The main objective of the current study was to characterize malnourished and undernourished elderly people admitted to an acute care ward and particularly to identify risk factors that could be the target for future intervention programs. Based on the MNA evaluation, we showed that 81.5% of the participants of this study were at risk for malnutrition and 18.5% were malnourished. Malnourished participants were less educated, had more depressive symptoms, and lower cognitive and functional status compared with participants at nutritional risk. Moreover, malnourished participants had fewer social contacts including visits and phone calls. The following problems were found to have a significantly higher prevalence among malnourished patients: chewing problems, nausea, and vomiting. These findings indicate the severe impact of these factors on the development of actual malnutrition.
In a study that was conducted among subacute care patients in St. Louis, the prevalence of undernutrition was evaluated, using the MNA. Among 837 patients consecutively admitted during 14 month, the prevalence of malnutrition was 28.8% and 62.5% were at 'nutritional risk' [19
]. In another group of very old hospitalized patients (mean age 84.8 ± 8.1 y), 33.2% were at risk for malnutrition and 49.4% were malnourished [21
]. In institutionalized women in Spain the prevalence of malnutrition was 7.9% and 61.8% were at risk for malnutrition [38
]. In our group, since we used participants who were already screened for nutritional risk and malnutrition the rates were different although it is quite clear that the rate of malnutrition is relatively low. The differences observed may reflect the type of elderly people being screened in each study.
The MNA is a dietary assessment tool that was validated in many different populations [11
] and was shown to be related to several outcomes including mortality, length of hospitalization and complications [19
]. In a study that assessed the impact of nutritional status measured by the MNA on pressure sores, the MNA provided advantages over using visceral proteins in screening [20
]. In our study the laboratory measurements were not related to the MNA results except for serum albumin which was slightly lower among the malnourished group. It is likely that poor nutrition takes considerable lag time until it is manifested in laboratory measurements. The decline of serum albumin is certainly a late phenomenon in terms of malnutrition. Additionally, albumin is a negative acute phase reactant which would likely be diminished in many hospitalized patients who do not suffer from malnutrition. Therefore its futility as an indicator of nutritional status is limited in this scenario.
Among the demographic parameters, we used country of origin as an important parameter. Israel is a multiethnic country with ongoing waves of immigration from various countries: 63.7% of the study population immigrated from Europe/America. The highest percent of malnourished participants immigrated from Africa/Asia. It is our assumption that the high prevalence of malnutrition in this population may stem from the poor living conditions and lower socioeconomic status (SES) highly prevalent in this population throughout their first years in Israel. Therefore, their retirement income is, on average, lower than people who emigrated from European/American countries. Data from the Central Bureau of Statistics indicate that older adults who emigrated from European/American countries are more educated and their retirement income is higher, compared with immigrants from African and Asian countries [39
Over 64% of the malnourished participants were widowed and over 39.6% were living alone. Marital status and social isolation, especially when combined with recent bereavement or poor social support, have been shown in previous studies to be major risk factors for malnutrition [40
]. In a case control study comparing health and nutritional values between widowed and married participants, Rosenbloom [42
] described reported lack of appetite as an important parameter associated with depression and weight loss in widowed elderly people. In another case control study [40
], widowed community dwelling participants lost significantly more weight compared with a control married group. Poor appetite was a significant risk factor for nutritional deterioration [40
Poor appetite is an important risk factor for nutritional risk. Payette et al. [43
], who evaluated a community living elderly population, showed that reported good appetite appeared to be a significant predictor for dietary intake of calories (p < .01) and protein (p < .05). In a previous study by Shahar et al. [44
], approximately 20% of the participants reported lack of appetite, or high frequency of feeling no wish to eat. These subjects had lower energy intake as well as lower intake of other nutrients, and thus were considered at risk for nutritional deterioration. In our study, severe loss of appetite was associated with malnutrition. Among the malnourished group, the prevalence of severe loss of appetite was significantly higher (16.7% vs. 8.5%).
The dietary assessment part of the MNA includes questions regarding protein, vegetable and fruit intake, appetite, fluid intake, and difficulties in eating. This part of the assessment has the highest sensitivity and specificity as indicated in the ROC curve. This further highlights the importance of dietary assessment as a mean of detecting nutritional risk.
Depression is the most common cause of unintentional weight loss and under-nutrition in older adults [45
]. Depression in the elderly is a frequent, treatable, but under-recognized and under-treated, disorder. Patients with depressive symptoms are not identified and thus are seldom treated for this condition [47
In a retrospective chart review to determine the cause of weight loss in nursing home residents, Morley and Kraenzle [48
] also concluded that depression was the most common cause for weight loss. In our study the average number of depressive symptoms was significantly higher among the malnourished participants. We also found that in the malnourished group 81.3% reported weight loss compared to 50.7% among the at risk group.
Nutritional risk is related to functional status [48
]. Our results indicate that the malnourished group suffered from more functional disabilities according to the Barthel Index and had a higher prevalence of impaired mobility. This observation, however, does not provide conclusive evidence regarding the causal relationship between ADL dependency and malnutrition, since each of these may be the cause of the other.
We did not find any difference in health status between the groups as measured by number of prescribed medications, number of hospitalizations during the year prior to the study, number of family or specialist physician visits, and duration of hospitalization, between the malnourished and at risk groups. However, subjective health evaluation compared with peers was significantly poorer among the malnourished group. The difference in subjective health evaluation may indicate a difference in severity of the disease. Subjective health evaluation in the elderly is considered one of the most accurate measures of health status; its association with malnutrition indicates a close relationship between health and nutritional status.
The malnourished group suffered significantly more from chewing problems, vomiting, and nausea compared to the at risk group. Eating problems and their relation to nutritional status clearly revealed the importance of identifying special problems related to eating and digestion. Mowe et al. showed that chewing problems can lead to a reduced dietary intake and thus to poor nutritional status [49
]. Therefore, these problems need to be given closer attention in patient care because of their cumulative effect on dietary intake. Earlier identification of these risk factors may allow a more efficacious intervention which may prevent actual malnutrition from occurring.
Our study suffers from several limitations. The study examined a selected population of hospitalized elderly patients at risk for malnutrition; however, characterization of these groups and the distinction between the levels of undernutrition is important for developing targeted interventions. In addition, the study is a cross-sectional survey and thus cannot serve to determine temporal relationships.
Our study evaluated the association between in-hospital malnutrition and several risk parameters. Of all the parameters studied, the difficulty in consuming foods was found to be highly associated with the development of malnutrition. The most important predictors of actual malnutrition in these patients were lower education, poorer cognitive status, and chewing problems. At least some of these parameters are amenable to pharmacological and non-pharmacological treatment modalities. Therefore, given the critical importance of nutritional status in the hospitalized elderly and its impact on mortality and morbidity [19
], an emphasis should be placed on correcting these problems. An example of such interventions may be withholding medications, performing speech therapy evaluation, or naturally changing food texture and constituents. We feel that our findings highlight the need for a nutritional intervention trial among at risk and malnourished hospitalized patients.