Anthropometric, body composition and biochemical indicators for the prediction of the HOMA-IR index in adult men
This study aimed to evaluate the effectiveness of anthropometric and body composition indicators and lipid profile biochemical indicators in predicting the HOMA-IR index in men. The study was conducted in a cross-sectional design, in which 138 healthy adults (20 - 59 years) were evaluated. The anthropometric evaluation consisted of determining weight, height, sagittal abdominal diameter (SAD) and waist (WP), hip and thigh perimeters. The WP and SAD were measured in four distinct anatomical sites. The body composition was assessed by bioelectrical impedance. The nine anthropometric and body composition indicators studied were analyzed by the type of obesity that is intended to assess: central obesity indicators (WP, SAD, conicity index and waist/height ratio), general obesity indicators (body mass index and body fat percentage) and body fat distribution indicators (waist/hip ratio, waist/thigh ratio and sagittal index). The lipid profile biochemical indicators examined were: total cholesterol (TC), HDL-C, LDL-C and triglycerides (TG). The ratios TC/HDL-C and TG/HDL-C were calculated. The HOMA-IR index (Homeostasis Model Assessment - Insulin Resistance), an indicator of insulin resistance (IR), was calculated by the formula: HOMA-IR = fasting insulin (amp;#956;U/mL) x fasting plasma glucose (mmol/L)/22.5, considering for the analyses the percentile 75 as the cut-off point. Statistical analysis consisted of intraclass correlation, analysis of variance with Tukey post-hoc test, Kruskall-Wallis test with Dunn's post-hoc test, the Spearman and Pearson correlation coefficients and ROC (Receiver Operating Characteristic) curves. There was high reproducibility for all WP and SAD measures, with an intraclass correlation coefficient ranging from 0.986 to 0.999 (p lt; 0.001). The WP measured in the lower waist and the SAD measured in the largest diameter differed from other locations. Among the anatomical sites tested, the lower waist between the chest and hip, for the SAD, and the midpoint between the iliac crest and the last rib, for the WP, were the sites that showed the strongest correlations with HOMA-IR (r = 0.482 and 0.464, p lt; 0.001) and the largest areas under the ROC curves (0.716 ± 0.051 and 0.746 ± 0.049, p lt; 0.001) respectively, and therefore, showed better performance in predicting IR risk. Among the nine indicators of obesity tested, the WP and SAD were the most promising for assessing the IR risk. The values of 89.3 cm (sensitivity = 80% and specificity = 66%) for the WP and 20.0 cm (sensitivity = 77.1% and specificity = 68%) for the SAD were the cut-off points that showed the most accurate prediction for HOMA-IR higher levels. For the lipid profile biochemical indicators, it was found that the TG/HDL-C ratio presented the strongest correlation (r = 0.334, p lt; 0.001) with HOMA-IR and largest area under the ROC curve (0.724 ± 0.046, p lt; 0.001), resulting in better performance for the prediction of the HOMA-IR index. The use of these three indicators as instruments for the IR prediction in clinical practice is advisable. However, the necessity of a greater number of investigations about the performance of these indicators in the IR prediction in larger samples should be pointed out, reaching other extracts of the Brazilian population, including women, adolescents and elderly, which would allow the use of these IR indicators in population screenings and in clinical practice, in a standardized way, respecting our population?s characteristics.
Advisor:Sylvia do Carmo Castro Franceschini; Lina Enriqueta F. Paez de Lima Rosado; Silvia Eloiza Priore; Gilberto Paixão Rosado; Rita de Cássia Lanes Ribeiro; Dirce Ribeiro de Oliveira
School:Universidade Federal de Viçosa
Source Type:Master's Thesis
Date of Publication:06/09/2008