Hinda and Arthur Marcus Institute for Aging Research, Hebrew SeniorLife, Boston, MA, USA.
Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, MA, USA.
J Bone Miner Res. 2021 Mar;36(3):469-479. doi: 10.1002/jbmr.4220. Epub 2020 Dec 18.
Genetic studies of bone mineral density (BMD) largely have been conducted in European populations. We therefore conducted a meta-analysis of six independent African ancestry cohorts to determine whether previously reported BMD loci identified in European populations were transferable to African ancestry populations. We included nearly 5000 individuals with both genetic data and assessments of BMD. Genotype imputation was conducted using the 1000G reference panel. We assessed single-nucleotide polymorphism (SNP) associations with femoral neck and lumbar spine BMD in each cohort separately, then combined results in fixed effects (or random effects if study heterogeneity was high, I index >60) inverse variance weighted meta-analyses. In secondary analyses, we conducted locus-based analyses of rare variants using SKAT-O. Mean age ranged from 12 to 68 years. One cohort included only men and another cohort included only women; the proportion of women in the other four cohorts ranged from 52% to 63%. Of 56 BMD loci tested, one locus, 6q25 (C6orf97, p = 8.87 × 10 ), was associated with lumbar spine BMD and two loci, 7q21 (SLC25A13, p = 2.84 × 10 ) and 7q31 (WNT16, p = 2.96 × 10 ), were associated with femoral neck BMD. Effects were in the same direction as previously reported in European ancestry studies and met a Bonferroni-adjusted p value threshold, the criteria for transferability to African ancestry populations. We also found associations that met locus-specific Bonferroni-adjusted p value thresholds in 11q13 (LRP5, p < 2.23 × 10 ), 11q14 (DCDC5, p < 5.35 × 10 ), and 17p13 (SMG6, p < 6.78 × 10 ) that were not tagged by European ancestry index SNPs. Rare single-nucleotide variants in AKAP11 (p = 2.32 × 10 ), MBL2 (p = 4.09 × 10 ), MEPE (p = 3.15 × 10 ), SLC25A13 (p = 3.03 × 10 ), STARD3NL (p = 3.35 × 10 ), and TNFRSF11A (p = 3.18 × 10 ) were also associated with BMD. The majority of known BMD loci were not transferable. Larger genetic studies of BMD in African ancestry populations will be needed to overcome limitations in statistical power and to identify both other loci that are transferable across populations and novel population-specific variants. © 2020 American Society for Bone and Mineral Research (ASBMR).
骨矿物质密度（BMD）的遗传研究主要在欧洲人群中进行。因此，我们进行了一项荟萃分析，共纳入了六个独立的非洲裔血统队列，以确定在欧洲人群中发现的先前报道的 BMD 位点是否可转移到非洲裔血统人群。我们纳入了近 5000 名既有遗传数据又有 BMD 评估的个体。使用 1000G 参考面板进行单核苷酸多态性（SNP）的基因型推断。我们分别在每个队列中评估了股骨颈和腰椎 BMD 的 SNP 关联，然后在固定效应（如果研究异质性高，I 指数 >60，则进行随机效应）逆方差加权荟萃分析中合并结果。在二次分析中，我们使用 SKAT-O 进行了罕见变异的基于基因座的分析。平均年龄范围为 12 岁至 68 岁。一个队列仅包括男性，另一个队列仅包括女性；另外四个队列中女性的比例在 52%至 63%之间。在 56 个 BMD 位点中，一个位点（6q25，C6orf97，p=8.87×10）与腰椎 BMD 相关，两个位点（7q21，SLC25A13，p=2.84×10）和 7q31（WNT16，p=2.96×10）与股骨颈 BMD 相关。效应与欧洲血统研究中先前报道的方向一致，并且达到了 Bonferroni 校正 p 值阈值，即可转移到非洲裔血统人群的标准。我们还发现了 11q13（LRP5，p<2.23×10）、11q14（DCDC5，p<5.35×10）和 17p13（SMG6，p<6.78×10）三个基因座的关联，这些关联达到了基于基因座的 Bonferroni 校正 p 值阈值，但是未被欧洲血统索引 SNP 标记。AKAP11（p=2.32×10）、MBL2（p=4.09×10）、MEPE（p=3.15×10）、SLC25A13（p=3.03×10）、STARD3NL（p=3.35×10）和 TNFRSF11A（p=3.18×10）中的罕见单核苷酸变异也与 BMD 相关。大多数已知的 BMD 位点不可转移。需要对非洲裔血统人群的 BMD 进行更大规模的遗传研究，以克服统计能力的限制，并确定可在人群间转移的其他位点以及新的人群特异性变异。
