[Study of validity in self-reported weight and height among pregnant women treated at municipal maternity hospitals in Rio de Janeiro, Brazil].

The main objectives of this article are the assessment of agreement between measure-based and report-based information on weight and height, besides the identification of the main characteristics related to 150 pregnant women who could report (versus could not report) their pre-gestational weight, weight at delivery, and height. According to the results, report-based information on previous weight is close to the measure-based information (Pearson coefficient correlation 0.96 and intra-class correlation 0.92; p < 0.05), with the same occurring for weight at delivery (Pearson coefficient correlation 0.96 and intra-class correlation 0.95; p < 0.05). Nevertheless, linear regression analysis detects that this relationship is troublesome for the height variable (beta0 = 45.5 and beta1 = 0.71; p < 0.05), and therefore the report-based information for height must be used cautiously when the measure-based values are not available. In relation to the characterization of the group of pregnant women who did not inform the three measures (n = 141), these women have the least schooling (63.9%), did not attend or attended fewer prenatal appointments (64.3%), and entered the prenatal program later, when they were already in the second or third trimester (59.3%).