Demographically unadjusted cognitive test scores may enhance the validity of mild cognitive impairment as a dementia prodrome.

Abstract

Mild cognitive impairment (MCI) precedes most forms of neurodegenerative dementias and is considered an important time window for treatment initiation. It is however unclear whether the widespread use of demographically adjusted norms improve detection of dementia-related impairment and identification of individuals at risk for dementia. In the current multicenter study (n = 561), we compared three Petersen/Winblad criteria-based MCI diagnoses with varying levels of demographic adjustments, in relation to their associations with biomarkers for Alzheimer's disease and neurodegeneration, as well as risk of conversion to future dementia. Our MCI diagnoses were (1) demographically unadjusted; (2) adjusted for sex and education; or (3) adjusted for age, sex, and education. In our cohort, we found the differences between the models to be small, albeit consistent, and our results show a tendency for an unadjusted MCI diagnosis to be more strongly associated to dementia biomarkers and conversion to dementia. Our results thus suggest that an MCI diagnosis based on unadjusted test scores may provide a more valid diagnosis of prodromal dementia, while no evidence is provided to support this in other contexts. This goes against the recommendations from the major diagnostic classification systems, indicating the need for further research to explore these relationships.