Serious mental illness and cardiovascular health: Observational and polygenic risk score analyses in a population-based cohort study

Aims: Individuals with a serious mental illness have a markedly shorter life expectancy compared to the general population. A major contributor to premature death in this population is cardiovascular disease. The present study investigates the association of depressive disorder, bipolar disorder and schizophrenia, as well as polygenic liability to these three disorders, with cardiovascular disease (risk) and assesses the degree to which these associations are driven by psychotropic medication use and other important confounders.
Methods: We included participants of the Dutch Lifelines cohort (N=147,695) with information on (self-reported) lifetime diagnosis of depressive disorder (N=14,735), bipolar disorder (N=646) and schizophrenia (N=130) and cardiovascular disease (risk) traits (self-reported or physically measured): diastolic blood pressure, systolic blood pressure, hypertension, arrhythmia, atherosclerosis, heart failure, QTc interval, QRS duration, bundle branch block, left ventricular hypertrophy and heart rate variability. For a subset of participants, genome-wide genotype data were available (N=46,423). We conducted two sets of analyses, employing linear mixed effects models to correct for familial clustering. First, we examined associations between mental illness diagnosis and cardiovascular disease, correcting for psychotropic medication use and demographic and lifestyle factors, and testing sex differences. Second, we repeated these analyses using polygenic risk scores (PRS) as the independent variables, capturing polygenic liability for mental illness. 
Results: There was strong evidence that depressive disorder is associated with a higher risk of hypertension, arrhythmia, and atherosclerosis and with lower diastolic blood pressure, systolic blood pressure, heart rate variability, and QRS. These associations were largely robust to correction for psychotropic medication use and demographic and lifestyle factors. Positive associations with hypertension, arrhythmia, and atherosclerosis were also found when using a PRS for depression as the independent variable. While bipolar disorder was associated with a higher risk of nearly all cardiovascular disease (risk) traits, the majority of these associations diminished after adjustment for confounders. Using a PRS for bipolar disorder there was strong evidence for a positive association with hypertension, even after confounder adjustment. There was no clear evidence for an association of schizophrenia diagnosis with any cardiovascular disease trait (likely due to a lack of power), but when using a PRS for schizophrenia there was strong evidence for positive association with arrhythmia and a negative association with heart rate variability. 
Conclusions: Our study shows widespread associations of (genetic liability to) mental illness (primarily depressive disorder) with cardiovascular disease, even after adjustment for various confounders. More longitudinal or otherwise causally informative research is needed to better understand the mechanisms underlying these relationships.
Key words: depression, bipolar disorder, schizophrenia, cardiovascular disease, polygenic risk score