“…In addition, we minimised the potential for confounding by controlling for numerous accepted risk factors for mortality. Since previous studies reported both greater (2,4,5,26,35,38,39) or lower mortality risk (11,20) among female former drinkers compared to female lifetime abstainers, we were concerned with potential bias through the inclusion of ill former heavy drinkers in the reference group, but found that the exclusion of former drinkers did not alter our results. Our study had limited power to explore associations with specific causes of mortality and mortality among women with alcohol intakes with over 20 grams per day.…”