Acute cannabis-related alterations in an fMRI time estimation task

Introduction: Cannabis is widely popular recreational drug of choice in the US. The drug is known to alter the subjective experience of time. However, its effects on time estimation at a brain level are still largely unexplored. Our goal was to investigate acute effects of cannabis on an fMRI time estimation task by evaluating brain activation differences between cannabis and placebo conditions. We hypothesized that participants’ time estimation accuracy and corresponding BOLD response would be altered during the cannabis condition in a dose-related manner, compared to placebo. Methods: In this placebo-controlled, double-blind randomized trial, a total of N=44 participants had 3 dose visits, at each of which they received either high-dose cannabis (0.5 gm of ~12.5% THC flower), low dose cannabis (0.5 gm of ~5.7% flower) or 0.5 gm placebo, using paced inhalation from a volcano via vaporizer. Drug material was supplied by NIDA/RTI. For the current study we analyzed fMRI data from the first of placebo and high dose fMRI sessions throughout each dosing day in which participants performed a time estimation task. Participants were asked to respond with a mouse click as to which box of two boxes displayed for different intervals was displayed on the screen longer. Both sub-second and supra-second temporal intervals were tested, with a range of easy to hard discriminations. We used the Human Connectome Project processing pipeline to prepare fMRI data for GLM modeling of activation using the FSL FEAT toolbox. This model estimated the unique effect sub-second (short) and supra-second (long) interval discrimination, their average effect, and their difference. From these contrasts, the mean activation amplitudes within 387 brain parcels from the Human Connectome cortical atlas were extracted. Robust statistics in R software estimated a paired t test equivalent using the bootdpci function to assess the difference between placebo and the high dose drug conditions for each contrast. Results: Only premotor cortex survived False Discovery Rate corrections for searching all 387 parcels across the entire brain for the average of short and long temporal estimation conditions. Numerous other brain regions differed between placebo and high doses at p<.05 uncorrected for various task contrasts: Short duration stimuli: Premotor cortex, posterior cingulate cortex, medial temporal cortex, visual area, somatosensory cortex, anterior cingulate and medial prefrontal cortex, paracentral and mid-cingular cortex, inferior frontal cortex. Long duration stimuli: Premotor cortex, visual areas, somatosensory motor cortex, paracentral and mid- cingulate cortex, the tempo-parieto-occipital junction, dorsolateral-prefrontal cortex, posterior opercular cortex, medial temporal cortex, posterior cingulate cortex, orbito-frontal cortex. Average of short and long duration stimuli: Premotor cortex, somatosensory and motor cortex, posterior cingulate cortex, visual are, medial temporal cortex, paracentral and midcingulate cortex, anterior cingulate and medial prefrontal cortex, inferior frontal cortex, tempo-parieto-occipital junction, premotor cortex, somatosensory motor cortex, posterior cingulate cortex, medial temporal cortex, orbital and polar frontal cortex, hippocampus. Difference of short and long duration stimuli: Anterior cingulate and medial prefrontal cortex, ventral stream visual cortex, dorsal stream visual cortex, early visual cortex. Conclusions: The current study elicited multiple brain activation differences for the initial, acute high-dose cannabis vs. placebo condition, but only premotor cortex region survived as significant following multiple comparison correction for short and long duration stimuli contrast. A post hoc power analysis showed that adding 10 additional subjects to this sample would achieve significance with multiple comparison correction for medium effect sizes at alpha=0.05. Future studies on a larger sample can help identify such significant activation differences, and examining all doses and tasks would elucidate unfolding of effects longitudinally post-dose, and dose-dependence of effects.