Determining Optimal Cut-Off Value of Pancreatic-Cancer-Induced Total Cholesterol and Obesity-Related Factors for Developing Exercise Intervention: Big Data Analysis of National Health Insurance Sharing Service Data

SIMPLE SUMMARY: This study aimed to investigate the impact of several variables affecting the development of pancreatic cancer through the analysis of a sample cohort of 1,108,369 individuals from the National Health Insurance Sharing Service (NHISS DB). Furthermore, we aimed to develop individualized evidence-based prognostic and preventive programs (e.g., tailored exercise-based programs) through the derivation of cut-off value results of significant parameters to improve pancreatic cancer. There were differences between the pancreatic cancer versus non-pancreatic cancer groups in terms of gender; for example, body mass index, systolic/diastolic blood pressure, fasting blood glucose, and total cholesterol concentrations were lower in women with pancreatic cancer than in women without pancreatic cancer (p < 0.01). Fasting blood glucose and total cholesterol concentrations were significantly different between men with versus without pancreatic cancer (p < 0.05). Logistic regression analysis identified more than 20 parameters, including the impact of exercise. Afterward, the ROC curve analysis revealed significant cut-off values, which differed between men and women (i.e., the ROC curve analysis showed that total cholesterol concentration was the only significant factor associated with pancreatic cancer in men). Our analysis of large-scale data from the NHISS DB suggests that identifying significant parameters helps in developing individualized evidence-based prognostic and preventive programs (e.g., tailored exercise-based programs) for ameliorating pancreatic cancer. ABSTRACT: This study aimed to examine the effects of multiple parameters on the incidence of pancreatic cancer. We analyzed data from 1,108,369 individuals in the National Health Insurance Sharing Service Database (NHISS DB; birth to death; 2002 to 2015) and identified 2912 patients with pancreatic cancer. Body mass index, systolic/diastolic blood pressure, and fasting blood glucose and total cholesterol concentrations were lower in women with than without pancreatic cancer (p < 0.01). Fasting blood glucose and total cholesterol concentrations were significantly different between men with and without pancreatic cancer (p < 0.05). In the logistic regression analysis, the total cholesterol concentration (odds ratio (OR), 1.007; 95% confidence interval (CI), 1.005–1.010) was significantly higher in men than women with pancreatic cancer (p < 0.05). Pancreatic cancer rates were highest in men who smoked for 5–9 years or more (OR, 5.332) and in women who smoked for 10–19 years (OR, 18.330). Daily intensive exercise reduced the risk of pancreatic cancer by 56% in men (95% CI, 0.230–0.896). Receiver operating characteristic curve analysis revealed a total cholesterol concentration cut-off point of 188.50 mg/dL (p < 0.05) in men with pancreatic cancer, with a sensitivity and specificity of 53.5% and 54.6%, respectively. For women, the cut-off values for weight and gamma glutamyl transpeptidase concentration were 58.5 kg and 20.50 U/L, respectively. The sex-specific differences in patients with pancreatic cancer identified herein will aid in the development of individualized evidence-based prognostic and preventive programs for the treatment of pancreatic cancer.