Skip to main content
. Author manuscript; available in PMC: 2024 May 1.
Published in final edited form as: Thorax. 2022 May 2;78(5):432–441. doi: 10.1136/thoraxjnl-2021-217674

Figure 3.

Figure 3

Expected survival curves of rs138717703, representative of the four genome-wide significant SNPs in linkage disequilibrium (rs138717703, rs77506063, rs116023293 and rs145325916) associated with poor ICS response as measured by OCS bursts in a meta-analysis across GERA, MGB Biobank, and the Rotterdam Study.1 (A) GERA (HR 1.73 (95% CI 1.39 to 2.16), p=7.91×10−7, MAF 0.01), (B) MGB Biobank (HR 1.48 (95% CI 0.75 to 2.90), p=2.55×10−1, MAF 0.01), (C) Rotterdam Study (HR 3.34 (95% CI 1.86 to 6.03), p=5.86×10−5, MAF 0.01). 1One subject each in GERA and MGB Biobank had the minor allele homozygous genotype (GG) for all four SNPs. The direction of effect for the homozygous minor allele genotype differed across these two cohorts due to low minor allele frequency. No subjects in the Rotterdam Study had the homozygous minor allele genotype. C, cytosine; G, guanineGERA, Genetic Epidemiology Research on Adult and Aging; ICS, inhaled corticosteroid; MGB, Mass General Brigham; OCS, oral corticosteroid; SNPs, single-nucleotide polymorphisms.