In a prospective study on occupational risk factors for bladder cancer, smoking habits and Body Mass Index (BMI) were investigated in the analysis as confounders. BMI was inversely related to risk of bladder cancer, but we initially assumed this association could be attributed to the generally lower BMI of smokers. As the inverse association between BMI and bladder cancer risk appeared not to be affected by smoking, we conducted an in-depth investigation. The Netherlands Cohort Study recruited 58 279 men aged 55-69 at baseline in 1986. A self-administered questionnaire was used to collect information on risk factors for cancer, including diet, other lifestyle factors, anthropometry, and occupational history. After 11.3 years of follow-up through the Cancer Registry, 517 incident invasive primary bladder cancer cases (ICD-O-3: C67) were detected. Data were analyzed using Cox regression in a case-cohort approach, in which the person time accumulated was estimated from a randomly sampled subcohort (n = 2314). BMI was categorized as: <23, 23-24.9, 25-26.9, ≥27 kg/m2. Mean (SD) BMI at baseline was 25.1 (2.6), 25.2 (2.5) and 24.7 (2.7) kg/m2 for never, former and current smokers, respectively. We observed age-adjusted inverse associations between BMI and bladder cancer, which were similar after adjustment for smoking variables (current/other + number of cigarettes + years of cigarette smoking). RRs (95% Confidence Interval (CI)) for increasing categories of BMI compared to the lowest category were 0.8 (0.6-1.1), 0.7 (0.5-1.0), 0.6 (0.4-0.8). The association was the same after exclusion of cases diagnosed in the first two years of follow-up. After stratification for cigarette-smoking status we found the strongest inverse association for never smokers and weakest for former cigarette smokers. In the smoker groups, RRs were smaller after adjustment for smoking quantity and duration and exclusion of cases diagnosed in the first two years (RRs (95% CI) per unit (kg/m2) BMI: 0.90 (0.77-1.06), 0.98 (0.92-1.05), and 0.94 (0.88-1.00) for never, former and current smokers respectively). However, this association was different for retrospectively reported BMI at age 20 (RRs (95% CI) per unit: 0.76 (0.62-0.92), 1.01 (0.94-1.10), and 1.04 (0.97-1.11) for never, former and current smokers respectively). We conclude that the observed inverse association between baseline BMI and bladder cancer does not seem to be due to residual confounding by smoking or to preclinical disease.

[Fifth AACR International Conference on Frontiers in Cancer Prevention Research, Nov 12-15, 2006]