By David Kiefer, MD, Editor
Clinical Assistant Professor, Department of Family Medicine, University of Wisconsin; Clinical Assistant Professor of Medicine, Arizona Center for Integrative Medicine, University of Arizona, Tucson
Dr. Kiefer reports no financial relationships relevant to this field of study.
SYNOPSIS: In Greenland Inuit, increasing serum hydroxyvitamin D3 was associated with a worsening of physiological measurements and definitions of glucose homeostasis, contrary to what was expected.
SOURCE: Nielsen NO, Bierregaard P, Ronn PF, et al. Associations between vitamin D status and type 2 diabetes measures among Inuit in Greenland may be affected by other factors. PLoS One 2016;11:e0152763.
- Fasting and non-fasting blood samples of Inuit in Greenland were used to demonstrate a positive association between vitamin D status and laboratory and clinical parameters such as fasting glucose, hemoglobin A1c, impaired glucose tolerance, and type 2 diabetes.
Sometimes, important discoveries in science begin with a perceived association between a variable (or variables) and an outcome, which leads to hypothesis generation, and more definitive testing. Along this vein, an intriguing story is developing between vitamin D and various physiological effects and clinical conditions, not the least of which is the possible connection between low serum vitamin D levels and type 2 diabetes. There has been a range of research trials investigating this, from epidemiological analyses to prospective cohort studies, and just as wide a range of results, but the connection remains dubious. Realizing that there was an opportunity to improve on the data, the researchers of the study being reviewed here mention both the methodological flaws of previous research and a unique population that they were able to study. With respect to the former, previous studies had sometimes just relied on dietary vitamin D intake to estimate levels, but that leaves out the crucial sunlight component. In addition, trials have relied on diabetes self-report, not always the most accurate way to corroborate that diagnosis. Which leads, then, to the demographic of focus for this study, the Inuit in Greenland.
The Greenland Inuit have seen an increase in type 2 diabetes and “low” levels of vitamin D, as per citations offered by the authors. These changes are thought to be caused, in part, by an erosion of their traditional diet, which had been high in vitamin D from sea mammals and fish. Another characteristic of the Greenland Inuit is the fact that, due to the high latitude (minimal sunlight) and cold temperatures necessitating protective clothing, for much of the year the sunlight that is present provides little to no skin conversion of vitamin D. The researchers postulated that this would be the perfect population in which to show an association between vitamin D and diabetes.
A random sampling of permanent Greenland residents (90% of whom are Inuit) yielded 3,108 Inuit for this study, although only 2,877 had blood testing results available from 2005-2010. In addition, blood tests were available from 1987 (and then frozen until the present day; n = 330). This blood was tested for fasting glucose and insulin. In addition, a two-hour oral glucose tolerance test was administered; another blood sample was taken after 75 g of glucose was ingested. Blood also was analyzed for serum 25-hydroxyvitamin D3 (vitD), and hemoglobin A1c. At the time of the blood collections in 2005-2010, information about possible confounders was collected, including ancestry, exercise, smoking, and use of supplements; this information was not available for the 1987 samples.
The median vitD for the 1987 sample was 59.0 nmol/L, and for the 2005-2010 sample was 46.6 nmol/L. See Table 1 for the percentage of people with various types of impaired glucose metabolism.
After adjusting for the confounders, interestingly increasing vitD by 10 nmol/L actually increased glucose by 0.03 mmol/L, two-hour glucose by 0.06 mmol/L, and hemoglobin A1c by 0.51%. Furthermore, increasing vitD decreased beta-cell function by 0.82%, a calculation not detailed here. With respect to glucose metabolism, the 2005-2010 samples showed that increasing vitD by 10 nmol/L increased the adjusted odds of impaired fasting glucose by 10%, and 7% for type 2 diabetes; no association was seen for impaired glucose tolerance. No statistically significant associations were seen for the 1987 samples.
How do we interpret findings that were opposite the expectation? The mechanism from other research groups does not support that vitamin D would have an adverse effect on type 2 diabetes parameters. The authors believe these results and implicate “other factors” that might be involved. That may be true, and, if anything, it prompts a continued search to define the physiological actions of vitamin D and its metabolites. At least, as the authors said, the effects on impaired fasting glucose, impaired glucose tolerance, and type 2 diabetes risk is “… of limited clinical importance.” The changes seen were minimal, thankfully, but still enough to make clinicians pause before globally recommending vitamin D supplementation for all patients with, or at risk for, type 2 diabetes or metabolic syndrome. Perhaps this is a lesson in the clinical limitations of statistical associations, as much as it is to embrace the wonder of science when results don’t turn out quite as you would expect.