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

Summary This trial compared the effects of daily treatment with vitamin D or placebo for 1 year on blood tests of vitamin D status. The results demonstrated that daily 4000 IU vitamin D3 is required to achieve blood levels associated with lowest disease risks, and this dose should be tested in future trials for fracture prevention. Introduction The aim of this trial was to assess the effects of daily supplementation with vitamin D3 4000 IU (100 μg), 2000 IU (50 μg) or placebo for 1 year on biochemical markers of vitamin D status in preparation for a large trial for prevention of fractures and other outcomes. Methods This is a randomized placebo-controlled trial in 305 community-dwelling people aged 65 years or older in Oxfordshire, UK. Outcomes included biochemical markers of vitamin D status (plasma 25-hydroxy-vitamin D [25[OH]D], parathyroid hormone [PTH], calcium and alkaline phosphatase), cardiovascular risk factors and tests of physical function. Results Mean (SD) plasma 25(OH)D levels were 50 (18) nmol/L at baseline and increased to 137 (39), 102 (25) and 53 (16) nmol/L after 12 months in those allocated 4000 IU, 2000 IU or placebo, respectively (with 88%, 70% and 1% of these groups achieving the pre-specified level of >90 nmol/L). Neither dose of vitamin D3 was associated with significant deviation outside the normal range of PTH or albumin-corrected calcium. The additional effect on 25(OH)D levels of 4000 versus 2000 IU was similar in all subgroups except for body mass index, for which the further increase was smaller in overweight and obese participants compared with normal-weight participants. Supplementation with vitamin D had no significant effects on cardiovascular risk factors or on measures of physical function. Conclusions After accounting for average 70% compliance in long-term trials, doses of 4000 IU vitamin D3 daily may be required to achieve plasma 25(OH)D levels associated with lowest disease risk in observational studies. Electronic supplementary material The online version of this article (doi:10.1007/s00198-016-3833-y) contains supplementary material, which is available to authorized users.


Randomization
Potentially eligible individuals were identified from a single general practice and were mailed a letter by their family doctor, together with a study information leaflet, to invite them to participate in the trial. Eligible individuals who agreed to participate in the trial were subsequently visited in their homes by a specially trained research nurse. After confirming eligibility and obtaining written informed consent, individuals were randomized using a central telephone randomization service. Allocation to study treatment (vitamin D3 4000 IU or 2000 IU or placebo daily) used a minimization algorithm that included age, body mass index [BMI], smoking history, ethnicity and history of fracture. Vitamin D3 and matching placebo administered in soft gel capsules were provided by Tischcon Corporation (Westbury, New York, USA).Participants were supplied with a six-month supply of study medication and were asked to take two capsules of either active vitamin D3 2000 IU, or matching placebo capsules daily.
Echocardiography to assess left ventricular function was undertaken on a subset of 150 participants at 12 months and these results will be reported separately.

Primary and secondary outcomes
The co-primary outcomes were mean plasma 25(OH)D levels and percentage of participants with 25(OH)D levels >90 nmol/L at 12 months. Secondary outcomes included: mean plasma 25(OH)D levels, and percentage of participants with 25(OH)D >90 nmol/L (36 ng/mL), at 1 and 6 months; percentage of participants with PTH in the normal range (1.1-6.8 pmol/L) at 1, 6 and 12 months; percentage of participants with albumin-corrected calcium levels above the normal range (2.15-2.55 mmol/L) at 1, 6 and 12 months; mean level at 6 and 12 months of albumin, phosphate, creatinine, alkaline phosphatase and lipids; and blood pressure recorded at 6 and 12 months. Additional secondary outcomes included heart rate, blood pressure and brachial and digital arterial stiffness in all participants at 6 and 12 months. Tertiary outcomes assessed at 12 months included all site and specific fractures, falls, muscle pain, joint pain, self-assessed physical activity, number of respiratory infections, geriatric depression score, weight, height, BMI, hand grip strength, physical performance measures, and bone density T-and Z-scores at the hand and wrist. Safety outcomes included all serious adverse events, irrespective of whether these were considered to be related to study treatment, reasons for stopping study treatment and biochemical safety data.

Laboratory measures
Plasma levels of 25(OH)D and parathyroid hormone (PTH) were measured on blood samples collected using lithium heparin tubes on an Access 2 Immunoassay System (Beckman Coulter (UK) Ltd, High Wycombe, England). The 25(OH)D assay used a two-step competitive binding immunoenzymatic method traceable to the National Institute of Standards and Technology (Gaithersberg, Maryland USA) standard reference material (NIST SRM) 2972 and the PTH assay used a two-site immunoenzymatic sandwich method traceable to the World Health Organisation International Standard 79/500. The between-run precision for 25(OH)D levels was 6.0% at a level of 30.1 ng/mL and 8.7% at a level of 16.7 ng/mL and for PTH was 3.2% at a level of 146.5 pg/ mL and 7.7% at a level of 19.7 pg/mL. To convert plasma 25(OH)D levels from nmol/L to ng/mL, divide by 2.5. To convert PTH from pmol/L to pg/ml divide by 0.11. Using a UniCel DxC 800 Synchron Clinical System (Beckman Coulter (UK) Ltd, High Wycombe, England) and lithium heparin plasma, end point absorbance methods were used to measure plasma levels of albumin (assay traceable to NIST 927a), calcium (assay traceable to NIST 956) and phosphate (assay traceable to NIST 3139a). A kinetic rate absorbance method was used to measure alkaline phosphatase. Calibration of the alkaline phosphatase procedure was based on a bichromatic extinction coefficient for p-Nitrophenol, which has a molar absorptivity of 17,900 at 410/480 nm. Between-run precision for albumin was 1.4% at a level of 31.1 g/L and 1.6% at a level of 6.2 g/L; for calcium 1.2% at levels of 2.49 mmol/L and 3.28 mmol/L, for phosphate 1.9% at levels of 1.37 mmol/L and 2.20 mmol/L and for alkaline phosphate 7.2% at a level of 23.0 IU/L and 3.7% at a level of 55.3 IU/L. Creatinine and albumin concentrations in urine were also assayed on the UniCel DxC 800 Synchron Clinical System using a modified Jaffe rate method and turbidimetric method, respectively, with the recommended manufacturers' reagents, calibrators and settings. Albuminuria was calculated from the urine creatinine and urine microalbumin ratio. Between-run precision for urine creatinine was 1.0% at a level of 17.3 mmol/L and 1.4% at a level of 7.15 mmol/L and for urine microalbumin 2.5% at levels of 183.0 mg/L and 5.0% at levels of 25.6 mg/L.
Using an AU680 Chemistry System (Beckman Coulter (UK) Ltd, High Wycombe, England) and EDTA plasma, end point absorbance methods were used to measure plasma levels of cholesterol (assay traceable to NIST 909b), LDL cholesterol (assay traceable to the CDC LDL cholesterol reference method), HDL cholesterol (assay traceable to the CDC HDL cholesterol reference method) and triglycerides (assay traceable to the Isotope Dilution Mass Spectrometry Reference Method); and turbidimetric methods were used to measure plasma levels of apolipoprotein A 1 (assay traceable to the WHO -IFCC International Reference Material SP1-01 and SP-03), apolipoprotein B (assay traceable to the WHO -IFCC International Reference Material SP3-07). The LDL cholesterol and HDL cholesterol methods used were the Genzyme direct methods manufactured by Sekisui Medical, Tokyo. The CTSU laboratories are accredited (ISO 17025:2005) by the UK Accreditation service for the lipid assays described above. Between-run precision for cholesterol was 1.5% at a level of 7.66 mmol/L and 1.3% at a level of 3.53 mmol/L; for LDL cholesterol 2.1% at levels of 3.63 mmol/L and 1.73 mmol/L, for HDL cholesterol 3.1% at a level of 2.88 mmol/L and 2.9% at a level of 1.01 mmol/L; for triglycerides 1.7% at a level of 2.93 mmol/L and 1.6% at a level of 1.11 mmol/L; for apolipoprotein B 1.9% at a level of 135.2 mg/dL and 1.7% at a level of 69.0 mg/dL; and for apolipoprotein A1 1.5% at a level of 124.1 mg/dL and 2.0% at a level of 61.2 mg/dL.
Plasma high sensitivity C-reactive protein levels were measured in EDTA plasma on a Siemens BN ProSpec System using an immunoassay method (assay calibrated with reference to ERM-DA470). Between-run precision for high sensitivity C-reactive protein was 3.4% at a level of 9.63 mg/L and 3.1% at a level of 1.22 mg/L.   6 Online Resource Counts are of the number of patients experiencing at least one event of the given type. There were 3 deaths among placebo-allocated patients (1 gastrointestinal death and 2 neoplastic deaths).