Abstract
Cognitive composites combine scores from multiple neuropsychological tests and demonstrate greater sensitivity to Alzheimer's disease (AD) related cognitive changes than individual tests. This review examines the development, composition and validity of cognitive composites to detect AD-related cognitive changes. The included cognitive composites were evaluated using four criteria: cognitive domains assessed; neuropsychological tests used, the inclusion of non-neuropsychological measures (e.g., Mini-Mental State Examination) that produce a global score themselves; and statistical methods used to calculate the composite. Existing composites fall into two categories: domain-specific (e.g., episodic memory, executive function, attention) or general composites combining multiple cognitive domains while incorporating clinical and functional measures. Psychometric properties were not consistently reported across all the studies. Therefore, a standardized validation framework is proposed to address these inconsistencies. Future work should focus on systematically evaluating optimized weighting, including data from clinical and functional measures and consistent psychometric reporting for assessing and monitoring cognitive dysfunction in early AD.