Beyond NACC's Uniform Data Set for cognition: the impact of additional items on measurement precision

We sought to determine the implications of integrating additional cognitive test items beyond the items of the Uniform Data Set (UDS) administered by Alzheimer's Disease Research Centers (ADRCs).

We considered the UDS 1&2 and UDS 3 cognitive batteries alone ("UDS-only") and compared them to batteries augmented with additional items ("UDS-augmented") administered by the University of Pittsburgh, ADRC. We used confirmatory factor analyses to co-calibrate and harmonize three cognitive domains (memory, executive functioning, and language). Then we compared the UDS-only and UDS-augmented with respect to measurement precision, sample sizes required to detect 25% differences in rates of decline, and association between baseline scores and the hazard of converting from mild cognitive impairment to AD dementia.

UDS-augmented substantially enhanced measurement precision across all cognitive domains for UDS 1&2 and UDS 3.

There is substantial value in integrating item content beyond the UDS.

We considered items from the UDS 1&2 and UDS 3 cognitive batteries alongside augmented site-specific content for the memory, executive functioning, and language domains. Augmented scales improved measurement precision in each case. As expected, improved measurement precision facilitated better performance for each scale. It would be useful to integrate multiple sources of cognitive data into comprehensive estimates of domain performance.