Vitamin D Supplementation and Type 2 Diabetes: A Substudy of a Randomised Placebo-controlled Trial in Older People (RECORD Trial, ISRCTN 51647438)

Alison Avenell; Jonathan A. Cook; Graeme S. MacLennan; Gladys C. McPherson


Age Ageing. 2009;38(5):606-609. 

In This Article


We did not find evidence that vitamin D3 in a daily dose of 800 IU, or in combination with 1000 mg calcium, was able to prevent the development of diabetes or an increase in the need for medication for diabetes. However, the confidence intervals do not rule out protective effects. The Women's Health Initiative (WHI) also found that 400 IU vitamin D3 and 1000 mg calcium were not associated with a reduced the risk of developing diabetes over a median follow-up period of 7 years in 33,951 women whose mean age was 62 years, where 50% of measured participants had a baseline 25(OH) vitamin D concentration of < 44 nmol/l.[5] In the RECORD trial, the 25(OH) vitamin D3 level from a small sample of 60 trial participants in Southampton and Newcastle, measured from February to July before supplementation averaged 38 (SD 16) nmol/l by high-performance liquid chromatography,[3] similar to the mean levels for men and women of similar age living in care homes in a recent English survey.[6] After 1 year of supplementation in these participants, 25(OH) vitamin D3 was 62 (SD 16) nmol/l.

Data were only collected by self-report, which was also used in the WHI trial.[5] The reported prevalence of diabetes at enrolment was 8%, lower than the figure of 10% reported for women aged 75–79 years and over in England.[7] This may reflect not only under-reporting but also a lower risk for type 2 diabetes in people with osteoporosis and lower body weight.

We did not collect data at baseline about body mass index or glycaemic control. The participants' mean weight was 65 kg, similar to the mean weight of 64.1 kg for a UK survey of women aged 75–84 years, for whom the mean body mass index was 26.7 kg/m2.[8] Again, this would suggest that this was not a particularly high-risk population for type 2 diabetes.

In a post hoc analysis, one randomised trial[9] found that supplemental 700 IU vitamin D3 and 500 mg calcium daily over 3 years in 314 men and women with a mean age of 71 years reduced the rise in fasting plasma glucose over time in a sub-group of people with pre-existing impaired fasting glucose, but not participants with normal fasting glucose. In that trial, supplementation achieved a 25(OH) vitamin D level of 99 nmol/l standardised to the vitamin D external quality assurance scheme.[10]

A recent randomised trial in the UK examined the effect of a single dose of 100,000 IU vitamin D2 or placebo in people with type 2 diabetes, a mean age of 64 years and a baseline 25(OH) vitamin D of 38 nmol/l.[11] Flow-mediated endothelial function was significantly improved by vitamin D2, but there was no significant difference between groups for glycaemic control or insulin sensitivity. An Australian trial utilised 100,000 IU vitamin D3 as two doses 2 weeks apart in people with a mean age of 55 years without type 2 diabetes with a mean 25(OH) vitamin D3 of 40 nmol/l.[12] Supplementation had no significant effect on insulin sensitivity or glucose tolerance. Thus, high-dose bolus oral supplementation was not found effective in people younger than the RECORD trial participants, who appeared to have similar pre-supplementation 25(OH) vitamin D3.

Based on data from studies of bone mineral density, lower extremity function, falls, fractures, colorectal cancer and dental health, Bischoff–Ferrrari and colleagues have argued that serum concentrations of 25(OH) vitamin D3 should be 75 nmol/l or more.[13] Neither the RECORD trial nor WHI attained these levels in people measured,[14] and poor compliance may have contributed in the RECORD trial.

The results suggest that randomised controlled trials should concentrate on evaluating the use of vitamin D3 in higher doses in groups at high risk of developing type 2 diabetes.