The impact of stress on cognitive performance in remote smartphone-based assessments: exploring within-person variability

Stress is a well-established risk factor for cognitive dysfunction in domains including processing speed (PS), working memory (WM), and episodic memory (EM). However, this research literature typically uses between-group comparisons. Advances in remote, smartphone-based cognitive assessments allow for examination of within-person variability, or testing how fluctuations in an individual's stress levels influence their cognitive performance. We examined the association between within-person variability in stress and performance on mobile-app assessments of processing speed, working memory, and episodic memory in cognitively unimpaired older adults.

123 cognitively unimpaired older adults (60-81 years old, 66.7% female, 86.2% White, mean education = 16.49 years) rated their stress using a single-item immediately before completing cognitive assessments on the Mobile Monitoring of Cognitive Change (M2C2) smartphone app. Data were gathered three times daily over eight days. Cognitive tasks included measures of PS, visual WM, and EM. Multilevel modeling examined the interaction between within-person stress and session number on cognitive performance.

Higher within-person stress was associated with faster PS (Est. = -3.82, p < .001), though these effects diminished across sessions (session × stress interaction, p = .003). By contrast, stress was not associated with EM or WM performance. Sensitivity analyses using different cumulative stress thresholds produced consistent findings.

Results demonstrate that momentary increases in stress were associated with faster PS, but did not impact WM or EM performances, though these effects changed over repeated testing. These findings highlight the importance of within-person stress variability in shaping cognitive performance on high-frequency smartphone-based assessments.