In this study, the authors assess the association of diabetes and obesity with urothelial cancer using a large Korean national health care dataset, the National Health Insurance System (NHIS). However, this is very clearly an association study, with no discussion of causation and biologic rationale.
Of approximately 1 million participants in the year 2004, 826,170 participants took the survey between 2004-2008. The authors excluded all women and men < 20 years of age – for unclear reasons. Female patients are at risk for bladder cancer and should have been included in the study.
They then excluded duplicate patients (~200K) and anyone with a prior urothelial carcinoma diagnosis at the time of the survey (117 patients).
Ultimately, 205,348 men aged 20 years and older who had a health examination in 2004 and 2008, and who did not have urothelial carcinoma in 2004, were included. They then assessed the development of urothelial carcinoma in that 4 year period. They completed multivariable analyses looking at predictors of developing bladder cancer, with obesity and diabetes specifically assessed.
Men with a higher BMI in 2004 were more likely to acquire urothelial cancer independent of variables. In the subset of the population with diabetes, there was an increasing trend in the overweight and obese patients towards developing bladder cancer, compared to the group without diabetes.
Unfortunately, their full multivariable analyses are not present, so it is hard to critique this analysis. However, I do think there are multiple flaws in the analysis.
- In the age-adjusted analysis, overweight and obese patients are at higher risk, but so are underweight patients (HR 1.11). No clear trend evident, suggesting something is wrong with the analysis.
- In the MVA (age, hypertension, dyslipidemia, smoking status, alcohol consumption, and exercise), underweight patients were also at higher risk (HR 1.20).
Additionally, they only allowed for the development of bladder cancer in a 4 year time frame, even when including patients 20-40, who are very unlikely to develop bladder cancer. Perhaps a better population or longer follow-up is needed.
This study has numerous flaws, and should not be taken at face value. While the theory is nice, the analysis is not appropriate.
Presented by: H.W. Moon, MD
Co-Authors: Yang J., Kang S., Lee K.W., Jeong H.C., Choi J.B., Choi S.W., Park Y.H., Bae W.J., Cho H.J., Hong S., Lee J.Y., Kim S.W., Ha U-S.
Written by: Thenappan Chandrasekar, MD Clinical Fellow, University of Toronto, twitter: @tchandra_uromd at the 2018 European Association of Urology Meeting EAU18, 16-20 March, 2018 Copenhagen, Denmark