Study objective: To evaluate (1) the prevalence of protein-energy malnutrition in elderly patients; (2) the changes in nutritional status during the hospital stay; and (3) (main objective) the relationship between simple nutritional parameters and short-term in-hospital mortality.
Design: Prospective time series at admission and on the 15th day of hospitalization.
Setting: Medical care unit in a teaching hospital.
Participants: Consecutive sample of 324 hospitalized patients greater than or equal to 70 years (86.4% of eligible patients). Norms of measurements were obtained from a referred sample of healthy control subjects (26 males and 36 females).
Main outcome measures: Mid-arm circumference, triceps skinfold thickness, serum albumin, prealbumin, and retinol-binding protein levels were measured in patients at admission and on the 15th day.
Results: (1) Prevalence of PEM was 30% in male and 41% in female patients. (2) Both mid-arm circumference and serum albumin level decreased over the first 15 days of hospital stay (53 patients, paired t test, P less than 0.05). Triceps skinfold thickness did not change. (3) A step-wise discriminant-function analysis determined the utility of the parameters at admission as predictors of in-hospital mortality before the 15th day. Mid-arm circumference, triceps skinfold thickness, albumin, and prealbumin levels, as well as age, are predictors of in-hospital mortality, with 73% sensitivity, 69% specificity, and 70% of correctly classified patients of both sexes.
Conclusions: Parameters used are predictors for short-term in-hospital mortality of elderly patients hospitalized in an acute medical unit. The lean body mass is preferentially mobilized for energy during hospitalization.