OBJECTIVES: to analyze housing features (types of floor, walls, ceiling, number of persons per room, water supply and sewage facilities) as malnutrition risk markers according to height/age ratio (<-2 scores-z of the National Center for Health Statistics standard (NCHS) in the metropolitan area of Recife (712) children, of the urban interior (684 children) and in the rural environment (644 children) from the database of the II Health and Nutrition Survey of the State of Pernambuco (1997). METHODS: bivariate analysis was used to establish associations of various housing features and nutritional status as well as multiple logistic regression analysis to discard confusing variables, to internally adjust association effects and to identify factors remaining in the final multivariate analysis. RESULTS: bivariate analysis demonstrate risk markers in the identification in the three sample spaces. In the logistic regression analysis, only the ratio of "number of persons per room" remained as a malnutrition risk factor in the different geographical spaces analyzed. CONCLUSIONS: these results demonstrate situations that should be focused by the health sector to achieve a better understanding of the nutritional issue, as well as other citizenship related aspects.
Child nutrition disorders; Risk factors; Logistic regression
