The New England Journal of Medicine August 08, 2019

Vitamin D Supplementation and Prevention of Type 2 Diabetes

Anastassios G. Pittas, Bess Dawson-Hughes, Patricia Sheehan, et al., for the D2d Research Group

Bottom Line

In adults at high risk for type 2 diabetes, supplementation with 4000 IU per day of vitamin D3 did not significantly lower the risk of developing diabetes compared to placebo.

Key Findings

1. A total of 2,423 participants with prediabetes were randomized to receive either 4,000 IU/day of vitamin D3 (n=1,211) or placebo (n=1,212).

2. Mean serum 25-hydroxyvitamin D levels increased from 27.7 to 54.3 ng/mL in the vitamin D group, while remaining relatively unchanged in the placebo group (28.2 to 28.8 ng/mL).

3. Over a median follow-up of 2.5 years, new-onset diabetes occurred in 293 participants in the vitamin D arm (9.39 events per 100 person-years) and 323 in the placebo arm (10.66 events per 100 person-years).

4. The hazard ratio for incident type 2 diabetes with vitamin D compared to placebo was 0.88 (95% CI, 0.75 to 1.04; P=0.12), which did not reach statistical significance.

5. Adverse events, including hypercalcemia and nephrolithiasis, did not differ significantly between the two groups.

Study Design

Design

RCT

Double-Blind

Sample

2,423

Patients

Duration

2.5 yr

Median

Setting

Multicenter, US

Population Adults at high risk for type 2 diabetes who met at least two out of three glycemic criteria for prediabetes (fasting plasma glucose 100-125 mg/dL, 2-hour post-load glucose 140-199 mg/dL, and HbA1c 5.7-6.4%), with no prior diagnosis of diabetes, regardless of baseline serum 25-hydroxyvitamin D levels.

Intervention Oral vitamin D3 (cholecalciferol) at a dose of 4,000 IU per day.

Comparator Matching oral placebo.

Outcome Time to new-onset type 2 diabetes, based on annual oral glucose-tolerance testing and semiannual fasting plasma glucose or HbA1c testing.

Study Limitations

• The trial enrolled participants regardless of their baseline vitamin D level; less than 5% had frankly deficient levels (<12 ng/mL), limiting the ability to assess benefits in populations with severe vitamin D deficiency.

• The median follow-up of 2.5 years may have been too short to capture long-term physiological changes in insulin resistance or beta-cell preservation.

• The participants were at a very high risk for type 2 diabetes, and high underlying incidence rates of the disease may have masked modest metabolic benefits of vitamin D.

Clinical Significance

Vitamin D supplementation at 4,000 IU daily should not be routinely recommended as a primary prevention strategy for type 2 diabetes in the general prediabetic population. Lifestyle modifications (diet and exercise) and established pharmacotherapies (such as metformin) remain the cornerstone of diabetes prevention.

Historical Context

Numerous epidemiological studies over the past decades demonstrated a strong inverse association between low serum 25-hydroxyvitamin D levels and an increased risk of developing type 2 diabetes. Prior intervention trials often lacked specific diabetes endpoints, were underpowered, or used lower doses of vitamin D (e.g., 400-800 IU). The D2d trial was designed as the largest and most definitive randomized trial specifically powered to test whether high-dose vitamin D supplementation could prevent progression from prediabetes to type 2 diabetes.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

Based on basic science principles, what is the theoretical mechanism by which Vitamin D was hypothesized to prevent type 2 diabetes, and why might an exogenous supplement fail to demonstrate this effect in a general prediabetic population?

Key Response

Vitamin D receptors are present on pancreatic beta cells and skeletal muscle. Mechanistically, Vitamin D is thought to improve insulin secretion (facilitated by calcium influx) and enhance peripheral insulin sensitivity. However, supplementing individuals who already have adequate baseline 25(OH)D levels (as seen in the majority of the D2d cohort) often yields no additional physiological benefit, highlighting the difference between correcting a clinical deficiency and using a nutrient in pharmacological dosing.

Resident

A patient with prediabetes asks if they should start taking high-dose Vitamin D to prevent progression to type 2 diabetes, noting they read it 'boosts metabolism.' Based on the D2d trial, how do you counsel them regarding Vitamin D supplementation versus other preventive strategies?

Key Response

The D2d trial demonstrated that 4000 IU/day of Vitamin D3 does not significantly reduce the risk of progressing to type 2 diabetes in high-risk patients. Residents should use this evidence to steer patients away from unproven supplement strategies and instead recommend evidence-based interventions like the Diabetes Prevention Program (DPP) intensive lifestyle modifications (target 7% weight loss, 150 minutes of exercise weekly) or metformin for appropriate candidates.

Fellow

In the D2d trial, the primary intention-to-treat outcome was negative, but how does baseline 25-hydroxyvitamin D status influence the interpretation of these results, and what does this imply for targeted metabolic risk screening in endocrinology clinics?

Key Response

Nearly 80% of the D2d cohort had baseline 25(OH)D levels at or above 20 ng/mL (considered sufficient). Post-hoc and secondary analyses suggested a potential signal for benefit in the small subgroup with very low baseline Vitamin D (e.g., <12 ng/mL). Fellows should recognize that while universal supplementation is ineffective for diabetes prevention, correcting true, severe deficiency might still play a nuanced role in restoring baseline metabolic health, emphasizing targeted treatment over universal prophylaxis.

Attending

The D2d trial is one of several large RCTs failing to show significant benefits of Vitamin D supplementation on chronic disease outcomes, despite decades of strong observational data. As an attending, how do you use this discrepancy to teach trainees about confounding in nutritional epidemiology?

Key Response

Observational studies consistently link low Vitamin D with incident diabetes, cardiovascular disease, and mortality. However, RCTs like D2d and VITAL demonstrate that low Vitamin D is often a biomarker of poor overall health, obesity (since it is a fat-soluble vitamin sequestered in adipose tissue), or a lack of outdoor physical activity, rather than a causal driver of the disease. This is a critical teaching point for evaluating nutritional literature and understanding residual confounding.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The D2d trial observed a hazard ratio of 0.88 (95% CI, 0.75 to 1.04) for incident diabetes. How does the choice of broad inclusion criteria regarding baseline Vitamin D levels impact the trial's statistical power to detect a true biological effect, and how might a 'nutrient trial' design differ from a standard pharmacological RCT?

Key Response

In nutrient trials, the control group is not truly 'unexposed' because they have endogenous Vitamin D and dietary intake. By including mostly Vitamin D-sufficient participants, the trial's effect size was likely heavily diluted. A classic pharmacological RCT design assumes a baseline exposure of zero; researchers must critique whether future nutrient trials should restrict enrollment exclusively to severely deficient populations to adequately test biological efficacy, despite the logistical challenges of recruitment.

Journal Editor

As an editor reviewing the D2d manuscript, how would you evaluate the impact of off-protocol use of OTC Vitamin D in the placebo group ('drop-in') and non-adherence in the active group on the trial's intention-to-treat analysis and its risk of a Type II error?

Key Response

A seasoned reviewer would flag that over a multi-year trial, placebo participants frequently start taking over-the-counter supplements due to public health trends, and active participants may discontinue the study drug. This bidirectional crossover dilutes the separation in mean 25(OH)D levels between the two study arms, biasing the result toward the null. Assessing whether per-protocol or as-treated sensitivity analyses support the primary null finding is crucial for editorial confidence in publishing a 'negative' trial.

Guideline Committee

Given the findings of the D2d trial, should current American Diabetes Association (ADA) Standards of Care recommend routine Vitamin D supplementation or screening for individuals with prediabetes to delay disease progression?

Key Response

Current ADA guidelines do not recommend routine Vitamin D supplementation for glycemic control or diabetes prevention. The D2d trial provides high-quality (Level A) evidence reinforcing this stance. The committee would conclude that while severe deficiency should be treated per general Endocrine Society guidelines for bone health, Vitamin D should not be incorporated into the prediabetes algorithmic pathway for glycemic benefits, keeping the guideline focus strictly on intensive lifestyle intervention and pharmacotherapy like metformin.

Clinical Landscape

Noteworthy Related Trials

2002

Diabetes Prevention Program (DPP)

n = 3,234 · NEJM

Tested

Intensive lifestyle modification or Metformin

Population

Adults with impaired glucose tolerance (prediabetes)

Comparator

Placebo

Endpoint

Incidence of type 2 diabetes

Key result: Lifestyle intervention reduced the incidence of type 2 diabetes by 58%, and metformin by 31%, compared with placebo.

2017

VIDA Trial

n = 5,110 · Lancet Diabetes Endocrinol

Tested

Monthly Vitamin D3 (100,000 IU)

Population

Adults aged 50-84 years

Comparator

Placebo

Endpoint

Incident cardiovascular disease

Key result: Monthly high-dose vitamin D supplementation did not prevent cardiovascular disease, nor did it significantly alter the incidence of type 2 diabetes.

2018

VITAL Trial

n = 25,871 · NEJM

Tested

Vitamin D3 (2000 IU/day) and Omega-3 fatty acids

Population

Healthy men aged >=50 and women aged >=55 years

Comparator

Placebo

Endpoint

Major cardiovascular events and invasive cancer

Key result: Supplementation with vitamin D3 at 2000 IU per day did not result in a lower incidence of invasive cancer or cardiovascular events than placebo.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis