“…9,23,25,27,30,31,34,42,60 The reasons for such disagreement may lie in the study designs, particularly in the sampling of controls. In six out of nine studies controls were not age-matched to cases, 9,23,25,31,42,60 and that is exceptionally important, since older age is an established risk factor for low vitamin D concentrations. 64,65 The three age-and sex-matched studies had other important limitations, which have been already emphasized.…”