The association between alcohol consumption and mortality: the Swedish Women's Lifestyle and Health cohort study

Abstract

Introduction: Alcohol consumption is common in many Western societies. Its role in women's health is, however, not well understood, in particular the role of light-to-moderate alcohol intake, which has been proposed to protect against cardiovascular disease (CVD) mortality but not against cancer mortality. Methods: We studied the relations of alcohol consumption to total mortality and mortality from specific causes in the Swedish Women's Lifestyle and Health Study, a cohort of 47,921 women aged 29 to 49 years residing in Sweden at baseline in 1991/1992 who were followed up to 2006. We estimated relative risks (RR) and corresponding 95% confidence intervals (CI) of mortality using Cox regression adjusted for age, smoking, BMI, saturated fat intake, physical activity, and education. Results: During 713,295 person-years of follow-up, 1,119 deaths occurred, including 158 deaths from CVD, 673 deaths from cancer, and 288 deaths from other causes. We observed a non-linear association between alcohol intake and total mortality with a nadir at intake of 1.5–4.9 grams per day of alcohol. Compared with abstainers, the RRs (95% CI) for increasing categories of total alcohol intake (0.1–1.4, 1.5–4.9, 5.0–9.9, and 10.0 grams per day or more) were 0.87 (0.72–1.05), 0.81 (0.67–0.96), 0.83 (0.68–1.02), and 0.85 (0.66–1.10), respectively. Compared with abstainers, the association between alcohol and CVD mortality was maximally inverse (RR=0.57, CI=0.36–0.91) at 1.5–4.9 grams per day of alcohol consumption, while that between alcohol intake and cancer mortality was not statistically significant with nadir at 5.0–9.9 grams per day (RR=0.84, CI=0.64–1.10). Conclusion: In this cohort of young to middle-aged women in Sweden, we found a non-linear relation of alcohol to total mortality and CVD mortality that was maximally inverse at intake of 1.5–4.9 grams per day. In contrast, alcohol intake showed no statistically significant association with cancer mortality in this group of women with predominantly light-to-moderate alcohol consumption.