Objective: The aim of this study was to determine whether a modified version of the Mini Nutritional Assessment (MNA) without body mass index (BMI) can effectively identify individuals at risk of malnutrition among patients with neuropsychiatric disorders. The study also compares two versions with each other.
Methods: The study involved 105 patients in the acute phase of confirmed neuropsychiatric disorders in a regional hospital in central Taiwan. All subjects were cognitively able to have effective verbal
communication. The study included serum biochemical and anthropometric measurements, and an on- site, in-person interview using a questionnaire survey to elicit participants' sociodemographic status, anthropometric measurements, healthcare-related biochemical indicators and answers to the MNA.
Subjects' nutritional statuses were graded with a version of MNA that adopted population-specific anthropometric cut-points or one further with the BMI question removed and its assigned score redistributed to other anthropometric questions. Results were analyzed with Pearson's Correlation
Analysis, and with multiple regression analysis to determine the association of the nutritional status with the major functional indicators.
Results: The two versions identified the same (7.6%) patients malnourished. The MNA-TI predicted 21.9%
and the MNA-TII predicted 20.0% patients at risk of malnutrition. Post hoc analysis indicated that results with the original version were different from that predicted with the two modified versions whereas the patterns predicted by the two modified versions were not different from each other. Among the
parameters examined, serum albumin concentrations showed the weakest correlations with MNA scores.
The total MNA scores of both modified versions were very highly significantly correlated with subjects' BMI, duration of psychiatric disease, number of prescribed drugs, self-assessed appetite status, stress status and % weight loss. Total MNA scores were also significantly correlated with serum albumin concentrations.
Both versions of the MNA had comparable r values. The MNA-TII version shows better or at least equal functionality compared to MNA-TI in predicting the nutritional status of neuropsychiatric patients.
Conclusion: Adoption of population-specific anthropometric cut-points improves the functionality of the MNA in neuropsychiatric patients, a non-Caucasian population, and the improved functionality is maintained in a version without BMI (with adjusted MAC and CC scores). The MNA is a valuable and effective tool for identifying individuals at risk of malnutrition among neuropsychiatric patients. A MNA without BMI has greater applicability and can enhance professional efficiency of healthcare workers.
