Data-driven cognitive composites track Aβ-related decline and predict MCI: ADOPIC study

Background

Cognitive composites are expected to detect amyloid beta (Aβ)‑related decline more sensitively than individual tests¹. Researchers developed and validated data‑driven composites from the harmonized ADOPIC study, which tracks the onset and progression of Alzheimer's dementia across international cohorts¹.

Methods

• The study enrolled 1,964 cognitively unimpaired participants at baseline, with 1,110 of them having longitudinal follow‑up¹.
• Domain‑specific and general composites were created using principal component analysis¹.
• They were evaluated for psychometric properties, sensitivity to (Aβ)‑related decline, and prediction of mild cognitive impairment, in comparison with the ADCS‑PACC¹.

Results

• Psychometric properties ranged from good to excellent, and the composites separated Aβ trajectories in CU participants (p < 0.001)¹.
• General cognition showed longitudinal sensitivity comparable to the ADCS‑PACC (d = -0.20) and similar classification accuracy, with an area under the receiver operating characteristic curve of 0.70¹.
• All composites significantly predicted progression to mild cognitive impairment (p < 0.001)¹.

Interpretation

The authors report that the composites perform comparably to the ADCS‑PACC in longitudinal sensitivity, while also offering comprehensive psychometric characterization and domain‑specific measurement¹. They view this as support for using these composites as outcome measures in preclinical Alzheimer's disease research¹.