Background
Increased mortality in patients with schizophrenia compared to the general population has been consistently reported [
1‐
4]. Previous studies have suggested that antipsychotic medications are associated with significant rates of lethal arrhythmias and instances of sudden death [
4‐
7]. Given that both typical and atypical antipsychotic drugs affect the autonomic nervous system (ANS) via neuroleptic effects on various neurotransmitter receptors, antipsychotic drugs are thought to be a trigger factor for autonomic dysregulation. In addition, evidence now exists for an association between decreased ANS activity and increased risk of sudden cardiac death [
8‐
12]. Therefore, it is considered important that clinicians assess ANS activity in patients with schizophrenia treated with antipsychotic drugs to estimate subclinical signs of adverse effects.
Decreased ANS activity in schizophrenia has been reported. Yet, some researchers have suggested an association between low ANS activity and the effects of severity of schizophrenia instead of the effects of antipsychotic drugs [
13‐
17]. These researchers report lower ANS activity in medication-free patients with schizophrenia and lower ANS activity in patients with schizophrenia with severe psychotic symptoms. Only a few studies have attributed decreased ANS activity to medication [
18,
19]. Those studies suggest that various antipsychotic drugs, which have wide-ranging receptor affinity profiles, decrease parasympathetic nerve activity. However, in those studies, the sample sizes were relatively small or the authors did not consider the dose-dependent effect of antipsychotic drugs on ANS activity or did not assess the influence of coadministered drugs.
The aim of this study was to investigate the dose-dependent effect of antipsychotic drugs on ANS activity in an adequate sample size. We targeted more than 200 patients with schizophrenia under medication for long periods and investigated various clinical factors, including psychotic symptoms, duration of illness, and other psychotropic drugs, that might affect ANS activity.
Discussion and conclusion
In this study, we found a markedly lower power of both the LF and HF components of HRV, but not a lower LF/HF ratio in patients with schizophrenia compared to control subjects. The LF component of HRV represents both sympathetic and parasympathetic nerve activity, the HF component represents almost entirely parasympathetic nerve activity, and the LF/HF ratio represents relative sympathetic nerve activity [
22,
31]. Thus, we found decreased parasympathetic nerve activity, but not decreased sympathetic nerve activity, in patients with schizophrenia compared to healthy controls. Based on HRV power spectral analysis, autonomic dysregulation is a consistent finding in schizophrenia, regardless of medication status [
13‐
15,
32]. Our present results are concordant with those of previous studies [
18,
19,
22,
33].
Some researchers have suggested an association between decreased ANS activity and the severity of schizophrenia. These authors reported parasympathetic hypoactivity in medication-free patients with schizophrenia [
13,
34,
35] and that parasympathetic nerve activity is significantly decreased when the psychotic state is more pronounced [
15‐
17,
36]. A few studies have suggested that suppression of parasympathetic nerve activity is associated with disturbances in cortical-subcortical circuits and other central nervous system pathologies in schizophrenia [
13,
37]. However, patients in those studies were on medication or had been treated with antipsychotic drugs 1 to 4 weeks before ANS activity measurements. Interrupted or readministered antipsychotic drugs could possibly have had more or less of an effect on ANS activity in patients in those studies.
In contrast, a few studies state that medication affects ANS activity in patients with schizophrenia. Studies have reported that some antipsychotic drugs, such as clozapine, with wide-ranging receptor affinity profiles, show imbalance between sympathetic and parasympathetic nerve activity [
18,
19] and that other antipsychotics drugs, such as haloperidol, which are relatively free of significant effects on neurotransmitter receptors except dopamine receptors, have no significant effect on ANS activity [
19,
38]. The results of those studies suggest that the effects of antipsychotic drugs on ANS function are derived from anticholinergic and antiadrenergic activity. In the present study, levomepromazine and chlorpromazine, which have potent anticholinergic and antiadrenergic effects, were commonly used, and many of our subjects were treated with multiple antipsychotic drugs.
Considering the results of previous studies and receptor affinity profiles, it is a consensus view that antipsychotic drugs have more or less of an effect on ANS activity. However, there is little evidence to indicate the dosage of antipsychotic drugs necessary to affect ANS activity. We investigated the dose-dependent effect of antipsychotic drugs on ANS activity. Our results suggest the following. If the daily dose of antipsychotic drugs is ≤500 mg/day, ANS activity in patients with schizophrenia is not significantly lower than that in controls. If the daily dose of antipsychotic drugs is 501 to 1000 mg/day, antipsychotic drugs exert a large effect on ANS activity and decrease ANS activity. If the daily dose of antipsychotic drugs is ≥1001 mg/day, antipsychotic drugs exert a very large effect on ANS activity and decrease ANS activity significantly. In short, the effect might be large if the dose of antipsychotic drug is high. There was a statistically significant difference in mean GAF scores among three subgroups. However, the result of the multiple regression analysis showed that there was no association between the spectral components of HRV and GAF score after adjustment for variables such as age, sex, BMI, duration of illness, CPZeq, BPDeq, and DZPeq (Table
3).
Multiple regression analysis of the schizophrenia group showed a statistically significant association between daily dose of antipsychotic drugs and parasympathetic nerve activity. However, there was no correlation between ANS activity and daily dose of anticholinergic antiparkinsonian or anxiolytic drugs. In the present study, the anticholinergic antiparkinsonian drug dose was 3.2 ± 1.5 mg/day (BPDeq). Anticholinergic antiparkinsonian drugs potentially affect ANS activity; however, the effect might be small unless at a high dose. There was also no association between ANS activity and the severity of the disease or duration of illness. In addition, multiple regression analysis showed that ANS activity was not associated with age, although there was a statistically significant difference in mean age between the patient group and the control group.
To summarize our results, we interpreted our findings as follows. 1) Antipsychotic drugs decreased ANS activity in medicated patients with schizophrenia, presumably mediated by the parasympathetic nervous system. 2) If the antipsychotic drug dose was ≥501 mg/day, antipsychotic drugs had a dose-dependent effect on ANS activity and decreased ANS activity significantly, at least, in medicated patients. 3) Anticholinergic antiparkinsonian drugs did not significantly affect ANS activity unless at a high dose. 4) The severity of schizophrenia did not have an effect on ANS activity.
The present study has some potential limitations. First, although the patient sample size was greater than 200, the number of subjects was still relatively small. Second, we did not analyze the influence of each drug separately because most of our subjects were treated with multiple drugs. Third, because there were no medication-free patients, it is unclear whether the pathomechanisms and/or the severity of schizophrenia directly affect ANS activity. Additional investigations are necessary to clarify these issues.
Our present results suggest that antipsychotic drugs exert a dose-dependent effect on ANS activity and decrease ANS activity significantly and that HRV power spectral analysis, as a diagnostic measurement of ANS, may allow for the identification of patients at high risk for sudden cardiac death. Therefore, we suggest that low ANS activity might be a biomarker for adverse effects of antipsychotic drugs rather than severity of schizophrenia. Prospective studies will be aimed at carefully collecting data from at-risk patients to elucidate the association between subclinically decreased ANS activity and the occurrence of serious adverse effects. These investigations will provide clinical evidence for HRV power spectral analysis as a potential monitoring system to protect psychiatric patients from severe adverse events such as fatal ventricular arrhythmias.