Abstract
Despite its widespread use and popularity, cannabis is known to impact higher order cognitive processes such as attention and executive function. However, far less is known about the impact of chronic cannabis use on cognitive flexibility, a component of executive function, and this is especially true for the underlying functional connectivity dynamics. To address this, we enrolled 25 chronic cannabis users and 30 demographically matched non-users who completed an interview probing current and past substance use, a urinalysis to confirm self-reported substance use and a task-switch cognitive paradigm during magnetoencephalography (MEG). Time-frequency windows of interest were identified using a data-driven statistical approach, and spectrally specific neural oscillatory responses were imaged using a beamformer. The resulting maps were grand-averaged across all participants and conditions, and the peak voxels in these maps of neural oscillatory activity were used as seeds to compute connectivity using a whole-brain cortical-coherence approach. Whole-brain neural switch cost connectivity maps were then computed by subtracting the connectivity map for the no-switch condition from that of the switch condition per participant. These switch cost functional connectivity maps were then correlated with the behavioural switch cost per group and probed for group differences in the neuro-behavioural associations. Our behavioural results indicated that all participants had slower reaction times during switch compared to no-switch trials. Regarding the MEG data, cannabis users exhibited altered associations between functional connectivity switch costs and behavioural switch costs along pathways connecting visual cortices and regions in the ventral attention network, within the theta, alpha and gamma frequency ranges. These results indicate modified multispectral associations between functional connectivity and behavioural switch costs among visual cortices and key brain regions underlying executive function in cannabis users.