Skip to main content

Anthropometry - Adiposity (includes Physical Activity)

Genome-wide association study of height in Samoan individuals

Authors
Ryan L. Minster
Jenna C. Carlson
Jerry Z. Zhang
Nicola L. Hawley
Samantha L. Rosenthal
Hong Cheng
Take Naseri
Muagututi‘a Sefuiva Reupena
Satupa‘itea Viali
John Tuitele
Ranjan Deka
Stephen T. McGarvey
Daniel E. Weeks
TOPMed Anthropometry–Adiposity Working Group
Name and Date of Professional Meeting
2020 Annual Meeting of ASHG
Associated paper proposal(s)
Working Group(s)
Abstract Text
Background: To date, more than 2,300 independent genetic loci have been associated with height in individuals of European ancestries. Additional loci and additional insights into factors that affect stature might be discovered in heretofore unexamined populations. Here, we performed the first genome-wide association study of height in individuals of Polynesian ancestries with a sample of 3,075 participants from the Independent State of Samoa.

Methods: We genotyped 659,492 SNVs genome-wide on an Affymetrix 6.0 array. We then imputed 9,619,694 additional SNVs using a Samoan-specific haplotype reference panel derived from 1,284 Samoans whole-genome sequenced as part of the TOPMed Program. We tested inverse-normalized height for association via linear mixed models as implemented in GENESIS. Age, age², sex, age × sex, and age² × sex were included in the model as covariates. We adjusted for relatedness and population stratification with the first three principal components from PC-AiR and a kinship matrix from PC-Relate, respectively.

Results: There was one independent signal associated with height at p < 5 × 10⁻⁸, overlapping with the previously known height-associated locus near HHIP. The estimated effect size of the peak variant is 1.44 cm per copy of the minor allele. One variant in linkage disequilibrium (r² = 0.69) with the peak SNV is rs781199748 (HHIP:p.I582T), which is predicted by SIFT to be deleterious. Its minor allele frequency in this Samoan sample is 8.5%; whereas in gnomAD, the minor allele has been observed just once in 143,212 alleles. If this missense variant is the causal variant underlying the statistical association, this could indicate a novel allele associated with height, independent of other variation that has been previously observed to be associated with height at this locus.

Additionally, we observed six independent signals associated with height at p < 1 × 10⁻⁶, near BTRC, VPS37C, PAX2, OTUD7B, UQCRHL, and DIS3L2. SNVs near all but BTRC have been associated with height in other populations.
Back to top