Enhancing Interpretation of Cognitive Changes in Alzheimer's Intervention Trials: Application of Standardized Regression-Based Change Scores.

Abstract

Objective: Interpreting results from intervention trials for cognition in Alzheimer's disease (AD) can be complicated due to comparison to an active control group. In some cases, there are improvements in cognition across both conditions, but no between-condition differences. These results might indicate that both interventions positively impact cognition; conversely, improvements might be attributable to other factors, such as practice effects. One way to differentiate between these possibilities is to use standardized regression-based (SRB) change indices. We present an example of applying this technique to understand equivocal behavioral intervention trial results.

Method: 214 participants (ages: 45-75) were randomly assigned to a 2-year health coaching or health education intervention to reduce AD risk and improve cognition.

Results: Linear mixed-effects models showed improvements in cognitive domains over time, without a differential treatment effect (p < 0.001). SRB change analyses showed that improvements on the Repeatable Battery for Assessment of Neuropsychological Status were likely not meaningful, with ≤5 participants demonstrating significant changes over time beyond what would be expected with no intervention.

Conclusions: SRB indices provide a nuanced interpretation of results from behavioral intervention trials for improving cognition, which may translate into improved patient care decisions.