The rapid proliferation of the legal cannabis market has provoked producers to cultivate an array of novel products to satisfy consumers' growing interests, including products dominant in CBG. Pre-clinical research indicates CBG may have a broad spectrum of therapeutic effects that have recently been supported by self-report data from a large sample of CBG users¹². The present study represents the first human clinical trial to examine the acute effects of CBG on anxiety, stress, and mood. Results indicate that CBG reduces global feelings of anxiety and stress and that it may enhance memory in the absence of intoxication, impairment, or subjective drug effects.

Most notably, there was a significant main effect of CBG on subjective state ratings of anxiety. Specifically, there was a mean decrease in anxiety ratings of 0.95 on a 10-point scale which represents a 26.5% reduction in the already low levels of baseline anxiety (M = 3.59 on a 0–10 VAS) in the CBG condition. In contrast there was a 0.66-point reduction in anxiety ratings in the placebo condition (which represents a 22.5% reduction in anxiety in the placebo condition). These findings are in line with our survey findings wherein 51% of CBG users reported using CBG to manage anxiety and 78% claimed it was more effective than conventional anxiety medications¹². These effects may stem from CBG’s effects on 5-HT_1A^28,29 and/or GABA³⁰. Nevertheless, enthusiasm for these findings is dampened by the lack of significant effect of CBG on STAI state anxiety ratings. This pattern of findings may reflect a tendency for CBG to reduce the global impression of feelings of anxiety, rather than affecting the specific sub-components of anxiety tapped by the STAI (e.g., jittery, confused, indecisive) and further indicate that replication of these findings is necessary.

Consistent with these effects, there was also a significant effect of CBG on subjective stress ratings at T1 (prior to the stressor). While our power analysis indicated we were sufficiently powered to detect medium-sized effects, a larger sample size and/or higher dose may be needed to robustly detect acute effects of CBG on subjective stress ratings. Our use of a field trial precluded our ability to examine physiological indicators of stress (e.g., cortisol, electrodermal activity, alpha-amylase) but we plan to conduct a follow-up laboratory study which will include such measures. As no human clinical trials with CBG had been published at the time we commenced the study our dose selection was guided by anecdotal reports of people using the same CBG tincture, observations that CBG is commonly sold in 10 mg units, as well as by common doses of CBG reported in clinical observations^31,32. However, given recent clinical trials using larger doses (25 mg¹² and 50 mg¹³), our dose may have been somewhat conservative.

Our prior survey study showed that over 30% of CBG users were self-medicating for depression and 80% reported CBG is more effective than conventional antidepressants¹² which is consistent with preclinical research demonstrating that CBG has antidepressant-like effects in a rodent tail suspension model¹⁰. Nonetheless, results from the present clinical trial failed to support our hypothesis that CBG would enhance mood. It is possible that our sensitivity to detect such effects was reduced by the administration of a single-item indicator of mood to a non-clinical sample. Indeed, the overall mean baseline depression levels on the DASS indicates that the sample was in the normal to mild range of depression¹⁹ and the overall baseline mood ratings suggest participants were in a positive mood state prior to drug administration. Future research employing a more comprehensive measure of depression and a larger sample of clinical patients with higher baseline depression is needed to attempt to reconcile these apparently contradictory findings.

CBG did not produce cognitive or motor impairments on the DRUID app. Prior placebo-controlled research reported DRUID scores around 57 following oral administration of 25 mg of THC and vaporized administration of 20 mg of THC²⁴ which is on par with the threshold of 57 for detecting significant impairment using this app²⁵. In contrast, participants’ mean DRUID score following administration of CBG was much lower (43.1) and was remarkably consistent with the mean DRUID score detected at baseline and in the placebo condition. As shown in Fig. 7, a small number of people surpassed the 57-score threshold for impairment at baseline and in both the CBG and placebo conditions which is likely attributable to preexisting conditions (e.g., one participant reported issues with balance, another had a leg injury).

One of the most robust detrimental effects of THC is on verbal memory³³ which guided our decision to include a test of verbal memory in the present study. We hypothesized that CBG would not impair memory, but our finding that CBG significantly enhanced verbal memory was unexpected. Inspection of Fig. 6 shows that the effect was most pronounced on trials 2 & 3 with participants in the CBG condition recalling on average 0.5 words more on trial 2 and 1-word more on trial 3, than they recalled in the placebo condition. Trials 2 & 3 represents learning trials which indicates that CBG may enhance learning. Nevertheless, these surprising findings warrant further corroboration. Future studies attempting to replicate these findings should include tests of verbal memory with additional learning trials, as well as other tests of memory (e.g., working memory) and cognition (e.g., executive functioning, attention) to further elucidate the effects of CBG on cognition. It would also be interesting to examine whether CBG might offset the detrimental effects of THC on verbal memory, as CBD was initially purported to do^34,35, until multiple attempts to replicate those findings failed^36–38.

Consistent with the lack of impairment detected in the CBG condition, participants reported no intoxicating effects of CBG. The overall mean intoxication rating (combined across the three time points) was less than 1 on a 0–10 VAS in both the CBG and placebo conditions. Similarly, overall mean ratings of subjective drug effects were low in both conditions and drug liking effects were neutral in both conditions at all timepoints suggesting a low potential for abuse. Further, there were no significant effects of CBG on changes in dry eyes, dry mouth, sleepiness, appetite, or heart palpitation ratings, with ratings changing less than 1-point from baseline over the course of the experiment. This suggests that CBG was well tolerated and did not elicit adverse acute effects that are typically associated with THC administration. Our decision to assess effects on these outcomes was informed by our prior survey study¹² that revealed that while the plurality (44%) indicated CBG has no side effects, small percentages of participants endorsed experiencing dry eyes, dry mouth, sleepiness, and increased appetite after using CBG-dominant cannabis, perhaps because some were using cannabis containing both CBG and THC. It is also important to note that five individuals who completed our screening survey indicated they had previously experienced a severe adverse reaction to CBG. While the extent to which these events were triggered by CBG taken in combination with THC are unknown, they suggest we should be cautious in our interpretation of these null findings. Additional caution should be taken when interpreting results from the present study because we conducted this study remotely via Zoom so there were no physical exams or assessments of vital signs, participants were limited to reporting on a set of pre-determined potential adverse events and were not asked about these in an open-ended manner, these assessments were limited to the first 20–60 min after dosing and only a single, relatively modest dose was assessed which may have missed some of the peak effects¹⁴. As such, these results cannot be extrapolated to other potential adverse effects, later timepoints, higher doses, or repeated/chronic dosing. Future research should examine a broader array of potential drug effects and side effects using physiological measures (heart rate monitors, blood pressure monitors) across longer periods of time prior to reaching strong conclusions that CBG is safe and not associated with side effects.

While our use of a gold-standard double-blind, placebo-controlled cross-over design increases internal validity and our use of Zoom to naturalistically test participants in their home environments enhances ecological validity, the study is not without its limitations. These limitations include our use of a non-clinical sample of cannabis users, a lack of correction for multiple statistical comparisons, modest effects of the stress manipulation, as well as the use of a relatively modest dose and the early timing of assessments. Specifically, as a prudent first step we recruited a moderate-sized community sample rather than a clinical sample of patients with anxiety or depression. A larger sized clinical sample may have increased our sensitivity to detect significant effects of CBG on mood, stress, and state anxiety but may have also increased risks as potential drug interactions with SSRIs are unknown. We also used a sample of experienced cannabis users (for ethical reasons) and as such must be very cautious in generalizing the findings to cannabis naïve individuals. Although potential interactions between THC and CBG are unknown, our use of experienced cannabis users may have also diminished our sensitivity to detect effects due to a general level of tolerance with cannabinoids. We also did not correct for the multiple statistical comparisons that were conducted as the study was not powered to do so. As such, some findings may represent Type I errors and replication is required to corroborate them. Further, while the acute stress manipulation significantly increased stress from T1 to T2, average subjective stress ratings remained below the baseline (T0) after the stressor. While the online version of the TSST has been previously validated²², a stress manipulation performed in the lab (rather than in participants’ comfortable home environments via Zoom) may have produce larger increases in subjective stress. However, the relatively modest increases in subjective stress following the TSST may also reflect our use of a sample of experienced cannabis users, as prior research has shown that cannabis users demonstrate blunted cortisol and subjective stress responses to acute stressors even in a laboratory environment³⁹.

At the time we commenced the study there had been no published clinical trials of CBG in humans, as such, we had little to guide us on the most appropriate dose or timing of assessments. As indicated previously, our dose selection was guided by anecdotal evidence from people using this tincture, standard doses of CBG in marketed products, and clinical observations of efficacy at similar doses^31,32. However, given recent clinical trials using larger doses of CBG (25 mg¹² and 50 mg¹³) our dose may have been too conservative to reveal optimal effects. Repeated dosing may have also increased potential effects on stress, anxiety, and mood. Further, while recent data from a human pharmacokinetic study of CBG¹⁴ demonstrated rises in plasma concentrations as early as 20 min after oral administration, these concentrations peaked from 45 min to nearly 2 h. (depending on dietary fat). As such our assessments from 20 to 60 min after dosing may have failed to capture peak effects. Additionally, we did not restrict participant’s diets in anyway, but recent evidence indicates that eating a meal high in dietary fat prior to ingestion of CBG increases plasma levels of CBG¹⁴. Our failure to control diet may have resulted in random variability in the potency of the product that could have contributed to random error that further diminished sensitivity to detect significant effects.

In conclusion, results of this double-blind, placebo-controlled, cross-over field trial indicate that 20 mg of hemp-derived CBG reduces subjective ratings of anxiety and stress in healthy cannabis-using adults in the absence of motor or cognitive impairment, intoxication, or other subjective drug effects (e.g., heart palpitations, dry mouth). Additional research is needed to corroborate these novel findings as well as to extend them to a clinical population of patients with anxiety disorders.