Vitamin D3 supplementation was associated with a 19% lower hazard of progression to diabetes among patients with prediabetes carrying ApaI AC or CC genotypes, but not among those with the ApaI AA genotype, according to a post hoc genetic association analysis of the Vitamin D and Type 2 Diabetes trial. The reduction among AC/CC carriers was modest and narrowly statistically significant (hazard ratio [HR], 0.81; 95% CI, 0.66-0.99).
The parent Vitamin D and Type 2 Diabetes (D2d) trial was a multisite, randomized, double-blind, placebo-controlled trial conducted from 2013 to 2018 in patients with prediabetes. Participants received 4,000 IU/d of vitamin D3 or placebo and were followed for a median of 2.5 years. In the original intent-to-treat analysis, vitamin D3 supplementation did not statistically significantly reduce incident diabetes compared with placebo (HR, 0.88; 95% CI, 0.75-1.04).
In the current analysis, researchers evaluated common vitamin D receptor polymorphisms to determine whether genotype modified the association between vitamin D exposure and diabetes risk. The study used a 2-phase analytic approach: an initial discovery analysis examining diabetes risk across achieved intratrial 25-hydroxyvitamin D levels by genotype, followed by a test-phase analysis evaluating treatment response in the broader genotyped cohort.
The analysis included 2,098 D2d participants with available ApaI genotype data. Eligible patients were aged 30 years or older and met at least 2 of 3 American Diabetes Association glycemic criteria for prediabetes without meeting criteria for diabetes at baseline. Incident diabetes was defined as meeting at least 2 glycemic thresholds for diabetes involving fasting plasma glucose, 2-hour plasma glucose, or hemoglobin A1c.
In the test-phase analysis, researchers classified patients with the ApaI AA genotype as proposed nonresponders and those with ApaI AC or CC genotypes as proposed responders. Among 618 patients with the ApaI AA genotype, vitamin D3 supplementation was not associated with reduced diabetes risk compared with placebo (HR, 1.02; 95% CI, 0.72-1.44). Among 1,480 patients with ApaI AC or CC genotypes, vitamin D3 supplementation was associated with a 19% lower hazard of progression to diabetes.
In the discovery analysis, patients with ApaI AC and CC genotypes had progressively lower diabetes risk at higher intratrial 25-hydroxyvitamin D levels, particularly at levels of 40 ng/mL or higher. Patients with the ApaI AA genotype did not show the same pattern.
Beyond ApaI, researchers also observed a similar pattern for the BsmI polymorphism, although there was near-complete overlap between the ApaI AA and BsmI TT genotypes. Because 260 of 261 patients with the BsmI TT genotype also carried the ApaI AA genotype, the investigators concluded that ApaI testing alone appeared sufficient for classifying proposed responders and nonresponders in this analysis.
Findings for the FokI polymorphism were less consistent. Patients with the FokI AG genotype showed lower diabetes risk at higher intratrial vitamin D levels, whereas GG and AA genotypes did not, producing a non-graded pattern that investigators considered insufficient for defining responder and nonresponder groups.
Models were adjusted for study site, race and ethnicity, sex, baseline age, body mass index, usual physical activity, statin use, and intratrial weight change. Follow-up was additionally censored when patients stopped trial pills, started diabetes or weight-loss medication, or exceeded the trial limit for outside vitamin D supplementation.
The researchers emphasized that the findings were exploratory and require confirmation in a future supplementation trial. The study was not powered to assess genotype-associated treatment response within racial or ethnic groups and did not evaluate biologic mechanisms underlying the observed associations.
“These exploratory genetic association findings support our hypothesis that a common VDR variant modulates the link between high intratrial 25(OH)D levels and diabetes risk,” wrote lead study author Bess Dawson-Hughes, MD, of the Jean Mayer USDA Human Nutrition Research Center on Aging at Tufts University, and colleagues.
In an invited commentary, Michael F. Holick, MD, and Arash Shirvani, MD, of Chobanian & Avedisian Boston University School of Medicine, said the findings may help explain heterogeneous responses to vitamin D supplementation and noted that vitamin D receptor signaling may influence pancreatic beta-cell function and insulin secretion. The commentators also noted that the active form of vitamin D acts on receptors in pancreatic beta-cells to stimulate insulin gene transcription, promote beta-cell survival, and suppress proinflammatory cytokines implicated in beta-cell dysfunction.
The commentators also described substantially larger apparent risk reductions among patients achieving higher intratrial vitamin D levels, although those estimates were derived from exploratory exposure analyses rather than the randomized treatment comparison reported in the primary analysis.
Disclosures: Dr Pittas reported receiving travel expenses from Abiogen outside the submitted work. No other disclosures were reported among the primary study authors. The study was supported by the US Department of Agriculture Agricultural Research Service and the National Institutes of Health.
Dr Holick reported grants, personal fees, and a pending patent related to vitamin D-associated products and therapies outside the submitted work.
Source: JAMA Network Open