In this study, we evaluated neurophysiological indicators via novel measures of brain perfusion and metabolism in a large and well-characterized group of 72 h abstinent cannabis users relative to nonusers. We found greater regional CBF in the right pallidum/putamen in users compared with nonusers, in line with reported functional and structural changes in this critical region in long-term cannabis users (Filbey et al, 2014). These findings of regional but not global CBF differences between users and nonusers are concordant with region-specific effects described in studies of both chronic (Bloom et al, 1997; Mathew and Wilson, 1993) and acute effects of cannabis use (O'Leary et al, 2002). Increases in resting CBF in basal ganglia structures have been linked to elevated salience and motivation processes that are mechanisms that contribute to compulsive use of cannabis (Volkow et al, 1996). The pallidum plays a critical role as the interface between the ventral striatum and the thalamus with the ventral pallidum, specifically, mediating reward and motivation processes (Smith et al, 2009). The putamen is dense with cannabinoid receptors (Miller and Walker, 1996) and has been shown to have functional and structural alterations that are associated with cannabis use (Filbey et al, 2016; Wetherill et al, 2015; Yip et al, 2014). Because THC relaxes arterial walls and increases blood flow to tissues, our finding adds to the literature by demonstrating that the pathomechanism underlying the reported functional and structural alterations in the pallidum/putamen may be due to altered perfusion that is closely coupled with neural activity.

Notably, other studies have reported decreases rather than increases in CBF in chronic cannabis users (Martin-Santos et al, 2010; Mathew and Wilson, 1991; Lundqvist et al, 2001). The discrepancy with existing reports may be because of heterogeneity both within and between study samples. For example, Mathew et al (1989) reported dissociable CBF effects where experienced users had greater CBF whereas inexperienced users had decreased CBF following THC exposure. Length of abstinence can also affect CBF effects. Tunving et al (1986) noted changes during the course of abstinence whereby rCBF level in cannabis users was significantly (11%) lower than that of age- and sex-matched healthy controls, although (12%) increases after 9–60 days of detoxification of the CBF level was noted at follow-up. Given that our sample of cannabis users were heavy, long-term users and that data were collected following 3 days of required abstinence, this observation may be an ‘overshoot of recovery’ of neurophysiological function during the early stages of abstinence in chronic cannabis users. In other words, these findings reflect the residual effects of both long-term cannabis exposure and adaptive changes that occur after a short period of abstinence. Future studies should measure the trajectory of these effects to fully determine whether this phenomenon is indeed a recovery of typical function.

It is also important to note that in addition to chronicity and cessation from cannabis use that complicate the understanding of THC effects on CBF, individual variation in response to THC also plays a role (Volkow et al, 1991). One important individual factor to consider is sex-specific effects of cannabis (see review by Ketcherside et al, 2016). Although we found concordant findings of a main effect of sex on CBF and CMRO₂ where females had greater CBF and CMRO₂ (Lu et al, 2011), we did not find a group × sex interaction in any of the brain measures. This could be due, in part, to the smaller number of females in the cannabis using group. Another possibility for the absence of a group × sex effect may be that any increased neuronal needs reflected by greater OEF/CMRO₂ and CBF in the pallidum/putamen in the cannabis users relative to nonusers may be compensated for by the typically elevated global CBF and CMRO₂ in females. Greater global CBF and CMRO₂ in females have been suggested to reflect hormonal, muscle, and tissue content differences between sexes (Fattore and Fratta, 2010).

We found positive associations between global and regional CBF and THC metabolite levels that further support the relationship between changes in blood flow and exposure to cannabis (Mathew et al, 1997; Volkow et al, 1996). Alterations in the superior frontal gyrus have previously been reported in cannabis users in terms of reduced thickness (Lopez-Larson et al, 2011) and enhanced activation during memory and inhibitory tasks (Smith et al, 2010; Tapert et al, 2007). This finding highlights the differential effects of acute and prolonged THC exposure that brings neural adaptations in long-term cannabis users to bear.

These findings have several implications. First, because CBF and brain function are closely coupled, it is likely that widely observed changes in brain response (eg, BOLD fMRI) may be partly explained by these latent cerebrovascular changes. There is emergent literature that has begun to shed light on the interplay between neural and vascular components in substance use disorders. Similar investigations have linked changes in EEG patterns with cerebral perfusion deficits in long-term cannabis users during abstinence that were interpreted as under-arousal of these systems (Herning et al, 2008). Using CMRO₂, long-term cocaine users have also been shown to have reduced CMRO₂ relative to nonusing controls that was associated with lifetime frequency of cocaine use (Liu et al, 2014).

It is worth noting that at our selected threshold, no group effects emerged in cortical areas. However, a less conservative threshold of p<0.01 and K>200 revealed that nonusers had higher regional CBF in cortical areas including right middle temporal, left middle frontal, and right superior frontal cortices compared with users. Thus, although differences were observed in other regions, our findings suggest that neurophysiological changes as a result of cannabis use are most robust in the pallidum/putamen area. Indeed, structural and functional alterations in the putamen have been reported to be a risk factor for cannabis use disorder (Wetherill et al, 2015), predict development of problems related to cannabis use (Vingerhoets et al, 2016), and are related to treatment success (Yip et al, 2014).