Abstract
The present study examines how alcohol intake from wine and non-wine alcoholic beverages (non-wine) in g/d, as well as cups of coffee and tea included as continuous covariates and mutually adjusted are associated with all-cause, cancer, non-cancer and CVD mortality. Consumption was assessed in 354 386 participants of the UK Biobank cohort who drank alcohol at least occasionally and survived at least 2 years after baseline with 20 201 deaths occurring over 4·2 million person-years. Hazard ratios (HR) for mortality were assessed with Cox proportional hazard regression models and beverage intake fitted as penalised cubic splines. A significant U-shaped association was detected between wine consumption and all-cause, non-cancer and CVD mortality. Wine consumption with lowest risk of death (nadir) ranged from 19 to 23 g alcohol/d in all participants and both sexes separately. In contrast, non-wine intake was significantly and positively associated in a dose-dependent manner with all mortality types studied except for CVD in females and with the nadir between 0 and 12 g alcohol/d. In all participants, the nadir for all-cause mortality was 2 cups coffee/d with non-coffee drinkers showing a slightly increased risk of death. Tea consumption was significantly and negatively associated with all mortality types in both sexes. Taken together, light to moderate consumption of wine but not non-wine is associated with decreased all-cause and non-cancer mortality. A minor negative association of coffee consumption with mortality cannot be excluded whereas tea intake is associated with a consistently decreased risk of all mortality types studied.</p>