[Utility of self-reported anthropometric data for evaluation of obesity in the Spanish population; study EPINUT-ARKOPHARMA]

Abstract

Introduction: Epidemiological studies frequently use weight and height collected by questionnaires, but the inaccuracy of self-reported data may bias the evaluation result. The aim of this study is to validate the self-report in Spanish adult population emphasizing the effect of age and nutritional status of the subjects.

Methods: The sample consist of 9,294 adults (8,072 women and 1,222 men) recruited from dietetic counselling centers dependents of Arkopharma laboratories in 46 Spanish provinces. Weight and height were asked and subsequently measured, calculating differences between self-reported and real parameters. Error in the estimation of body mass index (BMI) was evaluated considering the effect of age and nutritional status (T-test and multiple linear regression). Correlation between the classification based on self-reported BMI and anthropometric was analyzed using the Kappa test.

Results and discussion: Using the self-reported data, BMI was underestimated (2.62% in men, 3.10% in women). The error increases with age and extreme nutritional categories. The agreement between self-reported and real BMI in the nutritional assessment was good and correctly classified 74.71% of the males and 89.5% of women (Kappa: 0.695 and 0.782 respectively).

Conclusions: Considering the effect of age and nutritional status on the self-awareness of body size, we recommend caution in the use of questionnaires for epidemiological assessment.