Cannabidiol (CBD) is a non-intoxicating, plant-derived cannabinoid that has demonstrated considerable therapeutic potential [1]. CBD has well-established anticonvulsant effects [2–4], with the Food and Drug Administration (FDA) recently approving the oral CBD solution, Epidiolex®, for the treatment of intractable paediatric epilepsy [5]. Early-stage clinical trials have also demonstrated anxiolytic [6, 7] and antipsychotic [8, 9] effects. These effects are typically observed at oral doses of ~ 300–1500 mg CBD [10], although acute doses up to 6000 mg also appear to be safe and well-tolerated in humans, albeit with occasional mild side effects (e.g. diarrhoea, nausea, headache) [11, 12].

Alongside its emergent clinical use [13], interest in CBD has increased among general (non-clinical) populations [14, 15], including athletes [16]. While healthy individuals are usually ineligible to access regulated, prescription CBD (e.g. Epidiolex®), a wide range of low-dose “nutraceuticals” (e.g. oils, capsules, topicals and edibles typically containing between ~ 5 and 150 mg CBD·dose⁻¹), including some products marketed specifically to athletes (e.g. cbdMD™, fourfivecbd™), are readily available over-the-counter in certain countries (e.g. UK, Canada, USA) [17]. Within the context of elite sports, the use of CBD has been further facilitated by its recent removal from the World Anti-Doping Agency’s “Prohibited List” [18]. In fact, 26% of British professional rugby players surveyed in a recent study (n = 517; 39% of those aged ≥ 28 years) reported either currently using, or having previously used, CBD [16]. The most common reasons for use were to enhance recovery (80%), improve sleep (78%), reduce anxiety (32%), and for “other” medical purposes (14%; e.g. concussion) [16].

Despite its growing popularity [16], only two interventional studies, both randomised, double-blind, placebo-controlled crossover trials, have so far investigated the effects of CBD on outcomes relevant to athletic performance. The first found no effect of CBD (150 mg·d⁻¹ for 3 days) on non-invasive measures of muscle damage following eccentric exercise [19]. The second found CBD (60 mg; acute) decreased blood serum concentrations of creatine kinase and myoglobin, and increased one-repetition maximum back squat performance 72 h, but not 24 or 48 h following resistance exercise [20]. In any case, a recent review of preclinical studies and clinical trials (involving non-athlete populations) outlined the potential for CBD to exert anti-inflammatory, analgesic, anxiolytic and neuroprotective effects that could have utility in treating inflammatory pain (e.g. delayed onset muscle soreness, injuries), head injuries (e.g. concussion) and sports performance anxiety in athletes [21]. Of course, if CBD is to be used within the sporting context, it is important to understand how it influences key physiological and psychological responses during exercise, particularly given its complex pharmacology [22].

The current randomised, placebo-controlled exploratory pilot trial investigated the effects of acute, oral CBD treatment on physiological and psychological responses to submaximal and exhaustive running exercise in a small sample of endurance-trained males. It should be noted that, as a pilot study, this investigation was not designed nor formally powered to assess “effect” [23]. Rather, its intent was to gain a preliminary understanding of CBD’s effects on exercise physiology and to determine whether these are worthy of further investigation in larger, fully powered trials.

This investigation was approved by the Sydney Local Health District’s Human Research Ethics Committee (2020/ETH00226), registered with the Australia and New Zealand Clinical Trials Registry (ACTRN12620000941965) and conducted at the Charles Perkins Centre—Royal Prince Alfred Hospital Clinic, Sydney, Australia, in accordance with Good Clinical Practice guidelines, the Declaration of Helsinki (1983) and local regulations.

This pilot study investigated the effects of acute, oral CBD treatment (300 mg) versus placebo on physiological and psychological responses to aerobic exercise in a small sample of male endurance-trained runners. Effects were considered worthy of further investigation if the 85% CI around Cohen’s d_z included ± 0.5 but not zero. Results suggest that CBD may alter several key parameters, including submaximal (V̇O₂) and maximal (V̇O_2max) oxygen consumption, feelings of pleasure during submaximal exercise and markers of exercise-induced inflammation. However, it is important to stress that the analytical approach employed is susceptible to Type I error (i.e. identifying “false” effects) and was selected to inform the design of future “definitive” trials [23]. These initial observations should, therefore, be interpreted with caution and not taken as compelling evidence of “effect”.

CBD appeared to increase V̇O₂ (and V̇CO₂) during 60 min of submaximal exercise (RUN 1). An increase in V̇O₂ (at a fixed workload) may suggest a reduction in running economy, and therefore, a detrimental effect of CBD on aerobic exercise performance. However, the difference was relatively subtle (37_EX: + 38 ± 48 mL·min⁻¹, 50_EX: + 26 ± 55 mL·min⁻¹) compared to that of other interventions known to influence V̇O₂ (e.g. ketogenic diets: + 90—140 mL·min⁻¹ [40]; dietary nitrates: − 143 mL·min⁻¹ [41]), corresponding to only a 2.3% increase in estimated energy expenditure (+ 82 ± 121 kJ; based on the equations of Frayn [42]). Furthermore, it was not accompanied by an increase in HR or RPE, nor a decrease in TTE as one might anticipate [40]. In fact, in addition to V̇O₂, CBD appeared to increase V̇O_2max (+ 119 ± 206 mL·min⁻¹; + 1.5 ± 2.8 mL·min⁻¹·kg⁻¹). Thus, if present, any detrimental effect of CBD on running economy appears to be small and seems unlikely to impair aerobic exercise performance.

The potential mechanism by which CBD might have affected V̇O₂ responses in the present study is difficult to predict given its multiple pharmacological actions [22] and the limited amount of research in this field. Nonetheless, some mechanisms can be excluded. First, the current findings (and some previous research [43]) suggest that alterations in breathing patterns (i.e. RR, V_T, V_E) are unlikely to explain the observed increase in V̇O₂. Second, while participants could have modified their running biomechanics on the treadmill (e.g. stride frequency, strike pattern, ground contact time), changing muscle recruitment and therefore V̇O₂, this would not be expected to increase V̇O_2max. Third, while preferential utilisation of fat as a metabolic substrate can increase V̇O₂ [40], CBD did not appear to decrease submaximal RER. Fourth, while impairments in mitochondrial function (e.g. increased non-energetic proton leak across the inner membrane) can increase V̇O₂ [44], this would not be expected to increase V̇O_2max [45] (and would likely decrease TTE). Furthermore, most in vivo studies find that CBD improves aspects of mitochondrial function [46–49], although some in vitro (cellular) studies, usually involving high concentrations of CBD (e.g. > 5 µM), have reported detrimental effects [50–54].

One possible explanation for the observed increase in V̇O₂ and V̇O_2max is that CBD increased tissue blood flow via vasodilation. Indeed, studies have shown that nitrate- and ascorbic acid-induced vasodilation can increase skeletal muscle blood flow and V̇O₂ during hand-grip exercise [55, 56]. Findings from several preclinical studies also suggest that in vitro CBD treatment (i.e. ≤ 2-h exposure to 1–10 µM) can induce vasodilation in isolated arteries of rats [57–59] and humans [60]. A small clinical trial (n = 9) further reported that CBD (600 mg) reduced resting systolic BP (~ 6 mmHg) and total peripheral resistance during hand-grip exercise in normotensive males [61]. This mechanism could also explain why the observed increase in V̇O₂ was not accompanied by a clear increase in HR or RPE; nor decrease in TTE, as described above. That said, BP did not appear to be affected in the current trial.

In terms of psychological effects, CBD appeared to improve affect (i.e. ratings of pleasure) during the first 40-min of submaximal exercise (20_EX: + 0.7 ± 0.9; 40_EX: + 0.8 ± 1.1) but with no difference at 60_EX. It is important to note that the ratings obtained at 60_EX may have been influenced by the expectation of completing the exercise bout [62]; that is, participants “felt better” at 60_EX because they knew they were close to finishing the exercise task. Indeed, affect, which often decreases as exercise progresses [62], tended to increase from 40 to 60_EX on both treatments (Placebo: + 0.8 ± 1.5; CBD: + 0.1 ± 0.3), particularly placebo. Thus, the results at 60_EX should be interpreted with caution. The absence of any anxiolytic or other mood-altering effect of CBD should also be interpreted with caution as all participants recorded very low absolute scores on the STAI-S and POMS.

The mechanism via which CBD may improve affect during exercise is also difficult to predict. CBD has been shown to interact with a complex network of receptors and signalling pathways involved in mood regulation (e.g. 5-HT_1A, TRPV₁, PPARy, the cannabinoid type 1 receptor [CB1] [22]). An alternative possibility is that CBD, which has demonstrated some analgesic potential in humans [63], might have attenuated subjective feelings of pain, thereby increasing feelings of pleasure during exercise in the present study.

Exercise increased the serum concentrations of various biomarkers of systemic inflammation (i.e. TNF-α, IL-1β, IL-6), muscle damage (i.e. Mb, CK) and gastrointestinal damage (i.e. LPS) in the current trial. CBD appeared to suppress the exercise-induced increase in pro-inflammatory cytokine, IL-1β, Post-RUN 2 and 1-h Post-RUN 2. This is consistent with the robust anti-inflammatory effects of CBD in animal models [64]. However, it should be noted that (even post-exercise) serum IL-1β concentrations were very low (often needing to be extrapolated from the standard curve); effects should therefore be interpreted with caution. In addition, neither IL-6 nor TNF-α concentrations appeared to be impacted by CBD. Regarding muscle damage, CBD appeared to increase Mb concentrations 1-h Post-RUN-2, suggesting an exacerbation of muscle damage. This unexpected change could have been driven by a small increase in TTE (+ 39 ± 85 s) observed on the CBD treatment. While there is preliminary evidence suggesting that CBD may protect against some forms of GI damage [65], it did not appear to influence LPS concentrations in the current trial. Further research, employing more demanding exercise protocols (e.g. heat stress, eccentrically loaded exercise), may be required to better understand the anti-inflammatory and protective effects of CBD.

Several aspects of the current trial appeared to influence plasma AEA concentrations. First, all nine participants’ plasma AEA concentrations: (1) decreased from Baseline to Pre-RUN 1; and (2) increased from Pre-RUN 1 to Post-RUN 1, regardless of the treatment administered. These effects are likely to be due to breakfast consumption and the completion of submaximal exercise (RUN 1), respectively. Indeed, previous studies report that circulating AEA concentrations decrease post-prandially [66] and increase following submaximal exercise (e.g. ~ 70–85% HR_max) [67, 68] with the endocannabinoid system (in general) believed to contribute to the regulation of energy intake and storage [69]. Second, CBD appeared to decrease plasma AEA concentrations Post-RUN 2 relative to placebo. The only other study to have investigated the effect of CBD on circulating endocannabinoids, in fact, observed the opposite effect; that is, chronic CBD treatment (800 mg∙d⁻¹; 4 weeks) increased (resting) serum AEA concentrations in patients with schizophrenia [8]. Of course, the participant population, dosing regimen and experimental paradigm differed greatly between studies. While it is difficult to predict the mechanism by which CBD might have influenced AEA (and we cannot “rule out” an effect of the small increase in TTE observed on the CBD treatment), these findings add to a small body of evidence suggesting that CBD may modulate endocannabinoid tone.

This investigation does contain several limitations. First, as indicated above, the pilot trial was not formally powered to assess “effect”. It also generated a number of uncertain results; however, it is important to recognise that some degree of “uncertainty” is also often present in p-values > 0.05. Second, only male participants were recruited in this initial pilot study. As various physiological processes are influenced by the menstrual cycle, including substrate metabolism [70], this decision was made to minimise normal session-to-session variability and therefore maximise our capacity to detect effect (if present), given the limited sample size. Future studies should investigate the impact of CBD on both male and female physiology. Third, some participants had detectable albeit low levels of CBD (n = 2; < 0.8 ng·mL⁻¹) and(or) 7-COOH-CBD (n = 5; < 19.1 ng·mL⁻¹) in plasma on their placebo trial after receiving the active treatment ≥ 7-days prior. Although these low concentrations are unlikely to have had a meaningful effect on our results, future studies should consider extending the washout period between sessions (bearing in mind that this may require further standardisation of training and exercise behaviour). Fourth, plasma CBD concentrations remained relatively low during the initial stages of submaximal exercise (RUN 1); future studies, using similar formulations, may therefore benefit from delaying the start of exercise to better capture the observed T_max. Fifth, the CBD dose used in this investigation (300 mg) was relatively high (for a healthy population) and was selected to gain initial insights into CBD’s effects. Future investigations may consider using doses that more closely resemble products available to (and used by) the general and athlete population.

These preliminary results suggest that acute, oral CBD treatment has the potential to alter key physiological and psychological responses during aerobic exercise. Indeed, its effects on V̇O₂ responses, feelings of pleasure during exercise and exercise-induced inflammation appear worthy of further investigation. The absence of a clear detrimental effect on RPE, TTE and V̇O_2max also suggests that CBD is unlikely to impair aerobic exercise performance in endurance-trained males and may therefore have utility within the sporting context. Further research, involving a larger participant sample and different dosing regimens (e.g. chronic treatment, lower doses), is required to confirm and better understand these initial observations.

The authors would like to sincerely thank all participants for their contribution to the study. We would also like to thank Anastasia Suraev for her assistance with participant randomisation.

AS, CI, AJC, BD, NAJ, MH, PSH, ISM and DM contributed to the study conception and design. Data collection was performed by AS, AS and DM. Blood sample analyses were conducted by PTD, RCK and AJC. Data analysis was performed by AS and DM. Results of the experiment were interpreted by AS, CI, PTD, RCK, AJC, BD, NAJ, AS, ISM and DM. The first draft of the manuscript was written by AS and DM, and all authors contributed to previous versions of the manuscript. All authors read and approved the final manuscript.

This research project was funded by the Lambert Initiative for Cannabinoid Therapeutics, a philanthropically funded centre for medicinal cannabis research at the University of Sydney.

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.