Food intake and mortality in the frail elderly

Background. Adequate qualitative and quantitative food intake is a major determinant of health. However, nutritional requirements in the elderly are unknown, and even more so in the frail elderly. The aim of the study was to evaluate the influence of energy and macro-nutrients on health in the frail nursing home elderly. Methods. Food intake of 72 not severely diseased elderly patients was assessed with direct weighing method. Outcome measure was survival over 28-month follow-up period. Confounders of the association of food intake with survival were: age, gender, body-mass index, daily function, somatic health, energy, and nutritional status. Crude association of food intake with survival was assessed with Kaplan-Meyer method, and adjusted association with multiple Cox regression models. Results. Patients of the study had good average food intake. Mortality rate was relatively low (.20 per year). Low levels of energy, protein, lipid, and carbohydrate intake were negatively associated with survival even after adjustment for confounders. When compared to high intake, adjusted relative risks for mortality of low intake were 4.74, 3.75, 4.71, and 2.04, respectively. Medium levels of energy, protein, and lipid, but not carbohydrate, intake yielded intermediate mortality risk. Conclusions. Food intake is a strong predictor of survival even in moderately diseased elderly patients, suggesting possible low-cost interventions.