Detection bias may be the main cause of increased cancer incidence among diabetics: Results from the Rotterdam Study

Kirstin Bruijn, Roel Ruiter, Toke de Keyser, Bert Hofman, Bruno Stricker, Casper van Eijck

Research output: Contribution to journalArticleAcademicpeer-review

35 Citations (Scopus)


Aim: Type 2 diabetes is associated with an increased cancer risk. Most studies on this topic analyse diabetes as a risk factor without adjusting for diabetes duration before cancer occurrence. This study aimed to investigate the association between diabetes duration and cancer risk in more detail. Methods: In this prospective cohort study, diabetes diagnosis was based on clinical information and use of glucose lowering medication. Details on incident cancers were obtained via general practitioners and linkage to pathology registers. Cox proportional hazards models were used with onset and duration of diabetes as time-varying determinants. Results: The study comprised 10,181 individuals. Diabetes was associated with an increased overall risk of incident cancers (hazard ratio (HR) 1.2, 95% confidence interval (CI) 1.07-1.39) and pancreatic cancer (HR 2.9, 95% CI 1.75-4.89). A diagnosis of diabetes less than three months before the diagnosis of cancer was associated with strongly increased risks of all- (HR 3.3, 95% CI 2.50-4.32) and pancreatic cancers (HR 28.7, 95% CI 6.32-130.58). Conclusion: The magnitude of the association between diabetes and an increased risk of cancer seems to be inflated by detection-or protopathic bias. Future studies investigating this association should adjust for diabetes duration and include a plausible aetiological risk window. (C) 2014 Elsevier Ltd. All rights reserved.
Original languageUndefined/Unknown
Pages (from-to)2449-2455
Number of pages7
JournalEuropean Journal of Cancer
Issue number14
Publication statusPublished - 2014

Research programs

  • EMC MM-03-47-02-A
  • EMC NIHES-01-64-03

Cite this