MMSE cross-domain variability predicts cognitive decline in centenarians.

<p><b>BACKGROUND: </b>Early detection of dementia is one of the key issues in cognitive gerontology. However, so far the detection of early stages in cognitive decline has been rather unreliable. One central limitation of current assessment strategies is that they rely on information about a person's level of performance obtained from a single assessment.</p><p><b>OBJECTIVE: </b>In the first part of the present paper, we propose three strategies for overcoming this limitation by using information from several measurement occasions to improve diagnostic reliability.</p><p><b>METHODS: </b>In the second part, we present one exemplary empirical analysis of one of these approaches reporting data from the Heidelberg Centenarian Study.</p><p><b>RESULTS: </b>In this sample of very old individuals tested at baseline and at a follow-up 18 months later, the intra-individual cross-domain variability across MMSE subscales [1]at baseline predicted cognitive decline (i.e. change in an external cognitive functioning rating from baseline to follow-up) much better than baseline mean MMSE scores.</p><p><b>CONCLUSIONS: </b>In conclusion, we discuss the advantage of performance variability as a predictor for cognitive decline.</p>