Abstract
BACKGROUND: Neurocognitive impairment (NCI) is a common comorbidity among aging people with HIV (PWH), despite effective antiretroviral therapy (ART). Processing speed is often the earliest affected cognitive domain and may be linked to disrupted functional brain network organization. This study investigated whether the balance of segregation and integration in large-scale functional networks is associated with processing speed in middle-aged and older PWH. METHODS: In a prospective, cross-sectional study, 26 virologically suppressed PWH aged ≥ 50 years underwent neuropsychological testing and resting-state functional MRI (rs-fMRI). Functional brain networks were constructed using a 300-node cortico-subcortical parcellation. System segregation index and node-level participation coefficient (PC) were calculated to quantify the global and local balance between integration and segregation, respectively. Associations with age-adjusted Wechsler Adult Intelligence Scale (WAIS-III) Symbol Search (WAISsys) T-scores were assessed using regression and correlational analyses. RESULTS: Higher system segregation within associative networks was significantly associated with better WAISsys T-scores (β = 0.544, p = 0.004), whereas segregation in sensorimotor networks was not. The majority of nodal PC values were negatively correlated with WAISsys T-scores, indicating that lower processing speed was associated with less segregated and more integrated connectivity. Nodes showing the strongest negative associations with WAISsys T-scores were disproportionately located in the default mode and frontoparietal networks. CONCLUSIONS: In middle-aged and older PWH, greater segregation within associative networks is linked to better processing speed in our preliminary findings. Disruptions in network segregation and modularity, especially in cognitive control systems, may be associated with processing speed deficits despite viral suppression. These findings highlight the importance of functional brain network topology and organization as a potential biomarker for cognitive aging in HIV. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12883-026-04749-0.