Smoking and adult leukemia: a case-control study.

Existing data from a large case-control study of tobacco-related cancers were used to examine the relationship between smoking history and leukemia as a whole, as well as specific morphologic types of leukemia. A total of 342 male and 220 female leukemia cases and two series of patients with non-tobacco related conditions (non-cancer controls: N = 9349 and cancer controls: N = 9846) were available for study. Analyses were carried out for all leukemia cases combined and for ANLL (N = 249), ALL (N = 52), CML (N = 78), and CLL (N = 57) separately vs both control groups. In addition to crude and age-adjusted odds ratios and their 95% confidence intervals, logistic regression models including potential confounding variables were used. No positive association with smoking was seen either for all cases combined or for any of the four specific morphologic types.