Rickets has been cited as the commonest non-communicable disease of childhood worldwide. It is certainly a disease that has been well described over the last four centuries, initially by Glisson (1648), who differentiated it from infantile scurvy, Trousseau (p 38), who identified sunlight and cod liver oil as effective treatments, Mellanby (1919), who created an animal model of rickets that was cured by cod liver oil, and Hess and Unger (1921), who cured rickets in children by exposing them to sunlight on a rooftop in New York.

The epidemiology of rickets—lack of vitamin D from diet and/or the action of sunlight on exposed skin, or of the bone mineral substrates calcium and phosphate—is also well established. The contribution of such deficiencies to other human ill health problems is less clear. There is no doubt that vitamin D is important for function of the musculoskeletal and immune systems, neurological function, and possibly mental health. But how much vitamin D is necessary to support human health and should vitamin D supplementation be targeted or universal? How should vitamin D sufficiency be assessed? Are there critical periods during which attention should be concentrated? Two papers in this issue of Archives make useful contributions to the debate.

The first paper, by Robinson et al, examines the effects of vitamin D deficiency in infancy.¹ This is not new territory, but the message of increasing frequency of presentation over the period of a decade in a developed country makes salutary reading. The diagnosis of vitamin D deficiency in the paper by Robinson et al depended on clinical identification of ill health in the infants. Hypocalcaemia leading to convulsions (which may be accompanied by cardiac failure in severe cases), bone softening leading to deformity, metaphyseal swelling, proximal myopathy, and growth failure are all clear clinical indicators of deficiency. These immediate manifestations should be apparent to all practising paediatricians, and such signs should be actively sought in breast fed non-white infants whatever the reason for their initial presentation.

Furthermore, children and young adults who had low intakes of vitamin D during infancy not necessarily resulting in such florid clinical manifestations may be at increased risk of problems later. In a retrospective study of a Finnish birth cohort from 1966, Hypponen et al found an eight-fold increase in the risk of diabetes in those who had not received the recommended 2000 IU/day of vitamin D during the first six months of life.²

The second paper in this issue, by Das et al, focuses on vitamin D deficiency in the adolescent period, when at least one third of the mass of the adult skeleton accrues.³ The authors show that vitamin D insufficiency is common in apparently healthy adolescent girls, with non-white girls not surprisingly more severely affected. Despite the low levels of vitamin D, however, there were no clear functional consequences with immediate impact. Nevertheless, the concern remains that failure to supply an essential nutrient during a period of rapid growth and development is likely to result in problems across the population as a whole. Severely deficient girls do develop alterations in pelvic shape that can lead to obstructed labour. Failure to accrue bone within the skeletal envelope, or to have a smaller envelope as a result of reduced skeletal growth, are factors that could predispose to osteoporotic fracture in later life. These girls are likely to have persistent deficiency of vitamin D unless their lifestyle or diet changes significantly. This may impact in turn on the future vitamin D status of any infants that they subsequently bear themselves.

The pressure to increase vitamin D intake is strong in some quarters; the references cited in Robinson and colleagues’ paper to justify the use of serum 25-hydroxyvitamin D 50 nmol/l as a cut-off for deficiency come from the work of Holick,⁴ a vigorous pro-vitamin D and sunlight exposure campaigner. It is not entirely clear whether maintaining such a level is necessary to sustain health, however, at all ages. Using this cut-off, vitamin D insufficiency was found in 11/23 white adolescent girls in Maine during the period September 2000 to March 2003;⁵ no clinically relevant consequences appear to have been suffered. There has been no comparable work in infants until now. Previous work in the UK has taken the cut-off as being 25 nmol/l.⁶ In that study, while there was an average rate of 25% of insufficiency among children under age 2 years from South Asian ethnic backgrounds, the comparable rate in white young children was 1%. It would hardly be surprising if those numbers increased if the goalposts were moved. In fact, low levels of vitamin D in cord blood were reported as common in white Caucasians at the end of winter in Sheffield; in a study of 106 women (90% white Caucasian) and their infants, 70% of infants had serum 25-hydroxyvitamin D <20 nmol/l.⁷

How much vitamin D is needed to maintain human health? Robinson et al note the recent recommendation of the American Academy of Paediatrics that all should receive 200 IU vitamin D/day. In the Maine study, the average stated vitamin D intake was close to this value. It is interesting to note from Robinson’s paper, however, that 31/106 had not been breast fed. If we assume that this means they received infant formula, is the vitamin D content of infant formula always adequate to maintain vitamin D sufficiency? Although not stated in the paper, it might be that infants born to vitamin D deficient mothers need more vitamin D to restore homoeostasis.

In the UK, the Department of Health has made recommendations, based on the reports of COMA, for the vitamin D supplementation of breast fed infants and infants “at risk” of vitamin D deficiency, as well as for “at risk” pregnant mothers. These latter recommendations have been challenged by NICE who stated that “in the absence of evidence of benefit, vitamin D supplementation should not be routinely offered to pregnant women” (NICE, Compilation Issue 7, October 2003, p. 138). One might observe that absence of evidence for an effect is not evidence of absence; common sense would suggest that vitamin D deficiency, widely reported in the communities whose ethnic origins lie in sunnier climes, can affect pregnant as well as non-pregnant individuals. The high proportion of non-whites (96%) and the early presentation with hypocalcaemic seizures in one third in Robinson and colleagues’ study clearly suggests that vitamin D deficiency should be actively sought in non-white pregnant women. Indeed, the advice offered in a recent edition of the CMO Update (Summer 2005, Issue 42, p. 6) is that supplements of 400 IU/day (10 μg/day) should be offered to pregnant women, and 280 IU/day (7 μg/day) to children under the age of 5 from at-risk communities.

Taken together, the two papers published in this issue of Archives indicate a widespread insufficiency of vitamin D in children. Worryingly, from the perspective of current UK strategies to prevent vitamin D deficiency, virtually all of the vitamin D metabolite measured in the serum of the Manchester cohort was sunlight derived. Taken together with the earlier data from Sheffield, there may be a need to revise our current approaches to the prevention of vitamin D deficiency and insufficiency.

For now, however, we need to take simple, practical measures to reduce the burden of early bone disease and other later problems. These measures need to be part of our routine practice and recommendations, particularly in the at-risk groups of infants and young children from non-white backgrounds.

Education is a major part of this. Paediatricians, obstetricians, midwives, health visitors, and GPs all need to remind patients that there is virtually no vitamin D in breast milk, and that totally breast fed babies should be supplemented (irrespective of the colour of their skin) until receiving a full mixed diet. The other recommendations from the CMO are clear and simple and should now be applied universally in respect of infants and young children. We do still need to decide how best to meet the needs of older children and adolescents in a culture that covers up, either with clothing or with sunscreen. Perhaps more exercise outdoors would help deal with this problem.