Mobile cognitive testing captures divergent longitudinal trajectories of verbal learning in adults with and without HIV

Introduction Advanced cognitive aging remains a major concern for people living with HIV (PWH), even in the context of viral suppression. This underscores the need for sensitive tools that can detect subtle cognitive change. Mobile cognitive assessments offer a scalable and ecologically valid approach, yet their sensitivity to longitudinal change in clinical populations is not well established. Methods We examined longitudinal performance and predictors of change on a 14-day mobile Verbal Learning Test (mVLT) administered remotely at baseline and again 12–46 months ( M = 26.7) later in 24 PWH and 13 HIV-negative controls aged 51–74, and compared these trajectories with change on standard in-person neuropsychological testing. Results Aggregate mean mVLT performance improved over time among controls, but this was not evident among PWH (i.e., a significant group X time interaction). In contrast, longitudinal trajectories did not differ by group on the standard in-person Hopkins Verbal Learning Test-Revised, suggesting greater sensitivity of the mobile measure in this sample. Age moderated mVLT trajectories, such that increasing age was associated with worse longitudinal trajectories in PWH, whereas age was unrelated to change in controls. Among PWH, worse mVLT trajectories were associated with higher cerebrovascular risk, lower social functioning, and poorer baseline global learning performance, but not depressive symptoms, HIV disease markers, or other medical comorbidities. Discussion These preliminary findings suggest that the mVLT captures group-level differences in longitudinal learning trajectories and heterogeneity in performance over time among PWH, in line with contemporary models of cognitive aging in HIV. With replication in larger samples, mobile assessments could support scalable monitoring of cognitive function in PWH.