Department of Population Health Sciences, 215 Morris Street, Duke University School of Medicine, Durham, NC 27701, United States.
Department of Sociology, 102 Emerson Drive, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599, United States.
Hum Mol Genet. 2024 Oct 7;33(20):1748-1757. doi: 10.1093/hmg/ddae112.
Obesity and poverty disproportionally affect African American persons. Epigenetic mechanisms could partially explain the association between socioeconomic disadvantage and body mass index (BMI). We examined the extent to which epigenetic mechanisms mediate the effect of socioeconomic status (SES) on BMI. Using data from African American adults from the Atherosclerosis Risk in Communities (ARIC) Study (n = 2664, mean age = 57 years), education, income, and occupation were used to create a composite SES score at visit 1 (1987-1989). We conducted two methylation-wide association analyses to identify associations between SES (visit 1), BMI and cytosine-phosphate-guanine (CpG) sites measured at a subsequent visit (1990-1995). We then utilized structural equation modeling (SEM) to test whether identified sites mediated the association between earlier SES and BMI in sex-stratified models adjusted for demographic and risk factor covariates. Independent replication and meta-analyses were conducted using the Jackson Heart Study (JHS, n = 874, mean age 51 years, 2000-2004). Three CpG sites near MAD1L1, KDM2B, and SOCS3 (cg05095590, cg1370865, and cg18181703) were suggestively associated (P-value < 1.3×10-5) in ARIC and at array-wide significance (P-value < 1.3×10-7) in a combined meta-analysis of ARIC with JHS. SEM of these three sites revealed significant indirect effects in females (P-value < 5.8×10-3), each mediating 7%-20% of the total effect of SES on BMI. Nominally significant indirect effects were observed for two sites near MAD1L1 and KDM2B in males (P-value < 3.4×10-2), mediating -17 and -22% of the SES-BMI effect. These results provide further evidence that epigenetic modifications may be a potential pathway through which SES may "get under the skin" and contribute to downstream health disparities.
肥胖和贫困不成比例地影响非裔美国人。表观遗传机制可以部分解释社会经济劣势与体重指数（BMI）之间的关联。我们研究了表观遗传机制在多大程度上介导了社会经济地位（SES）对 BMI 的影响。使用来自社区动脉粥样硬化风险研究（ARIC）的非裔美国成年人的数据（n=2664，平均年龄=57 岁），在第一次访问（1987-1989 年）时使用教育、收入和职业来创建一个综合 SES 评分。我们进行了两次全基因组甲基化关联分析，以确定 SES（第一次访问）、BMI 和在下一次访问（1990-1995 年）测量的胞嘧啶-磷酸-鸟嘌呤（CpG）位点之间的关联。然后，我们利用结构方程模型（SEM）在性别分层模型中测试了在调整人口统计学和风险因素协变量后，确定的位点是否介导了早期 SES 与 BMI 之间的关联。利用杰克逊心脏研究（JHS，n=874，平均年龄 51 岁，2000-2004 年）进行了独立复制和荟萃分析。在 ARIC 中，三个位于 MAD1L1、KDM2B 和 SOCS3 附近的 CpG 位点（cg05095590、cg1370865 和 cg18181703）与 ARIC 的全基因组显著性（P 值<1.3×10-7）显著相关。SEM 显示，这三个位点在女性中有显著的间接效应（P 值<5.8×10-3），每个位点都介导了 SES 对 BMI 总效应的 7%-20%。在男性中，MAD1L1 和 KDM2B 附近的两个位点观察到名义上显著的间接效应（P 值<3.4×10-2），分别介导了 SES-BMI 效应的-17%和-22%。这些结果进一步证明，表观遗传修饰可能是 SES 可能“深入皮肤”并导致下游健康差异的潜在途径。
