Response to Comment on Krul-Poel et al. Effect of Vitamin D Supplementation on Glycemic Control in Patients With Type 2 Diabetes (SUNNY Trial): A Randomized Placebo-Controlled Trial. Diabetes Care 2015;38:1420–1426

We thank Muscogiuri et al. (1) for their interest and kind comments on our recent study showing that treatment with vitamin D₃ (50,000 IU/month) for 6 months did not improve glycemic control in patients with well-controlled type 2 diabetes despite an adequate increase of serum 25-hydroxyvitamin D [25(OH)D] in the treated group (60.6 nmol/L at baseline and 101.4 nmol/L after 6 months) (2).

However, in a small subgroup of patients with severe vitamin D deficiency [serum 25(OH)D <30 nmol/L], a small but statistically significant decrease in HbA_1c was seen. In a similar intervention study (3), the influence of vitamin D supplementation (Vigantol oil once a week corresponding to a daily dose of 1,904 IU vitamin D) in patients with noninsulin-requiring type 2 diabetes suggests the potential benefits of vitamin D as an intervention. In addition, another intervention study with daily calcium carbonate (1,200 mg) and vitamin D (2,000–6,000 IU) supplementation for 6 months to a target of serum 25(OH)D >75 nmol/L in 95 individuals with serum 25(OH)D ≤55 nmol/L and at risk for type 2 diabetes detected an improvement in insulin sensitivity, suggesting that vitamin D supplementation might be most effective in “patients” with prediabetes (4).

One of the major concerns of vitamin D deficiency is that it is mainly accompanied by other risk factors, such as obesity, older age, and non-Caucasian origin. The aforementioned studies corrected for these confounding factors. In addition, the proposed link between vitamin D deficiency and obesity was recently challenged by the study by Clemente-Postigo et al. (5). People with vitamin D deficiency appeared to have an increased risk for type 2 diabetes even with a normal BMI, whereas those with (morbid) obesity who did not have diabetes had higher vitamin D levels than those with diabetes. Thus, the authors concluded that vitamin D levels are more closely linked to blood glucose levels than BMI.

In contrast, the results of a Mendelian randomization study became recently available (6) that used different European populations, including data from people with (28,144 patients) and without (76,344 individuals) type 2 diabetes. The authors used single nucleotide polymorphisms (SNPs) in DHCR7, CYP2R1, DBP, and CYP24A1. These SNPs were independently associated with reduced circulating levels of vitamin D (the so-called vitamin D–lowering alleles). Following a Mendelian randomization approach, Ye et al. (6) showed that for each 25 nmol/L reduction in a genetically determined 25(OH)D level, the odds ratio for type 2 diabetes was 1.01 (95% CI 0.75–1.36; P = 0.94). In the same article, a meta-analysis of 22 prospective observational studies (8,492 cases of type 2 diabetes and 86,698 cases without diabetes) found that a 1-SD (25.0 nmol/L) lower 25(OH)D level was associated with an increased risk of type 2 diabetes (relative risk of 1.21, 95% CI 1.16–1.27; P = 7.3 × 10⁻¹⁹). Probably, the Mendelian randomization method was not applicable to serum 25(OH)D, which is influenced by exposure to sunlight and to a much lesser extent by SNPs.

Taken together, despite the cumulative data that vitamin D supplementation does not improve the glycemic control in patients with type 2 diabetes, studies including patients with prediabetes and severe vitamin D deficiency will probably shed more light into the clouded sky.