Rare genetic variants impact muscle strength

Abstract

Muscle strength is highly heritable and predictive for multiple adverse health outcomes including mortality. Here, we present a rare protein-coding variant association study in 340,319 individuals for hand grip strength, a proxy measure of muscle strength. We show that the exome-wide burden of rare protein-truncating and damaging missense variants is associated with a reduction in hand grip strength. We identify six significant hand grip strength genes, KDM5B, OBSCN, GIGYF1, TTN, RB1CC1, and EIF3J. In the example of the titin (TTN) locus we demonstrate a convergence of rare with common variant association signals and uncover genetic relationships between reduced hand grip strength and disease. Finally, we identify shared mechanisms between brain and muscle function and uncover additive effects between rare and common genetic variation on muscle strength.

Fig. 1. Genome-wide rare variant burden associations with hand grip strength.

a Rare variant burden associations with hand grip strength stratified on PTV, very damaging missense (CADD > 30), damaging missense (CADD 20–30), other missense (CADD 0–20), and synonymous variants. Associations were tested for all, LoF intolerant (pLI > 0.9) and LoF tolerant (pLI < 0.9) genes, respectively. P-values were derived from two-sided t-tests of linear regression coefficients. A Bonferroni correction was used to adjust for multiple testing. No. of genes: number of genes included in each rare variant burden association test; P.adj: Bonferroni-adjusted P-values. b, c Manhattan and QQ plot of gene-based PTV-burden associations with hand grip strength. P-values were derived from two-sided t-tests of linear regression coefficients. Bonferroni and false discovery rate (FDR) correction were used to adjust for multiple testing. The Red dashed line indicates the Bonferroni significance threshold, the red circle indicates significant genes that passed the Bonferroni threshold, and orange circles indicate genes with FDR < 0.05. Source data are provided in Supplementary Data 1 and 2.

Fig. 2. Converging rare and common variant associations of TTN with hand grip strength.

a Single PTV association of TTN with hand grip strength. The average effect sizes of each PTV in TTN were plotted against their genomic positions, with exon numbers and protein domains visualized at the bottom. ClinVar pathogenic/likely pathogenic PTVs were colored in red with different shapes indicating their respective associated disease phenotypes. b hand grip strength GWAS locus plot of TTN. P-values were derived from two-sided t-tests implemented in regenie. The purple diamond indicates the sentinel variant at the TTN GWAS locus. SNP IDs label the top 95% credible set identified by SuSiE. Colors indicate different levels of LD with the sentinel variant. Source data are provided in Supplementary Data 10 and the Source Data file.

Fig. 3. KDM5B loss-of-function causes a dose-dependent reduction in hand grip strength in humans and mice.

a Normalized hand grip strength stratified for KDM5B PTV carrier status in N = 340,319 independent UK Biobank participants. Hand grip strength was normalized against standing height, then residualized on age, sex, and PCs. Z-scores were generated through inverse-rank normalization, then compared between KDM5B PTV carriers vs. non-carriers (WT). P-values were derived using two-sided t-tests. ***P = 2.86 × 10⁻¹². b Forelimb grip strength of mice stratified on Kdm5b genotype in N = 55 biologically independent animals. FGS was normalized against femur length, then corrected for the cohort effect. Z-scores calculated from residuals were then compared across heterozygous (HET) and homozygous (HOM) Kdm5b mutant and wild-type (WT) animals. P-values were derived using two-sided t-tests. *P = 0.016; ***P = 0.0002; center line: median; box limits: upper and lower quartiles; whiskers: minimum or lower quartile minus 1.5 times IQR and maximum or upper quartile plus 1.5 times IQR; Source data are provided in the Source Data file.