Emotionally supportive relationships have been found to be cardio protective. To identify new genetic loci that influence serum lipid levels, we searched genome-wide for gene-social support interactions in data from six studies of African ancestry with social support measures: [JHS (n=1,571); ARIC (n=1,773); CARDIA (n=921); CHS (n=726); MESA (n=1,641) and WHI (n=7,415)]. We hypothesized that interaction of low social support with genetic variants would lead to the discovery of novel lipid-loci in individuals of African ancestry that would be missed by studying main effects alone. Low social support was measured in each cohort using validated questionnaires and dichotomized at the lowest quartile. To combine the information across studies (N=14,047) we performed meta-analysis using METAL-patch for 3 serum lipids: fasting low density lipoprotein (LDL) cholesterol, fasting high density lipoprotein (HDL) cholesterol and fasting triglycerides (TG). Loci were defined by SNPs separated by physical distance greater than 1000kb. Each study contributed two interaction tests, the joint 2 degree of freedom (2df) test of the main and interaction effects and the traditional 1 df interaction.
We identified 16 loci previously known to be associated with HDL, LDL and TG and an additional 9 new lipid loci that were associated with LDL and TG only. Of the 9 new loci, six were significantly associated with LDL [CCR6; LNPK; LINC02005; PALLD; PACRG; PTPRN2) and 3 with TG (rs140102131; rs189825167; rs145542007). The 1df interaction test was instrumental in identifying PACRG, rs140102131 and rs145542007; but the rest were found by the joint 2df test. While loci associated with LDL have been reported previously to be associated with rheumatoid arthritis (CCR6); stabilization of the three-way junction in the endoplasmic reticulum (LNPK); leprosy and Parkinson Disease (PACRG); various cancers, and myocardial infarction (PTPRN2; PALLD), those associated with TG have not been associated with any phenotype. Replicating these results is the next step in demonstrating that incorporating gene-social support is a viable strategy for identifying genetic variants with novel effects on plasma lipids. *This work is supported in part by the NHLBI grant HL 118305.