Michaëlsson and colleagues’ proposed mechanism for the effect of milk intake on the risk of mortality and fractures is based on the assumption that fermented dairy products (which had the opposite effects to those of non-fermented milk) are free of galactose.1 For most fermented dairy products, however, this is untrue. The paper quoted by the authors shows a decrease in lactose from 4.8 g/100 g to 2.3 g/100 g (loss of 2.5 g/100 g) in yoghurt compared with milk,2 with an increase in galactose of 1.3 g/100 g. Because lactose is composed of glucose and galactose, which have equal molecular weight, this decrease in lactose would lead to 1.25 g/100 g of galactose formed (equal to the 1.3 g/100 g in the quoted paper). Yoghurt therefore contains the same amount of galactose as milk (and in lactase non-persistent people may lead to even higher galactose intakes). The galactose content of (semi)hard cheeses is somewhat lower because the curd is washed during production, but cheese is not usually free from galactose. Overall, the galactose intake from fermented dairy products (soured milk and yoghurt in the paper) is equal to that from regular dairy products, which makes the authors’ proposed mechanism highly unlikely.