Vitamin C Supplementation and Respiratory Infections: a Systematic Review

Military Medicine, Nov 2004 by Hemilä, Harri

In this review, the vitamin C trials with military personnel and with other subjects living under conditions comparable to those of military recruits are analyzed to find out whether vitamin C supplementation affects respiratory infections. For this systematic review, we identified seven trials with military personnel, three trials with students in crowded lodgings, and two trials with marathon runners. Eight of these trials were double blind and placebo controlled and seven were randomized. Five small trials found a statistically significant 45 to 91% reduction in common cold incidence in the vitamin C group. These trials were short and the participants were under heavy exertion during the trial. Furthermore, three other trials found a statistically significant 80 to 100% reduction in the incidence of pneumonia in the vitamin C group. The large number of positive findings seems to warrant further consideration of the role of vitamin C in respiratory infections, particularly in military recruits.

Introduction

A high incidence of pneumonia and other respiratory infections is a common problem among military recruits, possibly caused by the crowding together of young adults from widely dispersed geographic areas.1-4 Navy and Marine recruits were at 30 times higher risk of hospital admission for pneumonia than were nonrecruits,1 and Navy and Marine personnel with less than 1 year of service were at five times higher risk of pneumonia than their peers with 4 or more years of service.4 Consequently, factors affecting susceptibility to respiratory infections in military recruits are of considerable practical importance.

The notion that vitamin C affects susceptibility to various infections-and respiratory infections in particular-is an old one,5,6 but the topic came to wider popularity only in the 1970s, when Pauling7,8 suggested that vitamin C supplementation would reduce the incidence and severity of colds. Trials carried out since then have consistently found that vitamin C alleviates common cold symptoms, but yields only modest benefit.9-12 Vitamin C had no effect on the incidence of the common cold in the largest trials,6,12,13 but trials with British men13 and with subjects under heavy, acute physical stress14 found reduction in common cold incidence, suggesting that vitamin C affects susceptibility to respiratory infections in restricted groups of people. Although Pauling was considerably overoptimistic about the potential benefits of vitamin c,7,8,11-13 another problem in this area of interest has been the negative bias against vitamin C. Three influential reviews concluded that vitamin C is ineffective against colds; however, the reviews presented data inconsistent with the original reports, overlooked several highly relevant findings, and analyzed data inappropriately.9,10

The purpose of the present review was to analyze the findings of vitamin C trials with military personnel and of trials with participants under conditions similar to those of military recruits to find out whether vitamin C affects the incidence or severity of respiratory infections.

Methods

Selection of the Trials

I previously searched the literature on vitamin C and respiratory infections using various MEDLINE, EMBASE, and SCISEARCH database searches, and I inspected the reference lists of reviews and original reports.6,11,13 One other author independently searched for trials on vitamin C and the common cold and published the identified bibliography.15 For this systematic review, further MEDLINE searches covering the years 1999-2002 were carried out.

The present review covers only controlled trials in which vitamin C was administered to one study group; the control group may or may not have received a placebo. This analysis focused on trials reporting respiratory tract infection outcomes. Two groups of trials were selected on the basis of subjects used in the trials: military personnel (seven trials; Table I) and students accommodated in crowded lodgings and marathon runners (five trials; Table II). Except for one trial,18 the trials in the tables are prophylactic such that participants were healthy at the trial's outset and supplementation continued over the occurring respiratory infection episodes. Although one trial examined patients hospitalized for influenza A, it was nevertheless prophylactic with respect to the occurrence of pneumonia after the initiation of vitamin C supplementation.18

Statistical Methods

Tables I and II show the p values calculated by the current author for the differences in the outcome values between the two groups. For dichotomous data, the mid-p modification28 of the Fisher's exact test29 was used to calculate p. For continuous variables, the exact p value was calculated using the t test,29 when mean and SD were reported in the original articles. The χ^sup 2^ test29 was used to test whether the distribution of participants with correct and incorrect answers in the Pitt and Costrini trial16 is explained by random variation. Assuming pure guessing, the expected number of correct and wrong answers is equal; both are one-half of all of the answers. Consequently, with 316 answers in the Pitt and Costrini trial, the expected number of correct and wrong answers is 158.