Risk of malnutrition in retirement homes elderly persons measured by the "mini-nutritional assessment".

BACKGROUND: The combined influence of age-associated factors such as general health, degree of dependency, diminished odor perception, and poor oral health on the risk for malnutrition was explored.

METHODS: A total of 81 persons living in retirement homes took part in the study (mean age 83.4 years, SD = 6.6, range 61-98). The Mini-Nutritional Assessment (MNA) was used to evaluate the risk of malnutrition. Odor perception was measured by the detection threshold for isoamylacetate. The number of drugs taken by each person was counted. General health status was determined by the Medical Outcome Study (MOS) scores. Oral examinations were carried out to count the number of natural teeth and type of dentures.

RESULTS: On average, women had slightly, but significantly, lower MNA scores than men (respectively, 23.4, SD = 2.8; and 24.6, SD = 2.6; p = .048). The correlations between age and MNA score and between odor perception and MNA score were not significant. Significant correlations were found between age and number of natural teeth (r = -.26, p = .001) and between MNA score and number of natural teeth (r = .27, p = .001). The mean MNA score of complete denture wearers (22.8, SD = 2.9) was significantly lower than that of partial denture wearers (25.8, SD = 2.9; p = .0005). The total MOS and MNA scores were not correlated, but a significant correlation was found with the subscales mental functioning (r = .29, p = .003), social functioning (r = . 19, p = .045), and perceived health (r = .19, p = .047). No relation was found between the activities of daily living (ADL) and MNA scores. A significant negative correlation was observed between number of drugs taken and the MNA score (r = -.34, p = .001). When participants without risk of malnutrition (MNA > or = 24) were compared with those at risk (MNA = 17-23.5), again, the number of drugs taken was significantly different (on average, respectively, 4.5, SD = 2.9; and 7.0, SD = 2.6; p < .0005). Using multiple regression to test the separate effects of the different independent variables, the number of drugs taken showed a significant negative regression coefficient (beta = -.31, p = .008), as did the mental health score (beta = .27, p =.02), giving a total R2 = .32. The other parameters did not contribute significantly.

CONCLUSION: Among the elderly in retirement homes, the health state (as measured by the MOS subscale mental health and by the medication use) appears to be the most clinically relevant parameter to explain the risk for malnutrition. Loss of natural teeth and perceived health are less independently contributing, whereas no contribution derives from decline of odor perception, degree of dependency, and age itself.