Self-assessing weight status: A comparison of self-assessment and measurement of weight among Los Angeles County adults

Introduction: Overweight and obesity are arguably the most pressing public health issues of our time. Validity of self-reported weight data is questionable. As people become more overweight, they increasingly under-report their weight. As a result, researchers may be underestimating overweight and the burden of disease. Systematic bias in self-reporting weight may be due to a) intentional underestimates of weight or b) lack of information about weight. Both hypotheses are explored. Methods: A representative sample of 2051 Los Angeles adults is used to compare measured height and weight to self-reported height and weight. OLS regression models predict the relationship between observed weight and error in reporting weight. Covariates include: insurance status, education, gender, smoking status, past diagnosis of overweight, mental health status, and race/ethnicity. Results: Observed BMI is a strong predictor of error in self-reported weight and is robust to the inclusion of control variables. Results suggest that stigma from overweight may lead some respondents to intentionally underreport their weight, but evidence is stronger that simple lack of information about weight produces bias in self-reporting error. Conclusion: It is critical to consider the validity of self-reported weight data, because there is a bias toward underreporting weight as overweight increases. Providing biofeedback about weight may be a way to promote behavior change. To capture ability to accurately self-report weight, we should ask survey respondents whether they have access to a scale or when they last learned their weight. The role of stigma in producing bias requires further exploration.