[Multidimensional profiles of lifestyle risk factors: application of the Grade of Membership method]

The aim of this study was to identify multidimensional profiles of lifestyle risk factors, verifying sociodemographic characteristics and self-rated health. Data were obtained from the Brazilian National Health Survey (2013) for 45,881 individuals 30 years and older. Identification of the profiles used the Grade of Membership (GoM) method, with 12 variables. Analysis of associated characteristics used Pearson's chi-square test and logistic regression. Two profiles were identified: "healthy profile" (61.6%), characterized by adequate consumption of fruits, vegetables, and fish, non-regular consumption of sodas, meat with fat, and beans, normal weight, and recommended leisure-time physical activity, and "risk profile" (38.4%), characterized by lack of consumption of healthy markers, except consumption of beans, consumption of all the unhealthy dietary markers, replacing meals with snacks, alcohol consumption, smoking, and leisure-time physical inactivity. The healthy profile was associated with female gender, old age, white skin color, residents in the North and Northeast of Brazil, widows/widowers, married marital status, more schooling, and better self-rated health. The risk profile was associated with male gender, young adults, residents in the Central and South, single marital status, less schooling, and worse self-rated health. Although more prevalent, the percentage of individuals with complete classification in the healthy profile was low. Lifestyle risk factors are associated with sociodemographic characteristics.