Abstract Body: Introduction
Observational studies have shown that coffee consumption relates to a decreased risk of heart failure (HF). Our recent findings indicate that caffeine metabolism, not caffeine itself, relates to the onset of HF. The causal effect of circulating caffeine levels on HF and cardiac function is largely inconclusive.
Hypothesis
We hypothesize that circulating caffeine level is not causally associated with cardiac function and HF.
Methods
We extracted the summary results from genome-wide association (GWAS) studies of plasma caffeine, HF, and 15 cardiac structure and function measures for left atrium, left ventricle, and right ventricle from public databases. Genetic instruments of caffeine were selected if the variants were associated with caffeine at p<5×10^-8. After clumping at r<0.001, two independent lead variants on gene CYP1A2, rs2472297, and AHR, rs4410790, were included in the analysis. Using the two-sample Mendelian Randomization (MR) approach, we examined the associations between caffeine level, cardiac function measurements, and HF on each selected variant, and pooled estimates were generated using the inverse-variance weighted method.
Results
The summary of caffeine, cardiac function, and HF was from GWAS of European descent, consisting of 9,876 participants from six cohorts, up to 29,506 participants from UK Biobank, and 47,309 cases and 93,014 controls from the Heart Failure Molecular Epidemiology for Therapeutic Targets study, respectively. CYP1A2 and AHR were not causally associated with HF individually or jointly (Odds Ratio_meta=1.001, 95% CI = 0.999, 1.004). Both genes also yield insignificant results between caffeine levels and measures of the left ventricle, left atrium, and right ventricle, key parameters underlying HF (Figure).
Conclusion
We observed no causal association between circulating caffeine and the risk of HF, suggesting health-beneficial effects of coffee consumption may relate to the improved molecular pathways instead of circulating caffeine levels.