Introduction: Heart failure (HF) is commonly subgrouped into HF with preserved and reduced ejection fraction (HFpEF and HFrEF). Genome-wide association studies (GWAS) of all-HF have identified few risk loci. No GWAS of HF subtypes have been reported and whether genetic risk for HF subtypes is shared is unknown. We hypothesized that GWAS could be used to elucidate unique and shared genetics mechanisms for HF subtypes.
Methods: We assigned HF status in 6,149 European ancestry individuals ascertained for cardiovascular phenotypes in BioVU, the Vanderbilt biobank linked to de-identified electronic health records. HF subtype was assigned using echocardiographic data when available (HFpEF: LVEF≥50%; HFrEF: ≤40%). Our dataset contained 3,043 HF cases (965 HFpEF; 1,214 HFrEF) and 3,106 non-HF controls. Subjects were genotyped on the Illumina Multi-Ethnic Genotyping Array and imputed to 1000Genomes. The association of SNPs with HFpEF and HFrEF was assessed with logistic regression, assuming an additive model and adjusting for sex, age at HF diagnosis (cases) or age at last record (controls), BMI, MI, hypertension, diabetes, eGFR, and principal components 1-3.
Results: One SNP (rs149963827) in a long noncoding RNA on chromosome 1 associated with HFrEF at genome wide significance (p=4.04e-08, Figure 1). No SNPs reached this threshold with HFpEF. Of 1,821 SNPs moderately associated (<1e-04) with either subtype, only 3 were shared. Shared SNPs were in SGCZ, a component of the dystrophin-glycoprotein complex, VDR, a gene previously implicated in HF, and FGF14, one of a family of proteins with wide biologic activity.
Conclusions: These findings suggest a possible role for non-coding RNAs in HFrEF risk. Strongly associated genetic modifiers of HFpEF and HFrEF are largely distinct, but shared SNPs annotate to biologically plausible genes. Larger studies are needed to replicate these findings, further elucidate subtype-specific risk loci, and evaluate shared polygenic risk.