Towards a trial-based definition of vitamin D deficiency

Ian R Reid
2016 The Lancet Diabetes and Endocrinology  
Vitamin D is not a vitamin (ie, an essential micronutrient), but rather the parent compound of an endocrine system primarily involved in the regulation of intestinal calcium absorption. It is synthesised in the skin as a result of the action of ultraviolet (UV) light on 7-dehydrocholesterol, and oral intake (other than supplement use) makes a negligible contribution to vitamin D concentration in most people. It is not biologically active, but two metabolites, 25-hydroxyvitamin D (25[OH]D) and
more » ... 25-dihydroxyvitamin D, bind to the vitamin D receptor (the latter with a 1000 times the affi nity of the former) and stimulate the active absorption of calcium in the upper small bowel. Both metabolites circulate bound to vitamin D binding protein. Severe vitamin D defi ciency results in hypocalcaemia and hypophosphataemia, resulting in undermineralisation of bone (osteomalacia). Before skeletal maturity, this defi ciency presents as rickets. Circulating concentrations of 25(OH)D are aff ected by sunlight exposure, and are reduced in people who are seldom outdoors (eg, those who are frail), women who are permanently veiled, and in people with dark skin (in whom melanin absorbs the UV light). Obesity also reduces 25(OH)D, since fat-soluble vitamin D is sequestered into adipose tissue, as do infl ammatory states (a poorly understood phenomenon that is partly due to changes in vitamin D binding protein). 1 Circulating concentrations of 25(OH)D are reduced in almost every medical or psychiatric disorder in which they have been studied, 2 and have a U-shaped association with mortality. 3 In pregnancy, low maternal 25(OH)D concentrations have been associated with gestational diabetes, pre-eclampsia, infants who are small for their gestational age, and lower off spring bone mass. 4 From these associations has arisen the widespread belief that vitamin D defi ciency is involved in the causation of a wide range of pathological abnormalities, resulting in advocacy for widespread supplementation. After a decade of publications demonstrating vitamin D's many disease associations, we are now entering a new era in which trials of vitamin D supplementation are being completed. These off er the possibility of determining which associations represent causation, and of guiding clinical practice. The MAVIDOS study, presented in The Lancet Diabetes & Endocrinology by Cyrus Cooper and colleagues, 5 shows that cholecalciferol 1000 IU/day does not aff ect neonatal bone mineral content (BMC) when given to pregnant women with baseline 25(OH)D concentrations between 25 nmol/L and 100 nmol/L. In a prespecifi ed subgroup analysis, an interaction was noted between the eff ects of treatment allocation and season of birth on off spring BMC (p interaction =0·04): for births occurring in winter, vitamin D supplementation increased neonatal BMC by almost 10% (p=0·004), and fat mass was increased by 17% (p=0·008). Increases in neonatal bodyweight after maternal vitamin D supplementation have been suggested in previous similar trials, 6 and the higher fat mass observed in the supplemented group in Cooper and colleagues' study might be mechanistically related to the bone changes, since fat mass is a dynamic determinant of bone mass in adult life. 7 The subgroup of women who gave birth in winter had a mean 25(OH)D concentration of 30 nmol/L (SD 16) at 34 weeks' gestation, compared with concentrations of 50-60 nmol/L in the summer and autumn birth subgroups, suggesting a threshold at which supplements might be of value. To the purist, MAVIDOS shows no benefi t from vitamin D supplementation in pregnancy. However, the subgroup analysis provides data that might help to address the vexed issue of what constitutes vitamin D defi ciency. The fi ndings suggest that when 25(OH)D concentration is lower than 30 nmol/L, neonatal fat and bone mass could be aff ected by maternal supplementation. This threshold is very similar to that for which benefi cial eff ects on adult bone density and for prevention of osteomalacia have been reported. 8 Similar post-hoc trial evidence has suggested that when 25(OH)D concentration is lower than 25-30 nmol/L, supplementation might reduce exacerbations of chronic obstructive pulmonary disease, 9 and reduce mortality in people who are critically ill. 10 By contrast, Bolland and colleagues 11,12 have produced persuasive evidence that supplements given to cohorts with higher 25(OH)D concentrations (generally >30 nmol/L), do not aff ect falls, fractures, myocardial infarction, stroke, or cancer. Together, these fi ndings suggest that vitamin D might be causally associated with some disorders when 25(OH)D concentration is lower than 25-30 nmol/L. The
doi:10.1016/s2213-8587(16)00079-6 pmid:26944420 fatcat:clw34zadmbbzxh4yov3nuai7uq