Modelos de regressão para o produto de acumulação lipídica em pessoas com síndrome metabólica

Abstract

The aim of the study was to analyze the relationship between anthropometric variables: waist circumference (WC), hip circumference (HC), body mass (BM), height (HGT) and hemodynamic variables: systolic (SBP) and diastolic (DBP) blood pressure with the lipid accumulation product (LAP), and describe a mathematical model that best represented it. Data from 430 subjects of both sexes, who presented at least three risk factors for Metabolic Syndrome (MetS) concomitantly, were analyzed, according to the recommendations of the Third Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation and Treatment of High Blood Cholesterol in Adults (NCEP-ATPIII). Descriptive statistical analysis was performed, followed by regression analysis, using the stepwise and enter methods to adjust the multiple linear regression model (MRL). The dependent variable for the MLR model was the LAP, and the predictor variables were WC, BM, HC, DPB, SBP and HGT. In the model obtained by the stepwise method, the percentage of explanation for LAP variation was 42% and the explanatory variables are WC (b = 1.946; p ≤ 0.001), DBP (b = 0.540; p ≤ 0.001) and HC (b = -0.594; p = 0.002). In the model generated by the enter method, the percentage of explanation for LAP variation was 39%, and the explanatory variables are WC (b = 1.567; p ≤ 0.001) and the DBP (b = 0.547; p ≤ 0.001). Therefore, the regression model that best described the relationship between anthropometric and hemodynamic variables with the LAP was that generated by the stepwise method. For the purpose of identifying a more simplified method of initial screening to detect likely candidates for the development of MetS, despite the limitations in the predictive power in the LAP variation, it is considered that the chosen model satisfies the need, from an analysis that also involved the health sciences.