Atividade física e comportamento sedentário em mulheres climatéricas: validação de instrumento utilizando técnicas de inteligência computacional

Abstract

Climacteric is a period in a woman's life characterized by hormonal changes, irregularity and cessation of menstrual cycles, which affects the health of this population. Thus, regular practice of activity is important at this stage of life, since it can minimize the deleterious symptoms resulting from this transition, which is still marked by an increase in sedentary behavior. In this context, studies on the practice of physical activity have gained notoriety in the health agenda, however, the biggest challenges pointed out refer to the lack of validated instruments to evaluate this practice in the population of climacteric women, and in recent years, new methods Nonlinear methods, such as computational intelligence techniques, have been used for the validation of data collection instruments. This thesis was conducted based on two interrelated, but distinct, studies. The first study aimed to assess the validity and accuracy of the International Physical Activity Questionnaire (IPAQ) short version, for climacteric women using computational intelligence techniques. The second study evaluated sedentary behavior in climacteric women and the associated factors. This is an epidemiological, cross-sectional and analytical study. Its population involved 30,018 climacteric women, assisted by the 73 units of Family Health Strategies in the city of Montes Claros, Minas Gerais, Brazil. Probabilistic sampling by conglomerate was adopted, whose sample consisted of 873 women. The data were obtained through the application of questionnaires, anthropometric evaluation and collection of peripheral venous blood for the analysis of biochemical parameters. In the validation process, considering the proposal to regroup the data set regarding the level of physical activity of climacteric women using the IPAQ, 2 algorithms were used: Kohonen and kMeans and, to evaluate the validity of these groupings, 3 indexes were used: Silhouette, PBM and Dunn. The questionnaire was tested for validity (factor analysis) and accuracy (Cronbach's alpha). The Random Forests technique was used to evaluate the importance of the variables that make up the IPAQ and to classify these variables, 4 algorithms were used: the Support Vector Machine (SVM), Artificial Neural Network (RNA) and Decision Tree (DT) and Random Forests (RF). To assess the factors associated with sedentary behavior, multivariate analysis was performed, using the Classification and Regression Tree algorithm to classify sedentary behavior into groups, based on exposure variables. In study 1, it was observed that the recommended for IPAQ, when applied to climacteric women, is to categorize the results in two groups (sufficiently active and insufficiently active). The results also showed that the instrument has reliability and validity for this specific population. And in the classification of the items that make up the IPAQ using the SVM, RNA, DT and RF algorithms, they point out that the values of Accuracy, Recall, Precision and area under the ROC curve did not differ with the removal of variables 1 and 8. Study 2 evidenced a high prevalence of sedentary behavior in the investigated population and the decision tree pointed out that the variables: total cholesterol, diastolic blood pressure, body mass index and triglycerides were associated with sedentary behavior. From the results, an educational pamphlet on sedentary behavior during the climacteric period and a pitch on the importance of physical activity and the promotion of the health of climacteric women were also made, which will be shared on social networks. Through this analysis it is concluded that the IPAQ, short version, has adequate measurement properties for the investigated population. There was a high prevalence of sedentary behavior and clinical, biochemical and anthrompometric factors are associated with this prevalence.