Polygenic risk for psychotic disorders in relation to cardiac autonomic dysfunction in unmedicated patients with schizophrenia.

Abstract

Cardiac autonomic dysfunction (CADF), mainly characterized by increased heart rate, decreased heart rate variability, and loss of vagal modulation, has been extensively described in patients with schizophrenia (SCZ) and their healthy first-degree relatives. As such, it represents an apparent physiological link that contributes to the increased cardiovascular mortality in these patients. Common genetic variation is a putative underlying mechanism, along with lifestyle factors and antipsychotic medications. However, the extent to which CADF is associated with genetic factors for SCZ is unknown. A sample of 83 drug-naive SCZ patients and 96 healthy controls, all of European origin, underwent a 30-minute autonomic assessment under resting conditions. We incorporated parameters from several domains into our model, including time and frequency domains (mean heart rate, low/high frequency ratio) and compression entropy, each of which provides different insights into the dynamics of cardiac autonomic function. These parameters were used as outcome variables in linear regression models with polygenic risk scores (PRS) for SCZ as predictors and age, sex, BMI, smoking status, principal components of ancestry and diagnosis as covariates. Of the three CADF parameters, SCZ PRS was significantly associated with mean heart rate in the combined case/control sample. However, this association was was no longer significant after including diagnosis as a covariate (p = 0.29). In contrast, diagnostic status is statistically significant for all three CADF parameters, accounting for a significantly greater proportion of the variance in mean heart rate compared to SCZ PRS (approximately 16% vs. 4%). Despite evidence for a common genetic basis of CADF and SCZ, we were unable to provide further support for an association between the polygenic burden of SCZ and cardiac autonomic function beyond the diagnostic state. This suggests that there are other important characteristics associated with SCZ that lead to CADF that are not captured by SCZ PRS.