1] Department of Epidemiology, Gillings School of Global Public Health, University of North Carolina, Chapel Hill, NC, USA [2] Carolina Population Center, University of North Carolina, Chapel Hill, NC, USA.
Nutr Diabetes. 2012 Sep 24;2(9):e47. doi: 10.1038/nutd.2012.20.
The contribution of genetic variants to body mass index (BMI) during adolescence across multiethnic samples is largely unknown. We selected genetic loci associated with BMI or obesity in European-descent samples and examined them in a multiethnic adolescent sample.
In 5103 European American (EA), 1748 African American (AfA), 1304 Hispanic American (HA) and 439 Asian American (AsA) participants of the National Longitudinal Study of Adolescent Health (Add Health; ages 12-21 years, 47.5% male), we assessed the association between 41 established obesity-related single-nucleotide polymorphisms (SNPs) with BMI using additive genetic models, stratified by race/ethnicity, and in a pooled meta-analysis sample. We also compared the magnitude of effect for BMI-SNP associations in EA and AfA adolescents to comparable effect estimates from 11 861 EA and AfA adults in the Atherosclerosis Risk in Communities study (ages 45-64 years, 43.2% male).
Thirty-five of 41 BMI-SNP associations were directionally consistent with published studies in European populations, 18 achieved nominal significance (P<0.05; effect sizes from 0.19 to 0.71 kg m(-2) increase in BMI per effect allele), while 4 (FTO, TMEM18, TFAP2B, MC4R) remained significant after Bonferroni correction (P<0.0015). Of 41 BMI-SNP associations in AfA, HA and AsA adolescents, nine, three and five, respectively, were directionally consistent and nominally significant. In the pooled meta-analysis, 36 of 41 effect estimates were directionally consistent and 21 of 36 were nominally significant. In EA adolescents, BMI effect estimates were larger (P<0.05) for variants near TMEM18, PTER and MC4R and smaller for variants near MTIF3 and NRXN3 compared with EA adults.
Our findings suggest that obesity susceptibility loci may have a comparatively stronger role during adolescence than during adulthood, with variation across race/ethnic subpopulation.
在多民族样本中,遗传变异对青少年时期体重指数(BMI)的贡献在很大程度上尚不清楚。我们选择了与欧洲血统样本中的 BMI 或肥胖相关的遗传位点,并在多民族青少年样本中对其进行了研究。
在全国青少年健康纵向研究(Add Health;年龄 12-21 岁,47.5%为男性)的 5103 名欧洲裔美国人(EA)、1748 名非裔美国人(AfA)、1304 名西班牙裔美国人(HA)和 439 名亚裔美国人(AsA)参与者中,我们使用加性遗传模型评估了 41 个已确定的与肥胖相关的单核苷酸多态性(SNP)与 BMI 之间的关联,按种族/族裔进行分层,并在合并的荟萃分析样本中进行了比较。我们还比较了 EA 和 AfA 青少年中 BMI-SNP 关联的效应大小与社区动脉粥样硬化风险研究(Atherosclerosis Risk in Communities study)中 11861 名 EA 和 AfA 成年人中可比的效应估计值(年龄 45-64 岁,43.2%为男性)。
41 个 BMI-SNP 关联中有 35 个与欧洲人群的已发表研究方向一致,18 个达到了名义显著性(P<0.05;每个效应等位基因使 BMI 增加 0.19-0.71kg/m2),而在 Bonferroni 校正后(P<0.0015),4 个(FTO、TMEM18、TFAP2B、MC4R)仍然显著。在 AfA、HA 和 AsA 青少年中,有 41 个 BMI-SNP 关联,分别有 9、3 和 5 个方向一致且具有名义显著性。在合并的荟萃分析中,41 个效应估计中有 36 个方向一致,其中 21 个具有名义显著性。在 EA 青少年中,与 EA 成年人相比,TMEM18、PTER 和 MC4R 附近的变异体的 BMI 效应估计值更大(P<0.05),而 MTIF3 和 NRXN3 附近的变异体的 BMI 效应估计值更小。
我们的研究结果表明,肥胖易感基因座在青少年时期可能比成年时期具有更强的作用,并且在不同种族/族裔亚群中存在差异。