Genetic and epigenetic correlates of behavioral risk and protective factors in aging

Epigenetic changes play a role in the biological pathways associated with psychosocial and cognitive outcomes. Little research, however, has been devoted to understanding whether differences in methylation for one psychosocial phenotype also accounts for differential methylation of other phenotypes. The objective of the proposed research is to estimate the associations between epigenetic modifications (i.e., DNA methylation) and behavioral and psychosocial risk factors (e.g., depressive symptomatology, loneliness, low physical activity) previously shown to predict cognitive decline in older adulthood. Our principal analyses include epigenome-wide association studies of the principal phenotypes of interest to quantify differential methylation at specific CpG sites for various phenotypes of Health and Retirement Study participants who were enrolled in the 2016 Venous Blood Study. We subsequently will estimate correlations between the beta weights generated from these EWASs to quantify whether differences in methylation underlying psychosocial and cognitive phenotypes covary. Follow-up analyses will include gene-enrichment analyses to identify pathways that contain genes that are overexpressed for two phenotypes (e.g., loneliness and cognition). In addition, we will construct DNA methylation age acceleration measures using principal components methodologies that improve the reliability of DNAm age variables. The phenotypic characteristics proposed for the present study are broad and include self-report measures of current and previous smoking use, frequency of social interactions, loneliness, depressive symptomatology, self-reported physical activity, subjective memory rating, and cognitive ability measures (e.g., immediate and delayed word recall, backward counting, serial 7s, verbal fluency, vocabulary, quantitative reasoning, and verbal reasoning).

The general purpose of the study is to identify epigenetic correlates of behavioral risk factors and cognitive decline in older adults. For example, our planned analyses will investigate epigenome-wide associations with loneliness, depressive symptomatology, self-reported physical activity, and cognition. Further, we will identify overlapping CpG sites between behavioral risk factors and cognition in older adulthood, and use gene ontology to identify the differentially expressed genes that are enriched in biological processes relevant to disease risk and aging.