Background: A large number of epidemiological studies have reported positive associations between bladder cancer and sales occupations. We investigated whether these findings are likely to be due to chance, confounding or publication bias, or may involve causal associations.
Methods: Studies reporting bladder cancer risk-estimates for sales occupations were reviewed. Using meta-analyses we assessed heterogeneity and publication bias, and derived summary estimates.
Results: Eighteen publications were identified, reporting 85 risk-estimates for sales-work. Meta-estimates were elevated for men (odds ratio (OR) 1.11, 95% confidence interval (CI) 1.01-1.21) and women (OR = 1.36, 95% CI = 1.11-1.67). The estimate was heterogeneous for men (p(Q-test) <0.01, women: 0.18) and indicated publication bias for women (p(Egger-test) <0.01, men: 0.40). When including only smoking-adjusted estimates reported irrespective of the strength of the association, the summary estimate for generic groups of sales workers was 0.99 (95% CI 0.90-1.08) for men, and 1.18 (95% CI = 0.99-1.39) for women, without statistically significant heterogeneity or publication bias. For women, risk was positively associated with longer duration of sales-employment in three studies.
Conclusions: Publication bias explained most of the reported increased bladder cancer risk, but sales-work still appeared to be associated with a small risk in women. Possible causal factors include lower frequency of urination and reduced fluid intake.
(c) 2006 Wiley-Liss, Inc.