Does alcohol consumption increase the risk of cutaneous melanoma? Comments on a recent meta-analysis: reply from the authors

DEAR EDITOR, We read with interest the letter by Drs Chen and Zhang1 about our recent investigation, ‘Alcohol drinking and cutaneous melanoma risk: a systematic review and dose–risk meta-analysis’, published in the British Journal of Dermatology.2 With reference to the literature search, excluding studies published in languages other than English has generally little effect on summary effect estimates of meta-analyses.3–5 In our meta-analysis, we considered for inclusion only articles published as originals in English up to 30 April 2012. The contribution by Kvaskoff et al.6 was published in French as a conference abstract, and thus it did not satisfy our inclusion criteria. The article by Asgari et al.7 reported the results of the American VITamins And Lifestyle (VITAL) cohort study. It was published in the June 2012 issue of the Journal of Investigative Dermatology, and appeared online ahead of print on 1 March 2012. Unfortunately, our literature search did not retrieve this article, probably because it was fully indexed in Medline only on 18 July 2012. Still, for the sake of completeness, we have updated our meta-analysis by including it. None of the results materially change; in particular, the overall pooled relative risks (RRs) are 1.20 [95% confidence interval (CI) 1.07–1.35], 1.11 (95% CI 0.99–1.25) and 1.20 (95% CI 1.04– 1.38) for any (Fig. 1), light and moderate-to-heavy alcohol drinking, respectively, vs. no/occasional drinking. With the addition of the Asgari et al. study,7 stratified analyses from studies (n = 8) conducted in America showed an RR of 1.34 (95% CI 1.18–1.52), while no differences appeared in the RR of studies that did not adjust for sun exposure (RR 1.27, 95% CI 1.20–1.35), as in the study of Asgari et al.7 We disagree with the existence of a relevant bias by study quality. In fact, large cohort studies such as the U.K. Million Women Study8 and the U.S. Women’s Health Initiative9 reported an increased risk of cutaneous malignant melanoma (CMM) in alcohol drinkers, and the pooled risk estimates of our meta-analysis were similar for case–control and cohort studies. We agree with Drs Chen and Zhang that current evidence is insufficient to conclude for a causal association of CMM with alcohol consumption. As we pointed out in our review,2 caution in interpreting our results is required, as residual confounding by sun ultraviolet exposure cannot be ruled out. Future well designed studies with specific control for this potential confounder are needed to confirm, or not, our findings.