METHODS: Using Alzheimer\'s Disease Neuroimaging Initiative (ADNI)-1/GO data, signal-to-noise ratios (SNRs) were calculated to quantify the sensitivity of a measure for detecting disease progression and hypothetical treatment effects. All possible combinations of selected sensitive measures were assessed for developing composite scores. The analyses were performed in the MCI population and subpopulations enriched by apolipoprotein E4 (APOE ε4), hippocampal volume, and cerebrospinal fluid β-amyloid.
RESULTS: The best composite score was \"Word Recall + Delayed Word Recall + Orientation + CDR-SB + FAQ\", more sensitive than 13-item Alzheimer\'s Disease Assessment Scale-cognitive subscale or CDR-SB.
CONCLUSIONS: The proposed composite score derived from the existing clinical endpoints demonstrated higher sensitivity in the MCI population and is easy to implement and standardize across studies.