Discussion
In this multi-institutional, nested, case–control study, we investigated risk factors for early circulatory mortality in patients with schizophrenia. A previous study measuring the 25-year mortality of schizophrenia reported there was an apparent increase in circulatory mortality relative to the general population, and cigarette smoking, one of traditional CVD risk factors, may account for majority of the excess natural mortality in the cohort (Brown et al., 2010). However, the cohort study failed to identify other non-traditional risk factors. The present study yielded two major findings; compared with the living patients with schizophrenia, the deceased ones had 1) a higher count of leukocytes and 2) a shorter duration of all antipsychotic treatments, after adjustment for clinical illness variables, cardiovascular variables, and laboratory data at the index admission. First, the elevated leukocyte counts in the patients with schizophrenia were more associated with circulatory mortality than with traditional cardiovascular risk factors, including smoking, BMI, hyperlipidemia, and ECG measurements. Elevated leukocyte counts, as an indicator of systemic inflammation, was reported to predict coronary heart disease progression in patients with preexisting vascular diseases (Danesh et al., 1998). Moreover, the leukocyte count may serve as an independent predictor of adverse events following intervention for myocardial infarction (Kojima et al., 2004). Therefore, systemic inflammation may be strongly associated with early circulatory mortality in patients with schizophrenia as repeatedly observed in patients with bipolar disorder (Tsai et al., 2005). Our results are consistent with a previous study reporting that an unbalanced immune response may be associated with the inflammatory process of the central nervous system, in which dopaminergic, serotonergic, noradrenergic, and glutamatergic neurotransmission is influenced by immune alternations in schizophrenia (Muller et al., 2015).