The epidemiology of the common cold IV. The effect of weather

Abstract

An investigation has been made of the association between weather and the numbers of colds reported on a given day. The seasonal trends were eliminated by working with the differences between the observed values on any day and the expected values derived from smooth curves fitted to the averages for the time of year. Examination of nine weather variables for the day on which the colds were reported and for each of the 29 preceding days showed that only two, mean day temperature and water-vapour pressure at 9 a.m., were significantly correlated with the numbers of colds. Partial correlation studies showed that the strongest association was with lowered mean day temperature between 2 and 4 days before the reported onset of symptoms. Regression analysis demonstrated that the magnitudes of the associations were sufficient to account for the greater part of the seasonal variation in the incidence of the common cold in both London and Newcastle. A small effect of atmospheric pollution appeared in this analysis. These results suggest that some effect of low outdoor temperature promotes transmission of the virus or the development of disease. © 1965, Cambridge University Press. All rights reserved.