The validity and reproducibility of clinical assessment of nutritional status in the elderly.

Malnutrition is an important predictor of morbidity and mortality. In the non-elderly, a subjective global assessment (SGA) has been developed. It has a high inter-rater agreement, correlates with other measures of nutritional status, and predicts subsequent morbidity. The purpose of this study was to determine the validity and reproducibility of the SGA in a group of patients older than 70 y of age. Consecutive patients from four geriatric/rehabilitation units were considered for the study. Each patient underwent independent nutritional assessments by a geriatrician and senior medical resident. At the completion of the assessment, skinfold caliper measurements were obtained and the patient reclassified according to the results, which were then compared with objective measures of nutritional status. Six-month follow-up was obtained on all patients. The agreement between the two clinicians was 0.48 +/- 0.17 (unweighted kappa), which represents moderate agreement and is less than the reported agreement in nonelderly subjects. Skin calipers improved the agreement between clinicians but did not improve the correlation with other nutritional markers or prediction of morbidity and mortality. There was a correlation between a patient's severely malnourished state and mortality. In addition, patients with a body mass index (BMI) of <75% or >150% age/sex standardized norms had an increased mortality. The SGA is a reproducible and valid tool for determining nutritional status in the elderly. The reproducibility is less than in the nonelderly, which may relate to changes in body composition or ability to obtain an accurate nutritional history.