UC San Diego

Rationale: Heavy alcohol use is prevalent among people with HIV (PWH) and increases risk for neurocognitive and everyday functioning impairments. Although self-reports of cognitive difficulties are often used clinically to screen for neurocognitive impairment, such retrospective measures are subject to recall error and response bias. Thus, in a sample of PWH who were heavy alcohol drinkers, this study aimed to: 1) evaluate psychometric properties of real-time self-reported cognitive difficulties assessed via smartphone-based ecological momentary assessment (EMA), and 2) examine temporal relationships among EMA self-reported cognitive difficulties, alcohol use, mood, and daily activities.Design: Participants were 23 PWH recruited from existing studies at the HIV Neurobehavioral Research Program who reported current heavy alcohol use. Participants completed two in-person visits separated by a 14-day EMA monitoring period with up to four surveys per day. Objective neurocognition was measured in person by the NIH Toolbox Cognition Battery. Multiple regressions examined whether the proportion of surveys on which participants reported cognitive difficulties related to objective neurocognitive functioning. Mixed effects logistic regressions examined whether EMA-reported alcohol use, depressive symptoms, and cognitively demanding activities related to concurrent EMA-reported cognitive difficulties. Results: Participants were 83% adherent to the EMA surveys on average. Higher proportions of surveys reporting cognitive difficulties were significantly associated with worse objective neurocognitive functioning (p = 0.040); however, EMA-reported real-time cognitive difficulties were not significantly more sensitive or specific in identifying objective neurocognitive impairment compared to an in-person retrospective measure (ps > 0.05). Greater EMA-reported alcohol use (OR = 1.37; p < 0.001) and depressive symptoms (OR = 1.80; p = 0.016) were significantly related to higher likelihood of concurrent cognitive difficulties within persons. EMA-reported engagement in cognitively demanding activities was related to a lower concurrent likelihood of attention difficulties within persons (OR = 0.54; p = 0.032). Conclusions: EMA-reported cognitive difficulties were strongly related to real-time psychological/behavioral factors within persons. Additionally, the association between EMA-reported cognitive difficulties and objective neurocognition suggests that assessing cognitive difficulties in real time via EMA may have some clinical utility to identify individuals from this population who need early intervention and potentially a higher level of care.