Caffeine consumption in a long-term psychiatric hospital: Tobacco smoking may explain in large part the apparent association between schizophrenia and caffeine use.

This study further explores the association between schizophrenia and caffeine use by combining two prior published Spanish samples (250 schizophrenia outpatients and 290 controls from the general population) with two Spanish long-term inpatient samples from the same hospital (145 with schizophrenia and 64 with other severe mental illnesses). The specific aims were to establish whether or not, after controlling for confounders including tobacco smoking, the association between schizophrenia and caffeine is consistent across schizophrenia samples and across different definitions of caffeine use. The frequency of caffeine use in schizophrenia inpatients was not significantly higher than that in non-schizophrenia inpatients (77%, 111/145 vs. 75%, 48/64) or controls but was significantly higher than in schizophrenia outpatients. The frequency of high caffeine users among caffeine users in schizophrenia inpatients was not significantly higher than in non-schizophrenia inpatients (45%, 50/111 vs. 52%, 25/48) or controls, but was significantly lower than in schizophrenia outpatients. Smoking was significantly associated with caffeine use across all samples and definitions. Between 2 and 3% of schizophrenia inpatients, schizophrenia outpatients and non-schizophrenia inpatients showed caffeinism (>700 mg/day in smokers). Several of these smoking patients with caffeinism were also taking other inducers, particularly omeprazole. The lack of consistent association between schizophrenia and caffeine use is surprising when compared with the very consistent association between tobacco smoking and caffeine use across all of our analyses (use and high use in users) and all our samples. The confounding effects of tobacco smoking may explain in large part the apparent association between schizophrenia and caffeine use.