Predictors of incident malnutrition—a nutritionDay analysis in 11,923 nursing home residents

BACKGROUND/OBJECTIVES: Malnutrition (MN) in nursing home (NH) residents is associated with poor outcome. To identify those at high risk of incident MN, predictors are needed. We investigated predictors of incident MN in older NH residents. SUBJECTS/METHODS: NH residents from the nutritionDay (nD) project (2007–2018), aged ≥65 years, with complete nutritional status data at nD and 6-month follow-up (FU) and without MN at baseline were included. Associations of 17 variables (general characteristics, function, nutrition, diseases, medication) with incident MN (weight loss ≥10% between nD and FU or BMI <20 kg/m² at FU) were analyzed using generalized estimating equations (GEE). Variables with p<0.1 in univariate models entered multivariate GEE. Effects are odds ratios (OR) with 99.5% confidence intervals (CI). RESULTS: Among 11,923 non-malnourished residents, 10.5% developed MN at FU. Predictors were no intake at lunch (OR 2.79 [1.56–4.98]), a quarter (2.15 [1.56–2.97]) or half (1.72 [1.40–2.11]) eaten (vs three-quarters to all), lowest BMI quartile 20.0–22.9 (1.86 [1.44–2.40]) vs highest (≥29.1), age 85–94 (1.46 [1.05–2.03]) vs 65–74, severe cognitive impairment (1.38 [1.04–1.84]) vs none, and immobility (1.28 [1.00–1.62]) vs mobile. CONCLUSION: 10.5% of non-malnourished NH residents develop MN within 6 months. High-risk groups include those with poor meal intake, low BMI, severe cognitive impairment, immobility, and older age.