In individuals with only mild or very mild cognitive attenuations (i.e., so-called pre-clinical AD), performance-based measures of function may be superior to informant-based measures because of increased sensitivity, greater reliability, and fewer ceiling effects.

We sought to determine if a performance-based measure of everyday function would demonstrate adequate psychometric properties and validity in the context of serial assessment over a one-year period in patients with Mild Cognitive Impairment (MCI) and early stage Alzheimer’s disease (AD).

Participants were assessed with the performance-based measure at baseline, six weeks, and one year.

A specialized center for the assessment and treatment of AD.

Three groups of subjects participated: a healthy subjects (HS) older cognitively intact group (N=43), an MCI group (N=20), and an AD group (N=26).

A three subtest short form of the UCSD Performance-Based Skills Assessment (UPSA) (called the UPSA-3) was the measure of interest. It consisted of the Communication, Planning, and Finance subtests. RESULTS: Mixed model repeated measures were used to assess performance over time. Large group effects were present (HS>MCI>AD). Additionally, the AD and MCI groups demonstrated declines over one year, while the HS group remained stable (group x time interaction p=.11). The MCI/AD group demonstrated adequate test-retest reliability and did not demonstrate ceiling or floor effects.

Our data indicate that the UPSA-3 is suitable for clinical trials in that it has adequate ecological coverage and reasonable psychometric properties, and perhaps most importantly, demonstrates validity in serial assessments.