Association of alcohol consumption with the onset of natural menopause: A systematic review and meta-analysis

Background: Early onset of menopause is associated with long-term health risks, including cardiovascular disease and premature death. Although alcohol intake has been suggested to affect the age at which natural menopause occurs, results from observational studies are not consistent. Objective and Rationale: In the view of the differing risks to the health of early menopause and the increasing trends in alcohol consumption in women, in this systematic review, we aimed to quantify the association between all levels of alcohol consumption and menopause onset. Search Methods: Six electronic databases (Medline, Embase, Cochrane, PubMed, Google Scholar and Web of Science) were systematically searched until 4 November 2015 to identify relevant studies assessing the association between alcohol consumption and onset of menopause. Two independent reviewers screened the titles and abstracts of all initially identified studies according to the selection criteria. Studies were sought if they (i) were observational cross-sectional, prospective and interventional studies, (ii) had reported on natural onset of menopause, (iii) had reported on alcohol consumption, (iv) had assessed the association between alcohol consumption and menopause onset, (v) were conducted in humans and (vi) were not conducted in patients with cancer. Data were extracted by two independent reviewers using a predesigned data-collection form. The primary exposure variable was the presence of active alcohol drinking at baseline compared with a reference group of non-drinkers. Pooled relative risks (RRs) were calculated. Outcomes: Of the 1193 references (all in English language) reviewed for eligibility, 22 articles based on 20 unique studies were included in the final analysis. A total of 41 339 and 63 868 non-overlapping women were included in the meta-analysis of cross-sectional and observational cohort studies, respectively. In cross-sectional studies, the pooled RR for earlier onset of menopause was 0.86 (95% confidence interval (CI): 0.78-0.96) between drinkers versus non-drinkers. Analysis of the levels of alcohol consumed showed that low and moderate alcohol consumption (more than one drink per week (RR = 0.60; 95% CI: 0.49-0.75) and three or fewer drinks per week (RR = 0.75; 95% CI: 0.60-0.94)) were associated with later menopause onset, compared to non-drinkers. In prospective studies, RR for earlier menopause onset was 0.95 (95% CI: 0.91-0.98) when comparing women who reported drinking alcohol versus women who did not. Analysis of the dose of alcohol consumed showed that low-to-moderate alcohol intake (0-8 g/day (RR = 0.95; 95% CI: 0.93-0.98), and 16 g/day (RR = 0.89, 95%CI: 0.86-0.92)) was associated with later menopause onset, compared to non-drinking. Wider Implications: The findings of this review indicate that alcohol consumption, particularly low and moderate alcohol intake, might be associated with later onset of menopause although the magnitude of the association is low. Further studies are needed to corroborate these findings, clarify the level of alcohol intake at which menopause is delayed and identify the potential mechanisms behind this association.