Skip main navigation

Effect of Vitamin D Supplementation on Markers of Vascular Function: A Systematic Review and Individual Participant Meta‐Analysis

Originally publishedhttps://doi.org/10.1161/JAHA.117.008273Journal of the American Heart Association. 2018;7:e008273

    Abstract

    Background

    Low 25‐hydroxyvitamin D levels are associated with an increased risk of cardiovascular events, but the effect of vitamin D supplementation on markers of vascular function associated with major adverse cardiovascular events is unclear.

    Methods and Results

    We conducted a systematic review and individual participant meta‐analysis to examine the effect of vitamin D supplementation on flow‐mediated dilatation of the brachial artery, pulse wave velocity, augmentation index, central blood pressure, microvascular function, and reactive hyperemia index. MEDLINE, CINAHL, EMBASE, Cochrane Central Register of Controlled Trials, and http://www.ClinicalTrials.gov were searched until the end of 2016 without language restrictions. Placebo‐controlled randomized trials of at least 4 weeks duration were included. Individual participant data were sought from investigators on included trials. Trial‐level meta‐analysis was performed using random‐effects models; individual participant meta‐analyses used a 2‐stage analytic strategy, examining effects in prespecified subgroups. 31 trials (2751 participants) were included; 29 trials (2641 participants) contributed data to trial‐level meta‐analysis, and 24 trials (2051 participants) contributed to individual‐participant analyses. Vitamin D3 daily dose equivalents ranged from 900 to 5000 IU; duration was 4 weeks to 12 months. Trial‐level meta‐analysis showed no significant effect of supplementation on macrovascular measures (flow‐mediated dilatation, 0.37% [95% confidence interval, −0.23 to 0.97]; carotid‐femoral pulse wave velocity, 0.00 m/s [95% confidence interval, −0.36 to 0.37]); similar results were obtained from individual participant data. Microvascular function showed a modest improvement in trial‐level data only. No consistent benefit was observed in subgroup analyses or between different vitamin D analogues.

    Conclusions

    Vitamin D supplementation had no significant effect on most markers of vascular function in this analysis.

    Clinical Perspective

    What Is New?

    • This is the first individual participant data meta‐analysis examining the effect of vitamin D analogues on markers of vascular function that are surrogates for cardiovascular events.

    • No consistent effect was found at trial level or on analysis of individual participant‐level data of supplementation on measures of endothelial function, arterial stiffness, or central blood pressure.

    • No subgroup benefited consistently on analysis of individual participant data.

    What Are the Clinical Implications?

    • This analysis did not find convincing evidence of benefit from Vitamin D supplementation on a range of markers of vascular function.

    Introduction

    Low circulating levels of 25‐hydroxyvitamin D (25(OH)D) have been associated with a wide range of illness states and physiological derangements. Within the field of cardiometabolic medicine, low 25(OH)D levels have been associated with higher levels of blood pressure (BP), with diabetes mellitus, stroke, myocardial infarction, and heart failure1, 2 in observational studies. Vitamin D affects hundreds of gene targets and has effects on a wide variety of cell types and organ systems, including the heart and vascular system.3, 4 Several pathophysiological pathways have been postulated to explain the observed associations between low 25(OH)D levels and cardiovascular disease, including effects on arterial stiffness, endothelial function, cytokine secretion, vascular endothelial growth factor, and cellular calcium influx.4

    Despite a sound rationale for improved cardiovascular health with vitamin D supplementation, results from intervention trials have been less encouraging. A recent individual participant data (IPD) meta‐analysis reported that vitamin D supplementation had no significant effect on BP,5 even in those participants with low baseline 25(OH)D levels or with high baseline BP. Similarly, only marginal effects were observed on glycemic control in a meta‐analysis of vitamin D supplementation in participants with diabetes mellitus.6 Meta‐analyses of cardiovascular outcomes show no effect of vitamin D supplementation on myocardial infarction or stroke, but suggest a possible effect in reducing new diagnoses of heart failure.7 It is important to note that these meta‐analyses include mostly trials performed in participants at risk for falls or with osteoporosis and may therefore not be generalizable. Several large trials of vitamin D supplementation with adequate power to detect reductions in cardiovascular events are due to report over the next few years, but the first of these trials did not report any reduction in cardiovascular event rates in a population of older people in New Zealand.8

    These results call into question the causal link between vitamin D status and vascular health. Results from trials investigating the effect of vitamin D supplementation on aspects of vascular health other than BP have shown mixed results. Arterial stiffness and endothelial function measures are validated markers of cardiovascular disease risk and major adverse cardiac events, but the beneficial impact of vitamin D supplementation on these markers is unclear. We therefore performed a systematic review with meta‐analysis of trial‐level and individual participant‐level data to ascertain whether (1) vitamin D supplementation improves measures of arterial stiffness and endothelial function and (2) certain subgroups of individuals are more likely to benefit.

    Methods

    Data Sharing Statement

    To preserve the rights of data owners, and as agreed with those who contributed data sets for this analysis, the data, analytical methods, and study materials will not be made available to other researchers for the purposes of reproducing the results or replicating the procedure.

    Review Design and Search Strategy

    We conducted a systematic review according to a prespecified protocol, which was registered on the PROSPERO database of systematic reviews. The protocol is accessible at: (http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42012002816). Ethics committee approval was not required because no new data were collected as part of this review. We included randomized controlled trials, which compared vitamin D or analogues with placebo, with a minimum exposure period of 4 weeks. The following databases were searched from inception to end of December 2016: Medline, EMBASE, CINAHL, ClinicalTrials.gov, and the Cochrane central register of controlled trials. Gray literature was sought using Google, and references of included studies were hand‐searched for further candidate trials. Only trials where a full published trial report was available were included; trials published in abstract form only were excluded.

    Trial Selection

    Trials with changes in the following vascular markers were included: brachial artery flow‐mediated dilatation; reactive hyperemia index measures using finger plethysmography; pulse wave velocity (PWV) and pulse wave analysis; central aortic BP derived from peripheral artery tonometry; microvascular function measured using acetylcholine iontophoresis; and laser Doppler perfusion imaging. Studies with any baseline 25(OH)D level were eligible for inclusion. The following interventions were eligible for inclusion: vitamin D2 (ergocalciferol), vitamin D3 (cholecalciferol), calcitriol (1,25 hydroxyvitamin D3), 1‐alpha‐vitamin D, paricalcitol, and doxerocalciferol. Control groups receiving placebo were used and those receiving placebo plus cointervention were included, provided both arms of the study received the cointervention. A minimum of 4 weeks of therapy was necessary for inclusion to ensure sufficient time for vascular markers to change. Studies from both primary and secondary care or population settings were included; no restrictions were placed on sex or ethnicity. Studies recruiting participants less than 16 years old were not included, but in contrast to our previous review,5 we did include studies of participants on renal replacement therapy (hemodialysis or peritoneal dialysis) given their very high cardiovascular risk and the current interest in using vitamin D supplementation therapy in this group.

    Data Extraction

    Data were extracted independently by 2 researchers (L.A.B. and M.D.W.) and differences resolved by consensus. Baseline trial population data were identified, including age, sex, ethnicity, diabetes mellitus, kidney function, history of cardiovascular events, history of hypertension, baseline BP, and baseline use of angiotensin‐converting enzyme (ACE) inhibitors, angiotensin receptor blockers, and statins. For each measure of vascular function, we recorded change in the outcome in each group between baseline and the last follow‐up visit. Study authors were contacted if data were incomplete or ambiguous in primary reports.

    IPD Collection

    Lead authors for each included trial were contacted and invited to contribute individual‐level participant data. Data were anonymized and transferred using a standard template before cleaning and incorporation in the final data set. Individual‐level participant data were sought for age, sex, body mass index, baseline and follow‐up 25(OH)D level, baseline medication use including ACE inhibitors and angiotensin receptor blockers, baseline estimated glomerular filtration rate (eGFR), total cholesterol, serum calcium and parathyroid hormone (PTH), presence of diabetes mellitus and previous vascular events, baseline and follow up BP and cholesterol, and baseline and follow‐up measures of vascular function.

    Risk of Bias Assessment

    Risk of bias was evaluated by 2 authors independently, with discrepancies resolved by consensus. We assessed each included study for risk for bias using the following fields from a risk of bias checklist9: quality of random allocation concealment, intention‐to‐treat analysis, blinding of outcome assessors, treatment and control group comparability, clear definition of inclusion and exclusion criteria, participant blinding to allocation, and description of withdrawals and dropouts. Funnel plots were generated and inspected for evidence of publication bias, supplemented by Egger's test for funnel plot asymmetry.

    Statistical Analysis

    Meta‐analysis at the trial level was performed using RevMan 5.3 software (Cochrane Collaboration). For all analyses, random‐effects and fixed‐effects meta‐analyses using a weighted least‐squares approach were performed. For outcomes measured with the same technique and same units (most brachial‐artery flow‐mediated dilatation [FMD] measures, reactive hyperemia index, augmentation index [AIx], central BP, and subgroups of PWV), results were expressed as mean difference between groups. For comparisons where dissimilar units were combined, results were expressed as standardized mean difference (SMD). Heterogeneity was assessed using the I2 statistic. Trial‐level meta‐regression was undertaken for FMD, PWV, and augmentation index outcomes, regressing treatment effect on daily dose equivalent (for trials using vitamin D3) and trial duration in months. Metaregression was not used for other outcomes because there were too few to produce reliable results. Metaregression was undertaken using Comprehensive Meta Analysis tools software (version 3; Biostat, Englewood, NJ).

    A 2‐stage analysis was used for IPD.10 For each trial, or subgroup within each trial, mean outcome values at follow‐up in each group were calculated and adjusted for baseline outcome values using ANCOVA (SPSS version 24; IBM, Armonk, NY). These values were then combined using RevMan software as described above. For those trials using more than 1 type or dose of vitamin D, the vitamin D arms were analyzed as a single arm. The following prespecified subgroup analyses were performed: diabetes mellitus versus no diabetes mellitus; baseline systolic BP of no greater than 140 mm Hg versus greater than 140 mm Hg; diastolic BP of no greater than 90 mm Hg versus greater than 90 mm Hg; baseline PTH level of above versus below median level for the IPD set; baseline adjusted serum calcium level and baseline total cholesterol above versus below median level for the individual participant data set; estimated glomerular filtration rate (eGFR) 60 mL/min per 1.73 m2 or above versus subgroups of eGFR below 60 mL/min per 1.73 m2; baseline 25(OH)D level of less than 25, 25 to 50, and greater than 50 nmol/L; and baseline ACE inhibitor versus no ACE inhibitor use. For analyses of ACE inhibitor use, participants taking angiotensin receptor blockers were excluded given their similar, but not identical, biological effects. Subgroups for analysis were selected on the basis of possible mechanisms by which vitamin D might act (eg, through effects on the renin‐angiotensin system or by suppressing PTH),4 to explore groups thought to be most likely to benefit (eg, high BP, low 25(OH)D levels), and to identify disease populations that might be targeted by future studies (diabetes mellitus, chronic kidney disease [CKD]). For CKD, subgroups of eGFR 45 to 59, 30 to 44, 15 to 29, and <15 mL/min per 1.73 m2 were used where sufficient numbers of participants were available for the outcome; for outcomes with small numbers of participants, these subgroups were collapsed into a category eGFR <60 mL/min per 1.73 m2. Participants on dialysis were included in the <15 mL/min per 1.73 m2 subgroup.

    Results

    A total of 31 trials, involving 2751 participants, were eligible for inclusion in the review.11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41 Of these, 29 trials (2641 participants) had data suitable for inclusion in the trial‐level meta‐analyses; IPD were obtained from 24 trials (2051 participants). The PRISMA flowchart is shown in Figure. One study that did not include data in the published article suitable for trial‐level meta‐analysis36 provided IPD data and was included in the IPD analyses. Study size ranged from 24 to 305 participants; vitamin D3 was the most common intervention, being used in 23 of 31 (74%) of trials. The daily dose equivalent given in trials of vitamin D3 ranged from 900 to 5000 IU, and the duration of administration ranged from 4 weeks to 12 months. Paricalcitol was the only activated vitamin D analogue used in studies included in this review. Table 1 shows which vascular outcomes were measured in each included trial, and baseline trial characteristics are shown in Table 2.

    Figure 1.

    Figure 1. PRISMA diagram showing trial selection.

    Table 1. Measurements From Included Studies

    Title and YearPreparation TestedFMDPWVAIxRHICentral BPMicrovascular FunctionIncluded in IPD Analysis?
    Alborzi 200811ParicalcitolXNo
    Sugden 200812Vitamin D2XYes
    Witham 201013Vitamin D3XYes
    Harris 201114Vitamin D3XYes
    Gepner 201215Vitamin D3XXXXYes
    Larsen 201216Vitamin D3XXXYes
    Marckmann 201217Vitamin D3XXXYes
    Sokol 201218Vitamin D2XYes
    Stricker 201219Vitamin D3XXXYes
    Witham 201220Vitamin D2XYes
    Breslavsky 201321Vitamin D3XYes
    Hewitt 201322Vitamin D3XNo
    Witham 201323Vitamin D3XXYes
    Witham 201324Vitamin D3XYes
    Witham 201325Vitamin D3XXXXXYes
    Yiu 201326Vitamin D3XXNo
    Dreyer 201427Vitamin D2XXYes
    Martins 201428Vitamin D3XNo
    Mose 201429Vitamin D3XXXYes
    Ryu 201430Vitamin D3XXNo
    Zoccali 201431ParicalcitolXYes
    Garg 201532Vitamin D3XXYes
    Pilz 201533Vitamin D3XYes
    Thethi 201534ParicalcitolXNo
    Witham 201535Vitamin D3XXXYes
    Barchetta 201636Vitamin D3XYes
    Borgi 201737Vitamin D2XNo
    Bressendorff 201638Vitamin D3XXXYes
    Dalan 201639Vitamin D3XXYes
    Forouhi 201640 Vitamin D2 Vitamin D3 XYes
    Hin 201741Vitamin D3XXYes

    AIx indicates augmentation index; BP, blood pressure; FMD, flow‐mediated dilatation of the brachial artery; IPD, individual participant data; PWV, pulse wave velocity; RHI, reactive hyperemia index.

    Table 2. Details of Included Studies

    StudyNLatitudeStudy PopulationMean Age, y% Male25(OH)D Range for Inclusion (nmol/L)Mean Baseline 25(OH)D (nmol/L)Mean Baseline SBP (mm Hg)ControlInterventionDuration
    Alborzi,11 USA, 20082440°NCKD and on ACE‐I or ARB7083No restriction34125.4 (24 hours BP)Placebo Paricalcitol 1 μg daily Paricalcitol 2 μg daily 1 mo
    Sugden,12 Scotland, 20083456°NType 2 diabetes mellitus6453<5038141Placebo Ergocalciferol 100 000 IU single dose 8 wks
    Witham,13 Scotland, 20106156°NType 2 diabetes mellitus6567<10045146Placebo 100 000 IU Cholecalciferol 200 000 IU cholecalciferol single dose 16 wks
    Harris,14 USA, 20114533°NBlack adults with no overt cardiovascular, pulmonary or metabolic disease3047No restriction36124PlaceboCholecalciferol 60 000 IU/4 weekly16 wks
    Gepner,15 USA, 201211443°NHealthy community dwelling postmenopausal females640>25 and <15078119.4PlaceboCholecalciferol 2500 IU/day4 mo
    Larsen,16 Denmark, 201213056°NHypertension6131No restriction58143PlaceboCholecalciferol 3000 IU/day20 wks
    Marckmann,17 Denmark, 20125255°NCKD6775<5033138PlaceboCholecalciferol 40 000 IU weekly8 wks
    Sokol,18 USA, 20129041°NAngiographically confirmed coronary artery disease5673<50?133PlaceboErgocalciferol 50 000 IU weekly12 wks
    Stricker,19 Switzerland, 20126246°NChronic peripheral vascular disease and vitamin D deficiency7461<7542137PlaceboCholecalciferol (vitamin D3) 100 000 IU single dose1 mo
    Witham,20 Scotland, 20125856°NOlder adults with previous stroke6772<7538128Placebo100 000 IU Ergocalciferol single dose16 wks
    Breslavsky,21 Israel, 20134732°NType 2 diabetes mellitus with cardiovascular risk factors6647No restriction30153PlaceboCholecalciferol 1000 IU daily12 mo
    Hewitt,22 Australia, 20136034°SAdults on hemodialysis?48<6043152PlaceboCholecalciferol 50 000 IU weekly for 8 wks, then monthly for 4 mo6 mo
    Witham,23 Scotland, 201315956°NIsolated systolic hypertension in over 70 year olds7752<7545163PlaceboCholecalciferol 100 000 IU 3 monthly12 mo
    Witham,24 Scotland, 20137556°NRecent myocardial infarction6669No restriction47127.5PlaceboCholecalciferol 100 000 IU/2 monthly6 mo
    Witham,25 Scotland, 20135056°NSouth‐East Asian women living in UK for 10 y410<7527120PlaceboCholecalciferol 100 000 IU single dose8 wks
    Yiu,26 Hong Kong, 201310022°NType 2 diabetes mellitus with suboptimal vitamin D status6550<7554146PlaceboCholecalciferol 5000 IU daily12 wks
    Dreyer,27 England, 20143851°NCKD stage 3 to 44774<4025116PlaceboErgocalciferol 50 000 IU weekly for 1 mo, then monthly for 5 mo6 mo
    Martins,28 USA, 201413034°NOverweight blacks with hypertension18 to 706125 to 6343127PlaceboCholecalciferol 100 000 IU monthly3 mo
    Mose,29 Denmark, 20145056°NHemodialysis or peritoneal dialysis6864No restriction39136PlaceboCholecalciferol 3000 IU daily6 mo
    Ryu,30 Korea, 20146238°NType 2 diabetes mellitus aged 30 to 69 y56?<5029130Placebo+200 mg calcium dailyCholecalciferol 2000 IU daily+200 mg calcium daily24 wks
    Zoccali,31 Italy, 20148839°NCKD stage 3 to 4 and PTH >65 pg/mL6365No restriction36126PlaceboParicalcitol 2 μg daily12 wks
    Garg,32 India, 20153229°NWomen aged 18 to 35 y with polycystic ovary syndrome220No restriction36Not knownPlacebo+1.5 g metformin daily120 000 IU Cholecalciferol monthly+1.5 g metformin daily6 mo
    Pilz,33 Austria, 201518847°NHypertension6053<7553143PlaceboCholecalciferol 2800 IU daily8 wks
    Thethi,34 USA, 20155530°NCKD 3 to 4 and type 2 diabetes mellitus6367No restrictionNA136PlaceboParicalcitol 1 μg daily12 wks
    Witham,35 Scotland, 20155056°NChronic fatigue syndrome4924<7546128.5PlaceboCholecalciferol 100 000 IU/2 monthly6 mo
    Barchetta.36 Italy. 20165542°NType 2 diabetes mellitus with nonalcoholic fatty liver disease5870No restrictionNA131PlaceboCholecalciferol 2000 IU daily24 wks
    Borgi,37 USA, 20179342°NOverweight or obese nonhypertensives with vitamin D deficiency37?<50NA118PlaceboErgocalciferol 50 000 IU weekly8 wks
    Bressendorff,38 Denmark, 20164056°NNormotensive adults with vitamin D deficiency4358<5032118PlaceboCholecalciferol 3000 IU daily16 wks
    Dalan,39 Singapore, 2016641°NMultiethnic group with type 2 diabetes mellitus5452No restriction45139PlaceboCholecalciferol 2000 to 4000 IU daily depending on baseline 25(OH)D and response16 wks
    Forouhi,40 England, 2016a16052°NAdults with elevated risk of type 2 diabetes mellitus5357No restriction52128Placebo Ergocalciferol 100 000 IU monthly Cholecalciferol 100 000 IU monthly 4 mo
    Hin,41 England, 201730552°NAged ≥65 y7251No restriction50131Placebo Cholecalciferol 2000 IU daily Cholecalciferol 4000 IU daily 12 mo

    25(OH)D indicates 25‐hydroxyvitamin D; ACEi, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; BP, blood pressure; CKD, chronic kidney disease; NA, not available; SBP, systolic blood pressure.

    aData from 1 of 2 centers included as vascular function measured only at 1 center.

    Risk of Bias Assessment

    Overall risk of bias was low. Most trials (27 of 31) reported clear evidence of effective allocation concealment, and most trials reported clear evidence for masking of participants (30 of 31), healthcare professionals (30 of 31), and outcomes assessors (29 of 31). Groups were comparable at baseline in 27 of 31 trials, dropouts were clearly described in 27 of 31 trials, but analysis was clearly by intention to treat in only 16 of 31 trials. A full description of the quality assessment for each trial is shown in Table 3. Funnel plots showed no asymmetry for any of the vascular outcomes; Egger's test was calculated only for those outcomes with at least 10 trials to ensure reliability; this was nonsignificant for FMD (P=0.18), AIx (P=0.32), and PVW (P=0.70).

    Table 3. Risk of Bias Assessment of Included Studies

    Title and YearQuality of Allocation ConcealmentAnalysis on Intention to TreatNo. and Description of DropoutsBlinding—ParticipantsBlinding—Health Care ProvidersBlinding—Outcome AssessorsComparable Treatment and Placebo Groups
    Alborzi 200811++++++
    Sugden 200812++++++
    Witham 201013+U+++++
    Harris 201114+U+++++
    Gepner 201215+++++++
    Larsen 201216++++++
    Marckmann 201217++++++
    Sokol 201218++U++++
    Stricker 201219++++++
    Witham 201220+U+++++
    Breslavsky 201321++UUU+
    Hewitt 201322+U+++++
    Witham 201323++++++
    Witham 201324+++++++
    Witham 201325+++++++
    Yiu 201326+++++++
    Dreyer 201427+++++++
    Martins 201428U++++++
    Mose 201429+U+++++
    Ryu 201430U+++++
    Zoccali 201431+++++++
    Garg 201532U+++++
    Pilz 201533+++++++
    Thethi 201534UUU++U+
    Witham 201535+++++++
    Barchetta 201636++++++
    Borgi 201737++++++
    Bressendorff 201638+U+++++
    Dalan 201639++U+++
    Forouhi 201640+++++++
    Hin 201741+++++++

    − indicates high risk of bias; +, low risk of bias; U, unclear risk of bias.

    Trial‐Level Data

    Meta‐analysis of trial‐level data showed no significant treatment effect of vitamin D analogues on FMD (mean difference, 0.5%; 95% confidence interval [CI], −0.1 to 1.1; P=0.12), PVW (SMD, 0.02; 95% CI, −0.13 to 0.17; P=0.81), AIx (mean difference, 0.0%; 95% CI, −1.3 to 1.3; P=0.98), reactive hyperemia index (mean difference, 0.02 units; 95% CI, −0.11 to 0.14; P=0.79), or central BP. Microvascular function measured by laser Doppler iontophoresis (SMD, 0.43; 95% CI, 0.09–0.76; P=0.01) showed a modest improvement with vitamin D supplementation. Results are shown in Table 4. Fixed‐effects analyses showed similar point estimates, but narrower CIs, leading to a significant treatment effect for all vitamin D analogues on FMD (mean difference, 0.5%; 95% CI, 0.1–0.9; P=0.02). On subgroup analysis by treatment type, only paricalcitol showed a significant treatment benefit on FMD (mean difference, 1.7%; 95% CI, 0.6–2.8; P=0.002); analysis for interaction showed no significant difference between the paricalcitol treatment effect and that for vitamin D3 (P=0.62) or vitamin D2 (P=0.17). No significant difference was evident in the effects of daily dosing versus intermittent dosing either for FMD (mean difference, 0.85% [95% CI, 0.01–1.69] for daily dosing versus 0.26% [95% CI, −0.58 to 1.11] for intermittent dosing; P=0.06) or for PWV (SMD, 0.10 [95% CI, −0.04 to 0.23] for daily dosing versus −0.04 [95% CI, −0.34 to 0.26] for intermittent dosing; P=0.40). Metaregression results for daily dose equivalent and for trial duration are shown in Table 5. No association between these factors and treatment effect for PWV or AIx was found, but higher dose and shorter trial length were associated with a slightly greater treatment effect for FMD.

    Table 4. Trial‐Level Analysis of Effect Size: Vitamin D Supplementation and Markers of Vascular Function

    OutcomeInterventionNo. of StudiesnRandom‐EffectsFixed‐EffectsI2
    Treatment Effect (95% CI)P ValueTreatment Effect (95% CI)P Value
    FMD (%)All127850.49 (−0.13 to 1.11)0.120.48 (0.06–0.90)0.0246%
    D374950.17 (−0.49 to 0.84)0.610.19 (−0.30 to 0.67)0.4538%
    D231630.79 (−1.04 to 2.62)0.400.91 (−0.39 to 2.21)0.1745%
    Paricalcitol21031.72 (0.63–2.82)0.0021.72 (0.63–2.82)0.0020%
    AIx (%)All1410300.0 (−1.3 to 1.3)0.980.0 (−1.1 to 1.1)0.9825%
    D31410300.0 (−1.3 to 1.3)0.980.0 (−1.1 to 1.1)0.9825%
    D20··················
    Paricalcitol0··················
    RHI, unitsAll32170.02 (−0.11 to 0.14)0.790.02 (−0.11 to 0.14)0.790%
    D321300.02 (−0.18 to 0.21)0.860.04 (−0.10 to 0.18)0.6137%
    D2187−0.05 (−0.30 to 0.20)0.70−0.05 (−0.30 to 0.20)0.70···
    Paricalcitol0············
    PWV (all; SMD)All1613330.04 (−0.11 to 0.20)0.600.04 (−0.07 to 0.15)0.5044%
    D31513040.05 (−0.11 to 0.21)0.520.04 (−0.07 to 0.15)0.4547%
    D22a138−0.24 (−0.57 to 0.10)0.17−0.24 (−0.57 to 0.10)0.170%
    Paricalcitol0··················
    PWV (carotid‐femoral only; m/s)aAll106740.04 (−0.32 to 0.41)0.81−0.01 (−0.20 to 0.21)0.9458%
    D3106740.00 (−0.32 to 0.41)0.81−0.01 (−0.20 to 0.21)0.9458%
    D21107−0.53 (−1.34 to 0.28)0.20−0.53 (−1.34 to 0.28)0.20···
    Paricalcitol0··················
    PWV (others; SMD)All66590.11 (−0.06 to 0.28)0.220.11 (−0.04 to 0.27)0.158%
    D356300.12 (−0.06 to 0.30)0.190.13 (−0.03 to 0.29)0.1114%
    D2129−0.20 (−0.93 to 0.53)0.59−0.20 (−0.93 to 0.53)0.59···
    Paricalcitol0··················
    Microvascular function (SMD)All31400.43 (0.09–0.76)0.010.43 (0.09–0.76)0.010%
    D321110.37 (−0.01 to 0.75)0.050.37 (−0.01 to 0.75)0.050%
    D21290.65 (−0.10 to 1.41)0.090.65 (−0.10 to 1.41)0.09···
    Paricalcitol0··················
    Central SBP, mm HgAll5324−1.5 (−5.6 to 2.6)0.46−1.2 (−3.8 to 1.4)0.3647%
    D35324−1.5 (−5.6 to 2.6)0.46−1.2 (−3.8 to 1.4)0.3647%
    D20··················
    Paricalcitol0··················
    Central DBP, mm HgAll5324−0.8 (−2.2 to 0.6)0.28−0.8 (−2.2 to 0.6)0.280%
    D35324−0.8 (−2.2 to 0.6)0.28−0.8 (−2.2 to 0.6)0.280%
    D20··················
    Paricalcitol0··················

    AIx indicates augmentation index; CI, confidence interval; D2, vitamin D2 (ergocalciferol); D3, vitamin D3 (cholecalciferol); DBP, diastolic blood pressure; FMD, flow‐mediated dilatation of the brachial artery; PWV, pulse wave velocity; RHI, reactive hyperemia index; SBP, systolic blood pressure; SMD, standardized mean difference.

    aData from Forouhi et al40 contain comparisons of D3 vs placebo and D2 vs placebo. Only D3 analysis was included in “All” category for PWV analyses.

    Table 5. Results of Trial‐Level Metaregression

    OutcomeModerator VariableNo. of TrialsRandom‐EffectsFixed‐EffectsRegression Coefficient Units
    Regression Coefficient (95% CI)P ValueRegression Coefficient (95% CI)P Value
    FMD (%)Trial duration11−0.06 (−0.11 to −0.01)0.03−0.06 (−0.10 to −0.01)0.009% per mo
    Daily dose equivalent60.14 (0.01–0.27)0.030.14 (0.01–0.27)0.03% per 1000 units D3
    PWV (SMD)Trial duration160.01 (−0.04 to 0.06)0.770.02 (−0.01 to 0.05)0.18SD per mo
    Daily dose equivalent150.07 (−0.08 to 0.23)0.350.06 (−0.05 to 0.17)0.29SD per 1000 units D3
    AIx (%)Trial duration14−0.01 (−0.05 to 0.04)0.73−0.00 (−0.04 to 0.03)0.84% per mo
    Daily dose equivalent140.10 (−0.06 to 0.25)0.220.09 (−0.04 to 0.23)0.18% per 1000 units D3

    AIx indicates augmentation index; CI, confidence interval; FMD, flow‐mediated dilatation of the brachial artery; PWV, pulse wave velocity; SMD, standardized mean difference.

    Individual Participant Data

    Similarly, meta‐analysis of IPD showed no significant treatment effect on any of the vascular outcomes studied; no effect was evident when PWV analyses were confined to studies using carotid‐femoral PWV. The main results for IPD analysis are shown in Table 6. For most analyses, heterogeneity as shown by the I2 statistic was low to moderate. Prespecified subgroup analysis of IPD data for each vascular outcome (Tables 7, 8, 9, 10, 11, 12 through 13) did not show any subgroup consistently deriving significant benefit from vitamin D or analogues; treatment/subgroup interaction analyses suggested that participants with eGFR <60 mL/min per 1.73 m2 may be less likely to show improvements in FMD or aortic diastolic BP with treatment, and those with 25(OH)D levels <25 nmol/L may be more likely to show improvements in reactive hyperemia index and AIx with treatment. Participants with diabetes mellitus appeared to have a significantly greater rise in aortic BP than those without diabetes mellitus. For outcomes in both trial‐level and IPD analysis that were performed using SMD, standardizing SDs are given in Table 14 to facilitate interpretation at the individual trial level.

    Table 6. IPD Analysis of Effect Size: Vitamin D Supplementation and Markers of Vascular Function

    OutcomeNo. of StudiesnMean Baseline Value Across All Groups (SD)Random EffectsFixed EffectsI2
    Treatment Effect (95% CI)P ValueTreatment Effect (95% CI)P Value
    FMD (%)106555.6 (3.6)−0.03 (−0.78 to 0.71)0.93−0.17 (−0.58 to 0.25)0.4463%
    AIx (%)1183227 (16)0.1 (−1.4 to 1.6)0.910.0 (−1.0 to 1.1)0.9639%
    RHI, units32201.65 (0.81)−0.02 (−0.12 to 0.08)0.68−0.02 (−0.12 to 0.08)0.680%
    PWV (all; SMD)131154ND−0.01 (−0.16 to 0.13)0.85−0.04 (−0.15 to 0.08)0.5625%
    PWV (carotid‐femoral only; m/s)96527.9 (2.8)−0.01 (−0.31 to 0.30)0.96−0.04 (−0.25 to 0.17)0.7044%
    PWV (others; SMD)4502ND−0.02 (−0.20 to 0.16)0.83−0.05 (−0.36 to 0.26)0.750%
    Microvascular function (SMD)3129ND0.36 (0.01–0.71)0.050.36 (0.01–0.71)0.130%
    Central SBP, mm Hg7400120.7 (23.9)−0.6 (−3.2 to 1.9)0.63−0.4 (−2.4 to 1.6)0.6731%
    Central DBP, mm Hg740076.5 (9.9)−0.4 (−1.5 to 0.7)0.48−0.4 (−1.5 to 0.7)0.480%

    AIx indicates augmentation index; CI, confidence interval; DBP, diastolic blood pressure; FMD, flow‐mediated dilatation of the brachial artery; IPD, individual participant data; ND, not done because of heterogeneity of measurement methods; PWV, pulse wave velocity; RHI, reactive hyperemia index; SBP, systolic blood pressure; SMD, standardized mean difference.

    Table 7. IPD Subgroup Analyses for FMD

    SubgroupNo. of StudiesnMean Baseline Value (%) (SD)Random EffectsFixed EffectsI2
    Treatment Effect (%) (95% CI)P ValueP for InteractionTreatment Effect (%) (95% CI)P ValueP for Interaction
    SBP >140 mm Hg92425.2 (3.2)0.0 (−1.1 to 1.0)0.940.88−0.1 (−0.6 to 0.5)0.821.056%
    SBP ≤140 mm Hg93765.9 (3.8)0.1 (−0.9 to 1.0)0.90−0.1 (−0.6 to 0.5)0.7958%
    Baseline 25(OH)D <25 nmol/L6895.7 (4.0)−0.3 (−1.8 to 1.2)0.65···−0.1 (−1.0 to 0.9)0.92···43%
    Baseline 25(OH)D 25 to 50 nmol/L92925.3 (3.3)−0.1 (−1.2 to 1.0)0.870.83−0.1 (−0.7 to 0.5)0.661.066%
    Baseline 25(OH)D >50 nmol/L82225.4 (3.4)−0.2 (−1.1 to 0.7)0.660.91−0.2 (−0.9 to 0.4)0.470.8741%
    DM61985.2 (3.5)0.1 (−1.1 to 1.4)0.820.470.4 (−0.5 to 1.2)0.400.2737%
    No DM74545.7 (3.5)−0.4 (−0.9 to 0.1)0.14−0.3 (−1.2 to 0.6)0.5563%
    No ACEi or ARB92546.5 (3.6)0.1 (−2.0 to 2.2)0.92···−0.6 (−1.3 to 0.1)0.09···88%
    ACEi, no ARB71715.4 (3.5)−0.4 (−1.7 to 1.0)0.580.69−0.2 (−1.0 to 0.6)0.640.4647%
    ACEi or ARB72975.0 (3.4)−0.7 (−1.9 to 0.5)0.280.52−0.7 (−1.4 to −0.1)0.030.8460%
    PTH >5.0 pmol/L82545.4 (3.6)−0.3 (−1.5 to 0.9)0.610.39−0.8 (−1.5 to −0.2)0.020.0262%
    PTH ≤5.0 pmol/L72305.7 (3.1)0.3 (−0.3 to 1.0)0.300.3 (−0.3 to 1.0)0.300%
    Ca >2.30 mmol/L72266.0 (3.4)0.3 (−0.3 to 1.0)0.300.190.3 (−0.3 to 1.0)0.300.080%
    Ca ≤2.30 mmol/L73715.2 (3.4)−0.5 (−1.5 to 0.5)0.30−0.5 (−1.1 to 0.1)0.0956%
    eGFR ≥60 mL/min per 1.73 m263825.6 (3.4)0.5 (−0.2 to 1.2)0.18···0.6 (0.0–1.2)0.04···27%
    eGFR 45 to 59 mL/min per 1.73 m25595.3 (3.1)−0.1 (−2.2 to 2.1)0.940.600.3 (−1.0 to 1.5)0.660.6755%
    eGFR 30 to 44 mL/min per 1.73 m23384.5 (2.9)−1.0 (−2.7 to 0.7)0.270.11−1.0 (−2.7 to 0.7)0.270.080%
    eGFR 15 to 29 mL/min per 1.73 m21243.4 (3.0)−3.1 (−5.7 to −0.5)0.020.009−3.1 (−5.7 to −0.5)0.020.007···
    eGFR <15 mL/min per 1.73 m2a0···························
    Total cholesterol ≥4.60 mmol/L82235.4 (3.4)−0.9 (−1.9 to 0.1)0.070.23−1.0 (−1.6 to −0.4)0.0020.0651%
    Total cholesterol <4.60 mmol/L82835.6 (3.7)0.1 (−1.1 to 1.4)0.83−0.2 (−0.8 to 0.4)0.6069%

    25(OH)D indicates 25‐hydroxyvitamin D; ACEi, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; Ca, calcium; CI, confidence interval; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; FMD, flow‐mediated dilatation of the brachial artery; IPD, individual participant data; PTH, parathyroid hormone; SBP, systolic blood pressure.

    aOr on dialysis.

    Table 8. IPD Subgroup Analyses for PWV

    SubgroupNo. of StudiesNMean Baseline Value (m/s) (SD)aRandom EffectsFixed EffectsI2
    Treatment Effect (95% CI) (SMD)P ValueP for InteractionTreatment Effect (95% CI) (SMD)P ValueP for Interaction
    SBP >140 mm Hg93968.9 (3.2)−0.1 (−0.3 to 0.1)0.210.49−0.1 (−0.3 to 0.1)0.210.430%
    SBP ≤140 mm Hg117287.5 (2.6)0.0 (−0.2 to 0.2)1.00−0.0 (−0.2 to 0.1)0.6240%
    Baseline 25(OH)D <25 nmol/L111577.6 (2.8)0.2 (−0.2 to 0.5)0.36···0.2 (−0.2 to 0.5)0.36···0%
    Baseline 25(OH)D 25 to 50 nmol/L125088.0 (3.2)−0.1 (−0.3 to 0.2)0.670.17−0.1 (−0.3 to 0.1)0.380.1539%
    Baseline 25(OH)D >50 nmol/L104827.8 (2.6)0.0 (−0.2 to 0.2)0.810.330.0 (−0.2 to 0.2)0.760.3312%
    DM61308.5 (4.5)0.1 (−0.3 to 0.4)0.650.300.1 (−0.3 to 0.4)0.650.300%
    No DM1310217.8 (2.5)−0.1 (−0.2 to 0.1)0.32−0.1 (−0.2 to 0.1)0.320%
    No ACEi or ARB93997.7 (3.1)−0.1 (−0.3 to 0.2)0.70···−0.1 (−0.3 to 0.1)0.44···21%
    ACEi, no ARB72038.5 (3.3)−0.2 (−0.5 to 0.1)0.170.62−0.2 (−0.5 to 0.1)0.170.620%
    ACEi or ARB83648.1 (3.2)0.0 (−0.3 to 0.2)0.680.580.0 (−0.3 to 0.2)0.680.580%
    PTH >5.0 pmol/L124827.9 (2.9)−0.1 (−0.3 to 0.1)0.160.49−0.1 (−0.3 to 0.1)0.160.490%
    PTH ≤5.0 pmol/L125947.9 (3.2)0.0 (−0.2 to 0.2)0.970.0 (−0.2 to 0.2)0.9512%
    Ca >2.30 mmol/L94567.8 (3.5)0.0 (−0.2 to 0.2)0.910.120.0 (−0.2 to 0.2)0.910.120%
    Ca ≤2.30 mmol/L104968.0 (2.6)−0.2 (−0.3 to 0.0)0.11−0.2 (−0.3 to 0.0)0.102%
    eGFR ≥60 mL/min per 1.73 m2108977.8 (2.3)0.0 (−0.2 to 0.1)0.57···0.0 (−0.2 to 0.1)0.55···5%
    eGFR 45 to 59 mL/min per 1.73 m26767.2 (4.8)0.3 (−0.3 to 0.9)0.350.340.2 (−0.3 to 0.7)0.410.4512%
    eGFR 30 to 44 mL/min per 1.73 m23278.8 (2.0)−0.8 (−1.6 to 0.1)0.070.07−0.8 (−1.6 to 0.1)0.070.070%
    eGFR 15 to 29 mL/min per 1.73 m22813.7 (1.1)0.1 (−2.0 to 2.2)0.950.930.1 (−2.0 to 2.2)0.950.930%
    eGFR <15 mL/min per 1.73 m2b2549.0 (4.9)0.4 (−0.1 to 1.0)0.120.170.4 (−0.1 to 1.0)0.120.170%
    Total cholesterol ≥4.60 mmol/L116517.8 (2.3)−0.1 (−0.3 to 0.1)0.240.54−0.1 (−0.3 to 0.1)0.240.490%
    Total cholesterol <4.60 mmol/L113987.7 (3.0)0.0 (−0.2 to 0.3)0.830.0 (−0.2 to 0.2)0.8723%

    25(OH)D indicates 25‐hydroxyvitamin D; ACEi, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; Ca, calcium; CI, confidence interval; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; IPD, individual participant data; PTH, parathyroid hormone; PWV, pulse wave velocity; SBP, systolic blood pressure; SMD, standardized mean difference.

    aData only from studies measuring baseline carotid‐femoral PWV.

    bOr on dialysis.

    Table 9. IPD Subgroup Analyses for AIx

    SubgroupNo. of StudiesnMean Baseline Value (%) (SD)Random EffectsFixed EffectsI2
    Treatment Effect (%) (95% CI)P ValueP for InteractionTreatment Effect (%) (95% CI)P ValueP for Interaction
    SBP >140 mm Hg921433.3 (16.6)0.8 (−1.1 to 2.7)0.390.310.8 (−1.1 to 2.7)0.390.260%
    SBP ≤140 mm Hg1161524.9 (14.6)−0.6 (−2.5 to 1.4)0.57−0.5 (−1.7 to 0.7)0.4452%
    Baseline 25(OH)D <25 nmol/L89324.9 (14.9)−4.3 (−11.0 to 2.4)0.20···−4.8 (−7.8 to −1.7)0.002···76%
    Baseline 25(OH)D 25 to 50 nmol/L1035725.8 (16.7)0.1 (−1.9 to 2.1)0.920.220.0 (−1.5 to 1.5)0.990.00630%
    Baseline 25(OH)D >50 nmol/L834228.6 (15.2)0.5 (−1.1 to 2.0)0.580.170.5 (−1.1 to 2.0)0.580.0020%
    DM716326.1 (15.6)2.0 (−2.6 to 6.6)0.400.370.4 (−2.2 to 3.1)0.750.6861%
    No DM966627.3 (15.7)−0.2 (−1.4 to 1.1)0.80−0.2 (−1.3 to 0.9)0.7311%
    No ACEi or ARB920923.1 (14.3)−0.2 (−2.0 to 1.7)0.86···−0.2 (−2.0 to 1.6)0.85···4%
    ACEi, no ARB713725.4 (12.6)0.3 (−1.6 to 2.2)0.790.710.3 (−1.6 to 2.2)0.790.710%
    ACEi or ARB724326.4 (12.7)−0.1 (−2.1 to 1.9)0.940.94−0.4 (−1.9 to 1.2)0.630.8727%
    PTH >5.0 pmol/L928225.3 (15.9)0.1 (−2.4 to 2.5)0.960.89−0.3 (−2.1 to 1.5)0.730.6243%
    PTH ≤5.0 pmol/L944127.9 (16.9)0.3 (−1.3 to 1.8)0.720.3 (−1.3 to 1.8)0.720%
    Ca >2.30 mmol/L528628.3 (18.7)0.0 (−3.6 to 3.6)1.000.45−0.1 (−3.4 to 3.2)0.960.3911%
    Ca ≤2.30 mmol/L626526.2 (13.3)1.6 (−0.4 to 3.6)0.121.6 (−0.4 to 3.6)0.120%
    eGFR ≥60 mL/min per 1.73 m2967827.3 (15.7)−0.3 (−1.7 to 1.2)0.34···−0.2 (−1.4 to 0.9)0.72···25%
    eGFR 45 to 59 mL/min per 1.73 m266428.3 (16.7)−5.4 (−11.7 to 0.9)0.090.36−11.7 (−12.2 to −11.3)<0.001<0.00186%
    eGFR 30 to 44 mL/min per 1.73 m221729.7 (18.2)−0.8 (−8.8 to 7.2)0.850.43−0.8 (−8.8 to 7.2)0.850.880%
    eGFR 15 to 29 mL/min per 1.73 m20···························
    eGFR <15 mL/min per 1.73 m2a25122.9 (12.8)6.6 (2.1–11.2)0.0050.406.6 (2.1–11.2)0.0050.0050%
    Total cholesterol ≥4.60 mmol/L943427.8 (15.6)−0.5 (−2.7 to 1.7)0.660.950.0 (−1.4 to 1.5)1.000.6139%
    Total cholesterol <4.60 mmol/L929826.4 (16.2)−0.4 (−2.5 to 1.7)0.73−0.6 (−2.4 to 1.2)0.5220%

    25(OH)D indicates 25‐hydroxyvitamin D; ACEi, angiotensin‐converting enzyme inhibitor; Aix, augmentation index; ARB, angiotensin receptor blocker; Ca, calcium; CI, confidence interval; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; IPD, individual participant data; PTH, parathyroid hormone; SBP, systolic blood pressure.

    aOr on dialysis.

    Table 10. IPD Subgroup Analyses for RHI

    SubgroupNo. of StudiesnMean Baseline Value (Units) (SD)Random EffectsFixed EffectsI2
    Treatment Effect (Units) (95% CI)P ValueP for InteractionTreatment Effect (Units) (95% CI)P ValueP for Interaction
    SBP >140 mm Hg3731.71 (0.94)−0.21 (−0.52 to 0.10)0.180.21−0.14 (−0.33 to 0.05)0.160.2049%
    SBP ≤140 mm Hg31471.62 (0.74)0.01 (0.11–0.14)0.850.01 (0.11–0.14)0.850%
    Baseline 25(OH)D <25 nmol/L1211.82 (0.68)0.23 (−0.10 to 0.56)0.18···0.23 (−0.10 to 0.56)0.18······
    Baseline 25(OH)D 25 to 50 nmol/L31531.71 (0.83)−0.02 (−0.19 to 0.16)0.860.190.02 (−0.09 to 0.14)0.670.2446%
    Baseline 25(OH)D >50 nmol/L2401.33 (0.76)−0.21 (−0.42 to 0.00)0.050.03−0.21 (−0.42 to 0.00)0.050.030%
    DM31101.23 (0.74)0.01 (−0.10 to 0.11)0.910.170.01 (−0.10 to 0.11)0.910.170%
    No DM21012.07 (0.63)−0.15 (−0.35 to 0.06)0.16−0.15 (−0.35 to 0.06)0.160%
    No ACEi or ARB3491.57 (0.79)0.42 (−0.72 to 1.57)0.47···0.47 (0.29 to 0.66)<0.001···97%
    ACEi, no ARB31281.75 (0.82)−0.00 (−0.16 to 0.15)0.960.48−0.00 (−0.14 to 0.13)0.96<0.00127%
    ACEi or ARB31711.67 (0.81)−0.05 (−0.21 to 0.11)0.540.43−0.04 (−0.15 to 0.07)0.50<0.00146%
    PTH >5.0 pmol/L31171.77 (0.77)−0.02 (−0.17 to 0.13)0.800.85−0.02 (−0.17 to 0.13)0.800.850%
    PTH ≤5.0 pmol/L31021.52 (0.84)0.00 (−0.14 to 0.14)1.000.00 (−0.14 to 0.14)1.000%
    Ca >2.30 mmol/L31271.50 (0.81)−0.08 (−0.21 to 0.05)0.210.22−0.08 (−0.21 to 0.05)0.210.220%
    Ca ≤2.30 mmol/L3931.86 (0.76)0.05 (−0.11 to 0.21)0.550.05 (−0.11 to 0.21)0.550%
    eGFR ≥60 mL/min per 1.73 m231921.69 (0.80)−0.04 (−0.14 to 0.07)0.480.81−0.04 (−0.14 to 0.07)0.480.770%
    eGFR <60 mL/min per 1.73 m22271.41 (0.84)−0.10 (−0.58 to 0.39)0.690.00 (−0.25 to 0.25)1.0054%
    Total cholesterol ≥4.60 mmol/L3601.44 (0.85)−0.04 (−0.38 to 0.30)0.811.000.00 (−0.19 to 0.18)0.980.7343%
    Total cholesterol <4.60 mmol/L31411.72 (0.81)−0.04 (−0.16 to 0.09)0.57−0.04 (−0.16 to 0.09)0.570%

    25(OH)D indicates 25‐hydroxyvitamin D; ACEi, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; Ca, calcium; CI, confidence interval; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; IPD, individual participant data; PTH, parathyroid hormone; RHI, reactive hyperemia index; SBP, systolic blood pressure.

    Table 11. IPD Subgroup Analyses for Microvascular Function

    SubgroupNo. of StudiesnRandom EffectsFixed EffectsI2
    Treatment Effect (SMD) (95% CI)P ValueP for InteractionTreatment Effect (SMD) (95% CI)P ValueP for Interaction
    SBP >140 mm Hg1210.08 (−0.80 to 0.96)0.860.470.08 (−0.80 to 0.96)0.860.47···
    SBP ≤140 mm Hg31060.44 (0.05–0.82)0.030.44 (0.05–0.82)0.030%
    Baseline 25(OH)D <25 nmol/L3460.05 (−0.54 to 0.64)0.86···0.05 (−0.54 to 0.64)0.86···0%
    Baseline 25(OH)D 25 to 50 nmol/L3560.20 (−0.39 to 0.79)0.510.720.20 (−0.34 to 0.74)0.470.7111%
    Baseline 25(OH)D >50 nmol/L2261.04 (0.14–1.93)0.020.041.04 (0.14–1.93)0.020.040%
    DM1180.41 (−0.54 to 1.35)0.400.910.41 (−0.54 to 1.35)0.400.91···
    No DM31110.35 (−0.03 to 0.73)0.070.35 (−0.03 to 0.73)0.070%
    No ACEi or ARB2700.19 (−0.28 to 0.66)0.44···0.19 (−0.28 to 0.66)0.44···0%
    ACEi, no ARB2290.37 (−0.38 to 1.13)0.330.690.37 (−0.38 to 1.13)0.330.690%
    ACEi or ARB2570.65 (0.11–1.18)0.020.210.65 (0.11–1.18)0.020.210%
    PTH >5.0 pmol/L3700.24 (−0.24 to 0.72)0.330.540.24 (−0.24 to 0.72)0.330.540%
    PTH ≤5.0 pmol/L2540.47 (−0.08 to 1.02)0.090.47 (−0.08 to 1.02)0.090%
    Ca >2.30 mmol/L225−0.45 (−1.25 to 0.35)0.270.10−0.45 (−1.25 to 0.35)0.270.100%
    Ca ≤2.30 mmol/L2370.43 (−0.23 to 1.09)0.200.43 (−0.23 to 1.09)0.200%
    eGFR ≥60 mL/min per 1.73 m22860.22 (−0.21 to 0.64)0.320.460.22 (−0.21 to 0.64)0.320.460%
    eGFR <60 mL/min per 1.73 m22430.50 (−0.11 to 1.11)0.110.50 (−0.11 to 1.11)0.110%
    Total cholesterol ≥4.60 mmol/L3660.05 (−0.47 to 0.57)0.850.320.05 (−0.45 to 0.54)0.850.317%
    Total cholesterol <4.60 mmol/L3610.42 (−0.10 to 0.94)0.110.42 (−0.10 to 0.94)0.110%

    25(OH)D indicates 25‐hydroxyvitamin D; ACEi, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; Ca, calcium; CI, confidence interval; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; IPD, individual participant data; PTH, parathyroid hormone; SBP, systolic blood pressure; SMD, standardized mean difference.

    Table 12. IPD Subgroup Analyses for Aortic SBP

    SubgroupNo. of StudiesnMean Baseline Value (mm Hg) (SD)Random EffectsFixed EffectsI2
    Treatment Effect (95% CI) (mm Hg)P ValueP for InteractionTreatment Effect (95% CI) (mm Hg)P ValueP for Interaction
    SBP >140 mm Hg579141.4 (26.3)0.4 (−8.8 to 9.6)0.930.950.4 (−5.5 to 6.3)0.900.8257%
    SBP ≤140 mm Hg7319115.8 (19.1)0.1 (−3.3 to 3.5)0.96−0.3 (−2.4 to 1.9)0.8251%
    Baseline 25(OH)D <25 nmol/L677114.2 (28.7)−2.7 (−9.8 to 4.4)0.46···−1.7 (−5.9 to 2.5)0.42···50%
    Baseline 25(OH)D 25 to 50 nmol/L7148121.9 (24.7)−2.9 (−6.0 to 0.3)0.080.96−2.9 (−6.0 to 0.3)0.080.650%
    Baseline 25(OH)D >50 nmol/L6171123.0 (20.1)0.7 (−3.8 to 5.1)0.760.431.2 (−1.9 to 4.2)0.450.2724%
    DM445119.3 (35.3)9.4 (−0.2 to 19.1)0.060.177.9 (0.3 to 15.4)0.040.00626%
    No DM7355120.9 (22.0)−1.6 (−13.8 to 10.6)0.80−3.0 (−5.1 to −1.0)0.00497%
    No ACEi or ARB5138111.8 (24.1)−6.1 (−11.8 to −0.4)0.04···−5.8 (−9.1 to −2.1)<0.001···51%
    ACEi, no ARB378130.4 (22.1)−5.4 (−10.8 to 0.1)0.050.87−5.4 (−10.8 to 0.1)0.050.90%
    ACEi or ARB4173132.6 (23.7)−3.1 (−7.1 to 0.8)0.120.40−3.1 (−7.1 to 0.8)0.120.320%
    PTH >5.0 pmol/L6168121.0 (27.5)−1.0 (−4.3 to 2.3)0.550.52−1.0 (−4.3 to 2.3)0.550.670%
    PTH ≤5.0 pmol/L6154123.1 (24.0)2.8 (−8.2 to 13.8)0.62−1.9 (−5.6 to 1.8)0.3185%
    Ca >2.30 mmol/L230118.4 (20.0)0.3 (−5.5 to 6.0)0.930.610.3 (−5.5 to 6.0)0.930.610%
    Ca ≤2.30 mmol/L3120121.9 (25.9)2.0 (−0.9 to 5.0)0.182.0 (−0.9 to 5.0)0.180%
    eGFR ≥60 mL/min per 1.73 m25300120.0 (17.7)−1.3 (−4.8 to 2.2)0.470.70−0.6 (−2.8 to 1.5)0.560.9259%
    eGFR <60 mL/min per 1.73 m2444130.9 (20.2)0.8 (−9.2 to 10.9)0.87−0.9 (−6.3 to 4.6)0.7669%
    Total cholesterol ≥4.60 mmol/L5158122.2 (21.6)−2.3 (−5.8 to 1.2)0.200.68−2.3 (−5.8 to 1.2)0.200.760%
    Total cholesterol <4.60 mmol/L5138122.5 (27.5)−0.7 (−7.5 to 6.1)0.83−1.5 (−5.3 to 2.4)0.4566%

    25(OH)D indicates 25‐hydroxyvitamin D; ACEi, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; Ca, calcium; CI, confidence interval; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; IPD, individual participant data; PTH, parathyroid hormone; SBP, systolic blood pressure.

    Table 13. IPD Subgroup Analyses for Aortic DBP

    SubgroupNo. of StudiesnMean Baseline Value (mm Hg) (SD)Random EffectsFixed EffectsI2
    Treatment Effect (95% CI)P ValueP for InteractionTreatment Effect (95% CI)P ValueP for Interaction
    SBP >140 mm Hg57979.8 (11.1)−0.7 (−3.3 to 2.0)0.620.95−0.7 (−3.3 to 2.0)0.620.950%
    SBP ≤140 mm Hg732075.7 (9.3)−0.6 (−1.9 to 0.6)0.33−0.6 (−1.9 to 0.6)0.330%
    Baseline 25(OH)D <25 nmol/L67775.8 (9.5)1.0 (−2.4 to 4.5)0.56···1.0 (−2.4 to 4.5)0.56···0%
    Baseline 25(OH)D 25 to 50 nmol/L715177.2 (10.0)−0.7 (−2.8 to 1.3)0.480.41−0.7 (−2.8 to 1.3)0.480.410%
    Baseline 25(OH)D >50 nmol/L617176.3 (10.0)−0.5 (−2.0 to 1.0)0.520.43−0.5 (−2.0 to 1.0)0.520.430%
    DM44574.0 (11.5)3.8 (−1.6 to 9.1)0.170.122.8 (0.1 to 5.4)0.040.0268%
    No DM735576.9 (9.6)−0.6 (−1.9 to 0.6)0.29−0.6 (−1.9 to 0.6)0.290%
    No ACEi or ARB615275.4 (9.8)−0.5 (−2.6 to 1.5)0.61···−0.5 (−2.6 to 1.5)0.61···0%
    ACEi, no ARB49279.3 (11.4)0.6 (−1.9 to 3.1)0.620.500.6 (−1.9 to 3.1)0.620.500%
    ACEi or ARB417378.4 (10.5)−0.4 (−2.2 to 1.4)0.690.94−0.4 (−2.2 to 1.4)0.690.940%
    PTH >5.0 pmol/L616977.3 (10.2)0.1 (−2.3 to 2.5)0.930.620.1 (−2.3 to 2.5)0.930.590%
    PTH ≤5.0 pmol/L615477.2 (10.4)−0.7 (−2.7 to 1.3)0.52−0.7 (−2.4 to 1.0)0.3917%
    Ca >2.30 mmol/L23179.3 (9.8)0.4 (−4.7 to 5.5)0.880.940.4 (−4.7 to 5.5)0.880.940%
    Ca ≤2.30 mmol/L312077.0 (10.2)0.2 (−1.6 to 1.9)0.860.2 (−1.6 to 1.9)0.860%
    eGFR ≥60 mL/min per 1.73 m2530077.3 (9.2)−0.6 (−1.8 to 0.7)0.350.07−0.6 (−1.8 to 0.7)0.350.030%
    eGFR <60 mL/min per 1.73 m244575.5 (9.2)4.6 (−0.9 to 10.1)0.102.9 (−0.1 to 5.9)0.0646%
    Total cholesterol ≥4.60 mmol/L515877.3 (9.5)−0.6 (−2.6 to 1.3)0.520.76−0.6 (−2.6 to 1.3)0.520.890%
    Total cholesterol <4.60 mmol/L513874.5 (9.8)0.0 (−3.3 to 3.4)0.98−0.8 (−3.0 to 1.3)0.4549%

    25(OH)D indicates 25‐hydroxyvitamin D; ACEi, angiotensin‐converting enzyme inhibitor; ARB, angiotensin receptor blocker; Ca, calcium; CI, confidence interval; DBP, diastolic blood pressure; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; IPD, individual participant data; PTH, parathyroid hormone; SBP, systolic blood pressure.

    Table 14. Standardizing SDs Pertaining to Trials Combined Using SMD Method

    StudyTrial LevelIPD (Pooled SD)
    PWV (m/s) (Vitamin D, Placebo)Microvascular Function (Units) (Vitamin D, Placebo)PWV (m/s)Microvascular Function (Units)
    Gepner 2012150.9, 1.1···1.0···
    Larsen 2012160.9, 1.4···1.2···
    Marckmann 2012172.6, 2.6···3.1···
    Stricker 201219···16.4, 12.4···13.4
    Hewitt 2013225.3, 4.8·········
    Witham 2013231.4, 1.3···1.2···
    Witham 2013251.2, 1.61.1, 3.31.21.8
    Yiu 2013263.4, 4.1·········
    Dreyer 2014271.1, 0.8727, 3911.0591
    Mose 2014292.1, 2.4···2.3···
    Ryu 2014301.4, 1.8·········
    Garg 2015321.3, 1.2···1.1···
    Pilz 2015331.6, 1.6···1.9···
    Witham 2015352.5, 1.6···0.9···
    Bressendorff 2016380.9, 0.4···1.1···
    Forouhi 2016402.1, 1.7···1.7···
    Hin 2017411.4, 1.4···3.7···

    IPD indicates individual participant data; PWV, pulse wave velocity; SMD, standardized mean difference.

    Discussion

    The present meta‐analysis found little evidence to support the hypothesis that supplementation of vitamin D or use of vitamin D analogues can improve markers of cardiovascular health. Our results were broadly consistent across a range of vascular markers and interventions, and subgroup analyses using IPD did not identify a subgroup that was more likely to benefit from treatment—this remained true even for those participants with the lowest 25(OH)D levels, with high baseline BP, and with higher baseline PTH levels. Random‐effects and fixed‐effects analyses gave very similar results in the majority of analyses. Our results are consistent with our previous work that failed to find a beneficial effect of vitamin D therapy on BP5 and are also in accord with 2 recent, smaller meta‐analyses examining arterial stiffness and endothelial function.42, 43 One further recent meta‐analysis, examining only FMD, showed a slightly greater benefit (treatment effect of vitamin D was 1.27% for FMD),44 perhaps attributable to differences in both study selection and the data used; our analysis had the benefit of access to IPD, which allowed us to verify the accuracy of published data and data used in previous meta‐analyses. The results are also consistent with recent data suggesting no effect of vitamin D supplementation on plasma N‐terminal pro‐B‐type natriuretic peptide levels or echocardiographic indices in older people after 12 months of therapy.45

    Despite the large number of participants included in this analysis, it is not possible to completely refute the possibility that vitamin D or its analogues could still have a modest benefit on vascular health. The markers measured in studies included in this meta‐analysis are subject to changes attributed to differences in environment, diet, smoking, medications, and operator skill; such factors require careful use of protocols to standardize measurement and reduce variability.46, 47 The upper limit of the 95% CIs in our analyses encompasses a 1% improvement in FMD, a 1% improvement in AIx, a 0.3‐m/s improvement in PWV, and a 5.6 mm Hg improvement in aortic systolic BP. A 5 mm Hg reduction in aortic systolic BP would be consistent with significant clinical benefit, and a 1% improvement in FMD would be consistent with an 8% to 13% reduction in cardiovascular event rates.48, 49 A trial published too recently to be included in this systematic review suggested a large improvement in FMD in participants with nondialyzed CKD,50 and it therefore remains possible that individuals with nondialyzed CKD, particularly with baseline low 25(OH)D levels, might benefit, although results from trials enrolling nondialyzed CKD participants that we included in this review showed improvement in FMD in only 1 of 3 trials.11, 31, 34 Similarly, a recently published substudy using monthly high‐dose vitamin D3 showed an improvement in aortic BP and arterial stiffness measures in those with baseline 25(OH)D levels <50 nmol/L; effect sizes were consistent with our IPD analysis findings for this subgroup. No significant improvements were observed in the overall trial group, however.51

    Pooled observational data show that a loge difference in PWV (≈2.7 m/s) corresponds to a 35% to 45% increase in the risk of a cardiovascular event. A 0.3‐m/s improvement in PWV is therefore unlikely to be associated with a clinically important reduction in cardiovascular events.52 Furthermore, it is still possible that agents such as paricalcitol might provide a greater magnitude of benefit to selected markers such as FMD. Paricalcitol is an active analogue of vitamin D (ie, it does not require further hydroxylation before binding to and activating the vitamin D receptor), and it is possible that this pharmacological difference from vitamin D2 or D3 might account for the observed result. It is, however, more likely that this result is attributed to the play of chance given the large number of comparisons contained in our analysis.

    A modest improvement in microvascular function with vitamin D was noted in the trial‐level analysis, although this was of smaller magnitude in the IPD analyses and did not reach significance. The clinical significance of such an improvement in microvascular function is less clear than for changes in macrovascular markers, given that there are few long‐term prognostic studies evaluating microvascular markers. Differences in the physiological control of small and large blood vessels, particularly the role of local metabolic factors in determining microvascular tone, may underpin the difference in response to vitamin D observed here.

    A number of limitations of our analysis require discussion. Despite the large number of participants, power for subgroup analyses was limited by the available data; most trials measured only 1 or 2 vascular outcomes, and some baseline variables were not collected in all trials. Caution is warranted in overinterpreting the results of positive associations in the IPD subgroup analyses; the large number of comparisons poses a risk of type I statistical error. Conversely, our decision to combine results from active treatment arms in trials with more than 1 active treatment arm risks diluting the apparent size of any treatment effect, although the impact of this is likely to be minimal given the small number of trials with more than 1 active treatment arm. For some outcomes, heterogeneity of measurement techniques required use of SMDs. Use of SMD limits the clinical utility of the results, and the heterogeneity of measurements means that translating SMD results to clinically meaningful values is challenging. However, use of SMD does at least allow some inferences about possible effect direction and magnitude to be obtained. Despite an extensive series of hypothesis‐driven subgroup analyses and metaregressions to examine potential causes for heterogeneity, we were unable to identify subgroups of patients more likely to benefit from intervention, and heterogeneity in our IPD subgroup analyses remained high. Some of this heterogeneity may be attributable to the small number of trials in each analysis, but other, unmeasured sources of real difference between trials may still exist.

    Although the risk of bias in most trials was low, only half of the included trials analyzed data by intention to treat, and the inclusion of trials with non‐intention‐to‐treat analyses will tend to inflate observed effect sizes. It is also possible that not all eligible trials were found or included; for some trials, published trial reports had not been produced or could not be obtained from the authors. New trials continue to be published in this area, but most continue to use small numbers of participants and are likely to have limited impact on our conclusions. Not all authors were willing to share their IPD, and not all trial reports contained sufficient information to allow data to be extracted for meta‐analysis. A further limitation is inherent in the populations studied; most populations contained only a minority of participants with 25(OH)D levels below 25 nmol/L, a group that would be thought to be most likely to benefit. Similarly, some trials were conducted in groups without overt vascular disease, where again the possibilities for improving vascular function might have been limited. Most trials were conducted in white populations, which potentially limits the generalizability of the findings. In particular, few blacks were enrolled in the included studies; this group have particularly low 25(OH)D levels when living at high latitudes and may be more likely to show a reduction in BP with vitamin D supplementation.53

    A range of vitamin D doses were used in the included trials; debate continues as to what dose of vitamin D is optimum or indeed what the target level of 25(OH)D should be. If a level of 75 nmol/L is regarded as optimum as has been suggested from observational studies,54 doses at the upper end of the range included in this analysis are required to reach this level.40, 55, 56 Metaregression of vitamin D dose versus treatment effect suggested that higher doses of vitamin D were associated with a slightly greater treatment effect for FMD, but not for PWV or AIx. We found no evidence that daily dosing was more efficacious than intermittent dosing, despite previous work that has suggested that daily dosing provides more‐consistent tissue exposure to the parent compound, facilitating uptake and autocrine activation,57 and evidence that daily dosing may be more efficacious in some conditions (eg, respiratory disease).58 Our results are consistent with our previous analysis that did not find a difference between daily and intermittent dosing on BP.5 A final explanation that requires consideration is that the duration of therapy in most trials may simply have been too short to produce biological effects—particularly those trials intervening for only a few weeks. This explanation is plausible for outcomes such as arterial stiffness if biological effects are mediated by changes in vascular calcification, but seems less so for outcomes such as FMD and reactive hyperemia index, where other interventions are known to alter these parameters within days or weeks. Further evidence against this hypothesis is provided by the metaregression results, which suggest that longer trial duration was associated with a smaller treatment effect for FMD.

    The results of these analyses add to the growing body of evidence suggesting that vitamin D supplementation may not have any beneficial effects on cardiovascular health. The lack of effect on vitamin D supplementation on BP in most studies to date5, 53 and the lack of effect on vascular markers observed in the current analysis suggests that associations between 25(OH)D levels and cardiovascular events observed in observational studies may not be causal. Not all observational studies have been prospective in nature, and the degree of adjustment for confounders has been variable. There are several reasons why assumptions about causality may be incorrect, including reverse causality (where overt or preclinical illness leads to lower 25(OH)D levels through mechanisms such as immobility, obesity, or inflammation59, 60), and confounding by shared risk factors for both cardiovascular disease and low 25(OH)D levels; obesity, inactivity, smoking, and advanced age are all known to be associated with lower 25(OH)D levels, and such confounding is notoriously difficult to fully adjust for. Existing evidence from meta‐analyses of vascular events in osteoporosis trials using vitamin D does not support an effect of vitamin D in lowering cardiovascular event rates,1, 7 with the possible exception of heart failure, and the first of a new wave of large, population‐based vitamin D trials has recently reported, again showing no effect of vitamin D supplementation on cardiovascular event rates.8 Randomized trials of relatively short duration cannot exclude a benefit of vitamin D supplementation over the span of a lifetime; observational designs including Mendelian randomization studies61 may still be the only way to shed light on very long exposures to vitamin D, although even these designs are subject to bias and confounding.

    A number of other large vitamin D trials are due to report over the next 3 to 4 years,62 and most of these include cardiovascular events as key outcomes—outcomes which our analysis did not focus on. Existing evidence does not support the use of vitamin D to reduce cardiovascular risk, and the results of our analysis do not suggest a specific target group that is particularly likely to benefit from vitamin D supplementation, although we found no evidence of a deleterious effect on cardiovascular function. The relative lack of representation of some groups, particularly nonwhite groups, tempers the generalizability of this conclusion. In the absence of a subgroup with clear benefit, and with large trial reports expected soon, further small‐scale trials examining surrogate vascular end points are unlikely to advance this field of research significantly.

    Acknowledgments

    Professor Forouhi acknowledges support from MRC Epidemiology Unit core funding (MC_UU_12015/5).

    Disclosures

    None.

    Footnotes

    *Correspondence to: Miles D. Witham, PhD, School of Medicine, Ninewells Hospital, Dundee DD1 9SY, United Kingdom. E‐mail:

    References

    • 1 Autier P, Boniol M, Pizot C, Mullie P. Vitamin D status and ill health: a systematic review. Lancet Diabetes Endocrinol. 2014; 2:76–89.CrossrefMedlineGoogle Scholar
    • 2 Theodoratou E, Tzoulaki I, Zgaga L, John PA. Vitamin D and multiple health outcomes: umbrella review of systematic reviews and meta‐analyses of observational studies and randomised trials. BMJ. 2014; 348:g2035.CrossrefMedlineGoogle Scholar
    • 3 Jones G, Strugnell SA, DeLuca HF. Current understanding of the molecular actions of vitamin D. Physiol Rev. 1998; 78:1193–1231.CrossrefMedlineGoogle Scholar
    • 4 Beveridge LA, Witham MD. Vitamin D and the cardiovascular system. Osteoporos Int. 2013; 24:2167–2180.CrossrefMedlineGoogle Scholar
    • 5 Beveridge LA, Struthers AD, Khan F, Jorde R, Scragg R, Macdonald HM, Alvarez JA, Boxer RS, Dalbeni A, Gepner AD, Isbel NM, Larsen T, Nagpal J, Petchey WG, Stricker H, Strobel F, Tangpricha V, Toxqui L, Vaquero MP, Wamberg L, Zittermann A, Witham MD. Effect of vitamin D supplementation on blood pressure: a systematic review and meta‐analysis incorporating individual patient data. JAMA Intern Med. 2015; 175:745–754.CrossrefMedlineGoogle Scholar
    • 6 George PS, Pearson ER, Witham MD. Effect of vitamin D supplementation on glycaemic control and insulin resistance: a systematic review and meta‐analysis. Diabet Med. 2012; 29:e142–e150.CrossrefMedlineGoogle Scholar
    • 7 Ford JA, MacLennan GS, Avenell A, Bolland M, Grey A, Witham M; RECORD Trial Group . Cardiovascular disease and vitamin D supplementation: trial analysis, systematic review, and meta‐analysis. Am J Clin Nutr. 2014; 100:746–755.CrossrefMedlineGoogle Scholar
    • 8 Scragg R, Stewart AW, Waayer D, Lawes CMM, Toop L, Sluyter J, Murphy J, Khaw KT, Camargo CA. Effect of monthly high‐dose vitamin D supplementation on cardiovascular disease in the vitamin D assessment study: a randomized clinical trial. JAMA Cardiol. 2017; 2:608–616.CrossrefMedlineGoogle Scholar
    • 9 Verhagen AP, de Vet HC, de Bie RA, Kessels AG, Boers M, Bouter LM, Knipschild PG. The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol. 1998; 51:1235–1241.CrossrefMedlineGoogle Scholar
    • 10 Riley RD, Kauser I, Bland M, Thijs L, Staessen JA, Wang J, Gueyffier F, Deeks JJ. Meta‐analysis of randomised trials with a continuous outcome according to baseline imbalance and availability of individual participant data. Stat Med. 2013; 32:2747–2766.CrossrefMedlineGoogle Scholar
    • 11 Alborzi P, Patel NA, Peterson C, Bills JE, Bekele DM, Bunaye Z, Light RP, Agarwal R. Paricalcitol reduces albuminuria and inflammation in chronic kidney disease: a randomized double‐blind pilot trial. Hypertension. 2008; 52:249–255.LinkGoogle Scholar
    • 12 Sugden JA, Davies JI, Witham MD, Morris AD, Struthers AD. Vitamin D improves endothelial function in patients with type 2 diabetes mellitus and low vitamin D levels. Diabet Med. 2008; 25:320–325.CrossrefMedlineGoogle Scholar
    • 13 Witham MD, Dove FJ, Dryburgh M, Sugden JA, Morris AD, Struthers AD. The effect of different doses of vitamin D(3) on markers of vascular health in patients with type 2 diabetes: a randomised controlled trial. Diabetologia. 2010; 53:2112–2119.CrossrefMedlineGoogle Scholar
    • 14 Harris RA, Pedersen‐White J, Guo DH, Stallmann‐Jorgensen IS, Keeton D, Huang Y, Shah Y, Zhu H, Dong Y. Vitamin D3 supplementation for 16 weeks improves flow‐mediated dilation in overweight African‐American adults. Am J Hypertens. 2011; 24:557–562.CrossrefMedlineGoogle Scholar
    • 15 Gepner AD, Ramamurthy R, Krueger DC, Korcarz CE, Binkley N, Stein JH. A prospective randomized controlled trial of the effects of vitamin D supplementation on cardiovascular disease risk. PLoS One. 2012; 7:e36617.CrossrefMedlineGoogle Scholar
    • 16 Larsen T, Mose FH, Bech JN, Hansen AB, Pedersen EB. Effect of cholecalciferol supplementation during winter months in patients with hypertension: a randomized, placebo‐controlled trial. Am J Hypertens. 2012; 25:1215–1222.CrossrefMedlineGoogle Scholar
    • 17 Marckmann P, Agerskov H, Thineshkumar S, Bladbjerg EM, Sidelmann JJ, Jespersen J, Nybo M, Rasmussen LM, Hansen D, Scholze A. Randomized controlled trial of cholecalciferol supplementation in chronic kidney disease patients with hypovitaminosis D. Nephrol Dial Transplant. 2012; 27:3523–3531.CrossrefMedlineGoogle Scholar
    • 18 Sokol SI, Srinivas V, Crandall JP, Kim M, Tellides G, Lebastchi AH, Yu Y, Gupta AK, Alderman MH. The effects of vitamin D repletion on endothelial function and inflammation in patients with coronary artery disease. Vasc Med. 2012; 17:394–404.CrossrefMedlineGoogle Scholar
    • 19 Stricker H, Tosi BF, Guidicelli‐Nicolosi S, Limoni C, Colucci G. Effect of a single, oral, high‐dose vitamin D supplementation on endothelial function in patients with peripheral arterial disease: a randomised controlled pilot study. Eur J Vasc Endovasc Surg. 2012; 44:307–312.CrossrefMedlineGoogle Scholar
    • 20 Witham MD, Dove FJ, Sugden JA, Doney AS, Struthers AD. The effect of vitamin D replacement on markers of vascular health in stroke patients—a randomised controlled trial. Nutr Metab Cardiovasc Dis. 2012; 22:864–870.CrossrefMedlineGoogle Scholar
    • 21 Breslavsky A, Frand J, Matas Z, Boaz M, Barnea Z, Shargorodsky M. Effect of high doses of vitamin D on arterial properties, adiponectin, leptin and glucose homeostasis in type 2 diabetic patients. Clin Nutr. 2013; 32:970–975.CrossrefMedlineGoogle Scholar
    • 22 Hewitt NA, O'Connor AA, O'Shaughnessy DV, Elder GJ. Effects of cholecalciferol on functional, biochemical, vascular, and quality of life outcomes in hemodialysis patients. Clin J Am Soc Nephrol. 2013; 8:1143–1149.CrossrefMedlineGoogle Scholar
    • 23 Witham MD, Price RJ, Struthers AD, Donnan PT, Messow CM, Ford I, McMurdo ME. Cholecalciferol treatment to reduce blood pressure in older patients with isolated systolic hypertension: the VitDISH randomized controlled trial. JAMA Intern Med. 2013; 173:1672–1679.MedlineGoogle Scholar
    • 24 Witham MD, Dove FJ, Khan F, Lang CC, Belch JJ, Struthers AD. Effects of vitamin D supplementation on markers of vascular function after myocardial infarction—a randomised controlled trial. Int J Cardiol. 2013; 167:745–749.CrossrefMedlineGoogle Scholar
    • 25 Witham MD, Adams F, Kabir G, Kennedy G, Belch JJ, Khan F. Effect of short‐term vitamin D supplementation on markers of vascular health in South Asian women living in the UK—a randomised controlled trial. Atherosclerosis. 2013; 230:293–299.CrossrefMedlineGoogle Scholar
    • 26 Yiu YF, Yiu KH, Siu CW, Chan YH, Li SW, Wong LY, Lee SW, Tam S, Wong EW, Lau CP, Cheung BM, Tse HF. Randomized controlled trial of vitamin D supplement on endothelial function in patients with type 2 diabetes. Atherosclerosis. 2013; 227:140–146.CrossrefMedlineGoogle Scholar
    • 27 Dreyer G, Tucker AT, Harwood SM, Pearse RM, Raftery MJ, Yaqoob MM. Ergocalciferol and microcirculatory function in chronic kidney disease and concomitant vitamin d deficiency: an exploratory, double blind, randomised controlled trial. PLoS One. 2014; 9:e99461.CrossrefMedlineGoogle Scholar
    • 28 Martins D, Meng YX, Tareen N, Artaza J, Lee JE, Farodolu C, Gibbons G, Norris K. The effect of short term vitamin D supplementation on the inflammatory and oxidative mediators of arterial stiffness. Health (Irvine Calif). 2014; 6:1503–1511.MedlineGoogle Scholar
    • 29 Mose FH, Vase H, Larsen T, Kancir AS, Kosierkiewic R, Jonczy B, Hansen AB, Oczachowska‐Kulik AE, Thomsen IM, Bech JN, Pedersen EB. Cardiovascular effects of cholecalciferol treatment in dialysis patients—a randomized controlled trial. BMC Nephrol. 2014; 15:50.CrossrefMedlineGoogle Scholar
    • 30 Ryu OH, Chung W, Lee S, Hong KS, Choi MG, Yoo HJ. The effect of high‐dose vitamin D supplementation on insulin resistance and arterial stiffness in patients with type 2 diabetes. Korean J Intern Med. 2014; 29:620–629.CrossrefMedlineGoogle Scholar
    • 31 Zoccali C, Curatola G, Panuccio V, Tripepi R, Pizzini P, Versace M, Bolignano D, Cutrupi S, Politi R, Tripepi G, Ghiadoni L, Thadhani R, Mallamaci F. Paricalcitol and endothelial function in chronic kidney disease trial. Hypertension. 2014; 64:1005–1011.LinkGoogle Scholar
    • 32 Garg G, Kachhawa G, Ramot R, Khadgawat R, Tandon N, Sreenivas V, Kriplani A, Gupta N. Effect of vitamin D supplementation on insulin kinetics and cardiovascular risk factors in polycystic ovarian syndrome: a pilot study. Endocr Connect. 2015; 4:108–116.CrossrefMedlineGoogle Scholar
    • 33 Pilz S, Gaksch M, Kienreich K, Grübler M, Verheyen N, Fahrleitner‐Pammer A, Treiber G, Drechsler C, Ó Hartaigh B, Obermayer‐Pietsch B, Schwetz V, Aberer F, Mader J, Scharnagl H, Meinitzer A, Lerchbaum E, Dekker JM, Zittermann A, März W, Tomaschitz A. Effects of vitamin D on blood pressure and cardiovascular risk factors: a randomized controlled trial. Hypertension. 2015; 65:1195–1201.LinkGoogle Scholar
    • 34 Thethi TK, Bajwa MA, Ghanim H, Jo C, Weir M, Goldfine AB, Umpierrez G, Desouza C, Dandona P, Fang‐Hollingsworth Y, Raghavan V, Fonseca VA. Effect of paricalcitol on endothelial function and inflammation in type 2 diabetes and chronic kidney disease. J Diabetes Complications. 2015; 29:433–437.CrossrefMedlineGoogle Scholar
    • 35 Witham MD, Adams F, McSwiggan S, Kennedy G, Kabir G, Belch JJ, Khan F. Effect of intermittent vitamin D3 on vascular function and symptoms in chronic fatigue syndrome—a randomised controlled trial. Nutr Metab Cardiovasc Dis. 2015; 25:287–294.CrossrefMedlineGoogle Scholar
    • 36 Barchetta I, Del Ben M, Angelico F, Di Martino M, Fraioli A, La Torre G, Saulle R, Perri L, Morini S, Tiberti C, Bertoccini L, Cimini FA, Panimolle F, Catalano C, Baroni MG, Cavallo MG. No effects of oral vitamin D supplementation on non‐alcoholic fatty liver disease in patients with type 2 diabetes: a randomized, double‐blind, placebo‐controlled trial. BMC Med. 2016; 14:92.CrossrefMedlineGoogle Scholar
    • 37 Borgi L, McMullan C, Wohlhueter A, Curhan GC, Fisher ND, Forman JP. Effect of vitamin D on endothelial function: a randomized, double‐blind, placebo‐controlled trial. Am J Hypertens. 2017; 30:124–129.CrossrefMedlineGoogle Scholar
    • 38 Bressendorff I, Brandi L, Schou M, Nygaard B, Frandsen NE, Rasmussen K, Ødum L, Østergaard OV, Hansen D. The effect of high dose cholecalciferol on arterial stiffness and peripheral and central blood pressure in healthy humans: a randomized controlled trial. PLoS One. 2016; 11:e0160905.CrossrefMedlineGoogle Scholar
    • 39 Dalan R, Liew H, Assam PN, Chan ES, Siddiqui FJ, Tan AW, Chew DE, Boehm BO, Leow MK. A randomised controlled trial evaluating the impact of targeted vitamin D supplementation on endothelial function in type 2 diabetes mellitus: the DIMENSION trial. Diab Vasc Dis Res. 2016; 13:192–200.CrossrefMedlineGoogle Scholar
    • 40 Forouhi NG, Menon RK, Sharp SJ, Mannan N, Timms PM, Martineau AR, Rickard AP, Boucher BJ, Chowdhury TA, Griffiths CJ, Greenwald SE, Griffin SJ, Hitman GA. Effects of vitamin D2 or D3 supplementation on glycaemic control and cardiometabolic risk among people at risk of type 2 diabetes: results of a randomized double‐blind placebo‐controlled trial. Diabetes Obes Metab. 2016; 18:392–400.CrossrefMedlineGoogle Scholar
    • 41 Hin H, Tomson J, Newman C, Kurien R, Lay M, Cox J, Sayer J, Hill M, Emberson J, Armitage J, Clarke R. Optimum dose of vitamin D for disease prevention in older people: BEST‐D trial of vitamin D in primary care. Osteoporos Int. 2017; 28:841–851.CrossrefMedlineGoogle Scholar
    • 42 Hussin AM, Ashor AW, Schoenmakers I, Hill T, Mathers JC, Siervo M. Effects of vitamin D supplementation on endothelial function: a systematic review and meta‐analysis of randomised clinical trials. Eur J Nutr. 2017; 56:1095–1104.CrossrefMedlineGoogle Scholar
    • 43 Upala S, Sanguankeo A, Congrete S, Jaruvongvanich V. Effect of cholecalciferol supplementation on arterial stiffness: a systematic review and meta‐analysis. Scand Cardiovasc J. 2016; 50:230–235.CrossrefMedlineGoogle Scholar
    • 44 Mazidi M, Karimi E, Rezaie P, Vatanparast H. The impact of vitamin D supplement intake on vascular endothelial function; a systematic review and meta‐analysis of randomized controlled trials. Food Nutr Res. 2017; 61:1273574.CrossrefMedlineGoogle Scholar
    • 45 Tomson J, Hin H, Emberson J, Kurien R, Lay M, Cox J, Hill M, Arnold L, Leeson P, Armitage J, Clarke R. Effects of vitamin D on blood pressure, arterial stiffness and cardiac function in older people after 1 year: BEST‐D trial. J Am Heart Assoc. 2017; 6:e005707. DOI: 10.1161/JAHA.117.005707.LinkGoogle Scholar
    • 46 Onkelinx S, Cornelissen V, Goetschalckx K, Thomaes T, Verhamme P, Vanhees L. Reproducibility of different methods to measure the endothelial function. Vasc Med. 2012; 17:79–84.CrossrefMedlineGoogle Scholar
    • 47 Ghiadoni L, Faita F, Salvetti M, Cordiano C, Biggi A, Puato M, Di Monaco A, De Siati L, Volpe M, Ambrosio G, Gemignani V, Muiesan ML, Taddei S, Lanza GA, Cosentino F. Assessment of flow‐mediated dilation reproducibility: a nationwide multicenter study. J Hypertens. 2012; 30:1399–1405.CrossrefMedlineGoogle Scholar
    • 48 Inaba Y, Chen JA, Bergmann SR. Prediction of future cardiovascular outcomes by flow‐mediated vasodilatation of brachial artery: a meta‐analysis. Int J Cardiovasc Imaging. 2010; 26:631–640.CrossrefMedlineGoogle Scholar
    • 49 Ras RT, Streppel MT, Draijer R, Zock PL. Flow‐mediated dilation and cardiovascular risk prediction: a systematic review with meta‐analysis. Int J Cardiol. 2013; 168:344–351.CrossrefMedlineGoogle Scholar
    • 50 Kumar V, Yadav AK, Lal A, Kumar V, Singhal M, Billot L, Gupta KL, Banerjee D, Jha V. A randomized trial of vitamin D supplementation on vascular function in CKD. J Am Soc Nephrol. 2017; 28:3100–3108.CrossrefMedlineGoogle Scholar
    • 51 Sluyter JD, Camargo CA, Stewart AW, Waayer D, Lawes CMM, Toop L, Khaw KT, Thom SAM, Hametner B, Wassertheurer S, Parker KH, Hughes AD, Scragg R. Effect of monthly, high‐dose, long‐term vitamin D supplementation on central blood pressure parameters: a randomized controlled trial substudy. J Am Heart Assoc. 2017; 6:e006802. DOI: 10.1161/JAHA.117.006802.LinkGoogle Scholar
    • 52 Ben‐Shlomo Y, Spears M, Boustred C, May M, Anderson SG, Benjamin EJ, Boutouyrie P, Cameron J, Chen CH, Cruickshank JK, Hwang SJ, Lakatta EG, Laurent S, Maldonado J, Mitchell GF, Najjar SS, Newman AB, Ohishi M, Pannier B, Pereira T, Vasan RS, Shokawa T, Sutton‐Tyrell K, Verbeke F, Wang KL, Webb DJ, Willum Hansen T, Zoungas S, McEniery CM, Cockcroft JR, Wilkinson IB. Aortic pulse wave velocity improves cardiovascular event prediction: an individual participant meta‐analysis of prospective observational data from 17,635 subjects. J Am Coll Cardiol. 2014; 63:636–646.CrossrefMedlineGoogle Scholar
    • 53 Forman JP, Scott JB, Ng K, Drake BF, Suarez EG, Hayden DL, Bennett GG, Chandler PD, Hollis BW, Emmons KM, Giovannucci EL, Fuchs CS, Chan AT. Effect of vitamin D supplementation on blood pressure in blacks. Hypertension. 2013; 61:779–785.LinkGoogle Scholar
    • 54 Bischoff‐Ferrari HA, Giovannucci E, Willett WC, Dietrich T, Dawson‐Hughes B. Estimation of optimal serum concentrations of 25‐hydroxyvitamin D for multiple health outcomes. Am J Clin Nutr. 2006; 84:18–28.CrossrefMedlineGoogle Scholar
    • 55 Gallagher JC, Sai A, Templin T, Smith L. Dose response to vitamin D supplementation in postmenopausal women: a randomized trial. Ann Intern Med. 2012; 156:425–437.CrossrefMedlineGoogle Scholar
    • 56 Aloia JF, Patel M, Dimaano R, Li‐Ng M, Talwarr SA, Mikhail M, Pollack S, Yeh JK. Vitamin D intake to attain a desired serum 25‐hydroxyvitamin D concentration. Am J Clin Nutr. 2008; 87:1952–1958.CrossrefMedlineGoogle Scholar
    • 57 Hollis BW, Wagner CL. The role of the parent compound vitamin D with respect to metabolism and function: why clinical dose intervals can affect clinical outcomes. J Clin Endocrinol Metab. 2013; 98:4619–4628.CrossrefMedlineGoogle Scholar
    • 58 Martineau AR, Jolliffe DA, Hooper RL, Greenberg L, Aloia JF, Bergman P, Dubnov‐Raz G, Esposito S, Ganmaa D, Ginde AA, Goodall EC, Grant CC, Griffiths CJ, Janssens W, Laaksi I, Manaseki‐Holland S, Mauger D, Murdoch DR, Neale R, Rees JR, Simpson S, Stelmach I, Kumar GT, Urashima M, Camargo CA. Vitamin D supplementation to prevent acute respiratory tract infections: systematic review and meta‐analysis of individual participant data. BMJ. 2017; 356:i6583.CrossrefMedlineGoogle Scholar
    • 59 Reid D, Toole BJ, Knox S, Talwar D, Harten J, O'Reilly DS, Blackwell S, Kinsella J, McMillan DC, Wallace AM. The relation between acute changes in the systemic inflammatory response and plasma 25‐hydroxyvitamin D concentrations after elective knee arthroplasty. Am J Clin Nutr. 2011; 93:1006–1011.CrossrefMedlineGoogle Scholar
    • 60 Wortsman J, Matsuoka LY, Chen TC, Lu Z, Holick MF. Decreased bioavailability of vitamin D in obesity. Am J Clin Nutr. 2000; 72:690–693.CrossrefMedlineGoogle Scholar
    • 61 Afzal S, Nordestgaard BG. Vitamin D, hypertension, and ischemic stroke in 116 655 individuals from the general population: a genetic study. Hypertension. 2017 Jul 31. pii: HYPERTENSIONAHA.117.09411. DOI: 10.1161/hypertensionaha.117.09411. [Epub ahead of print].LinkGoogle Scholar
    • 62 Kupferschmidt K. Uncertain verdict as vitamin D goes on trial. Science. 2012; 337:1476–1478.CrossrefMedlineGoogle Scholar

    eLetters(0)

    eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

    Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.