Low mitochondrial copy number drives atherogenic cardiovascular disease: evidence from prospective cohort analyses in the UK Biobank combined with Mendelian Randomization

Background Mitochondrial DNA (mtDNA) content might be involved in the risk of cardiovascular disease. We aimed to investigate the association of mtDNA copy number (mtDNA-CN), as a proxy of mtDNA content, and coronary artery disease (CAD) and heart failure (HF) using multivariable adjusted and Mendelian Randomizations (MR) analyses. Methods Multivariable-adjusted analyses were conducted using Cox-proportional hazard models in 273,619 unrelated European descendants from UK Biobank (UKB). MtDNA-CN in peripheral blood cells was computed based on the weighted intensities of the mitochondrial genome probes. For the two-sample MR analyses, single nucleotide polymorphisms (SNPs) associated with mtDNA-CN were retrieved from genome-wide association studies in UKB. SNP-outcome associations were obtained for CAD from CARDIoGRAMplusC4D, UKB and FinnGen, comprising 902,538 participants (134,759 cases), and for HF from the HERMES consortium and FinnGen, collectively having data on 1,195,531 participants (70,706 cases). MR analyses were performed per database and results were subsequently meta analyzed using fixed-effects models per study. Results During a median follow-up of 11.8 years, participants in the lowest quintile of mtDNA-CN had higher risk for CAD (hazard ratio [95% CI]: 1.08 [1.03, 1.14]) and HF (hazard ratio [95% CI]: 1.15 [1.05, 1.24]) compared to those in the highest quintile. In MR analyses, the pooled odds ratios of genetically predicted per one-SD decrease in mtDNA were 1.16 (95% CI: 1.05, 1.27) for CAD and 1.00 (95% CI: 0.90, 1.10) for HF, respectively. Conclusion Our findings support a possible causal role of lower mtDNA-CN in higher CAD risk, but not in higher HF risk.