Effect of Vitamin D₃ Supplementation on Upper Respiratory Tract Infections in Healthy Adults:  The VIDARIS Randomized Controlled Trial FREE

The role of vitamin D in infectious diseases and the potential impact of vitamin D supplementation as a preventive measure for infectious diseases are unclear.¹ Vitamin D has a role in both innate and adaptive immune responses. Of particular interest is the induction by vitamin D of cathelicidins, a group of antimicrobial peptides produced by neutrophils, macrophages, and epithelial cells. Epidemiological studies show an association between low vitamin D levels and a variety of respiratory tract infections. For example, vitamin D insufficiency is associated with increased risk of developing tuberculosis.¹

The association of vitamin D insufficiency and susceptibility to viral respiratory tract infections is also unclear. Several observational studies report an inverse association between serum 25-hydroxyvitamin D (25-OHD) levels and incidence of upper respiratory tract infections (URTIs).^2- 4 Observational studies are potentially limited by reverse causality and residual confounding, and randomized controlled trials are necessary to properly assess whether there is an effect of vitamin D on risk of infection.⁵ To date, however, the few clinical trials of vitamin D supplementation in adults to prevent URTIs have had conflicting results^6- 8 and were limited in their ability to test the effect of vitamin D supplementation on URTI occurrence. For example, some were of short duration, used a relatively low dose of vitamin D, or were not designed to address URTI outcomes.

The current trial was established to determine the effect of vitamin D supplementation on incidence and severity of URTIs in healthy adults.

The study was a randomized, double-blind, placebo-controlled trial in Christchurch, New Zealand (latitude 43°S). Participants were staff or students of the Canterbury District Health Board, the regional publicly funded health care organization, or the University of Otago, Christchurch, who were aged 18 years or older, were able to give written informed consent, and who anticipated that they would be a resident of the Christchurch region for the study period. Following an advertising campaign, we screened and enrolled volunteers during February through April 2010.

Exclusion criteria were (1) use of vitamin D supplements other than as part of a daily multivitamin preparation (in which the daily intake was ≤400 IU); (2) use of immunosuppressants or medications that interfere with vitamin D metabolism (eg, thiazide diuretics, phenytoin, carbamazepine, primidone, phenobarbital, doses of prednisone >10 mg/d, methotrexate, azathioprine, cyclosporin); (3) history of hypercalcemia or nephrolithiasis; (4) sarcoidosis; (5) kidney disorders requiring dialysis or polycystic kidney disease; (6) cirrhosis; (7) current malignancy diagnosis in which the cancer was aggressive and prognosis was poor; (8) baseline plasma calcium (corrected for plasma albumin concentration) greater than 10.4 mg/dL or less than 8.4 mg/dL; (9) enrollment or planned enrollment in other research that would conflict with full participation in the study or confound the observation or interpretation of the study findings (eg, in which 25-OHD levels were tested and results known by the participant; in which the participant was required to take conflicting medications; any investigations of viruses and antiviral treatments); and (10) pregnancy or planned pregnancy during the study period.

The study was approved by the Upper South B Regional Ethics Committee and all participants provided written informed consent.

Participants were assigned using computer-generated randomization to receive either vitamin D₃ or placebo. Participants randomized to the active treatment received oral vitamin D₃, 200 000 IU, then 200 000 IU 1 month later, then 100 000 IU monthly thereafter for a total of 18 months. This dosing regimen was chosen with the aim of achieving mean 25-OHD levels of about 40 ng/mL,^9- 10 which, at the time of the study, was in the 25-OHD range with the lowest risk of disease in observational studies. (To convert 25-OHD to nmol/L, multiply by 2.496.) Monthly dosing also helped ensure adherence to treatment. Those randomized to placebo received matching inactive tablets administered in a dosing regimen identical to the active treatment.

Both vitamin D₃ and placebo tablets were sourced from Tishcon Corp and were identical in appearance. The randomization process and bottling of tablets were performed in Auckland, New Zealand, under the supervision of the study biostatistician (A.W.S.) to ensure that those running the study, including outcome assessors and those administering the intervention, were blinded to allocation. Research staff directly administered all treatments to participants during monthly visits throughout the study period.

Information on baseline characteristics was obtained by interviewer-administered questionnaire at the screening visit and included data on demographics, occupation, medical history, smoking, current medications, and supplement use.

After randomization, follow-up visits occurred monthly until the end of the study. During these visits, dedicated study staff met the participants in person, administered their monthly dose of vitamin D or placebo, and administered a brief questionnaire. This questionnaire asked about episodes of respiratory tract illness during the preceding month that had not already been reported to study personnel and also noted any changes in medications or supplement use and adverse events.

In addition to the monthly visits, participants were asked to contact study staff whenever they experienced a URTI, defined as the sudden onset of 1 or more of runny nose, nasal stuffiness, sore throat, or cough that the participant did not attribute to allergy. A research staff member then visited the participant to complete a symptom survey (Wisconsin Upper Respiratory Symptom Survey 24 [WURSS-24])¹¹ and collect a nasopharyngeal swab sample. The WURSS-24 is an instrument for assessing the severity and functional impact of URTIs; it contains 24 items, each based on 7-point Likert-type severity scales, and comprises the same items as an earlier version (WURSS-21^12- 15) plus items for headache, body ache, and fever. The WURSS-24 was completed daily over subsequent days through telephone interview by research staff or by self-reported questionnaire until the individual had reported 2 consecutive days of “not sick,” or until 14 days after the onset of the URTI. For those whose URTI symptoms ended before 14 days, the duration of the episode was recorded as ending the day before 2 consecutive “not sick” days. The number of days of missed work as a result of the URTI was also recorded.

Nasopharyngeal swab samples were tested for the following respiratory viruses by real-time polymerase chain reaction (Fast Track Diagnostics): adenovirus; bocavirus; coronaviruses 229E, OC43, NL63, and HKU; enterovirus; influenza A and B viruses; human metapneumoviruses A and B; human rhinovirus; parainfluenza viruses 1 to 4; parechovirus; and respiratory syncytial viruses A and B.

The primary end point was number of URTI episodes. Secondary end points were number of days of missed work as a result of URTI episodes, duration of URTI episodes, severity of URTI episodes, and detection of respiratory viruses in nasopharyngeal samples.

Plasma calcium and serum 25-OHD levels were measured at baseline and at 2, 6, 12, and 18 months after enrollment. Plasma calcium was measured in real time to monitor safety (Abbott c8000 analyzer, Abbott Laboratories). Serum samples for 25-OHD measurement were stored at −112°F (−80°C) until each participant completed the study. All samples from each participant were batched and analyzed within the same run. The 25-OHD levels were measured by liquid chromatography–tandem mass spectrometry (ABSciex API 4000) and the results were unknown to the study team until after completion of the study.

On the assumption that participants would have an average of 1.6 URTIs per year^16- 17 and follow-up of 18 months and that the intervention would need to reduce the mean number of infections by 20% to have clinical relevance, we calculated that a sample of 240 participants would be required to observe this effect with a power of 80% at the .05 level of significance. This number was increased to 320 to compensate for the potential influence of influenza vaccination and loss to follow-up.

The numbers of URTI events for each participant were summed and then compared between the treatment and placebo groups using a negative binomial model that included a dispersion parameter. This model was then extended to include variables that might indicate participant subgroups with different patterns of effect. These binary variables included winter season (May-September), positive nasopharyngeal swab, and season-adjusted baseline serum 25-OHD value less than 20 ng/mL (50 nmol/L). Analysis of the number of days of missed work also used a negative binomial model but because there were multiple events for many participants, a generalized estimating equation model with an exchangeable correlation matrix was used.

A comparison of the sum of the WURSS-24 scores in the first 7 days of the URTI event was made using a general linear mixed model and modeling the participants as random effects. For participants who had incomplete data on the WURSS-24, missing observations were estimated using multiple imputation (5 imputations) using the Markov chain Monte Carlo method on natural log scores (with a constant of 1 added).¹⁸

Duration of URTI events was assessed using the Cox proportional hazard model with multiple events per participant being treated as clustered events and using robust sandwich covariance matrix estimates.

The α level/hypothesis testing was 2-sided with statistical significance set at P <.05. The statistical software used was SAS, version 9.2 (SAS Institute Inc).

Figure 1 shows the study flow diagram. Of 351 potential participants screened, 322 were eligible for inclusion and were randomly assigned to a treatment group. Two hundred ninety-four participants (91%) completed the study treatment and follow-up, 18 (6%) withdrew from the study altogether, and 10 (3%) withdrew from treatment but completed the 18-month follow-up. All 322 participants were included in the intention-to-treat analysis. There were only 3 missed appointments throughout the study.

Table 1 shows the baseline characteristics of the participants. Overall, the mean age at recruitment was 47 years and 241 (75%) were female. The groups were evenly balanced on all characteristics.

Characteristics of the URTI episodes are shown in Table 2. There were a total of 1204 URTI episodes reported throughout the study period (mean, 3.7 [range, 0-17] and median, 3 [interquartile range {IQR}, 2-5] episodes per person), including 593 URTI episodes in the vitamin D group (mean, 3.7 [range, 0-17] and median, 3 [IQR, 2-5]) and 611 episodes in the placebo group (mean, 3.8 [range, 0-12] and median, 4 [IQR, 2-5]). Of the 1204 URTI episodes, 762 were spontaneously reported to study staff and 442 episodes were reported during regular monthly follow-up visits. Only 7 participants in the vitamin D group and 6 participants in the placebo group reported no URTI episodes throughout the entire study period. There were no statistically significant differences in the duration or severity of URTI episodes, in the number of days of missed work as a result of URTI episodes, or in the number of URTI episodes associated with positive nasopharyngeal swabs (Table 2).

The severity findings were unchanged when the data analyzed using WURSS-24 scores were summed over 14 days rather than 7 days, or when the WURSS-21 was used to measure severity (Table 2). The 442 URTI episodes that were not spontaneously reported were evenly distributed between the vitamin D and placebo groups (213 and 229, respectively). Fewer than half of URTI episodes resulted in any missed work (154 [41%] of 380 spontaneously reported episodes in the vitamin D group and 157 [41%] of 382 spontaneously reported episodes in the placebo group).

Overall, respiratory viruses were detected in 342 (50%) of 686 URTI episodes for which nasopharyngeal swabs were collected; the main reason for noncollection of swabs was that the participant was out of town when the URTI was reported. The most common viruses detected were human rhinoviruses and coronaviruses (Table 2). Influenza viruses were detected in only 4 episodes.

There was no interaction of treatment effect and URTIs between winter and summer (P = .52); mean number of URTI episodes during summer months was 1.3 for both treatment groups and during winter was 2.5 and 2.3 for the placebo and vitamin D groups, respectively. Both influenza seasons during the study period were relatively mild.

The mean baseline 25-OHD level was 29 (SD, 9) ng/mL, and only 5 participants (1.6%) had levels less than 10 ng/mL. Figure 2 shows mean 25-OHD levels among the intention-to-treat population. Vitamin D supplementation resulted in a steep increase in 25-OHD levels that was maintained throughout the study period. No statistically significant differences were noted for any outcome when the data were reanalyzed by baseline 25-OHD levels less than 20 ng/mL. Only 13 participants (all in the placebo group) had 25-OHD levels consistently less than 20 ng/mL throughout the study duration.

Mean corrected plasma calcium levels did not differ between the vitamin D and placebo groups at any time during the study period (mean, 9.2 [SD, 0.4] mg/dL for both treatment groups and at all time points). There were no cases of asymptomatic hypercalcemia (corrected plasma calcium >10.4 mg/dL). There were 40 serious adverse events (21 in the vitamin D group and 19 in the placebo group) and 1492 other adverse events (700 in the vitamin D group and 792 in the placebo group) recorded during the study, none of which was thought to be related to vitamin D supplementation (Table 3).

The main finding from this study is that a monthly dose of 100 000 IU of vitamin D₃ in healthy adults did not significantly reduce the incidence or severity of URTIs. This result remained unchanged when the analysis included winter season or baseline 25-OHD levels.

This finding is consistent with 2 other randomized controlled trials that were specifically designed to assess whether vitamin D supplementation prevents acute respiratory infections in adults. Li-Ng et al⁸ found no decrease in the incidence or severity of URTIs with vitamin D supplementation during winter in 162 adults. However, this study was of short duration (12 weeks), was underpowered, and first administered vitamin D₃ during (rather than before) winter. Laaksi et al⁷ also found no difference in their primary outcome (number of days absent from duty owing to respiratory tract infection) in 164 soldiers randomized to vitamin D₃ (400 IU/d) or placebo for 6 months during winter, although the overall proportion of participants who had no days absent from duty was higher in the vitamin D group. This study was also underpowered and the relatively low dose of vitamin D₃ resulted in only 29% of those in the intervention group obtaining 25-OHD levels greater than 32 ng/mL. In an adjunct to a trial investigating vitamin D supplementation as an intervention to prevent fractures in elderly people, there was no statistically significant relationship between treatment group and having had an infection or taking antibiotics during a week in winter.⁶

Two recent studies showed differing effects of vitamin D supplementation in children. A randomized trial of daily vitamin D supplementation in Mongolian schoolchildren in winter, a population with an average 25-OHD level of less than 10 ng/mL, found a 50% reduction in acute respiratory infections.¹⁹ In contrast, vitamin D supplementation did not affect the incidence of first episodes of pneumonia in 1524 infants from Afghanistan, another population with a high prevalence of vitamin D deficiency.²⁰

Strengths of our study include the relatively large sample size, the 18-month duration, and the high dose of vitamin D₃ administered (with a loading dose), thus avoiding the shortcomings of previous adult studies. Our dosing regimen, started during summer/fall, resulted in sustained mean 25-OHD levels greater than 48 ng/mL throughout the study period in those in the intervention group. Other strengths are the rigorous efforts to capture URTI episodes and the collection of virological data.

We did not show a benefit of vitamin D supplementation in our study population; however, it is possible that vitamin D may prevent URTIs in other populations. The mean baseline 25-OHD level was 29 ng/mL, and the mean level decreased to about 20 ng/mL during the winter in the placebo group; only 5 participants (1.6%) had baseline levels less than 10 ng/mL. It is possible that an effect may be observed in a population with a higher prevalence of vitamin D deficiency, as occurred in a recent trial of vitamin D supplementation to reduce exacerbations of chronic obstructive pulmonary disease.²¹ In that trial, vitamin D supplementation significantly reduced exacerbations only in patients with baseline 25-OHD levels less than 10 ng/mL.

We were also unable to assess the effect of vitamin D supplementation on prevention of infection caused by individual viruses. Of particular note, there were few cases of confirmed influenza infection among our partly vaccinated group of participants. Although adult data are unavailable, a randomized controlled trial in Japanese schoolchildren, set up to assess the effect of vitamin D supplementation on “doctor-diagnosed influenza,” did not report on that outcome but did report a statistically significant reduction in laboratory-confirmed influenza A infection (relative risk, 0.58; P = .04).²²

Would the results of our study have been different if we had given participants vitamin D, 3300 IU/d, as opposed to 100 000 IU monthly? Opposite outcomes have been documented for trials of 4-monthly vs annual dosing regimens of vitamin D supplementation for risk of fractures.^23- 24 Several mechanisms have been proposed to explain how various dosing regimens may have different effects on immune function.²⁵ However, it is purely speculative at this stage as to whether some conditions (eg, infections) require a smaller steady dose of vitamin D supplementation for benefit. Alternatively, genetic variation in vitamin D metabolism or signaling may modify the anti-infective effects of vitamin D. Vitamin D receptor polymorphisms have been linked to both susceptibility to tuberculosis²⁶ and response to vitamin D supplements in patients with tuberculosis.²⁷

With regard to safety, our regimen of monthly 100 000-IU doses of vitamin D₃ was more than 5 times the recommended daily allowance of 600 IU for adequacy in adults,²⁸ yet we documented no episodes of hypercalcemia and no adverse events attributed to vitamin D. Another recent study of 182 adults with moderate to severe chronic obstructive pulmonary disease reported 4 cases of mild asymptomatic hypercalcemia in those receiving monthly 100 000-IU doses of vitamin D without a loading dose.²¹ All 4 cases were noted 4 months after starting treatment and spontaneously resolved with normal calcium levels at 8 and 12 months, despite continuation of study medication.

In conclusion, we report that monthly administration of 100 000-IU doses of vitamin D₃ did not reduce the incidence or severity of URTIs in healthy, predominantly European adults with near-normal vitamin D levels. Further research is required to clarify whether there is benefit from supplementation in other populations and with other dosing regimens.