Physician-evaluated and self-reported morbidity for predicting disability

Objectives. This study compared the predictive validity of physician-evaluated morbidity and self-reported morbidity on disability among adults. Methods. Subjects from a large national survey (n = 6913) received a detailed medical examination by a physician and were asked about the presence of 36 health conditions at baseline. Disability measured 10 and 15 years later was regressed on the morbidity measures and covariates with tobit models. Results. Although physician-evaluated morbidity and self-reported morbidity were associated with greater disability, self-reports of chronic nonserious illnesses manifested greater predictive validity. Disability was also higher for obese subjects and those of lower socioeconomic status. Conclusions. The findings demonstrate the predictive utility of self-reported morbidity measures on functional disability.