Validation of regression-based change formulae for mild cognitive impairment and Alzheimer's disease.

ABSTRACT

INTRODUCTION: Identification of cognitive decline is critical in older adults at risk for dementia. In a 2020 study reported in Archives of Clinical Neuropsychology, Kiselica and colleagues developed standardized regression-based (SRB) change formulae for the Uniform Data Set 3.0 Neuropsychological Battery in cognitively unimpaired older adults. However, validation of their applicability in impaired individuals is needed. METHODS: Using longitudinal data on 5974 participants (cognitively unimpaired, mild cognitive impairment, dementia) from the National Alzheimer's Coordinating Center, SRB change scores were calculated for each individual and compared across groups. RESULTS: Across 6 to 24 months, minimal cognitive change was observed in cognitively unimpaired participants. Modest declines were seen in those with mild cognitive impairment and substantial declines in those with dementia. Change scores were negatively correlated with the Clinical Dementia Rating scale. In impaired individuals, SRB scores indicated more decline in those with positive amyloid scans. DISCUSSION: Validation of SRB scores affords greater confidence in employing them in clinical and research settings. Highlights Validation of regression-based cognitive change scores in impaired samples.Clear differences on change scores across three groups (intact, MCI, dementia).Largely stable scores in intact participants, but notable decline in MCI and dementia.Moderate to strong relationship between change scores and the Clinical Dementia Rating scale sum of boxes.