Quantifying placebo responses in clinical evaluation of neuropsychiatric symptoms in Alzheimer's disease.

PURPOSE

This study aimed to establish a non-linear mixed effects model to quantitatively analyze the placebo responses of neuropsychiatric symptoms (NPS) in Alzheimer's disease (AD).

METHODS

A comprehensive literature search was conducted in public databases. Placebo-controlled randomized AD clinical trials using the neuropsychiatric inventory (NPI) score as the primary or secondary outcome were included. Non-linear mixed effects model was used to describe the time course of the placebo responses of NPS in AD clinical trials. Potential affecting factors were tested as covariates.

RESULTS

A total of 32 clinical studies (involving 3942 subjects) were included in model-based analysis. We found that the maximal placebo responses of NPS were reached at week 4 approximately, after which rebound effects appeared. The baseline NPI score had a significant impact on the placebo responses. Higher baseline NPI score tended to cause greater reductions in NPI score at week 8 and a smaller degree of rebound. For AD patients whose normalized baseline NPI score was 10 points and 30 points, the reduction in normalized NPI score at week 8 was estimated to be 0.83 and 7.43 points, respectively; and the rebound rate after week 8 was estimated to be 0.1 points/week and 0.08 points/week, respectively.

CONCLUSIONS

The duration of 4 weeks is sufficient to determine the drug efficacy for assessing NPS in AD clinical trials. The baseline NPI score was a key factor associated with placebo responses of NPS, which should be considered when designing future clinical trials and conducting comparisons across trials.