CONFIABILIDADE DO USO DE MEDIDAS ANTROPOMÉTRICAS AUTO-REFERIDAS PARA O DIAGNÓSTICO DO ESTADO NUTRICIONAL EM GESTANTES.

Objective: Check the reliability of using self-reported measures of height and weight to obtain body mass index in women. Methods: Epidemiological study census, transversal in four pregnant women registered with the Family Health Units. We used a structured questionnaire containing questions about sociodemographic and anthropometric measures referred. Appealed to the Mann-Whitney U - Kruskal Wallis test to analyze differences between means in terms of categorical variables. For the agreement and obtain a summary measure of the values measured and reported, we used the kappa index and the correlation coefficient of Lin (CCL) respectively. Results: The women tended to underestimate both the weight and height, but more severe in weight (-0.90), which led to an overestimation of BMI (0.01). The weight was found to be above the upper limit (0.93) second correlation coefficient Lin, on the other hand the height and BMI are below the lower limit (0.89). 45.5% of women fall into the appropriate weight, representing a difference of approximately 10% compared to that (36.4%). Conclusion: BMI was calculated from the information provided does not fit the criteria of reliability values enough to indicate the reliability of these measures. [ABSTRACT FROM AUTHOR]