Alcohol Screening Scores and All-Cause Mortality in Male Veterans Affairs Patients

Abstract

Extensive research demonstrates a J- or U-shaped association between in-depth interview measures of alcohol consumption and mortality. Little is known about the relationship between alcohol screening questionnaires and mortality. This study examined the association between scores (ranging from 0 to 12) on the three-item Alcohol Use Disorder Identification Test-Consumption Questionnaire (AUDIT-C) and mortality. This cohort study included male outpatients from seven Veterans Affairs (VA) medical centers who completed the AUDIT-C on mailed surveys (1997-2000; n=34,292) and who were followed for a mean of 2.5 years. Death was ascertained from VA databases. In analyses adjusted for demographics, a U-shaped relationship was observed between AUDIT-C scores and all-cause mortality. Nondrinkers (AUDIT-C 0) and heavy drinkers (AUDIT-C 8-9 and 10-12) had increased risk of death compared with drinkers who screened negative for alcohol misuse (AUDIT-C 1-3): hazard ratios (HR)=1.41 (95% confidence interval [CI]: 1.29-1.54), 1.32 (1.03-1.69), and 1.47 (1.14-1.91), respectively. This association differed across age groups (p=.003). For men younger than 50 years, AUDIT-C scores 10-12 were associated with increased mortality (HR=2.21, 95% CI: 1.29-3.77), whereas for men age 50 or older, self-reported abstinence (AUDIT-C 0) was associated with increased mortality, compared with drinkers with AUDIT-C scores 1-3: HR formen 50-64=1.45 (1.19-1.77); HR for men 65 or older=1.42 (1.28-1.58). A U-shaped association between the AUDIT-C and mortality was observed, with important differences by age group. This is the first study to demonstrate that a clinical scaled screening measure of alcohol use has a similar association with mortality to that observed in epidemiological research with lengthier measures.