Background/objectives: Patients with dementia present with feeding difficulties (FDs) since diagnosis, conditioning their progression. Early identification is vital for preventing deterioration due to nutritional problems. The Edinburgh Feeding Evaluation in Dementia Scale (EdFED) identifies the FDs of patients with dementia by studying their behaviours while eating or being fed. The aim of this study was to assess the responsiveness of the EdFED over time in older people with cognitive impairment and to compare its effectiveness in identifying malnutrition and risk with that of the gold standard Mini Nutritional Assessment (MNA) method.

Methods: This was a multicentre, prospective, observational, longitudinal, analytic study with a follow-up period of 18 months (with patients participating in nursing homes and in the community). Sociodemographic and nutritional data (body mass index (BMI), MNA, forearm circumference (FC), calf circumference (CC), and a nutritional blood test) were collected; EdFED score was reported by nurses, nursing assistants, and family caregivers.

Results: The total sample consisted of 359 individuals-60.7% residential participants and 39.3% community participants. In the last follow-up there were 149 remaining (41.5%). Malnutrition was more than 30%, and the risk was 40% at the three follow-ups. The results suggest that the EdFED scale is a useful tool for assessing feeding difficulties (FDs) in older persons with dementia. It demonstrated good sensitivity and specificity in detecting malnutrition, similar to the MNA, and, more importantly, detecting risk and also identifying changes in nutritional status over time.

Conclusions: The EdFED scale provides a means of evaluating nutritional problems, making it possible to work on prevention.