The accuracy of self‐reported pregnancy‐related weight: a systematic review. (March 2017)

• Record Type: Journal Article
• Authors: Headen, I.; Cohen, A.K.; Mujahid, M.; Abrams, B.
• Is Part Of: Obesity reviews. Volume 18:Number 3(2017:Mar.)
• Page Start: 350
• Page End: 369
• Publication Date: 2017-03
• Subjects: Postpartum period -- pregnancy -- validity -- weight; Obesity -- Periodicals; 616.398005
• DOI: 10.1111/obr.12486
• Languages: English
• ISSNs: 1467-7881

Abstract

Summary: Self‐reported maternal weight is error‐prone, and the context of pregnancy may impact error distributions. This systematic review summarizes error in self‐reported weight across pregnancy and assesses implications for bias in associations between pregnancy‐related weight and birth outcomes. We searched PubMed and Google Scholar through November 2015 for peer‐reviewed articles reporting accuracy of self‐reported, pregnancy‐related weight at four time points: prepregnancy, delivery, over gestation and postpartum. Included studies compared maternal self‐report to anthropometric measurement or medical report of weights. Sixty‐two studies met inclusion criteria. We extracted data on magnitude of error and misclassification. We assessed impact of reporting error on bias in associations between pregnancy‐related weight and birth outcomes. Women underreported prepregnancy (PPW: −2.94 to −0.29 kg) and delivery weight (DW: −1.28 to 0.07 kg), and over‐reported gestational weight gain (GWG: 0.33 to 3 kg). Magnitude of error was small, ranged widely, and varied by prepregnancy weight class and race/ethnicity. Misclassification was moderate (PPW: 0–48.3%; DW: 39.0–49.0%; GWG: 16.7–59.1%), and overestimated some estimates of population prevalence. However, reporting error did not largely bias associations between pregnancy‐related weight and birth outcomes. Although measured weight is preferable, self‐report is a cost‐effective and practical measurement approach. Future researchers should develop bias correction techniques for self‐reported pregnancy‐related weight.