Manousaki Despoina, Mokry Lauren E, Ross Stephanie, Goltzman David, Richards J Brent
From the Department of Epidemiology, Centre for Clinical Epidemiology, Lady Davis Institute for Medical Research, Jewish General Hospital, Montreal, Quebec, Canada (D.M., L.E.M., S.R., J.B.R.); Departments of Medicine (D.G., J.B.R.) and Human Genetics (J.B.R.), McGill University, Montreal, Quebec, Canada; and Department of Twin Research and Genetic Epidemiology, King's College London, United Kingdom (J.B.R.).
Circ Cardiovasc Genet. 2016 Aug;9(4):349-56. doi: 10.1161/CIRCGENETICS.116.001396. Epub 2016 Jul 14.
Observational studies support a possible association between decreased vitamin D levels and risk of coronary artery disease (CAD); however, it remains unclear whether this relationship is causal. We aimed to evaluate whether genetically lowered vitamin D levels influence the risk of CAD using a Mendelian randomization approach.
In this 2-stage Mendelian randomization study, we first identified single-nucleotide polymorphisms associated with 25-hydroxyvitamin D (25OHD) levels in the SUNLIGHT consortium (n=33 996), then tested them for possible violation of Mendelian randomization assumptions. A count of risk alleles was tested for association with 25OHD levels in a separate cohort (n=2347). Alleles were weighted by their relative effect on 25OHD and tested for their combined effect on CAD in the Coronary Artery Disease Genome-Wide Replication and Meta-Analysis (CARDIoGRAM) study (22 233 cases/64 762 controls). Four single-nucleotide polymorphisms were identified to be associated with 25OHD levels, all in or near genes implicated in 25OHD synthesis, transport or metabolism. A count of these risk alleles was strongly associated with 25OHD (n=2347, F-test statistic=49.7, P=2×10(-12)). None of the single-nucleotide polymorphisms associated with 25OHD levels were associated with CAD (all P values >0.6). The Mendelian randomization odds ratio (OR) for CAD was 0.99 (95% confidence interval, 0.84-1.17; P=0.93; I(2)=0) per SD decrease in log-transformed 25OHD levels. These results persisted after sensitivity analyses for population stratification and pleiotropy.
Genetically lowered 25OHD levels were not associated with increased risk of CAD in a large, well-powered study, suggesting that previous associations between circulating 25OHD levels and CAD are possibly confounded or due to reverse causation.
观察性研究支持维生素D水平降低与冠状动脉疾病（CAD）风险之间可能存在关联；然而，这种关系是否为因果关系仍不清楚。我们旨在使用孟德尔随机化方法评估基因水平降低的维生素D是否会影响CAD风险。
在这项两阶段孟德尔随机化研究中，我们首先在SUNLIGHT联盟中确定了与25-羟维生素D（25OHD）水平相关的单核苷酸多态性（n=33996），然后测试它们是否可能违反孟德尔随机化假设。在另一个队列（n=2347）中测试风险等位基因的计数与25OHD水平的关联。根据等位基因对25OHD的相对影响对其进行加权，并在冠状动脉疾病全基因组复制和荟萃分析（CARDIoGRAM）研究（22233例病例/64762例对照）中测试它们对CAD的综合影响。确定了四个与25OHD水平相关的单核苷酸多态性，均位于参与25OHD合成、转运或代谢的基因内或附近。这些风险等位基因的计数与25OHD密切相关（n=2347，F检验统计量=49.7，P=2×10-12）。与25OHD水平相关的单核苷酸多态性均与CAD无关（所有P值>0.6）。每标准差对数转换的25OHD水平下降，CAD的孟德尔随机化优势比（OR）为0.99（95%置信区间，0.84-1.17；P=0.93；I(2)=0）。在对人群分层和多效性进行敏感性分析后，这些结果仍然成立。
在一项规模大、功效强的研究中，基因水平降低的25OHD与CAD风险增加无关，这表明先前循环25OHD水平与CAD之间的关联可能存在混淆或由反向因果关系导致。
