Higher serum caffeine in smokers with schizophrenia compared to smoking controls.

Previous studies of high dietary caffeine intake in individuals with schizophrenia have not demonstrated biological evidence of higher intake or controlled smoking behavior. This study aimed to examine differences in serum caffeine levels in 104 smokers with schizophrenia/schizoaffective disorder (SCZ/SA) and compare them to 63 smokers without any mental illness (CON). Since we were interested in measuring caffeine levels, we excluded all non-caffeine users from the study. Blood draws were standardized to occur at mid-day on a usual smoking day. The mean serum caffeine level was significantly higher for SCZ/SA group compared to CON (2722 ng/mL vs. 1122 ng/mL; p < 0.001). This trend persisted in subsets of smokers who smoked less than 20 cigarettes per day (CPD; 2052 ng/mL vs. 587 ng/mL; p < 0.05), 20–30 CPD (2743 ng/mL vs. 1170 ng/mL; p < 0.001) or more than 30 CPD (3430 ng/mL vs. 1834 ng/mL; NSS). Linear backward stepwise regression analyses including demographic and smoking variables revealed that having a diagnosis of SCZ/SA (compared to CON) significantly predicted serum caffeine level (B = 1528.2; p < 0.001). In addition, SCZ/SA group had two times greater serum caffeine levels as compared to CON with similar smoking behavior. Clinical effects of smoking and caffeine intake are important and may complicate the interpretation of schizophrenia symptoms and antipsychotic medication side effects, thus warranting further research.