Predicting cognitive change within domains.

Area of Science:

• Neuroscience
• Psychology
• Gerontology

Background:

• Standardized regression based (SRB) formulas traditionally predict cognitive change using the same neuropsychological measure at baseline and follow-up.
• Situations like provider changes or practice effects necessitate alternative prediction methods when identical tests are unavailable.

Purpose of the Study:

• To expand SRB methodology by developing formulas to predict performance on different tests within the same cognitive domain.
• To assess the efficacy of using baseline scores to predict 1-year follow-up performance on both identical and alternate cognitive tests.

Main Methods:

• Developed two sets of SRBs: one for predicting performance on the same test, and another for predicting performance on a different test within the same cognitive domain (learning/memory, processing speed, language).
• Utilized data from 127 non-demented community-dwelling older adults assessed at baseline and after 1 year.

Main Results:

• Baseline scores significantly predicted 1-year follow-up scores across both sets of SRBs, particularly for learning and memory and processing speed.
• SRBs developed for different tests within the same domain showed comparable predictive power to those for the same test, though less variance was explained as test similarity decreased.

Conclusions:

• Preliminary findings support the development of SRBs for predicting cognitive change using different tests within the same domain.
• This domain-based approach offers a flexible alternative for cognitive change prediction when identical measures are not feasible.