Mental disorders and cardiovascular disease: not just an issue of older age

This editorial refers to ‘Increased cardiovascular events in young patients with mental disorders: a nationwide cohort study’, by C.S. Park et al., https://doi.org/10.1093/eurjpc/zwad102.

It has long been accepted that mental disorders are associated with increased incidence of cardiovascular diseases (CVDs) and their adverse prognosis.^1–4 While most research has been conducted on depression-associated cardiovascular risk, there is also some evidence that other mental disorders such as anxiety disorders,⁵ posttraumatic stress disorder (PTSD⁶), psychoses (e.g.⁷), and various other stress-related conditions have an adverse impact on the cardiovascular system.

Most studies on prognostic effects of mental disorders have focused on single conditions in middle-aged populations or older patients with manifest CVD. In contrast, in this issue Park et al.⁸ present data from a huge Korean population database of n = 6 557 727 young persons to show that virtually all examined mental disorders, whether analyzed in total or separately, were related to incident coronary heart disease (CHD) and stroke. The mental disorders studied included not only the ‘usual suspects’ mentioned above but effects were also found for insomnia as well as somatoform, personality, eating, and substance abuse disorders.

Across the various mental disorders, adjusted hazard ratios differed from 1.49 to 3.13 for predicting incident myocardial infarction and from 1.17 to 3.06 for ischemic stroke. These effects may even be underestimated as only persons with a routine diagnosis of a mental disorder were considered. Given the known underdiagnosis of mental disorders in medical routine, persons with undiagnosed mental disorders may have biased the results toward null effects. Still, all but two hazard ratios were statistically significant and the only two conditions that didn’t significantly predict ischemic stroke, i.e. PTSD and eating disorders, were at least clearly associated with incident CHD, which was predicted by each single mental diagnosis.

This finding may come as a surprise as it challenges the assumption of type and severity of mental disorders as the main factors explaining elevated CVD risk. Interestingly, PTSD often not considered a ‘severe’ mental illness conferred the highest CHD risk of all analyzed conditions. Depression that has been a main focus of research linking mental disorders to CVD during the last three decades only ranked 8th out of 10 specific conditions in terms of its predictive effect for CHD. Psychoses, substance use disorders, and even eating disorders had numerically higher adjusted hazard ratios for predicting CHD. For ischemic stroke, personality disorders had the highest adjusted hazard ratio, followed by schizophrenia and bipolar disorder, while depression only ranked 6th out of 10 disorders in predicting stroke risk.

One might wonder whether the cardiovascular effect of mental disorders can be explained by the higher traditional risk factor burden often found in the mentally ill. In the present sample, risk factor burden indeed differed significantly between persons with vs. without mental disorders. However, the high significance levels are mainly due to the huge sample size. Numerical differences are small and in many cases opposite to what one might expect. Furthermore, the Cox regression analyses were adjusted for a variety of established risk factors such as diabetes, smoking, dyslipidemia, hypertension, or physical inactivity. Adjustment resulted in only minor changes from univariate hazard ratios, suggesting that other mechanisms must be involved.

It might be possible that some uncontrolled factors confounded the effect. However, there are few possible confounders one could think of. One candidate could be psychotropic drugs taken to treat mental disorders. It is well-known that some antidepressants and antipsychotics can have unfavorable metabolic and/or autonomic effects. However, many psychotropic drugs are considered unproblematic with respect to their impact on the cardiovascular system. Unfortunately, Park et al. were unable to present data on psychotropic medications taken by their probands and more research looking at long-term cardiovascular endpoints in patients treated with different psychotropic drugs is certainly warranted in order to rule out a relevant drug-induced increase in CVD risk. Nevertheless, the wide variation in cardiometabolic side effects of psychotropic drugs and the different indications for the relevant drug classes makes a uniform adverse drug effect across the various mental disorders studied here quite unlikely.

Physiological mechanisms linking mental disorders to CVD include autonomic and neuroendocrine dysbalance, impaired vasodilation, hypercoagulable states, and inflammation. However, most such associations have been reported for depression and the question arises what the disorders studied by Park and colleagues have in common. Chronic or periodic distress is a common feature of most mental disorders. Chronic stress is related to similar physiological effects as depression, e.g. chronic low-grade inflammation. Although the relevant literature is not fully consistent, persons with mental disorders may be at increased risk of showing excessive physiological reactions to acute stressors as, for example, data from patients with Takotsubo Syndrome suggest. Acute stress is also a frequent trigger of cardiovascular events. Chronic and acute distress can therefore be assumed to play a major role in explaining the global effect.

Another possible explanation is a common risk factor predisposing affected persons to developing both, mental disorders and CVD. Genetic factors may be of relevance here but a general genetic factor predisposing to both, various mental and cardiovascular conditions is unknown. In addition, the effect of PTSD—which is by definition caused by environmental rather than genetic influences—argues against a major role of a single genetic factor.

It is more likely that a substantial proportion of variability can be attributed to adverse experiences during childhood or adolescence that are known to increase the risk for mental disorders, various cardiovascular risk factors, and CVD later in life.⁹

While most studies on cardiovascular effects of mental disorders or childhood adversities have looked at cardiovascular events in middle or older age, Park et al.⁸ studied young adults aged 39 or younger (mean age 30.9 years) at baseline who experienced their CVD events during the subsequent follow-up period lasting a median of 7.6 years. This indicates that events also occurred at an unusually young age, in the majority of cases at less than 40 years. The observed effects of some mental disorders were even stronger in the younger (20–29 years old) vs. older (30–39 years) subgroup, despite lower absolute cardiovascular event rates in the younger group. A causal factor starting to act early in life would be an obvious explanation for early-onset CVD. A few studies have looked at young adults to show that psychosocial factors such as poor perceived parenting or low resilience levels were followed by early-onset CVD (e.g.¹⁰). The same factors are also known to predispose to mental disorders.

The remaining question is how early adversity gets ‘under the skin’ to induce both, mental and cardiovascular conditions. A scientific statement from the American Heart Association⁹ provides a rough overview of suspected mechanisms. Besides unhealthy behaviors, the authors mention the effect on mental disorders (with their known effects) as mediators and briefly discuss biological alterations directly induced by early adversities. These alterations are similar to those thought to mediate the effect of mental disorders on the cardiovascular system and may be summarized as a learned biological hypersensitivity to stress, leading to both, direct short- and long-term effects on the cardiovascular system and compensatory behavior such as unhealthy eating or smoking in order to ward off mental distress. There is some evidence from the literature that such alterations in organismic stress sensitivity may be epigenetically anchored and can thus persist into adulthood.

However, exploring these mechanisms further requires more longitudinal research involving plausible behavioral biomarkers as predictors. This research can receive further support by epidemiological findings such as the paper by Park et al.⁸ that may serve as one more piece in the puzzle of longitudinal effects of psychological and social factors influencing CVD.

Author notes

The opinions expressed in this article are not necessarily those of the Editors of the European Journal of Preventive Cardiology or of the European Society of Cardiology.

Conflicts of interest: The author is receiving royalties from Hogrefe publishers for the German version of the Hospital Anxiety and Depression Scale. During the last three years, he has received a lecture honorarium from Novartis and research support from the German Ministry for Education and Research, the German Research Foundation, and the European Commission.