Abstract P255: Moderate Alcohol Consumption and the Risk of Acute Ischemic Stroke and All-Cause Mortality: The Million Veteran Program

Abstract

Introduction: Moderate alcohol consumption has been associated with lower risk of acute ischemic stroke (AIS) in the general population, but this relation has not been examined among U.S. Veterans. Furthermore, inconsistent data are available on the relation of alcohol and all-cause mortality. Methods: Million Veteran Program (MVP) participants who completed a lifestyle survey were eligible for analysis (n=335,754). We combined grams of ethanol in wine, beer, and spirits from the food frequency questionnaire to estimate grams/day (g/d). Participants were categorized into categories of: Never, former, or current drinkers of ≤6 g/d, >6-12 g/d, >12-24 g/d, >24-36 g/d, >36-48 g/d, and >48 g/d. We defined incident AIS as 1 inpatient or 2 outpatient codes (ICD-9 433-434.XX, 436.XX, 437.0, 437.6; ICD-10 codes I63.XX, I65-I66.XX, I67.2, I67.6, I67.8) using the VA electronic health record. Participants with prevalent CVD (n=93,496), missing alcohol data (n=24,871), no follow-up time (n=18,948) or missing age or sex (n=135) were excluded. We used a Cox Proportional Hazards model to relate alcohol consumption and AIS and mortality (separately) adjusting for age, sex, race, body mass index, smoking, education, exercise, DASH score, diabetes, and hypertension status. A sensitivity analysis using simulation reclassifying never drinkers with prevalent comorbidities assessed the impact of potential exposure misclassification on effect estimates. Results: Among 198,304 participants, the mean age was 64y (SD=12), and 10% were women. During a mean follow up of 3.3 years, 3,834 AIS and 7,020 deaths occurred. Using never drinkers as the reference group, we identified a 20-26% (95% CI range: 0.42-0.93) lower risk of AIS, and a 29-39% (95% CI range: 0.46-0.80) lower risk of mortality among drinkers of >0-36 g/d, respectively (Figure 1). In sensitivity analyses, HRs were attenuated but the effects remained robust. Conclusion: Our data show a lower risk of AIS and all-cause mortality with moderate alcohol consumption among MVP participants.