Study design and population
This is a cross sectional study on PGCS participant (PERSIAN Guilan Cohort Study), a prospective, population-based cohort study in Guilan has been described in detail elsewhere[17-19].Briefly, The Guilan cohort study(GCS) was conducted on 10 520 participants aged between 35-70 years in Guilan province, northern Iran, between October 8, 2014 and January 20, 2017 as part of the Prospective Epidemiological Research Studies in Iran (PERSIAN).Eligible subjects were contacted through phone by trained interviewers who can spoke the native language of the region and invited to participate the study. After signed informed constant all study data including demographic characteristics, socio-economic status, lifestyle and sleep habits, Anthropometric indices and blood pressure were recorded by a trained research assistants. Also biological samples were collected. In phase 2, annually active follow up was planed for next 15 years for all participants according to the PERSIAN cohort protocol . Present study data included 7989 non diabetic participants of the GCS study. Diabetic subjects were excluded. Subjects with DM in the GCS were defined as 1) history of diagnosed DM 2) history of anti diabetic medication consumption 3) fasting blood sugar (FBS)>126 in the initial cohort laboratory data .
Data collection and measurement
For every participant, we retrieved data from GCS database that were collected through interviews, physical examinations, and laboratory tests according to cohort protocol . For the present study, data included demographic factors like age, sex, living location (city or rural), Marital status, Occupation (employed, unemployed), anthropometric indices including weight, height, hip and waist circumference, waist to hip ratio (WHpR) and waist to height ratio (WHtR), history of hypertension (HTN), gestational DM in women subjects and any history of DM in their first degree family like father, mother, sister or brother and finally information about physical activity. All anthropometric indices including weight, height, Hip Circumference (HC), Waist Circumference (WC), WHpR, and WHtR were measured by trained research assistants according to GCS protocol. Body mass index (BMI) was categorized as underweight (BMI<18.5 kg/m2), normal weight (BMI= 18.5-24.99 kg/m2), overweight (BMI= 25-29.9 kg/m2) and obese (BMI≥30 kg/m2). The level of Physical activity was reported as metabolic equivalent rates (METs) based on self reported daily activity PERSIAN cohort questionnaire.
The risk of developing DM or prediabetes was calculated for every individual based on ADA risk prediction model through online calculator  the ADA risk prediction model was developed based on American population higher than 20 years without DM to identify high risk individuals for DM or prediabetes. ADA risk score included 7 questions like age, sex, race, weight, height, family history of DM, history of gestational DM, history of HTN and physical activity. Total score was calculated between 0-11. The higher score represent higher risk of diabetes. The cut point 5 or higher shows the high risk for DM and cut point 4 shows the high risk for prediabetes . All required data for calculating ADA risk were extracted from cohort study. Family history of DM in ADA risk score was defined any history of diabetes in mother, father, sister or brother. Gestational diabetes in PERSIAN cohort was considered yes if women answered yes to the question “did you have a history of diabetes in pregnancy or did you have given birth a baby with ≥4 kg?” For race, all participants were defined as white. For physical activity, the question in ADA risk score tool was “are you physically active? Yes or no” Low level of physical activity in PERSIAN cohort was defined as less than mean METs rates of participants (41 METs/hour/day) that have been previously descried in details .
This research project was approved by the Ethics Committee of the Gastrointestinal and Liver Disease Research Center and Guilan University of Medical Sciences (code number IR.GUMS.REC.1398.241). All participants expressed their consent for participation in the research.
In this study, continuous variables were expressed as mean ± standard deviation (SD) and categorical variables as frequency (percentage). One-way ANOVA and Chi-square test were used to compare demographic characteristics and anthropometric indices among normal, prediabetes, and diabetes groups. Receiver operating characteristic (ROC) curves were used to study diagnostic accuracy of the anthropometric indices for detecting patients with diabetes, represented by area under the curve (AUC). An AUC value of 0.5 indicates an entirely random classifier and an AUC value of 1 indicates perfect classifier. The best cut-off value was defined as the value with the highest accuracy that maximizes you den’s J statistic, i.e. J = sensitivity + specificity – 1. Data analysis was performed using IBM SPSS Statistics for Windows, version 26.0 (IBM Corp., Armonk, NY, USA), and a P <0.05 was considered statistically significant.