New England Journal of Medicine July 28, 2022

Supplemental Vitamin D and Incident Fractures in Midlife and Older Adults

Meryl S. LeBoff, Sharon H. Chou, Kristin A. Ratliff, Nancy R. Cook, Bharti Khurana, Eunjung Kim, Peggy M. Cawthon, Douglas C. Bauer, Dennis Black, J. Chris Gallagher, I-Min Lee, Julie E. Buring, JoAnn E. Manson

Bottom Line

In a large, randomized controlled trial of generally healthy midlife and older adults, daily vitamin D3 supplementation did not significantly lower the risk of total, nonvertebral, or hip fractures compared with placebo.

Key Findings

1. Over a median follow-up of 5.3 years, supplemental vitamin D3 did not significantly reduce the incidence of total fractures, which occurred in 769 of 12,927 participants in the vitamin D group compared to 782 of 12,944 in the placebo group (HR 0.98; 95% CI, 0.89 to 1.08; P=0.70).
2. The incidence of nonvertebral fractures was not significantly affected by vitamin D3 supplementation (HR 0.97; 95% CI, 0.87 to 1.07; P=0.50).
3. The risk of hip fractures was nearly identical between the intervention and control arms (HR 1.01; 95% CI, 0.70 to 1.47; P=0.96).
4. There was no evidence of treatment effect modification based on baseline characteristics, including age, sex, race, BMI, or baseline serum 25-hydroxyvitamin D levels, meaning even participants with lower baseline vitamin D levels experienced no fracture risk reduction.

Study Design

Design
RCT
Double-Blind
Sample
25,871
Patients
Duration
5.3 yr
Median
Setting
United States
Population Generally healthy, community-dwelling men ≥50 years and women ≥55 years, unselected for vitamin D deficiency, low bone mass, or osteoporosis.
Intervention Supplemental vitamin D3 (cholecalciferol) 2,000 IU per day
Comparator Placebo
Outcome Incident total, nonvertebral, and hip fractures

Study Limitations

Participants were community-dwelling and unselected for severe vitamin D deficiency, osteoporosis, or low bone mass; findings may not extrapolate to high-risk clinical populations or those with profound deficiency (e.g., osteomalacia).
The trial evaluated only one dose (2,000 IU/day) of vitamin D3; it cannot definitively rule out effects from different dosing strategies, though 2,000 IU is highly standard for supplementation.
The median follow-up of 5.3 years, while substantial for an RCT, might theoretically miss longer-term skeletal remodeling benefits, though the Kaplan-Meier fracture curves showed absolutely no divergence during the study period.
Approximately 43% of participants were permitted to continue their own low-dose personal vitamin D supplements (up to 800 IU/day) and 21% took calcium, which could have masked the intervention's effect, although subgroups not taking outside supplements also showed no benefit.

Clinical Significance

The VITAL-Bone ancillary study fundamentally challenged widespread clinical practices by providing decisive, rigorous evidence that routine vitamin D3 supplementation does not prevent bone fractures in typical community-dwelling older adults. It prompted accompanying editorials urging clinicians to abandon routine screening of 25-hydroxyvitamin D levels and cease recommending widespread vitamin D supplementation for fracture prevention in the general population. It shifts the primary focus of fracture prevention back toward known pharmacological interventions and fall prevention in high-risk patients.

Historical Context

For decades, epidemiological studies persistently linked low serum 25-hydroxyvitamin D levels with a higher incidence of fractures, cardiovascular disease, and cancer. This generated massive public and clinical enthusiasm for vitamin D supplementation, leading to a dramatic spike in screening and prescribing. However, early RCTs were often fraught with methodological issues, such as small sample sizes, concomitant calcium use, or specific population constraints. The 2011 IOM report highlighted these gaps, calling for massive, definitive trials. The primary VITAL trial (2019) demonstrated that vitamin D did not prevent cardiovascular events or cancer. This 2022 ancillary study on fractures effectively closed the loop, solidifying the paradigm shift away from vitamin D as a ubiquitous preventive panacea in evidence-based medicine.

Guided Discussion

High-yield insights from every perspective

Med Student
Medical Student

Physiologically, why was vitamin D traditionally thought to prevent fractures, and based on the VITAL study results, what might this suggest about the threshold effect of 25-hydroxyvitamin D levels in healthy adults?

Key Response

Vitamin D increases intestinal calcium absorption and regulates osteoclast and osteoblast activity. The null results in VITAL suggest a threshold effect: once baseline levels are sufficient to prevent osteomalacia and secondary hyperparathyroidism, additional supplementation provides no further mechanical benefit to bone strength in otherwise healthy individuals.

Resident
Resident

A healthy 60-year-old female with a normal DEXA scan asks if she should start taking over-the-counter Vitamin D3 to protect her bones. Based on the VITAL fracture study, how do you counsel her, and in what specific clinical scenarios would you still recommend supplementation?

Key Response

The VITAL study showed no fracture reduction in generally healthy adults, so universal supplementation is not indicated for her. Exceptions where supplementation is still indicated include patients with documented osteoporosis, institutionalized or homebound older adults at high risk for falls, malabsorption syndromes, or those with known severe deficiency.

Fellow
Fellow

The VITAL study used 2000 IU of Vitamin D3 daily without co-administered calcium. How does this monotherapy design compare to prior trials that evaluated Vitamin D and calcium combined, and how might the absence of supplementary calcium influence the interpretation of these negative outcomes?

Key Response

Prior meta-analyses suggested fracture benefits were primarily seen when Vitamin D was combined with calcium. Evaluating Vitamin D monotherapy in VITAL isolates its independent effect but raises the question of whether an adequate calcium substrate is the actual limiting factor. However, subgroup analyses in VITAL accounting for dietary calcium showed no benefit, solidifying the lack of independent efficacy for Vitamin D monotherapy.

Attending
Attending

Given the definitive lack of fracture benefit in the VITAL study, how should this evidence reshape our systemic approach to routine 25-hydroxyvitamin D screening in primary care, and what cognitive biases have perpetuated its widespread use despite mounting negative evidence?

Key Response

This study underscores the need to abandon routine 25(OH)D testing in asymptomatic, average-risk patients. The persistence of its use is driven by anchoring bias to early observational data, the 'more is better' fallacy, and a societal inclination toward simple supplement-based fixes over complex lifestyle modifications like weight-bearing exercise.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD
PhD

The VITAL cohort was not pre-screened for Vitamin D deficiency or osteoporosis. From an epidemiologic and trial design perspective, how does enrolling a primary prevention, unselected population impact the trial's statistical power to detect an effect on hip fractures, and how does 'risk dilution' apply here?

Key Response

Enrolling an unselected population means the majority already have sufficient baseline Vitamin D and normal bone density, creating a floor effect where the intervention cannot further reduce risk. This 'risk dilution' significantly lowers the event rate for hip fractures, potentially masking a true benefit that might exist for a severely deficient subgroup, despite the overall null finding.

Journal Editor
Journal Editor

As a peer reviewer evaluating the VITAL fracture manuscript, how would you critically appraise the impact of 'drop-in' use of out-of-trial Vitamin D up to 800 IU/day by the placebo group on the study's internal validity, and does this contamination flaw the null conclusion?

Key Response

Permitting up to 800 IU/day of non-study Vitamin D in both arms reflects a pragmatic design but biases the results toward the null by narrowing the biological gradient between groups. A rigorous appraisal acknowledges this limitation but recognizes that subgroup analyses showing no benefit even in those taking no outside supplements preserve the study's validity.

Guideline Committee
Guideline Committee

The USPSTF historically concluded that evidence was insufficient to assess the balance of benefits and harms of vitamin D supplementation for primary fracture prevention. How should the VITAL fracture data alter this specific guideline, and what grade recommendation is now justified?

Key Response

The robust, large-scale RCT data from VITAL provides high-certainty evidence of no benefit. The guideline committee should move from an 'I' statement to a 'D' recommendation, advising against the use of Vitamin D supplementation to prevent fractures in asymptomatic, community-dwelling adults, aligning guidelines with definitive negative evidence.

Clinical Landscape

Noteworthy Related Trials

2005

RECORD Trial

n = 5,292 · Lancet

Tested

800 IU vitamin D3 daily, alone or with 1000 mg calcium

Population

Older adults aged 70 years or older with a recent low-trauma fracture

Comparator

Placebo

Endpoint

Incidence of new, low-trauma fractures

Key result: Daily supplementation with vitamin D3, calcium, or both did not significantly prevent secondary fractures in this older, frail population.
2006

WHI Calcium and Vitamin D Supplementation Trial

n = 36,282 · NEJM

Tested

1000 mg elemental calcium plus 400 IU vitamin D3 daily

Population

Postmenopausal women aged 50 to 79 years

Comparator

Placebo

Endpoint

Hip fracture incidence

Key result: Supplementation resulted in a small but non-significant improvement in hip bone density but did not significantly reduce hip fractures and increased the risk of kidney stones.
2017

ViDA Study

n = 5,110 · Lancet Diabetes Endocrinol

Tested

Monthly high-dose vitamin D3 (100,000 IU)

Population

Community-dwelling adults aged 50 to 84 years

Comparator

Placebo

Endpoint

Incident non-vertebral fractures

Key result: Monthly high-dose vitamin D supplementation did not reduce the incidence of non-vertebral fractures over a median follow-up of 3.4 years.

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