Premium
A Meta‐Analysis of the Transferability of Bone Mineral Density Genetic Loci Associations From European to African Ancestry Populations
Author(s) -
Yau Michelle S,
Kuipers Allison L,
Price Ryan,
Nicolas Aude,
Tajuddin Salman M,
Handelman Samuel K,
Arbeeva Liubov,
Chesi Alessandra,
Hsu YiHsiang,
Liu ChingTi,
Karasik David,
Zemel Babette S,
Grant Struan FA,
Jordan Joanne M,
Jackson Rebecca D,
Evans Michele K,
Harris Tamara B,
Zmuda Joseph M,
Kiel Douglas P
Publication year - 2021
Publication title -
journal of bone and mineral research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.882
H-Index - 241
eISSN - 1523-4681
pISSN - 0884-0431
DOI - 10.1002/jbmr.4220
Subject(s) - cohort , single nucleotide polymorphism , femoral neck , bone mineral , medicine , snp , bonferroni correction , genetics , locus (genetics) , demography , biology , genotype , osteoporosis , gene , statistics , mathematics , sociology
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 2 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 −4 ), was associated with lumbar spine BMD and two loci, 7q21 ( SLC25A13 , p = 2.84 × 10 −4 ) and 7q31 ( WNT16 , p = 2.96 × 10 −5 ), 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 −4 ), 11q14 ( DCDC5 , p < 5.35 × 10 −5 ), and 17p13 ( SMG6 , p < 6.78 × 10 −5 ) that were not tagged by European ancestry index SNPs. Rare single‐nucleotide variants in AKAP11 ( p = 2.32 × 10 −2 ), MBL2 ( p = 4.09 × 10 −2 ), MEPE ( p = 3.15 × 10 −2 ), SLC25A13 ( p = 3.03 × 10 −2 ), STARD3NL ( p = 3.35 × 10 −2 ), and TNFRSF11A ( p = 3.18 × 10 −3 ) 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).