Modeling Life-Course Socioeconomic Changes and Blood Pressure in Adults: Time for Transdisciplinary Approach

As the disease burden of society shifts from infectious to noncommunicable or degenerative disorders, cause and effect relationship in the causal chain transits from proximal to more distal causes. Latter include life style choices, health behaviors, and dietary habits which account for significant variance in prevalence of chronic disorders like metabolic syndrome, diabetes, and cardiovascular diseases including hypertension.1 Indeed, a life-course approach is being widely explored in understanding the epidemiology of chronic diseases like hypertension.2 Latter is underpinned by various socioeconomic susceptibilities that evolve and are operational over the entire life span.3 In this regard, a study published by Glover et al. in the current issue has provided mechanistic insights into decoding socioeconomic differentials of hypertension over the life course among adult ethnic population of United States.4 Authors examined the associations of multiple measures of life course socioeconomic status (SES) with existing (prevalent) and new onset (incident) hypertension in the Jackson Heart Study (JHS), a large population-based study of African Americans, with >10 years of follow-up. The robustness of study results was further enhanced when authors tested sex- and age-adjusted (model 1) and other cardiovascular risk factor–adjusted (model 2) associations between life-course SES and hypertension. There was clear evidence that childhood SES showed inverse gradient with hypertension in this ethnic population. Conceptually, this finding is consistent with “latency effects” life-course SES model (also called “biological chain of risks” model).3 Latter can easily explain why early childhood socioeconomic adversity (a critical period for “development origins of health and disease” programming) has emerged as an independent predictor of adult onset hypertension even without intervening factors playing any major role. Nonetheless, adverse childhood events (ACE) as well as poor childhood health should also be factored along with early life socioeconomic challenges as a part of a composite index of childhood misfortune which drives emergence of late life adverse physical and mental health.5 Notably, subjects in whom improved adulthood socioeconomic status fails to mitigate the effect of early life adversity are the one who had suffered concomitant ACE and poor childhood health.5 In addition, upper middle-income subjects in present study had significantly higher prevalence as well as incidence of hypertension when adjusted for age, sex, physical activity, sodium intake, waist circumference, diabetes, smoking, drinking, fat, and total energy intake. This higher predilection of upper middle class for prevalent as well as incident hypertension is possibly driven by high coping stresses (a phenomenon called “John Henryism”) as middle-income population reflects the meeting point of downward social mobility in affluent and upward social mobility in poor. A study by Subramanyam et al. had already shown that male African-American subjects with high John Henryism scores had highest prevalence ratios for hypertension among subjects in middle income group (1.16; 95% CI: 1.00–1.35) as well as those in middle tertile of cumulative SES based on income, education, and childhood SES (1.15; 95% CI: 0.98–1.35).6 Surprisingly, the current study also showed an apparent beneficial influence of downward mobility on prevalence as well as incidence of hypertension (model 2 was statistically significant). One possible explanation cited frequently in literature for this apparently unexpected finding is that these participants might have only “subjectively” downward social mobility (high childhood and high “objective” adult SES, but low “subjective” adult SES).3 The various conceptual models used to capture the impact of life-course SES on emergence of adulthood hypertension are graphically summarized below (Figure 1). Life-course models of socioeconomic mobility showing that consistently low SES (cumulative burden), downward mobility, and latency effects models signal high risk of hypertension (* “latency effects” model reflects independent predictive role of low childhood SES irrespective of low or high adulthood SES) whereas consistently high SES and upward mobility models reflect lower risk of hypertension. However, there was discordance of findings in the current study when subjects with upward mobility as well as consistently high SES showed significantly higher prevalence but lower incidence of hypertension on follow-up in comparison to subjects with consistently low SES. Authors attributed higher prevalence to increasing allostatic load due to perceived every day and lifetime social stressors, but better access and adherence to medical care may have played a protective role in stonewalling newer cases of hypertension. Nonetheless, these divergent findings certainly reflect that currently available conceptual life-course SES models are suboptimal to establish a clear directional SES-hypertension gradient. In this context, the need for novel SES markers beyond education, income, and occupation had already been emphasized in the scientific statement issued by “American Heart Association” about social determinants of risk and outcomes for cardiovascular diseases.7 Therefore, a transdisciplinary “systems medicine” paradigm using newer tools of epigenetics and machine learning algorithms merits attention to unmask the directional causation as well as to minimize the statistical uncertainties of life-course approach in understanding hypertension. In this regard, recent advances in epigenetic (gene-environment alterations) can easily decode the mechanisms by which socioeconomic factors are biologically embedded to cause cardiovascular diseases in later life.8–10 Epigenetic methylation of promoter region of NR3C1 gene (also called glucocorticoid reception-GR gene) by childhood adversity has been linked with adult blood pressure. This association has been observed independent of current socioeconomic circumstances as well as life style practices.8,10 Elevated 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2) promoter methylation resulting in reduced placental 11β-HSD2 gene expression was also reported to be associated with later life hypertension and the prenatal socioeconomic adversity affects the promoter methylation of 11β-HSD2 gene.9,11 Furthermore, epigenetic mechanisms are also being held accountable for apparent paradox of lower prevalence of risk factors yet higher prevalence of cardiovascular diseases in low-income countries.12 Undeniably, epigenetics has transformed “early childhood” SES into an actionable metric by unmasking a “critical childhood window period” where interventions to mitigate or improve unfavorable SES can yield maximum adulthood benefits. For example, a study by Miller and Chen found that children whose families had their own houses by 2–3 years of age showed favorable regulation of inflammatory biomarkers with higher glucocorticoid receptor (GR) mRNA and lower toll-like receptor-4 (TLR-4) mRNA during adolescence independent of current lifestyle practices or socioeconomic vulnerabilities.13 In addition, adulthood lifestyle risk factors like drinking, smoking, pollution, and poor diet (low dietary folate) have been shown to induce epigenetic changes promoting cardiovascular disease and may possibly modulate the causal role of social mobility in adult hypertension as well as other cardiovascular disorders. For example, alcohol consumption raised the risk of hypertension by enhancing DNA methylation (CpG1 methylation) of the Adducin (ADD1) gene promoter.14 Similarly, significant epigenetic changes between smoking and genes associated with atherosclerosis had already been reported among African Americans in the “Intergenerational Impact of Genetic and Psychological Factors on Blood Pressure Study” (InterGEN) and the “Genetic Epidemiology Network of Arteriopathy” (GENOA).15 Also, altered DNA methylation levels of Angiotensin II Type I Receptor (AGTR1) gene due to smoking and irregular diet were associated with hypertension.16 Furthermore, reduced Alu and TLR-4 methylation caused by human exposure to concentrated ambient particles (CAPs) resulted in an increase in systolic blood pressure.17 In addition, epigenomics can provide further insights into the intergenerational biological embedding of social adversity resulting in pervasive health disparities, more so in marginalized ethnic groups.18 At the same time, it is also equally imperative to augment the diagnostic and predictive potential of social determinants using machine learning algorithms. With emergence of high throughput digital health data captured in real time, even few patients can generate huge number of social, environmental, and personal determinants or predictors of a health outcome like hypertension.19 In this context, machine learning algorithms will easily navigate the technical barriers of non-linearity, multi-collinearity, high predictor to event ratios, as well as interactions commonly encountered in such multi-dimensional data.19 Recently a study by Huang et al. showed superiority of artificial neural network (ANN) over logistic regression in predicting hypertension using risk factors, such as socioeconomic status and education level.20 Both supervised and unsupervised machine learning algorithms can outperform the conventional regression modeling techniques and usher in precision medicine by identifying new phenotypes of hypertensive disordersfrom large and globally represented datasets. To conclude, transdisciplinary approach based on epigenomic techniques and machine learning algorithms has immense potential to elucidate the biological underpinnings of life-course socioeconomic status and latter can easily emerge as an actionable metric of health vulnerability in marginalized and ethnic populations. None. The author declared no conflict of interest.