Smartphone-based cognitive assessment in older adults with depression: Feasibility and task performance using...

This study evaluated the feasibility of smartphone-based ecological momentary assessment (EMA) to test middle-aged and older adults' cognition - via tasks of processing speed and cognitive control -with and without depression history.

Participants aged 55-79 (n = 79; 31 with depression history, 48 never-depressed controls) completed a 14-day EMA protocol with three daily assessments. The relevant tasks were Symbol Match Task (SMT; processing speed) and Go/No-Go task (GNG; cognitive control). Feasibility was evaluated through completion rates and missingness. Generalized additive modeling was used to evaluate associations between depression and cognitive performance, including trajectories, controlling for age, sex, and education.

Of the 3318 possible response sessions, 2181 GNG and 2311 SMT were completed, nearly 70%. Controls demonstrated higher daily completion (GNG: 2.15 vs. 1.69 sessions, p = 0.035) and lower missingness (28% vs. 44%, p = 0.035) than depressed participants. Across tasks, depression history was associated with slower response times, less accuracy, and reduced efficiency. Controls averaged 2071 ms on SMT versus 2342 ms for depressed participants. Both groups showed performance improvements over time, suggesting learning effects, with parallel trajectories indicating a model consistent with depression's stable impact on cognitive functioning.

While adherence challenges do exist, especially for those with a history of depression, smartphone-based EMA appears feasible for cognitive assessment in older adults. Depression impacts processing speed and cognitive control in this these real-world settings, and deficits persist regardless of current presence of symptoms. These findings support EMA's ecological validity for evaluating cognitive function and its potential for informing targeted interventions in late-life depression.