Hin · Osteoporosis international : a journal established as result of cooperation between the European Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 2017 · randomized placebo-controlled trial · n=305

Optimum dose of vitamin D for disease prevention in older people: BEST-D trial of vitamin D in primary care.

Cited 78 times in the scientific literature.

Level 2 - randomized trial

Individual randomized controlled trial

PubMed 27986983 · doi:10.1007/s00198-016-3833-y · record verified 2026-08-30

What was done

This randomized, placebo-controlled trial allocated 305 community-dwelling adults aged 65 years or older in Oxfordshire, UK, to receive daily oral vitamin D3 at 4000 IU (100 µg), 2000 IU (50 µg), or placebo for 1 year. The study assessed changes in biochemical markers of vitamin D status (plasma 25[OH]D, parathyroid hormone [PTH], albumin-corrected calcium, and alkaline phosphatase), cardiovascular risk factors, and physical function.

What was found

Baseline mean (SD) plasma 25(OH)D was 50 (18) nmol/L. At 12 months, mean (SD) 25(OH)D levels increased to 137 (39) nmol/L with 4000 IU, 102 (25) nmol/L with 2000 IU, and 53 (16) nmol/L with placebo, with 88%, 70%, and 1% achieving the pre-specified target of >90 nmol/L, respectively. Neither active dose caused significant deviations outside the normal range for PTH or albumin-corrected calcium. The incremental increase from 4000 versus 2000 IU was smaller in overweight and obese individuals compared with normal-weight individuals. Supplementation showed no significant effects on cardiovascular risk factors or physical function.

Why it matters

This study establishes the human dose-response relationship for high-dose vitamin D3 in older adults, showing that 4000 IU daily is required to achieve plasma concentrations (>90 nmol/L) associated with lower disease risk in observational studies without biochemical safety issues.

Limits

The sample size is modest (n=305) from a single UK region, and the study was designed as a biochemical precursor rather than powered for hard clinical endpoints such as fractures or cardiovascular events. Quantitative data for cardiovascular and physical performance measures were omitted in the abstract.

Cited by