Use of Insulin and Insulin Analogs and Risk of Cancer — Systematic Review and Meta-Analysis of Observational Studies
Oystein Karlstad, Jacob Starup Linde, Peter Vestergaard, Vidar Hjellvik, Marloes T Bazelier, Marjanka K Schmidt, Morten Andersen, Anssi Auvinen, Jari Haukka, Kari Furu, Frank de Vries and Marie L De Bruin
Affiliation: Norwegian Institute of Public Health, P.O.Box 4404 Nydalen, N-0403 Oslo, Norway
Keywords: Cancer risk, diabetes mellitus, insulin, neoplasm, meta-analysis, systematic review.
Background: An association of insulin use and risk of cancer has been reported but evidence is conflicting and
methodological issues have been identified.
Objective: To summarize results regarding insulin use and cancer risk by a systematic review and meta-analysis of cohort
and case-control studies examining risk of cancer associated with insulin use in patients with diabetes.
Data Sources: Systematic literature search in 5 databases: PubMed, Embase, Web of Science, Scopus and Cochrane
Study Eligibility Criteria (PICOS): Population: diabetes patients. Exposure: Users of any exogenous insulin. Comparison:
Diabetes patients with or without use of antidiabetic drugs. Outcome: Any incident cancer. Study Design: Cohort and
Results: 42 eligible studies examined risk of any cancer and 27 site-specific cancers. Results of individual studies were
heterogeneous. Meta-analyses were significant for: Insulin vs No Insulin: Increased risk for pancreas, liver, kidney,
stomach and respiratory cancer, decreased risk for prostate cancer. Insulin vs Non-Insulin Antidiabetics: Increased risk for
any, pancreatic and colorectal cancer. Glargine vs Non-Glargine Insulin: Increased risk for breast cancer, decreased risk
for colon cancer.
Limitations: Few studies available for most cancer sites and exposure contrasts, and few assess effect of dose and duration
of exposure. Methodological issues in several studies. Availability of confounders.
Conclusions: Insulin use was associated with risk of cancer at several sites. Cautious interpretation of results is warranted
as methodological issues and limitations in several of the included studies have been identified. Choice of study design
may have a profound effect on estimated cancer risk.
Rights & PermissionsPrintExport