for the dynamic modulation of the genetic architecture of major components of MS by the SES environment. These findings highlight the importance of SES environmental effects on health-related physiology. Our study serves as a foundation for further research into transcriptomic and methylomic analysis to identify the specific proteins, canonical pathways, and molecular changes involved in the GxE interaction. We then can explore interventions that mitigate the negative health impacts associated with low SES, such as targeted genetic therapies or lifestyle interventions.