A ROLE OF MNA-SF AS A PREDICTOR FOR 30-MONTH MORTALITY IN A NURSING HOME IN JAPAN

Background: Japan is a super-aged society. Along with the aging, the number of dependent elderly has increased. The Mini Nutritional Assessment-Short Form (MNA-SF) has been reported to be one of the useful questionnaires to assess nutritional status in older people. However, studies reporting with MNA-SF for the dependent elderly are lacking. Therefore, we observed residents in nursing homes in Japan for 30 months to examine whether MNA-SF is useful for assessment of nutritional status. Methods: A total of 423 residents (333 women; 84.2 ± 8.6 years) in nursing homes in Japan participated in this investigation. We investigated the following characteristics of the participants: age, sex, height, weight, medical history, Barthel index, clinical dementia rating, and MNA-SF. MNA-SF consists of 6 items: food intake decline, weight loss, mobility, suffering from psychological stress or acute disease, Neuropsychological problems, and BMI. We performed Cox proportional regression analysis to examine the association of MNA-SF score with 30-month mortality. Results: Mean MNA-SF score was 8.9 ± 2.2. Among the participants, 104 (24.6%) were malnourished, 250 (59.1%) were at a risk of malnutrition, and 49 (11.6%) were in a normal nutritional status. After 30-month follow-up, 166 (39.2%) of the participants died. After adjustment for confounders, MNA-SF score was significantly associated with 30-month mortality (HR: 0.843, 95% CI: 0.770–0.922). Conclusion: MNA-SF score was significantly associated with 30-month mortality in Japanese nursing homes, even after adjustment for mortality-associated confounders. Therefore, the approach based on MNA-SF can be effective for nutritional assessment of dependent elderly.