부산대학교 도서관

OBJECTIVE:: This study is aimed at investigating in longitudinal epidemiological setting the effects, if any, of caffeine and coffee intake on the occurrence of heart failure (HF) across −163C>A polymorphism of CYP1A2 gene, controlling caffeine metabolism. DESIGN AND METHOD:: A cohort of 1,475 unselected subjects aged 60.0 ± 16.7 years from general population was studied. Daily caffeine intake was calculated from an anamnestic questionnaire and a 7-day dietary diary. Events due to HF were recorded during a 12-year follow up. All subjects were genotyped for CYP1A2 −163C>A polymorphism and divided into fast (AA homozygous) and slow caffeine metabolizers (C-carriers). Cox analysis adjusted for confounders was used to study the effect of caffeine or coffee on incident HF. RESULTS:: In the whole cohort, HF incidence decreased with increasing caffeine intake (estimate −0.002, standard error, SE 0.001, p = 0.02). After stratifying by sex and genotype, this effect was still detectable in C-carrier men only (estimate −0.006, SE 0.002, p = 0.01). When repeating the analysis using daily servings of coffee, the estimate was significantly higher than that of caffeine (−0.17, SE 0.08, p = 0.03 in the whole cohort; −0.48, SE 0.18, p < 0.01 in C-carrier men). Prognostic cut-off was >210 mg/day of caffeine intake or >2 cups/day of coffee.(Figure is included in full-text article.) CONCLUSIONS:: At a population level, caffeine intake is protective against HF occurrence in C-carrier men, and innocuous in AA men and in all women. The protective effect of coffee is even greater that that of mere caffeine.