The effect of β-carotene on common cold incidence is modified by age and smoking: evidence against a uniform effect in a nutrient–disease relationship

t of β-carotene on common cold incidence is modified by age and smoking: evidence against a uniform effect in a nutrient–disease relationship Original Research (2900) Total Article Views Authors: Harri Hemil Published Date October 2010 Volume 2010:2 Pages 117 - 124 DOI: http://dx.doi.org/10.2147/NDS.S13299 Harri Hemil Department of Public Health, University of Helsinki, Helsinki, Finland Background: Analyses in nutritional epidemiology usually assume that there is a uniform effect a nutrient. The purpose of this study was to test whether the effect of -carotene on common cold incidence is uniform over the population. Methods: The Alpha-Tocopherol Beta-Carotene Study, which recruited male smokers aged 50–69 years, was conducted in Finland in 1985–1993. The active follow-up lasted for 4.7 years (mean). This analysis is restricted to the β-carotene and placebo arms (n = 14,569). The rate ratio (RR) of the common cold was modeled as a function of age at follow-up in the β-carotene arm compared with the placebo arm using Poisson regression. Results: Separate regression models in four subgroups of participants were constructed on the basis of the age of smoking initiation (≤20 years versus ≥21 years) and baseline smoking level (5–14 versus ≥15 cigarettes/day). In three of the four subgroups, the effect of β-carotene was significantly modified by age. Among participants older than 70 years, the extent of smoking modified the effect so that β-carotene increased the incidence of colds in those who started smoking at an early age and smoked heavily: RR = 1.16 (95% confidence interval [CI]: 1.02–1.33), but decreased the incidence in those who started smoking at a later age and smoked less: RR = 0.76 (95% CI: 0.61–0.94). Conclusions: The strong evidence of heterogeneity in the β-carotene effect on the incidence of colds challenges the validity of cohort studies on nutrients, because they are usually based on the assumption of a uniform effect of the nutrient over the studied population.