Vitamin D: some perspective please

Vitamin D deficiency has been associated with an ever expanding list of diseases, and with this has come almost tonic-like claims for vitamin D supplementation. In observational studies, low vitamin D status has been associated with increased risk of multiple sclerosis, type 1 and type 2 diabetes, cardiovascular disease, colon cancer, breast cancer, autoimmunity, and allergy.1 The UK government has advised that all pregnant women, and children under 5 years, should take 400 IU vitamin D daily; a recent news story, however, reported a survey conducted by a charity which suggested that only 26% of pregnant women and 46% of healthcare professionals are aware of these guidelines.2 The most recent musculoskeletal trend seems to be the attribution of childhood problems such as Blount’s disease and slipped femoral epiphyses to vitamin D deficiency and the incorrect conflation of rickets with low serum calcidiol (25-hydroxyvitamin D₃) concentrations.3 So are health professionals causing ill health through their lack of awareness and advocacy of vitamin D supplementation?

The high profile news coverage and the enthusiastic promotion of the results of observational studies as though they proved causality might lead the undiscriminating observer to think so. We think, however, that some perspective is needed. That vitamin D deficiency may cause childhood rickets is indisputable. Rickets was widespread in the white population during the industrial revolution, but the current increase in rickets and other manifestations of vitamin D deficiency, such as neonatal hypocalcaemic tetany, is seen mostly in the dark skinned population of the United Kingdom. Serum calcidiol concentrations are lower in dark skinned than in white UK populations,4 independent of latitude. Among Asian women in the south of England median serum calcidiol concentrations were 24.9 nmol/L in summer and 16.9 nmol/L in winter, whereas in white women the corresponding values were 62.5 nmol/L and 39.9 nmol/L. The proportion of women with low concentrations of calcidiol across the year (<40 nmol/L) was 10-49% in white women and 89-91% in Asian women.4 The marked disparity in the incidence of clinical disease and severity of vitamin D deficiency between ethnic groups in the UK seems to have been overlooked in much lay and scientific reporting. This is not to say that we should ignore moderately low concentrations of vitamin D in up to 49% of the white population; the question is whether it is a health problem.

These differences also beg the question of how to define normality. Studies have reported the serum calcidiol concentration at which parathyroid hormone reaches a plateau to be between 25 nmol/L and 125 nmol/L, making it difficult to deduce a functional definition of vitamin D deficiency.5 Similar uncertainty surrounds estimates derived using fractures, bone turnover markers, and fractional calcium absorption.6 A recent postmortem based study found that a substantial proportion of people with serum values less than 25 nmol/L had normal bone histology.7 These observations suggest that it is difficult to extrapolate from a low serum vitamin D status even to bone disease with any degree of certainty on an individual basis. Indeed, a recent randomised controlled trial suggested that high dose vitamin D supplements might even increase the risk of fracture.8

Furthermore, much of the evidence linking low vitamin D status to non-bone outcomes has been derived from observational studies and is therefore subject to a range of interpretational difficulties (as recognised by the recent Institute of Medicine report). These include reverse causality (disease may cause reduced exposure to sun), confounding (low physical activity may cause low vitamin D, obesity, and increased risk of diabetes), classification bias (vitamin D status defined in terms of diet, not blood concentrations, which correlate poorly with nutritional intake),9 and differences in assay methods.10 Most basic science studies use the active form of vitamin D—calcitriol (concentrations of which are tightly regulated in vivo)—and the results of its pharmacological action on end organs is conflated with an effect of supplementation. In addition, reluctance to publish null or negative findings tends to bias the literature in favour of a positive effect. The result is that when the benefits of vitamin D supplementation suggested by such studies have been tested in randomised controlled trials they have often not been confirmed. A good example is chronic soft tissue pain, where observational studies showed an inverse association between serum calcidiol concentration and level of pain, but randomised controlled trials found no benefit from supplementation.11 Finally, the safety of population strategies incorporating vitamin D supplementation in large numbers of people over long periods of time also needs clarification.

In an age of evidence based medicine, we need to prove the benefit, lack of benefit, or even harm of our interventions. In addition to detailed laboratory based mechanistic studies and research to validate appropriate biomarkers for adequacy, large well designed randomised controlled trials with long term follow-up are urgently needed to clarify the role of vitamin D and the usefulness of supplementation.12 Only with such investigations will we be able to navigate through the mass of literature, both lay and scientific, that enthusiastically promotes vitamin D as a cure for almost all modern maladies.