The HOMA-IR cut-off values for diagnosis of insulin resistance and relationship of metabolic syndrome with insulin resistance in elderly people in Minhang District of Shanghai / 中华内分泌代谢杂志

ABSTRACT

Objective To investigate the threshold values of insulin resistance ( IR) assessed by homeostasis model and the prevalence of IR in elderly people over 60 years old in Minhang district of Shanghai, and to evaluate the relationship between IR and metabolic syndrome ( MS) . Methods A total of 3003 elderly people aged 60 and over in the Jiangchuan community of Minhang District, Shanghai, were recruited, including 1286 males and 1717 females. Blood pressure, waist circumference, BMI, blood routine, serum creatinine, blood lipids, glucose, and fasting insulin were measured in all populations studied. Homeostasis model assessment ( HOMA) was used to estimate IR, and MS, and defined according to three diagnostic criteria including NCEP-ATPIII, IDF, and CDS. Results 75th percentile, 80th percentile and 90th percentile of HOMA values in 268 subjects with normotensive and normal BMI, glucose tolerance were considered as the thresholds of IR. The cut-off values were 2. 78, 3. 01 and 3. 56, respectively. And the prevalence of IR were 50. 0％, 42. 1％, and 27. 2％, respectively. IR level was significantly higher in people with MS. Based on the receiver operating characteristic ( ROC ) curve analysis, HOMA-IR and QUICKI index predicted MS well, and the optimal thresholds for diagnosing MS of HOMA-IR were 3. 17 for NCEP-ATPⅢ, 3. 02 for IDF and 3. 03 for CDS. BMI was the best factor for diagnosing IR among different MS components. Logistic regression analysis showed that gender, WC, BMI, SBP, HDL-C, TG, FBG and WBC were independent risk factors for IR. FBG≥5.84 mmol/L was the most dangerous factor of IR (OR=3.603,P<0.01), followed by WC≥85.4 cm(OR=2.152, P<0.01) and BMI≥24.6 kg/m2(OR=2.150,P<0.01). Conclusion The cut-off values of IR estimated by HOMA and the prevalence of IR were higher in elder subjects than other populations. IR was significantly positively correlated with MS. Excluding the conditions that insulin measurement were affected by external factors, HOMA-IR may predict the risk of MS. The components of MS were relative specific measurements of IR, FBG, BMI and WC were important risk predictors of IR in the elderly.