Nutritional screening in the elderly
Associate Professor Julie Byles, Dr Lynne Parkinson, Dr Claire Collins (Hunter Area Health), Professor Kichu Nair, Dr Michael Dibley, Associate Professor Catherine D'Este, Dr Barbara Blades, Associate Professor Manohar Garg
Older people are at high risk of under nutrition due to age-related decrease in digestive function, reduced taste sensation, use of medications, reduced mobility and ability for self-care, and depleted social and economic resources. It is estimated that up to 30% of Australian people over 60 living independently in the community may be suffering from some degree of malnutrition, and rates are higher among institutionalised and hospitalised elderly. In 2002, a pilot study we conducted involving a random sample of 123 hospital in-patients aged 65 and over found that 37% of these were mildly or moderately malnourished and 11% were severely malnourished according dietitian assessment using the Subject Global Assessment.
Nutrition screening requires a tool that is accurate, quick, easy to implement and reliably classifies individuals with poor nutrition who are at risk of poor health outcomes. Currently, four potential tools are available, however, their reliability and validity have yet been determined. Our pilot study tested three of these tools: The Australian Nutritional Screening Initiative (ANSI), Malnutrition Screening Tool (MST),and the Nutrition Risk Assessment Tool (NRAT).
In our pilot study 86% were at risk of being malnourished according to the ANSI, 42% according MST, and 50% according to NRAT. However, analysis of the items shows poor discrimination for some variables, which limits the validity of these tools for use in hospital.
Other findings from this pilot study suggest that biochemical markers may have more discriminatory utility than previously considered. Moreover, some of these markers are routinely collected on admission and may be readily incorporated into a simple screening algorithm. We hypothesise that different subsets of screening questions, combined with simple biochemical markers, may demonstrate stronger correlations and better sensitivity and specificity, but a larger sample size is required to test this hypothesis.
The next phase of this project has been funded by the University of Newcastle Research Grants Committee. The aims of this new phase are:
- To identify the most parsimonious set of items and measures that can be used to accurately screen those classified as malnourished on the SGA (Subjective Global Assessment) and MNA (Mini Nutritional Assesssment);
- To explore the paradoxical relationship of red cell folate levels and other nutritional markers, and to identify the extent to which this can be explained by diet and past medical history;
- The findings will provide a much needed basis for targeting nutritional assessment and services in a growing population of older people, and may provide a basis for developing new dietary products for older people and other health promotion strategies.