Symptomatic and Asymptomatic Neurocognitive Impairment, ART Adherence and HIV Control: A 4-Year Observational Study.

We assessed whether symptomatic neurocognitive impairment (NCI) and asymptomatic NCI -of which the clinical relevance is debated- affect HIV control and the role of ART adherence in this relationship. Observational study on the relationship between NCI and viral control during the 2 years before and the 2 after the neurocognitive evaluation (NCE) of 322 PLWH on ART. Viral load (VL) was defined as undetectable, very low-level (VLLV), low-level (LLV), or high-level viremia (HLV), and classified overtime as persistent (p; ≥2 consecutive values in the same worst category), viral failure (VF; ≥1 HLV requiring ART changes), or optimal control. Adherence was the proportion of days covered by ART. Frascati criteria were used. Adjusted models were performed for factors associated with viral control. Mediation analyses informed causality in the path from NCI to viral control through adherence. Sensitivity analyses were focused on the year following NCE for only participants with optimal viral control before. Among the participants (53 ± 10 years, CD4 + T-cells 630/µL), 41.6% and 10.8% presented asymptomatic and symptomatic NCI. Over 3,304 VLs, 8.4% and 22.1% of participants had VF and pLLV/pVLLV. Both symptomatic and asymptomatic NCI were independently associated with VF (aRRR = 8.5; aRRR = 4.3) and pVLLV/pLLV (aRRR = 4.3; aRRR = 2.1). Specific cognitive domains showed independent associations with VL categories (models' P < 0.001). Adherence partially mediated these relationships (models' P < 0.001). Sensitivity analysis confirmed these findings. Prevalence and severity of poor viral control increased as the severity of NCI increased, with ART adherence mediating this relationship. The current "asymptomatic" attribution used by Frascati's criteria could overlook clinical risks.