BACKGROUND: The relationship between cognitive and functional impairment in Alzheimer’s disease (AD) is not well characterized. Recent analyses suggest that cognitive decline precedes and predicts functional decline throughout the progression of AD. The objective of this study was to better understand the relationship between cognitive decline and functional decline in mild AD using autoregressive cross-lagged panel analyses in four different multicenter Phase 3 clinical trials.

METHODS: Data from placebo patients with mild AD (Mini-Mental State Examination score 20 to 26) were pooled from two multicenter, double-blind, Phase 3 solanezumab studies (EXPEDITION and EXPEDITION2) and separately for two Phase 3 semagacestat studies (IDENTITY and IDENTITY2). Cognitive and functional outcome measures were assessed at baseline and 6 post-baseline time points every three months for 18 months using the AD Assessment Scale-Cognitive subscale (ADAS-Cog14) and the AD Cooperative Study-Activities of Daily Living instrumental subscale (ADCS-iADL). Patients who completed all scheduled assessments were included in analyses. Auto-regressive cross-lagged panel analyses were used to evaluate the relationships between cognitive and functional impairment over time, and potential causal relationships between them. Overall model fit was assessed using Comparative Fit Index (CFI) and Root Mean Square Error of Approximation (RMSEA). A model with CFI of 0.9 or greater and RMSEA of 0.06 or less is considered acceptable.

RESULTS: In EXPEDITION/EXPEDITION2, the results from the cross-lagged panel analyses demonstrated cognitive impairment significantly predicts future functional impairment in 5 of 6 time points (CFI=0.993, RMSEA=0.045). Results showed that functional scores predicted cognitive outcome in only 1 of 6 time points when the same analyses were performed to test the opposite hypothesis. The data from IDENTITY/IDENTITY2 were more limited due to early termination of trials and a smaller sample size, but the analyses yielded consistent results that cognition predicted subsequent function (CFI=0.998, RMSEA=0.025).

CONCLUSIONS: The analyses from all 4 clinical trials indicated that cognitive decline precedes and predicts subsequent functional decline. These findings are consistent with previously proposed hypotheses and with recent publications using similar methodologies. Cognition, in this research measured by ADAS-Cog14, can be used as a predictor for future functional impairment in mild AD.