Mood instability may be causally associated with the high risk of cardiovascular disease: evidence from a mendelian randomization analysis

BackgroundMental illness was identified associated with high risk of cardiovascular diseases (CVDs). However, few studies focused on the effect of personality traits, the causal relationships remain unknown. Here, we use mendelian randomization (MR) analyses to evaluate the causal association between mood instability (mood swings) and 5 common CVDs. MethodsLarge genome-wide association studies (GWAS) of mood swings (n= 373733) and 5 CVDs from two independent cohorts respectively including coronary artery disease (CAD) (n= 766053), myocardial infarction (MI) (n= 596436), heart failure (HF) (n= 1185501), atrial fibrillation (AF) (n= 2169833) and stroke (n = 627558). We performed a range of bidirectional two-sample MR and related sensitive analysis including MR-Egger regression, MR-PRESSO global test and "Leave-one-out" method. A Bonferroni-corrected significance level of p < 0.01 (0.05/5) was identified to be statistically significant, while p < 0.05 was considered to indicate suggestive evidence. Moreover, multivariable MR (MVMR) and mediation analyses were also conducted to adjust confounding factors as well as found potential mediators. ResultsThis MR analyses revealed the significant causal effects of mood swings on CAD (OR = 1.45, 95% CI 1.24-1.71; P = 5.52e-6), MI (OR = 1.60, 95% CI 1.32-1.95; P = 1.77e-6), HF (OR = 1.42, 95% CI 1.12-1.71; P = 2.32e-6) and stroke (OR = 1.48, 95% CI 1.19-1.83; P = 3.46e-4). However, no causal effects of mood swings on AF (P=0.16) were found. In the reverse MR, no causal relationships were observed. Additionally, hypertension may mediate the causal pathway from mood swings to CAD (proportion of mediation effect in total effect: 39.60%, 95% CI: 19.31%-59.89%), MI (35.37%, 95% CI: 17.10%-53.65%), HF (43.19%, 95% CI: 20.68%-65.69%) and stroke (55.47%, 95% CI: 27.00%-83.95%). ConclusionMood instability (mood swings) causally resulted in CAD, MI, HF and stroke, and these causal effects may be partly mediated by hypertension.