To the Editor.
—Dr Cohen et al1 state that "people who participate in more types of social relationships have less susceptibility to rhinovirus-induced colds." While this may be true, their explanation may be incorrect. The probability of developing a cold in an individual is composed of 3 factors: first, the probability of coming into contact with an infected person; second, given contact, the probability of infection; and third, the probability of developing disease given infection. Since Cohen and colleagues challenged all individuals and artificially introduced the virus directly into nasal passages, their study investigated the third stage of this process only—ie, the probability of developing disease following infection (although based on the figures presented in Tables 1 and 2 it appears that those individuals not infected were included in the denominator for assessing risk of developing disease).We have performed a number of studies in which we have attempted