Reduction of alcohol consumption and subsequent mortality in alcohol use disorders: systematic review and meta-analyses

Objective: To determine whether a reduction in drinking in individuals with alcohol use disorders resulted in reduced mortality risk.

Data sources: Electronic searches were performed of MEDLINE, EMBASE, and ISI Web of Science and references of identified articles were searched up to May 2012 using these keywords: (alcohol dependence OR alcohol abuse) AND (mortality) AND (cohort OR follow-up). Only English-language articles were included.

Study selection: Sixteen cohort studies were identified that reported all-cause mortality risk by drinking groups measuring change in alcohol intake among people with alcohol use disorders.

Data extraction: Numbers of participants and deaths in each group; odds ratios (ORs); and demographic, clinical, and methodological variables were extracted.

Results: In comparison to continued heavy drinking, a reduction below heavy levels of alcohol use (including abstention) was associated with a substantially reduced risk of mortality (random-effects pooled OR = 0.41; 95% CI, 0.34-0.50; P < .001). The OR was 0.35 (95% CI, 0.20-0.60; P < .001) for those who reached abstention and 0.61 (95% CI, 0.39-0.94; P = .026) for those who did not reach abstention but substantially reduced their consumption. The pooled OR for abstention compared to reduced consumption was 0.42 (95% CI, 0.19-0.92; P = .031). Meta-regression models did not reveal significant influences of study characteristics examined.

Conclusions: Reduction of drinking in alcohol use disorders was associated with a marked reduction in mortality risk for those who reached abstinence or reduced drinking compared to continued heavy drinkers. Those who reached abstention showed the smallest mortality risk, lower than the risk for reduced consumption without abstinence.