We performed gene-environment interaction genome-wide association analysis (G × E GWAS) to identify SNPs whose effects on metabolic traits are modified by chronic psychosocial stress in the Multi-Ethnic Study of Atherosclerosis (MESA). In Whites, the G × E GWAS for hip circumference identified five SNPs within the Early B-cell Factor 1 (EBF1) gene, all of which were in strong linkage disequilibrium. The gene-by-stress interaction (SNP × STRESS) term P-values were genome-wide significant (Ps = 7.14E − 09 to 2.33E − 08, uncorrected; Ps = 1.99E − 07 to 5.18E − 07, corrected for genomic control). The SNP-only (without interaction) model P-values (Ps = 0.011-0.022) were not significant at the conventional genome-wide significance level. Further analysis of related phenotypes identified gene-by-stress interaction effects for waist circumference, body mass index (BMI), fasting glucose, type II diabetes status, and common carotid intimal-medial thickness (CCIMT), supporting a proposed model of gene-by-stress interaction that connects cardiovascular disease (CVD) risk factor endophenotypes such as central obesity and increased blood glucose or diabetes to CVD itself. Structural equation path analysis suggested that the path from chronic psychosocial stress to CCIMT via hip circumference and fasting glucose was larger (estimate = 0.26, P = 0.033, 95% CI = 0.02-0.49) in the EBF1 rs4704963 CT/CC genotypes group than the same path in the TT group (estimate = 0.004, P = 0.34, 95% CI = − 0.004-0.012). We replicated the association of the EBF1 SNPs and hip circumference in the Framingham Offspring Cohort (gene-by-stress term P-values = 0.007-0.012) as well as identified similar path relationships. This observed and replicated interaction between psychosocial stress and variation in the EBF1 gene may provide a biological hypothesis for the complex relationship between psychosocial stress, central obesity, diabetes, and cardiovascular disease.
INTRODUCTIONAlthough mortality attributable to cardiovascular disease (CVD) has declined in the United States, the burden of disease remains high. 1 It remains the leading cause of illness and death worldwide. Hypertension, obesity, dyslipidemia, insulin resistance and type II diabetes mellitus, and physical inactivity are among the eight risk factors that account for 61% of cardiovascular deaths. These same risk factors account for over three quarters of ischemic heart disease. 2 These risk factors are influenced by both environmental exposures and genetic background and the heritability of these risk factors can be as high as 77%, 3 making it difficult to clearly separate CVD risk factors into genetic and nongenetic categories. The INTERHEART study has evaluated the effect of both physical and psychosocial factors on the risk of myocardial infarction and has shown that a higher prevalence of psychological stress and other psychosocial factors like depression account for 34% of the population attributable risk for myocardial infarction, independently of physical risk factors. 4 The co-occ...