Alexandra M Mueller1, Christoph R Meier2, Susan S Jick3, Cornelia Schneider1. 1. Basel Pharmacoepidemiology Unit, Division of Clinical Pharmacy and Epidemiology, Department of Pharmaceutical Sciences, University of Basel, Basel, Switzerland; Hospital Pharmacy, University Hospital Basel, Basel, Switzerland. 2. Basel Pharmacoepidemiology Unit, Division of Clinical Pharmacy and Epidemiology, Department of Pharmaceutical Sciences, University of Basel, Basel, Switzerland; Hospital Pharmacy, University Hospital Basel, Basel, Switzerland; Boston Collaborative Drug Surveillance Program, Lexington, MA, United States. Electronic address: christoph.meier@usb.ch. 3. Boston Collaborative Drug Surveillance Program, Lexington, MA, United States; Boston University School of Public Health, Boston, MA, United States.
Abstract
OBJECTIVES: To evaluate the potential of blood glucose levels and weight change before the onset of diabetes as predictors of pancreatic cancer among subjects with new-onset diabetes, that is, cancer-related diabetes versus normal type 2 diabetes. METHODS: We conducted a case-control study among subjects with new diabetes in the United Kingdom-based Clinical Practice Research Datalink. Cases were pancreatic cancer subjects with diabetes for ≤2 years before the cancer diagnosis (i.e., cancer-related diabetes). Controls were cancer-free, type 2 diabetic subjects matched to cases on age, sex, and diabetes duration. We calculated adjusted odds ratios (aORs) for pancreatic cancer as a function of both weight change and blood glucose before the onset of diabetes. RESULTS: Weight loss of 10.0%-14.9% at diabetes onset was associated with an aOR for pancreatic cancer of 3.58 (95% CI 2.31-5.54), loss of ≥15.0%, with an aOR of 4.56 (95% CI 2.82-7.36), compared with stable weight. Blood glucose levels of ≤5.1 mmol/L or 5.2-5.6 mmol/L before diabetes onset were associated with an increased risk of a pancreatic cancer diagnosis, with aORs of 2.42 (95% CI 1.60-3.66) and 2.20 (95% CI 1.45-3.35), respectively, when compared with blood glucose levels ≥6.3 mmol/L within >2-3 years before cancer detection. CONCLUSIONS: Weight loss as well as blood glucose levels in the normal range (and thus rapid development of hyperglycemia) before diabetes onset may be predictive of pancreatic cancer-related diabetes and may help target which subjects with new diabetes to refer for pancreatic cancer screening examinations.
OBJECTIVES: To evaluate the potential of blood glucose levels and weight change before the onset of diabetes as predictors of pancreatic cancer among subjects with new-onset diabetes, that is, cancer-related diabetes versus normal type 2 diabetes. METHODS: We conducted a case-control study among subjects with new diabetes in the United Kingdom-based Clinical Practice Research Datalink. Cases were pancreatic cancer subjects with diabetes for ≤2 years before the cancer diagnosis (i.e., cancer-related diabetes). Controls were cancer-free, type 2 diabetic subjects matched to cases on age, sex, and diabetes duration. We calculated adjusted odds ratios (aORs) for pancreatic cancer as a function of both weight change and blood glucose before the onset of diabetes. RESULTS:Weight loss of 10.0%-14.9% at diabetes onset was associated with an aOR for pancreatic cancer of 3.58 (95% CI 2.31-5.54), loss of ≥15.0%, with an aOR of 4.56 (95% CI 2.82-7.36), compared with stable weight. Blood glucose levels of ≤5.1 mmol/L or 5.2-5.6 mmol/L before diabetes onset were associated with an increased risk of a pancreatic cancer diagnosis, with aORs of 2.42 (95% CI 1.60-3.66) and 2.20 (95% CI 1.45-3.35), respectively, when compared with blood glucose levels ≥6.3 mmol/L within >2-3 years before cancer detection. CONCLUSIONS:Weight loss as well as blood glucose levels in the normal range (and thus rapid development of hyperglycemia) before diabetes onset may be predictive of pancreatic cancer-related diabetes and may help target which subjects with new diabetes to refer for pancreatic cancer screening examinations.