Influence of treatment components and diagnosis on parameters.

*subjects with treatment for diabetes excluded.**subjects with treatment for dyslipidemia excluded.1Corrected for age, sex, BMI, physical activity.2Corrected for age, sex, BMI, smoking, physical activity.3Corrected for age, sex, BMI, smoking.4corrected for age, sex, SES, BMI.5corrected for age, sex, physical activity. Glucose, insulin, HOMA and triglycerides levels were normally distributed after log-transformation and were expressed in percentages. ref = reference value (control subjects).Multiple linear regression analyses were performed using the following strategy: Model 1: the effects of both diagnoses (dummy variables) were added. Model 2: the effects of chemotherapy, nephrectomy, adrenalectomy, abdominal radiotherapy (dummy variables) were added. Additional linear regression analyses were performed according to the following strategy: Model 3: the effects of radiotherapy to the total pancreas and on part of the pancreas (Figure 1) (dummy variables) were added. Model 4: the effect of radiotherapy on the total liver and on part of the liver (Figure 1) (dummy variables) were added. Model 3 and 4 were additionally corrected for the treatment components that were significant in Model 2. P-values indicate the significance of the difference with control subject.