IMPERFECT MATCH: MODELING PHYSIOLOGICAL STRESS IN INDIVIDUALS OF DISCORDANT SELF-REPORTED RACE AND GENETIC ANCESTRY

Abstract

In the United States, racial classification of African Americans is derived from the ‘one-drop rule’ and currently practiced through phenotypic presentation, but African Americans, having undergone recent genetic admixture, do not consist of only one geographic ancestry group. While previous research has also concluded that individuals may self-identify as one race but contain contradictory genetic ancestry, there is no literature that discusses whether or not living one’s life with this inconsistency may be associated with an increased genetic predisposition to stress-related health consequences, such as high blood pressure and diabetes.

In this study, I sought to determine if ancestrally-deviant African Americans and White Americans had a higher risk and stronger predictive power for physiological determinants of chronic stress compared to their ancestrally-concordant counterparts. To do so, I constructed a cohort study consisting of ancestrally-deviant and concordant African Americans and Caucasians (n = 1,076, 1,122, 6,444, and 8,920 participants, respectively) using genomic, demographic, and physiological data from the Health and Retirement Study (HRS). I then chose nine dependent variables based on current research that links them to long-term stress. Cohort creation via genetic ancestry principal component thresholds, model construction, statistical analysis, and subsequent data presentation were all performed in R.

There was a significant robustness in the type 2 diabetes (T2D) and BMI models in both ancestrally-deviant and concordant African American cohorts, seen in their corresponding adjusted R2 values (T2D; R2 = 0.575, 0.230 | BMI; R2 =0.237, 0.117). The models also predicted a positive correlation between AFR ancestry and BMI and T2D, suggesting that these two particular dependent variables may be more integral to determining the effects of discordant genetic ancestry and self-identified race and ethnicity (SIRE).

While a majority of the models proved to be either inconclusive or have low predicting power, the contribution of this paper to both the literature concerning geographic ancestry as well as the current pharmacogenetics industry positions itself as a relevant guideline for future related works.