Serum 25-hydroxyvitamin d and the incidence of acute viral respiratory tract infections in healthy adults

Abstract

Background: Declining serum concentrations of 25-hydroxyvitamin D seen in the fall and winter as distance increases from the equator may be a factor in the seasonal increased prevalence of influenza and other viral infections. This study was done to determine if serum 25-hydroxyvitamin D concentrations correlated with the incidence of acute viral respiratory tract infections.

Methodology/findings: In this prospective cohort study serial monthly concentrations of 25-hydroxyvitamin D were measured over the fall and winter 2009-2010 in 198 healthy adults, blinded to the nature of the substance being measured. The participants were evaluated for the development of any acute respiratory tract infections by investigators blinded to the 25-hydroxyvitamin D concentrations. The incidence of infection in participants with different concentrations of vitamin D was determined. One hundred ninety-five (98.5%) of the enrolled participants completed the study. Light skin pigmentation, lean body mass, and supplementation with vitamin D were found to correlate with higher concentrations of 25-hydroxyvitamin D. Concentrations of 38 ng/ml or more were associated with a significant (p<0.0001) two-fold reduction in the risk of developing acute respiratory tract infections and with a marked reduction in the percentages of days ill.

Conclusions/significance: Maintenance of a 25-hydroxyvitamin D serum concentration of 38 ng/ml or higher should significantly reduce the incidence of acute viral respiratory tract infections and the burden of illness caused thereby, at least during the fall and winter in temperate zones. The findings of the present study provide direction for and call for future interventional studies examining the efficacy of vitamin D supplementation in reducing the incidence and severity of specific viral infections, including influenza, in the general population and in subpopulations with lower 25-hydroxyvitamin D concentrations, such as pregnant women, dark skinned individuals, and the obese.

Figure 1. Predicted vs. observed serum 25-hydroxyvitamin D concentrations.

Predicted from the equation [serum vitamin D ng/ml] = [427.4×(1+pigmentation grade)^0.5+(dose IU/day) ^0.5]/Body mass index, where body mass index = (weight lbs)(703)/height inches)². See text for pigmentation grade. The observed concentrations were from the first observation period. Model parameters were SSE = 18755.12; DFE = 197; RMSE = 9.75. The error associated with the parameter estimate was 427.4 +/− 10.4 (SEM) (406.8–448.0) 95% CI.

Figure 2. Length of time to viral infection related to initial serum concentration of 25-hydroxyvitamin D.

Shown are the results of the pharmacodynamic model relating 25-hydroxyvitamin D to length of time before a viral respiratory tract infection. The model equation is Log(time to event) = b₀+b₁/(Maximal 25OH-Vit D–serum 25-OH vitamin D+1) calculated using a Weibull loss function. The parameters associated with the regression are as follows, given with 95% CI and P-values: b₀ = 4.29 (3.50–4.96), <0.001; b₁ = 38.33 (9.64–76.48), <0.001; Sigma = 0.79 (0.65–0.97), <0.001. Overall P-value for this model was <0.0023. Curves are fitted 0.1, 0.5 and 0.9 quantiles as a function of the regressor. The x points represent individuals who developed viral infections, and the other points represent individuals who did not develop infections and the three who were not followed until the end of the observation period. For the individuals who developed viral infections (x points) the mean age was 46.9 years; there were 47% men; 76.5% had light pigmentation, 18.8% intermediate pigmentation, 4.7% dark pigmentation; the mean initial 25-hydroxyvitamin D concentration was 26.12 ng/ml; and the mean vitamin D supplementation 292.5 IU. For the individuals who did not develop viral infections (the other points) the mean age was 47.0 years; there were 39.8% men; 78.8% had light pigmentation, 14.2% intermediate pigmentation, 7.0% dark pigmentation; the mean initial 25-hydroxyvitamin D concentration was 30.03 ng/ml; and the mean vitamin D supplementation 601.4 IU.

Figure 3. Survival without respiratory viral infection over course of study, stratified by serum 25-hydroxyvitamin D concentrations.

Results of Cox proportional hazard analysis for time to viral infection, stratified by serum 25-hydroxyvitamin D concentrations. Parameters included 103 events (viral infections), 3 in the ≥38 ng/ml group and 100 in the <38 ng/ml group, and 114 censorings (15 without infection in the ≥38 ng/ml group, 99 in the <38 ng/ml group, and 3 lost to follow-up). The parameter estimate for 25-hydroxyvitamin D≥38 ng/ml was −0.66 (−1.36—0.17) 95% CI and the risk ratio 0.51 (0.25 to 0.84) 95% CI (p<0.0001).