Malnutrition risk and nutritional interventions among older hospitalised patients with and without dementia: a multicentre nine-year study

OBJECTIVES: To (1) describe the prevalence of malnutrition risk in older hospitalised patients with and without dementia across annual cross-sectional assessments between 2017 and 2025; (2) examine nutritional interventions among older hospitalised patients at malnutrition risk with and without dementia between 2017 and 2025; and (3) identify clinical and demographic factors associated with the co-occurrence of dementia and malnutrition risk. DESIGN: Multicentre, cross-sectional secondary data analysis. SETTING: Austrian acute care hospitals participating in the Nursing Quality Measurement 2.0 (LPZ) between 2017 and 2025. PARTICIPANTS: A total of 9,991 hospitalised patients aged ≥65 years were included; 616 (6.2%) had a documented dementia diagnosis, 2,369 (23.7%) were at risk of malnutrition, and 144 (1.4%) had both dementia and malnutrition risk. MEASUREMENTS: Data collection included a standardized questionnaire using patient interviews, clinical assessments, and review of medical records. Malnutrition risk was assessed using the Malnutrition Universal Screening Tool (MUST). Dementia diagnoses were based on ICD-10 documentation. Thirteen nutritional interventions were recorded dichotomously. Group comparisons and multivariable logistic regression analyses (with co-occurrence of dementia and malnutrition risk according to MUST as the dependent variable) were performed. RESULTS: The overall malnutrition risk was 23.7% and did not differ between patients with and without dementia (23.4% vs. 23.7%, = 0.840). Among patients at malnutrition risk, those with dementia more frequently received mealtime assistance (59.7% vs. 15.3%, p < 0.001), modification of the mealtime environment (57.6% vs. 26.3%, p < 0.001), texture-modified diets (47.9% vs. 16.7%, p < 0.001), enriched diets (47.5% vs. 16.0%, p < 0.001) and oral nutritional supplements (28.5% vs. 19.2%, p = .008) than those without dementia. Between 2017 and 2025, the use of several nutritional interventions varied. The co-occurrence of dementia and malnutrition risk was independently associated with female sex (OR 1.56, CI 1.07-2.29), higher age (OR 1.07 per year, CI 1.04-1.09), end-of-life condition (OR 1.89, CI 1.21-2.95), higher number of medical diagnoses (OR 1.37, CI 1.24-1.50), and greater care dependency (OR 0.95 per CDS point, CI 0.94-0.96) in the multivariable analysis. CONCLUSION: Older hospitalised patients with dementia and malnutrition risk received a broad range of nutritional interventions in routine clinical practice, particularly interventions supporting oral intake. Nevertheless, variation in nutritional care and the comparatively modest use of some guideline-recommended interventions indicate a need for further research on the appropriateness and implementation of nutritional care in this population.