Facial emotional processing deficits in long-term HIV-suppressed patients

Introduction: Emotional processing is basic for social behaviour. We examine for the first time the facial emotion processing in long-term HIV-suppressed patients. Materials and Methods: Cross-sectional study comparing (ANOVA) six facial emotional processing tasks (two discrimination, two memory and two recognition) between HIV-suppressed patients (HIV+) on effective antiretroviral therapy (>2 years) and matched (age, gender) healthy controls (HCs). Accuracy in the recognition of basic emotions (neutral, happiness, sadness, anger and fear) in each recognition task was also compared (Mann–Whitney U test) between HIV+ and HCs. In the subset of HIV+, we evaluate which factors were associated with impaired recognition of basic emotions (accuracy below 50%) by multiple logistic regression analysis. Overall performance in all six emotional tasks were separately compared between neurocognitive impaired and non-impaired HIV+. Results: We included 107 HIV+, mainly Caucasian (89%) male (72%) with a mean age of 47.4 years, neurocognitively non-impaired (75.5%), and 40 HCs. Overall discrimination (p=0.38), memory (p=0.65) and recognition tasks (p=0.29) were similar in both groups. However, HIV+ had lower sadness recognition in both recognition tasks and lower sadness, anger and fear recognition in the facial affect selection task (Figure 1). Only estimated pre-morbid functioning (WAIS-III-R vocabulary subtest score) was significantly associated with sadness (1.99 [95% CI 1.18–3.58]; p=0.01) and anger recognition deficits (2.06 [95% CI 1.14–3.45]; p=0.015) in the facial affect selection task. In HIV+ individuals, neurocognitive impairment was associated with worse memory task results (p<0.01, d=0.88; p