Validity assessment of self-reported weight and its correction process among Mexican adult women of reproductive age.

Abstract

We aimed to evaluate the agreement between self-reported weight (SRW) and measured weight (MW) in adult women of reproductive age, identify characteristics associated with the difference between SRW and MW (DW), and develop a correction procedure for SRW. We used data from 3,452 non-pregnant or non-lactating adult women who participated in the Mexican Family Life Survey. Standardized personnel asked women about their weight before measuring weight and height. We conducted a Bland-Altman analysis for agreement and adjusted linear regression models for sociodemographic characteristics. Mean DW was -0.59±3.21 kg. Difference varied according to Body Mass Index (BMI) and region of residence (p< 0.05). Correction model for log-MW, included the log-SRW, age group (18-34 and 35-49 years), interaction term (age × SRW), log-height, Southern region, and living with a partner. Based on self-reported weight, we observed an overestimation of underweight/normal weight prevalence and an underestimation of overweight or obesity prevalence. SRW has limitations to be considered as an alternative to MW among women of reproductive age with specific characteristics. Our proposed correction equation may decrease SRW imprecision improving the estimation of overweight and obesity. We suggest that studies consider and adjust the possible bias associated with weight misreporting on health outcomes.