Motor-Related Neural Dynamics are Modulated by Regular Cannabis Use Among People with HIV

Abstract

Recent work has shown that people with HIV (PWH) exhibit deficits in cognitive control and altered brain responses in the underlying cortical networks, and that regular cannabis use has a normalizing effect on these neural responses. However, the impact of regular cannabis use on the neural oscillatory dynamics underlying motor control deficits in PWH remains less understood. Herein, 102 control cannabis users, control nonusers, PWH who regularly use cannabis, and PWH who do not use cannabis performed a motor control task with and without interference during high-density magnetoencephalography. The resulting neural dynamics were examined using whole-brain, voxel-wise statistical analyses that examined the impact of HIV status, cannabis use, and their interaction on the neural oscillations serving motor control, spontaneous activity during the baseline period, and neurobehavioral relationships. Our key findings revealed cannabis-by-HIV group interactions in oscillatory gamma within the prefrontal cortices, higher-order motor areas, and other regions, with the non-using PWH typically exhibiting the strongest gamma interference responses. Cannabis-by-HIV interactions were also found for oscillatory beta in the dorsal premotor cortex. Spontaneous gamma during the baseline was elevated in PWH and suppressed in cannabis users in all regions exhibiting interaction effects and the left primary motor cortex, with spontaneous levels being correlated with behavioral performance. These findings suggest that regular cannabis use has a normalizing effect on the neural oscillations serving motor control and the abnormally elevated spontaneous gamma activity that has been widely replicated in PWH, which may suggest that cannabis has at least some therapeutic utility in PWH.

Fig. 1

Experimental Paradigm and Behavioral Results. (A) An illustration of the flanker arrow paradigm. Each trial had a fixation period lasting on average 1500 ms (variable ISI: 1450–1550 ms) and a stimulus-presentation period lasting 2500 ms, which consisted of one of the four shown options. (B) Reaction time (in ms) across all participants is shown on the y-axis with condition on the x-axis. Across all groups, participants responded slower during incongruent relative to congruent trials (i.e., classic flanker effect; p = 2.22 × 10^− 21). (C) Neither the main effects of cannabis nor HIV group were significant, but the condition-by-HIV group interaction was, and follow-up testing showed that PWH exhibited a larger flanker interference effect than seronegative participants (p =.007). ***p ≤.001

Fig. 2

Motor-related neural oscillatory responses. (Left): Grand-averaged time-frequency spectrograms from MEG sensors near the precentral gyrus (top: MEG0432; bottom: MEG0443) exhibiting a significant motor-related gamma response (top; 68–82 Hz, -75 to 75 ms) and a peri-movement beta ERD response (bottom; 16–24 Hz, -300 to 300 ms). The spectrograms display frequency (Hz) on the y-axis and time (ms) on the x-axis. Signal power is expressed as a percent difference from the baseline period, with the color legend shown to the right of the spectrogram. Note that the post-movement beta rebound (PMBR) response can also be seen, but this was not further examined since it happened after the movement, and we were interested in the motor planning and execution aspects of the task. (Right): Grand-averaged beamformer images (pseudo-t) across all participants for each significant time-frequency component

Fig. 3

Cannabis use and HIV status interact to modulate whole-brain gamma and beta oscillatory and spontaneous power. Voxel-wise 2 × 2 ANOVAs on flanker interference maps revealed significant cannabis-by-HIV status interactions on beta power in the right dorsal premotor cortex (dPMC), and on gamma power in the right ventral premotor cortex (vPMC), right insula, right dorsolateral prefrontal cortex (dlPFC), and left cerebellum. All maps have been thresholded at p <.005, corrected, with the scale bar shown to the right of the brain in the top row. (A): Post-hoc testing revealed that cannabis using PWH exhibited weaker beta interference responses (i.e., less negative) in the right dPMC relative to each of the other three groups. Meanwhile, gamma interference responses were aberrantly elevated (i.e., more positive) in non-using PWH relative to seronegative nonusers and to cannabis-using PWH across all four cannabis-by-HIV status interaction peaks. Other group comparisons were also significant in specific regions. (B, C): We probed the spontaneous power during the baseline period (i.e., -1800 to -1000 ms) within each region that had interaction effects for oscillatory gamma or beta activity. There were significant main effects of (B) cannabis use and (C) HIV status on spontaneous gamma power in all regions such that spontaneous gamma activity was reduced in participants who use cannabis compared to non-using participants, and elevated in PWH compared to seronegative participants. *p ≤.05; **p ≤.005; ***p ≤.001

Fig. 4

Whole-brain oscillatory beta and gamma activity differs by HIV status. Voxel-wise 2 × 2 ANOVAs revealed a significant effect of HIV status on (A) beta power in the left dorsolateral prefrontal cortex (dlPFC) and left prefrontal cortex (PFC), and on (B) gamma power in the left cingulate, right cerebellum, and left thalamus (not shown). All maps have been thresholded at p <.005, corrected, with the scale bar shown to the right of the first brain in the top row. Follow-up analyses revealed that, across all regions exhibiting an effect of HIV status, PWH exhibited significantly weaker flanker interference beta and gamma oscillatory power compared to controls, regardless of cannabis use status. **p ≤.005; ***p ≤.001

Fig. 5

Whole-brain oscillatory beta and gamma activity differs by cannabis use. Voxel-wise 2 × 2 ANOVAs revealed a significant effect of cannabis use on (A) beta power in the right postcentral gyrus, parietal cortex, and ventromedial prefrontal cortex (vmPFC; not shown), and on (B) gamma power in the right prefrontal cortex (PFC), right cerebellum (not shown), right superior parietal cortex (not shown), and right inferior frontal gyrus. All maps have been thresholded at p <.005, corrected, with the scale bar shown to the right of the first brain in the top row. In general, people who regularly use cannabis exhibited significantly weaker beta and gamma interference responses compared to non-using participants, regardless of HIV status. The only exception was for gamma in the right superior parietal where the effect was reversed. **p ≤.005; ***p ≤.001

Fig. 6

Group differences in spontaneous gamma power in the left primary motor cortex. (A): The grand-averaged map across both conditions and all participants indicated that the peak movement-related gamma response was centered on the motor hand knob of the left primary motor cortex (M1). The strength of the movement-related gamma oscillatory response (-75 to 75 ms) and the mean spontaneous gamma levels during the pre-movement baseline period were computed using data from the peak voxel of this cluster. (B): Spontaneous levels during the baseline period (-1800 to -1000 ms, gray shaded area) are plotted separately per group. (C): Spontaneous gamma activity was significantly weaker in chronic cannabis users relative to nonusers. (D): PWH exhibited significantly elevated spontaneous gamma activity relative to seronegative participants. No group differences or interactions were observed for the movement-related gamma response (-75 to 75 ms). *p ≤.05; **p ≤.005; ***p ≤.001