Dr Reid recently published an ‘‘Opinion’’ in our journal [1] reconsidering some of the actions of vitamin D on the basis of recently published paper [2]. He reaches to the conclusion that vitamin D is not a tonic for bone and that, therefore, small amounts are needed to maintain bone health. Experimental investigation supporting this hypothesis showed that adequate intestinal calcium absorption is crucial for normal bone mass accrual and that 1,25(OH)2D signalling is crucial to secure this process, also during normal calcium intake [2]. The finding that the osteomalacia of the vitamin D receptor (VDR) knockout mouse is completely corrected by selective expression of VDR in the gut enterocytes is in line with previous observations [1, 2]. These data furthermore emphasize the relative importance of intestinal vitamin D receptor compared with skeletal VDR. Following these premises, the physiological role of VDR in bone in normal conditions should be better defined, also in light of the finding that selective knockout of this receptor in bone is paradoxically associated with an increase in bone mass [3]. Dr Reid postulates a critical effect of VDR in facilitating calcium absorption; it appears therefore intuitive that the most favourable meta-analyses, for example on the reduction of hip fractures, are those in which vitamin D is co-administered with calcium salts. We agree with the conclusion that vitamin D should be administered only to those who are insufficient or deficient, even though such thresholds are still an issue of discussion [4, 5]; similar to what happens in the field of other nutrients, a unique term should probably be used [6]. Concerning the three studies cited linking high doses of vitamin D to adverse effects on bone resorption and skeletal mass, we should observe that one of them used the active metabolite [1,25(OH)2D] which from a pharmacological point of view is different from its precursor [7]. The other studies utilized high doses of vitamin D [8, 9]; apart from possible biases [10], there is another study showing that the annual administration of vitamin D2 indeed decreased the incidence of fracture [11]. The statement that individuals living independently in the community and not dark skinned or veiled will not need supplements at all, is not universally applicable. Indeed, it is well known that sunlight is responsible for producing 80 % of the nutritional supply of vitamin D in the body and maximizes serum 25(OH)D levels only in summer and autumn. In Italy, for example, from late October till early March, we do not have enough UV light from the sun to make vitamin D in the skin; therefore, the possibility of being insufficient in the winter–spring season, it is not an unrealistic hypothesis. We do not entirely agree with Dr Reid’s suggestion that measuring serum 25(OH)D levels has no value [12–14]. Serum 25(OH)D is at present a satisfactory indicator of total body stores of vitamin D, even though problems with assay do exist [4]. To further complicate the issue, we still do not exactly know if the free fraction of 25(OH)D should be measured to obtain meaningful information with respect to what we presently know [15]. Therefore, because some issues are not fixed, we believe that a word of caution should be offered on the problems previously addressed. S. Minisola (&) C. Cipriani S. Piemonte J. Pepe Department of Internal Medicine and Medical Disciplines, ‘‘Sapienza’’ Rome University, Via del Policlinico 155, 00161 Rome, Italy e-mail: salvatore.minisola@uniroma1.it