Validity of Self-Reports of Height and Weight among the General Adult Population in Japan: Findings from National Household Surveys, 1986

Background/Objectives

Growing evidence indicates that self-reported height and weight are biased, but little is known about systematic errors in the general adult population in Japan. This study takes advantage of the unique opportunity to examine this issue provided by the 1986 National Nutrition Survey.

Subjects/Methods

Individual-level data on a nationally representative sample aged 20–89 years from the National Nutrition Survey (November 1986) were merged with Comprehensive Survey of Living Conditions (September 1986) data to obtain a dataset containing both self-reported and measured data on height and weight for each person (n = 10,469). Discrepancies between self-reported and measured means of height, weight, and body mass index (BMI) were tested across measured BMI categories (<18.5, 18.5–24.9, 25.0–27.4, 27.5–29.9, and ≥30.0 kg/m²), age groups (20–44, 45–64, and 65–89 years), and sexes. Reporting bias in mean BMI was decomposed into the contributions of misreporting height and weight. The sensitivity and specificity of self-reported BMI categories were estimated.

Results

Mean self-reported BMI was substantially underestimated in older women (P<0.001; Cohen’s d, -0.4), and the major contributor to the bias was their over-reported height. Mean self-reported BMI was also considerably underestimated in both men and women who were overweight and obese (P<0.001; Cohen’s d, -1.0 to -0.6), due mainly to their underreported weight. In contrast, mean self-reported BMI was considerably overestimated in underweight men (P<0.001; Cohen’s d, 0.5), due largely to their over-reported weight. The sensitivity of self-reported BMI categories was particularly low for individuals who had a measured BMI of 27.5–29.9 kg/m² (40.9% for men and 26.8% for women).

Conclusions

Self-reported anthropometric data were not sufficiently accurate to assert the validity of their use in epidemiological studies on the general adult population in Japan in the late 1980s.