Objectives: To examine whether the Mini Nutritional Assessment-Short Form (MNA-SF) score and its individual items are predictors of mortality in a nursing home population.
Design: Prospective, secondary analysis from the Incidence of pNeumonia and related ConseqUences in nursing home Residents (INCUR) study with 1-year follow-up.
Participants: A total of 773 older persons (women 74.4%) living in 13 French nursing homes.
Measurements: At baseline, nutritional status was assessed with the MNA-SF. Overall mortality rate was measured over a 12-month follow-up period after the baseline assessment visit. Cox proportional hazard models were performed to test the predictive capacity of the MNA-SF score and its single components for mortality.
Results: Mean age of participants was 86.2 (standard deviation, SD 7.5) years. Mean MNA-SF score was 9.8 (SD 2.4). Among participants, 198 (25.6%) presented a normal nutritional status (12-14 points), 454 (58.7%) were at risk of malnutrition (8-11 points), and 121 (15.7%) were malnourished. After one year of follow-up, 135 (17.5%) participants had died. Age, female gender, baseline weight, BMI and MNA-SF were significant predictors of mortality whereas no specific chronic disease was. The total MNA-SF score was a significant predictor of mortality (Hazard Ratio=0.83; 95% CI 0.75-0.91; p<0.001), even after adjustment for potential confounders. Four individual items: weight loss, decrease in food intake, recent stress and BMI were independent predictors of mortality.
Conclusions: The MNA-SF appears to be an accurate predictor of one-year mortality in nursing home residents. Thus, this tool may be regarded not only as a nutritional screening tool, but also as an instrument for identifying the most-at-risk individuals in this population.