Prevalence of chest symptoms in nonexposed blue-collar workers.

Epidemiological investigations of lung disease induced by occupational exposures often require prevalence estimates for various respiratory symptoms in subjects without the exposure in question. A standard respiratory questionnaire was administered to carefully selected, nonexposed blue-collar workers (n = 1,372), and linear logistic prediction equations were developed for nine symptoms. For each smoking status, equations were fit employing race, sex, age, height, weight, and education as independent variables. These results can be used as descriptive statistics, as prediction equations from an external control group, and as a guide for selecting risk factors for adjustment of symptom prevalences in other population studies.