Publications Search - Abstract View
||INDEPENDENT PREDICTORS OF BASELINE FASTING SERUM INSULIN AND GLUCOSE AND THE RISK OF PANCREATIC CANCER IN MALE SMOKERS in: Abstracts of the 41st Annual Meeting of the Society for Epidemiologic Research
||Meinhold CL, de GonzaleÂ´z AB, Albanes D, Weinstein SJ, Taylor PR, Virtamo J, Stolzenberg-Solomon RZ
||Am J Epidemiol
||2008 Jun 1
||Diabetes mellitus and higher glucose concentrations are consistent riskfactors for pancreatic cancer with most prospective studies reporting associationsindependent of body mass index (BMI). Using linear and logisticregression models, the authors identified potentially modifiable factors thatwere independently associated with baseline fasting insulin, glucose, and/ordiabetes among 383 Finnish male smokers who had fasting serologic measurementsin the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study(ATBC). Cox proportional hazards models were subsequently used tocalculate adjusted hazard ratios (HRs) for incident pancreatic canceramong the full ATBC Study cohort with relevant baseline information(n = 27,035). Between 1985 and 2004, 305 participants developed pancreaticcancer. Of the 20 factors associated with baseline fasting insulin, glucose,and/or diabetes, self-reported diabetes (HR = 2.12) and intakes ofsucrose, total fat, and saturated fat were significantly related to pancreaticcancer risk (highest versus lowest quintiles: HR sucrose = 0.68; 95% confidenceinterval (CI): 0.47, 0.98; HR saturated fat = 1.40; 95% CI: 0.98, 2.01;p-trend = 0.02). BMI strongly predicted baseline insulin, glucose, and diabetesbut was not associated with cancer risk (highest versus lowest quintile:HR = 0.96; 95% CI: 0.68, 1.37). In conclusion, dietary influences oninsulin and glucose concentrations independent of BMI and other cardiovascularrisk factors may play a role in pancreatic carcinogenesis.