Vitamin D to Prevent Respiratory Tract Infections: Is the Evidence Dazzling or Disappointing?


Written By

Reviewed By


Martineau AR, Jolliffe DA, Hooper RL, et al. Vitamin D supplementation to prevent acute respiratory tract infections: systematic review and meta-analysis of individual participant data. BMJ 2017;356:i6583.

Among Medicare recipients, from 2000 to 2010, there was a whopping 83-fold increase (that’s an 8300% increase!) in the number of blood tests performed to determine their vitamin D status.1 What could possibly cause such a dramatic increase?  While the aging U.S. population is at a higher risk for fractures and falls and may benefit from vitamin D supplementation, it is unlikely that this is the main driver of this phenomenon.  A more likely explanation is the fascination among clinicians and patients about the potential uses of vitamin D beyond bone health.2-4


In addition to playing a potential role in preventing multiple sclerosis, type 1 diabetes, and asthma, vitamin D might protect against acute respiratory tract infections (ARTIs).  When vitamin D is converted into calcitriol, an anti-microbial peptide, cathelicidin, is produced.  This peptide has been shown to kill bacteria, including Mycobacterium tuberculosis.2 Observational studies and systematic reviews have shown an association between low vitamin D concentrations and ARTIs.5-6 An association is one thing, but can vitamin D supplementation reduce the risk of ARTIs?  That’s what a recent systematic review and meta-analysis attempted to determine.7


Martineau and colleagues used individual participant data (IPD) from published and unpublished studies to assess the effect of vitamin D supplementation on the risk of ARTIs and to identify factors driving the observed effect.7  Studies included in the meta-analysis were required to 1) be randomized, double blind, placebo-controlled trials evaluating the use of vitamin D3 or D2 supplementation, 2) have ARTIs as a pre-specified efficacy outcome, and 3) prospectively collect data on ARTIs.  Studies that reported long-term follow-up from a primary randomized controlled trials were excluded.  Two to three investigators (depending on the task; 5 unique individuals in total) determined study eligibility, IPD integrity, and independently assessed risk of bias.7 


The primary outcome analyzed was incidence of ARTIs, which was a composite of upper RTIs, lower RTIs, and ARTIs of unclassified location.  Secondary outcomes included incidence of upper RTIs, incidence of lower RTIs, incidence of emergency room visits/hospitalizations for ARTIs, use of antimicrobials to treat ARTIs, absence from work/school due to ARTIs, incidence of adverse events, and mortality.  The investigators also planned a number of pre-specified sub-group analyses intended to identify factors that influence the effects of vitamin D.  The subgroups were classified by baseline vitamin D status, vitamin D dosing regimen, dose size, age, BMI, previous influenza vaccination as well as presence of asthma or COPD.  The authors performed both a one-step IPD meta-analysis (analyzed data from all studies simultaneously) and a two-step IPD meta-analysis (analyzed data from all studies independently and then together) for each outcome.  The analyses were adjusted for age, sex, and study duration.7 


A total of 532 studies published through December 2015 were assessed for inclusion in the meta-analysis.  Twenty-five met eligibility criteria.  IPD for all 25 studies were obtained and included in the meta-analysis.  These studies enrolled a total of 11,321 participants and primary outcome data was available for 10,933 (96.6%) of the participants.  The age of participants ranged from birth to 95 years, with about half being less than 1 year of age.  Baseline serum 25-hydroxyvitamin D concentration was reported in 19 of the trials, representing 4172 participants.  Of participants with baseline levels reported, the majority had levels ≥10 ng/mL (86.8% in the control, 87.3% in the intervention).  Fifteen of the studies used weekly or daily doses of vitamin D in the intervention group, while 7 used bolus doses, and 3 used a combination of bolus and daily doses.  Incidence of ARTIs was the primary/co-primary outcome in 14 of the studies and a secondary outcome in the remaining studies.7


The investigators found a statistically significant reduction in the proportion of patients experiencing at least one ARTI (NNT=33) and the rate of ARTI for patients receiving vitamin D supplementation in both the one-step and two-step analyses.  See Table 1. There was no significant difference seen in the time to first ARTI.  There was also no statistically significant effect of vitamin D supplementation on any of the secondary outcomes.  The sub-group analyses revealed that vitamin D had a relatively strong protective effect among those patients with low baseline vitamin D concentrations, defined as <10 ng/mL (adjusted odds ratio 0.58, 95% CI 0.40 to 0.82, P = 0.002, NNT = 7).  In addition, patients on daily or weekly vitamin D regimens without bolus doses (adjusted odds ratio 0.81, 95% CI 0.72 to 0.91, P < 0.001, NNT = 20) also appeared to benefit from vitamin D supplementation.  None of the other sub-group analyses were statistically significant.  No differences in adverse events were observed.7


Table 1. Effects of Vitamin D on Acute Respiratory Tract Infections7


One-step Analysis

Two-step Analysis

Proportion with ≥1 ARTI

OR = 0.88,

95% CI 0.81 to 0.96

P=0.003, NNT=33

OR = 0.8

95% CI 0.69 to 0.93


Rate of ARTI

IRR = 0.96

95% CI 0.92 to 0.997


IRR = 0.91

95% CI 0.84 to 0.98


Time to first ARTI

HR = 0.95

95%CI 0.89 to 1.01


HR = 0.92

95% CI 0.85 to 1


OR = odds ration; CI = confidence interval; IRR = incidence rate ratio; HR = hazard ratio


The investigators went to great lengths to ensure the data were of high quality by obtaining and assessing integrity of IPD, determining risk of bias in individual studies, assessing quality across studies, and performing sensitivity analyses.  However, the authors found significant heterogeneity (I2 = 53.3%, P=0.001) among the included studies.  Of particular importance is the variability seen in the patient populations, dosing regimens used, duration of treatment, and definitions of ARTI.  For example, the participants in these studies were from 15 countries and four continents.  This is important since patient characteristics, such as sun exposure based on latitude, skin pigmentation, and nutritional status, can affect vitamin D concentrations.  Furthermore, about 60% of the participants were children under the age of 16 years.  While the authors adjusted for age, extrapolating these results to adults is questionable.  Another limitation is that small negative studies may not have been published due to publication bias.  Also, given the amount of missing data from the identified studies, it is likely there was a lack of power to adequately analyze secondary outcomes and compare sub-groups. 


It is important to note the definition of low vitamin D concentration in this study.  In the United States, most experts believe that either >20 ng/mL (based on the IOM’s 2010 report) or >30 ng/mL (based on the Endocrine Society’s 2011 clinical practice guideline) represents adequate vitamin D stores.2,3  However, the authors used a cut off of 10 ng/mL when examining the effect of baseline vitamin D concentrations on the primary outcome based on the UK Department of Health guidelines.7  As might be expected, the patients with the lowest baseline vitamin D concentrations (< 10 ng/mL) received the most benefit from vitamin D supplementation.  Vitamin D supplementation in patients with less severe or no vitamin D deficiency (concentrations of 10-20 ng/mL, 20-30 ng/mL, or ≥30 ng/mL) did not show a significant benefit.  However, this may be due to a lack of power given baseline vitamin D concentrations were missing for 63% of the patients.


While this meta-analysis supports the role of vitamin D in preventing ARTIs, the evidence probably shouldn’t alter clinical practice.  The patients benefitting most from the vitamin D supplementation were those patients with very low baseline vitamin D concentrations, but these are patients who rightfully should receive supplementation in order to prevent osteoporosis, osteomalacia, and rickets.3  The results also support supplementation with either a daily or weekly regimen rather than large bolus doses.  However, it is still unclear what dose of vitamin D must be given to prevent ARTIs.  What do you think?  Given that vitamin D is relatively safe, would you recommend vitamin D supplementation to patients at high risk for low vitamin D concentrations and/or ARTIs? Do we need to measure vitamin D concentrations before recommending supplementation?  Given that infants and children are disproportionately impacted by ARTIs, should vitamin D supplementation be routinely recommended in this population?