The Resting State Central Auditory Network: a Potential Marker of HIV-Related Central Nervous System Alterations

Abstract

Objective: HIV positive (HIV+) individuals with otherwise normal hearing ability show central auditory processing deficits as evidenced by worse performance in speech-in-noise perception compared with HIV negative (HIV-) controls. HIV infection and treatment are also associated with lower neurocognitive screening test scores, suggesting underlying central nervous system damage. To determine how central auditory processing deficits in HIV+ individuals relate to brain alterations in the cortex involved with auditory processing, we compared auditory network (AN) functional connectivity between HIV+ adults with or without speech-in-noise perception difficulties and age-matched HIV- controls using resting-state fMRI.

Design: Based on the speech recognition threshold of the hearing-in-noise test, twenty-seven HIV+ individuals were divided into a group with speech-in-noise perception abnormalities (HIV+SPabnl, 38.2 ± 6.8 years; 11 males and 2 females) and one without (HIV+SPnl 34.4 ± 8.8 years; 14 males). An HIV- group with normal speech-in-noise perception (HIV-, 31.3 ± 5.2 years; 9 males and 3 females) was also enrolled. All of these younger and middle-aged adults had normal peripheral hearing determined by audiometry. Participants were studied using resting-state fMRI. Independent component analysis was applied to identify the AN. Group differences in the AN were identified using statistical parametric mapping.

Results: Both HIV+ groups had increased functional connectivity (FC) in parts of the AN including the superior temporal gyrus, middle temporal gyrus, supramarginal gyrus, and Rolandic operculum compared to the HIV- group. Compared with the HIV+SPnl group, the HIV+SPabnl group showed greater FC in parts of the AN including the middle frontal and inferior frontal gyri.

Conclusions: The classical auditory areas in the temporal lobe are affected by HIV regardless of speech perception ability. Increased temporal FC in HIV+ individuals might reflect functional compensation to achieve normal primary auditory perception. Furthermore, increased frontal FC in the HIV+SPabnl group compared with the HIV+SPnl group suggest that speech-in-noise perception difficulties in HIV-infected adults also affect areas involved in higher-level cognition, providing imaging evidence consistent with the hypothesis that HIV-related neurocognitive deficits can include central auditory processing deficits.

Figure 1.

Brain mapping of the auditory network defined by ICA (voxel-wise FDR correction, p < 0.001, cluster size > 100). The color bar indicates the t value of one-sample t-test on the functional connectivity strength. L and R stands for left and right hemisphere, respectively.

Figure 2.

Brain mapping of two-sample t-test results within auditory network between two different groups. (a) Regions showing significant increased FC in HIV+SPabnl, compared to healthy subjects (voxel-level p < 0.01, cluster-wise AlphaSim threshold of p < 0.05); (b) Regions showing significant increased FC in HIV+SPnl, compared to healthy subjects (voxel-level p < 0.01, cluster-wise AlphaSim threshold of p < 0.05); (c) Significant differences in auditory network between HIV+SPabnl and HIV+SPnl. Warm color and cool color indicate increased and decreased FC in HIV+SPabnl, compared to HIV+SPnl (voxel-level p < 0.01, cluster-wise AlphaSim threshold of p < 0.05). There is also decreased FC in the left cerebellum in HIV+SPabnl (Table 2), which is not presented here. The color bar indicates the t value of the two-sample t-test. L and R stand for left and right hemisphere, respectively.