Background The purpose of this study was to investigate the association between the use of herbal medicine containing aristolochic acid (AA) and the risk of cancers among patients with diabetes.

Methods We conducted a population-based cohort study on patients older than 18 years who had a diagnosis of diabetes (ICD-9 codes 250) between January 1, 1997 and December 31, 2010. To ensure comparability, we included only patients with diabetes who had visited traditional Chinese medicine clinics between January 1, 1997 and one year before the diagnosis of cancer or the censor dates. The use of herbal medicine containing AA was identified from January 1, 1997 to October 31, 2003 (the ban of herbs containing AA in November 2003). Each patient was individually tracked to identify incident cases of cancer (140–208) between January 1, 1999 and December 31, 2013.

Findings A total of 4 30 377 male and 4 31 956 female patients with diabetes were identified by using the National Health Insurance Research Database in Taiwan. There were 37 554 and 31 535 cancers during the follow-up period. AA use increased the risks of incident liver (155.0), kidney (189.0), pelvis and ureter (189.1, 189.2), and bladder (188) cancer in male patients with diabetes in a dose-dependent relationship. Similarly, there were increased risks of incident liver, pelvis and ureter, and bladder cancer in female diabetic patients in a dose-dependent manner.

Interpretation Our study suggests that AA exposure plays an important role in the carcinogenesis of liver, kidney, pelvis, ureter, and bladder cancers in patients with diabetes.