Smartphone-Based Measurement of Executive Function in Older Adults with and without HIV

Objective: To examine feasibility, convergent validity and biases associated with a mobile color-word interference test (mCWIT) among older persons living with HIV (PLHIV). Method: Over a 14-day period, 58 PLHIV and 32 HIV-uninfected individuals (aged 50-74) completed the mCWIT on smartphones once per day in real-world settings. Participants also completed a comprehensive laboratory-based neuropsychological evaluation. Results: A high rate of compliance was observed (86%) in the repeated administration of the mCWIT. A practice effect was observed in the overall sample concerning mCWIT subscores, and these learning effects were greater for PLHIV. Stabilization of performance was observed after 6 (HIV+) and 7 days (HIV-) for completion time and after 2 (HIV-) and 3 days (HIV+) for mCWIT errors. A minor fatigue effect was observed in the overall sample which was unassociated with group status. Moderate to strong correlations were found between mCWIT completion time and mCWIT errors with global neurocognition and with all of the individual neurocognitive domains. The strongest associations were with mCWIT completion time and laboratory-based global neurocognition, executive function, and working memory scores. Conclusions: Cognitive testing administered within the context of a person's daily life provides qualitatively different data than neuropsychological testing completed in clinical settings, and it may constitute a more ecologically valid indicator of cognitive performance than traditional methods. Mobile cognitive testing has potential to help characterize real-time cognitive performance and serve as a complementary assessment tool to traditional methods.