Abstract

1. Introduction

Cannabis use has been associated with changes in cognitive task performance (Curran et al., 2002; Jacobus et al., 2009; Schoeler et al., 2016a; Schreiner and Dunn, 2012a; Scott et al., 2018; Solowij and Pesa, 2010) and altered brain function (Batalla et al., 2013; Tapert et al., 2007) involving various cognitive domains. Impaired task performance (Bhattacharyya et al., 2015a; Curran et al., 2002; D’Souza et al., 2004; Hindocha et al., 2015, 2017; Lawn et al., 2016; Ramaekers et al., 2006) and brain functional alterations (Batalla et al., 2014; Bhattacharyya et al., 2017, 2015b) involving different cognitive domains have also been observed during acute intoxication. Recent meta-analyses employing different analytic approaches have shown that persistent long-term use of cannabis is associated with functional alterations in key brain regions across different cognitive tasks (Blest-Hopley et al., 2018; Yanes et al., 2018; Yücel et al., 2008).

One of the key issues that can potentially confound the interpretation of current evidence is whether the effects of cannabis use on cognition and underlying brain function abstinence persist or recover following a period of abstinence. Following abstinence, cognitive performance has been found to improve in cannabis users (CU) (Hanson et al., 2010) to the level of controls after longer periods of abstinence (Schulte et al., 2014), with cognitive deficits possibly only detectable within the first 25 days of abstinence. Meta-analysis of cognitive task performance in continuing CU has shown significant impairment over a wide range of tasks, while abstinent users showed no significant difference to controls in any specific or global cognitive domain (Schreiner and Dunn, 2012b). In contrast, structural changes in CU have been observed (Batalla et al., 2013), in particular decreased volume in the hippocampus (Chye et al., 2018; Cousijn et al., 2012; Matochik et al., 2005), that persisted after a prolonged abstinence in some (Ashtari et al., 2011) but not all studies (Koenders et al., 2017).

However, whether the effects of recreational cannabis use on brain function persist not only beyond the acute intoxication stage typically lasting 2–3 h (Grotenhermen, 2003)(when used by the inhalation route), but even after the key metabolites of delta-9-tetrahydrocannabinol (THC) with psychotropic effects have been excreted from the body, is less well known. THC and its metabolites(11-hydroxy-Δ9-tertrahydrocannabinol and 11-nor-9-carboxy-Δ9-tertrahrdrocannabinol (Sharma et al., 2012) are of particular interest, as it is the main psychotropic ingredient in cannabis known to be associated with harmful effects on various cognitive domains(Pertwee, 2008). It is worth noting that the half-life of THC in frequent users is 5–13 days (Smith-Kielland et al., 1999) and THC is detectable in urine for up-to 2–4 weeks(Lowe et al., 2009). Interestingly, the upper limit for the period of detection of metabolites in urine is consistent with the period over which cognitive deficits are detectable following abstinence (Schreiner and Dunn, 2012b).

We have recently examined the residual effects of recreational cannabis use on brain function in adult and adolescent cannabis users by meta-analytically combining the data from 20 published studies employing functional MRI techniques (Blest-Hopley et al., 2018). While some of these studies investigated cannabis-using participants after a period of abstinence, several others allowed cannabis use up until, as short a period as, 3 h prior to scanning. Therefore, interpretation of the results of these studies may be confounded by residual acute effects of THC and its metabolites that may still be left in cannabis-using participants as well as effects of withdrawal from cannabis. On the other hand, brain functional alteration following a sustained period of abstinence has also been investigated, though the results of these studies are less consistent, with users showing both increased (Chang et al., 2006; De Bellis et al., 2013; Jacobsen et al., 2004; Schweinsburg et al., 2008; Tapert et al., 2007) and decreased (Chang et al., 2006; Schweinsburg et al., 2008) activation compared to controls. Studies comparing cannabis users with different periods of abstinence have found greater activation in the prefrontal cortex and insula in recently abstinent users compared to users with longer (at least 27 days) periods of abstinence, who in turn had greater activation in the precentral gyrus (Schweinsburg et al., 2010). Another study that investigated cannabis users at multiple time-points following abstinence, reported that 28 days of abstinence resulted in reduced activation difference to controls in some regions, but some differences in brain activation persisted (Pillay et al., 2008). Therefore, understanding differences in brain activation between currently using (or non-abstinent) and abstinent cannabis users (ACU), is of particular interest. However, to our knowledge existing evidence in this regard has not been systematically reviewed and summarized using meta-analytic approaches. Hence, we have carried out a meta-analysis complementary to that previously reported by us (Blest-Hopley et al., 2018) to investigate whether altered brain function associated with regular cannabis use persists even after a sustained period of abstinence from cannabis. Consistent with our approach previously (Blest-Hopley et al., 2018), we included fMRI studies that employed a wide range of cognitive activation paradigms engaging various cognitive processes rather than focusing only on task-specific approaches as in other work (Yanes et al., 2018). Our strategy was driven by two key considerations. Firstly, only a limited number of available studies have specifically employed comparable activation paradigms limiting our ability to meaningfully investigate the question of interest here. More importantly, as we have argued before (Blest-Hopley et al., 2018), the effects of cannabis use are unlikely to be limited to only those brain regions that sub-serve cognitive processes examined in studies conducted hitherto. Rather they are more likely to be widely distributed, consistent with ubiquitous distribution of cannabinoid receptors in the brain (Iversen, 2003). Therefore, we included fMRI studies employing a range of cognitive activation paradigms to investigate using a meta-analytic approach whether brain functional alterations associated with cannabis use persist even after periods of abstinence sufficiently long such that cannabis metabolites are no longer detectable in urine.

(…)