|| Checking for direct PDF access through Ovid
Background: Alcohol consumption is associated with negative consequences including liver disease, major trauma and various cancers. However, a paradoxical beneficial effect of alcohol has been documented for cardiovascular disease in a large meta-analysis. On the other hand, a similar relationship has not been consistently documented or modest for the risk of stroke due to its heterogeneity in subtypes and pathophysiology. In this context, more comprehensive approaches using a representative dataset and composite outcomes are required to clarify the differential effects of alcohol on cardio-cerebrovascular disease (CCVD).Methods: The Korean National Health Insurance Service (NHIS) is the sole provider of full-coverage health insurance for Korean citizens. NHIS-National Sample Cohort (NHIS-NSC) was a standardized cohort representative of Korean population. Based on biennial health exam at 2003 and 2004, we investigated the relationship between alcohol consumption and the incidence of CCVDs [ischemic stroke (IS), hemorrhagic stroke (HS) and myocardial infarction (MI)] and all-cause mortality until 2013, using multivariable Cox proportional hazards models.Results: Among 204,557 participants (median follow up, 9.1 years), the following events were observed: IS 3,274; HS 1,056; MI 1,065 ; and all-cause mortality 8,771 cases. The hazards for IS and MI remained below the null point for the moderate alcohol consumption, but the risk for HS increased in heavy drinkers. In composite outcome indices, moderate consumption was associated with a lower risk for CCVD and all-cause mortality; however, the risk for all-cause mortality was significantly elevated in cases of heavy drinkers.Conclusions: Alcohol consumption was associated with a decreased risk of both CCVD, specifically ischemic vascular disease, and all-cause mortality at moderate dose, but this protective relationship was offset by an increase in the risk for HS and mortality at heavy dose.