Agrupamento de fatores de risco cardiovasculares em população rural, por meio da análise fatorial

Abstract

Ever since Reaven introduced the concept know as Syndrome X, the clustering of several cardiovascular risk factors, such as: high blood pressure, dyslidaemia, excess of weight, and insulin resistance, has been the subject of keen debate. Evidence suggests that more thanone physiologic process mediates the simultaneous development of these metabolic disorders, each one producing a cluster of some metabolic markers. The factor analysis technique is used to interpret these clusters of risk factors, reducing a set of original variables in sub-sets of new latent variables, called as common factors, mutually notcorrelated. The objective of this study was, using factor analysis, to define the clustering of anthropometric and metabolic variables related to metabolic syndrome (MetS), by sex, in a Brazilian rural population. Anthropometric, bioquimical, hemodynamic and body composition data was collected from 579 subjects aged 18-94 years living in two rural areas of Brazil. The factor analysis technique was performed using the principal components analysis with varimax orthogonal rotation. For men, factor 1 was represented by waist-tohip-ratio (WHR), Body Mass Index (BMI), VLDL and TG/HDL; in factor 2, the variables were WHR, systolic blood pressure (SBP) and diastolic blood pressure (DBP); and factor 3,the variables were WHR, BMI, glucose and IR (HOMA-IR) - theses factors together explained 66.61% of the total variance in the data. For women, factor 1 was represented by WHR, VLDL and TG/HDL; factor 2 included WHR, SBP and DBP, and factor 3 include WHR, BMI, glucose and IR (HOMA-IR) - together these factors explained 68.98% of the total variance in the data. This study reduced a complex set of cardiovascular risk factor in 3 independent factors, each one reflecting a different aspect of MetS. The WHR, a central obesity marker, was correlated with the 3 factors in both men and women. This finding corroborates with the hypothesis that at least 3 physiopathologics domains are linked to central obesity in the mechanisms of the MetS in the population studied.