Diagnosing and Predicting Clinical and Para-clinical Cutoffs for Diabetes Complications in Lur and Lak Populations of Iran: A ROC Curve Analysis to Design a Regional Guideline
Description
Objectives: American diabetes association (ADA) updates its guideline every year. This guideline is obtained based on evidence-based medicine (EBM). However this guideline can be changed for different populations based on their cultural and genetic status. For this reason we intend to design a regional study in Lur and Lak of Iran, emphasizing on lipid profile cutoffs. Methods: A number of 133 diabetes mellitus (DM) patients were enrolled in this study. The collected information for each patient were sex, age, body-mass index (BMI), DM type, DM duration, fasting blood sugar (FBS), HbA1c, lipid profile, type of DM treatment, type of statin and dose, documented neuropathy, documented nephropathy, symptomatic retinopathy, peripheral vessel disease (PVD), documented cardiovascular disease (CVD), food ulcer history, dawn effect, systolic blood pressure (SBP), and diastolic blood pressure (DBP). Receiver operating characteristics (ROC) curve was used and area under curve (AUC) was reported. Results: For neuropathy, age was the most accurate diagnostic index (AUC= 79%). For nephropathy systolic SBP was the most accurate diagnostic index (AUC= 88%). For symptomatic retinopathy DM duration was the most accurate diagnostic index (AUC= 81%). For PVD HDL was the most accurate diagnostic index (reverse AUC= 100-33%). For CVD age was the most accurate diagnostic index (AUC= 81%). For foot ulcer history age was the most accurate diagnostic index (AUC= 85%). Conclusion: The final suggested guideline is like the international guidelines; however some unique points should be regarded. Blood pressure >165/110mmHg is diagnostic of diabetic nephropathy. HDL >48 is strongly suggested. A cohort study should be performed in our region in order to reevaluate predictive values.