A pilot study of plasma caffeine concentrations in a US sample of smoker and nonsmoker volunteers.

Abstract Even though 85% of adults drink caffeinated beverages daily, very limited studies on plasma caffeine concentration in the US population have been published. Smoking induces cytochrome P 450 1A2 (CYP1A2), which is the main enzyme involved in caffeine metabolism. The current naturalistic pilot study explores plasma caffeine concentrations in a US sample, and presents a mathematical model of the relationship between caffeine intake and plasma concentrations for smokers and nonsmokers. Caffeine intake and average plasma caffeine concentrations from morning (7:30–9:30 a.m.) and afternoon (2:00–4:00 p.m.) samples were studied in 69 volunteers (21 smokers and 48 nonsmokers). The mean caffeine intake obtained from caffeinated beverages was 3.02 mg/kg/day, which is similar to the intake in the US population. Almost all subjects in the present sample (99%; 95% confidence interval [CI]: 96–100) had detectable plasma caffeine concentrations. Smokers had significantly higher caffeine intake than nonsmokers. The ratio of concentration/dose of caffeine from caffeinated beverages was approximately four-fold higher in nonsmokers (1.33 kg×day/l) than in smokers (0.29 kg×day/l). According to the model, the median plasma caffeine concentration was two- to three-fold higher in nonsmokers for each level of caffeine intake. Our model improves our understanding of the interactions between caffeine and smoking. Additional studies are needed to replicate the model. This model may help epidemiologists to correct for the effects of smoking on caffeine intake and pharmacologists to screen for the activity of CYP1A2.