Evidence that intrinsic brain connectivity moderates the association between gait and neurocognitive function in people living with HIV

Technical Summary: Evidence that intrinsic brain connectivity moderates the association between gait and neurocognitive function in people living with HIV

Problem Statement
People living with HIV (PLWH) experience HIV-associated brain injury (HABI), often linked to accelerated aging and neurocognitive impairment (NCI). While slower gait speed has been identified as a potential clinical sign of NCI in PLWH, the neural mechanisms underlying this association remain poorly understood, particularly in low- and middle-income country (LMIC) settings where resource-intensive cognitive screening is challenging. Existing literature suggests that in older adults, gait speed and cognitive function converge due to the recruitment of higher-order functional brain networks, specifically the frontoparietal network (FPN), as automatic motor control declines. However, there is a lack of research directly comparing groups differing in disease exposure and age to determine if HIV status moderates the relationship between gait, cognition, and intrinsic functional brain connectivity. This study aims to test whether gait speed is associated with cognitive deficits in middle-aged PLWH in an LMIC setting, whether this association mirrors that seen in healthy older adults, and whether intrinsic functional connectivity abnormalities serve as a marker of HABI that moderates these associations.

Methodology
This study employed a secondary analysis of data from a cohort recruited from a community health centre in Khayelitsha, Cape Town, South Africa. The primary sample consisted of 39 participants (17 PLWH and 22 HIV-negative controls) with a mean age of approximately 39 years. To ensure the validity of the findings regarding HABI, PLWH participants were either antiretroviral therapy (ART)-naïve or had been on ART for less than one month.

• Clinical and Behavioral Assessments: Participants underwent a comprehensive neurocognitive battery translated into Xhosa, assessing seven domains: speed of processing, language, executive function, delayed recall, learning, working memory, and motor function. A global neurocognitive score was derived from these domains. Gait speed was measured using a timed 20-yard walk (10 yards out and back), with the average time of three trials recorded.
• Neuroimaging: A subset of participants (11 PLWH and 11 HIV-negative controls) underwent resting-state functional magnetic resonance imaging (rs-fMRI) on a 3T Siemens Skyra scanner. Data were preprocessed using the HALFpipe pipeline, including motion correction (ICA-AROMA, aCompCor) and spatial smoothing.
• Connectivity Analysis: Intrinsic functional connectivity was assessed using seeds corresponding to seven canonical resting-state networks (RSNs) defined by Yeo et al. (2011), excluding the limbic network due to signal attenuation issues. The study focused on within-network connectivity, specifically the FPN, using a multivariate regression framework (AFNI FATCAT toolbox) to test for group differences in the association between connectivity and gait/cognitive scores.
• Replication: To contextualize the findings, the association between gait speed and neurocognitive performance was replicated using a large, publicly available dataset of older US adults (National Social Life, Health, and Aging Project; NSHAP, N=3196).

Key Results

• Gait and Cognition Association: In the LMIC cohort, slower gait speed (longer time) significantly predicted lower global neurocognitive scores ($t = -3.12, p < 0.008$) within the PLWH subgroup. This association was also observed in the total sample after adjusting for age and sex. Crucially, this relationship was replicated in the NSHAP cohort of healthy older adults, confirming a robust negative association between gait time and cognitive performance across different populations.
• Domain Specificity: Bivariate analyses revealed that slower gait was significantly associated with poorer performance in memory, language, learning, and motor domains across the entire sample. After Bonferroni correction, the association with memory remained significant. In the PLWH subgroup specifically, significant associations were found for memory and motor function.
• Neural Moderation (FPN Connectivity): The study identified a significant interaction between HIV status and gait time regarding the intrinsic connectivity of the frontoparietal network (FPN). Specifically, in PLWH, slower gait speed and lower executive function scores were uniquely associated with altered connectivity patterns involving the right inferior frontal gyrus (r. IFG).
  • In PLWH, slower gait and poorer executive function correlated with less connectivity between the r. IFG and the right precentral gyrus.
  • Conversely, in PLWH, these same deficits correlated with greater connectivity between the r. IFG and the left inferior parietal lobe.
  • These specific connectivity-gait/cognition associations were not observed in the HIV-negative control group.
• Network Specificity: The moderating effect of HIV status on the relationship between connectivity and gait/cognition was specific to the FPN; no significant interactions were found for the other five RSNs assessed.

Significance and Claims
The authors claim that their findings provide the first evidence, across two independent datasets (one LMIC, one high-income), that gait speed is a robust indicator of neurocognitive function in PLWH, mirroring patterns seen in healthy aging. The study posits that the integrity of the FPN, particularly involving the right inferior frontal gyrus, mediates the relationship between gait and executive function in PLWH.

The paper suggests that the observed pattern of connectivity in PLWH—specifically the altered coupling between frontal and parietal regions—may reflect impaired compensatory mechanisms. The authors propose that successful performance in PLWH may rely on the disengagement of the frontoparietal network in favor of motor-specific regions, whereas compromised FPN function leaves patients at a disadvantage. Consequently, the study underscores the potential utility of simple, non-invasive gait tests as a screening tool for HABI in resource-limited settings. The findings support the hypothesis that HIV-associated accentuated aging may be a mechanism linking abnormal FPN connectivity to poorer executive function and slower gait speed in this clinical population.

The authors maintain a modest tone regarding causality, noting that while the findings are consistent with HABI, other factors not controlled for in the study may contribute to cognitive decline. They also acknowledge limitations, including the small sample size for the rs-fMRI analysis which limited statistical power for detecting interactions, and the inability to fully account for ART exposure effects despite selection criteria.