Development of a risk scale for postpartum weight retention

Date

2003-05

Authors

Milani, Tracey Joy

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The purpose of this research was to develop and va lidate a Risk Scale for Postpartum Weight Retention in low-income tri-ethnic women. In Study I, the scale was created in 187 subjects given six questionnaires and monitored for weight from 0 - 6 months postpartum. There were a total of 263 items derived from the questionnaires and 39 were related to weight change at 6 months postpartum (r = -0.20 to 0.61, p < 0.005). Expert review and exploratory multivariate analyses were used to reduce the questionnaire to six items that included weights, fat habits, weight control practices and self-esteem. Weight change was calculated as weight in kilograms retained from prepregnancy to 6 months postpartum. An initial method of evaluation, multiple regression, accounted for 45.8% of the variance in weight change (p < 0.001). A simpler method for clinicians, a ROC curve, discriminated between women who retained < 10 kg prepregnancy weight vs >̲ 10 kg. Area under the curve was 0.87; sensitivity was 76% and specificity, 80%. In Study II, validity was determined in a sample of 75 women who visited health clinics at 1.5 and 6 months postpartum. The risk scale and a demographic questionnaire were administered at 1.5 months postpartum; weight and waist circumference were measured at 1.5 and 6 months after delivering a child. When cross-validated in this sample, multiple regression explained 35.7% of the variance in weight. The ROC curve classified 81% of subjects into correct weight categories, with sensitivity of 65% and specificity, 89%, (p < 0.001). To enhance the predictive ability and reliability of the risk scale, weight at 1.5 months was replaced by waist circumference at 1.5 months postpartum. In the revised scale, multiple regression explained 41.5% of the variance in weight retention (p < 0.001). The updated ROC curve had an area under the curve of 0.83; with a sensitivity and specificity of 65 and 92%, respectively (p < 0.01). These results indicate that the risk scale is a valid measure of postpartum weight retention, using either complex multiple regression or simpler ROC methods. Additionally, waist circumference greatly increases its predictive power in Study II.

Description

text

LCSH Subject Headings

Citation