Cigarette smoking, bladder cancer, and changes in risk over 40 years.

A209

OBJECTIVE: To test the hypothesis that the risk of bladder cancer from cigarette smoking has increased over time. A result of the changing cigarette is an increase in the levels of aromatic amines, which are considered potent bladder carcinogens. METHODS: Patients who attended Roswell Park Cancer Institute between 1957 and 1965 and completed a detailed epidemiologic questionnaire were eligible for case-control analysis #1, while patients who attended Roswell Park Cancer Institute between 1982 and 1997 and completed a similarly detailed questionnaire were eligible for case-control analysis #2. Nine-hundred fifty cases (520, 430; analysis #1 and #2, respectively) of primary, incident, histologically confirmed bladder cancer were frequency matched to 2,590 malignancy free controls (1,560, 1,240; analysis #1 and #2, respectively; 3:1 ratio) on age, gender, and admission year (analysis #2 only). Logistic regression was used to calculate bladder cancer risk by various smoking exposures for each of the two analyses, while controlling for relevant covariates including race, alcohol use (ever, never), fruit consumption, vegetable consumption, and BMI. The data for analyses #1 and #2 were pooled to investigate potential differences in risk between the two analyses. RESULTS: Both analysis #1 and #2 found a significant association between bladder cancer and cigarette smoking. The relative risks for the various smoking exposures calculated in analysis #2 were approximately double the corresponding risks calculated in analysis #1. For analysis #1, those in the highest tertile of smoking duration (45 or more years) had a significant increase in bladder cancer risk, while those in the highest tertile of smoking duration (39 or more years) for analysis #2 had a greater risk of bladder cancer (analysis #1 OR = 1.87 vs. analysis #2 OR = 3.93; p<0.01 for both). The increase in risk was greater for males than females, although statistical power for females was limited by smaller sample size. CONCLUSIONS: The risk of bladder cancer from cigarette smoking significantly increased from the late 1950's to the late 1990's in these two datasets. The increase in risk may be due to changes in cigarette design, which may have resulted in greater concentrations of aromatic amines over the past 40 years.