This study sought to observe relative changes to eCBs and NAEs in the serum of HVs following oral phytocannabinoid administration in a retrospective investigation that measured matched time-points from two completed clinical trials (LOGIN-TS4, GEI-TCP-II). We postulate that the observed changes for CBD|800 mg (AEA, OEA and PEA), Δ⁹-THC (AEA) and in-combination (CBD|800mg+Δ⁹-THC|20 mg; AEA, OEA and PEA) represent mechanistically relevant responses, given no significant influences in their endogenous levels were detected prior to phytocannabinoid administration.

Increased levels of AEA from CBD|800 mg were consistent with our prior findings for CBD treatment of schizophrenia, leading to increased sera AEA associated with improved clinical symptoms.¹⁰ Reports suggest CBD-induced increase is caused by inhibiting FAAH, AEA’s principal metabolising enzyme, verified in rat brain membranes,¹⁰ mouse brain microsomes¹² and COS-7 cells expressing rodent FAAH.¹² However, CBD has not been reported to inhibit FAAH in humans, rather blocking fatty acid binding proteins (FABPs) intracellular transport of AEA to FAAH at the endoplasmic reticulum.¹² Although the relevance of this mechanism remains conjectural, CBD’s capacity to block FABPs-mediated transport to FAAH would also lead to increases in OEA and PEA, given these endogenous lipid mediators share this conserved hydrolysis pathway with AEA.¹² Significant increases in OEA and PEA were observed following CBD administration, aligning with the proposed FABPs–FAAH mechanism of inhibition (figure 2E,F).

Though evidently more complex, it is widely acknowledged that NAEs follow a canonical biosynthesis pathway initiated by the N-acylation of membrane glycerophospholipids.2022 This process results in the formation of N-acyl-phosphatidylethanolamine (NAPE) precursors, which undergoes specific phospholipase D type (NAPE-PLD) hydrolysis, with N-arachidonoyl-phosphatidylethanolamine, N-oleoyl-phosphatidylethanolamine and N-palmitoyl-phosphatidylethanolamine each being converted into AEA, OEA and PEA, respectively.^{23 24} Given that NAPE-PLD is integral to the NAE metabolic process, an alternative, although speculative, viewpoint is that CBD induces concomitant increases in AEA and its congeners through increased biosynthesis. Indeed, CBD has been shown to promote the upregulation of NAEs in multiple brain regions of rats via an NAPE-PLD-dependent mechanism,²⁵ as well as enhance mRNA expression of NAPE-PLD in syncytiotrophoblasts.²⁶

Increases in AEA, OEA and PEA occurred only from the administration of CBD at 800 mg, alone or in combination with Δ⁹-THC. This metabolic response aligns with the dose-dependent clinical efficacy of CBD. In our previous report, 800 mg of CBD significantly improved schizophrenia symptoms,¹⁰ while clinical trials reported CBD at 600 mg, as an add-on therapy, was insufficient to improve cognitive function or attenuate psychotic symptoms,²⁷ though the latter study did observe increased connectivity during resting state.²⁸ In support of a ≥800 mg threshold, CBD as an add-on to conventional antipsychotics does exhibit clinical benefit at 1000 mg,¹¹ while displaying no positive effects when given as a CBD/Δ⁹-THC mix (Bedrolite, with CBD ca. 300 mg/day) to patients with a psychotic disorder and comorbid cannabis use.²⁹

CBD’s poor oral bioavailability (estimated at 4%–6%³⁰) could explain why benefits and induced effects on AEA/NAEs require the high dosage, as 800 mg CBD exhibits a faster accrual, over 600 mg (figure 2A). This same rationale may also account for comparable effects towards AEA following Δ⁹-THC at different doses, where their accumulation rates and subsequent impact on AEA were similar (figure 2B). Analogous declines in AEA concentrations of the CSF have been reported in studies comparing heavy (>10 times/month) to light (<10 times/month) cannabis consumers without any diagnosis of psychopathology according to Diagnostic and Statistical Manual of Mental Disorders, version IV (DSM-IV) criteria,⁶ as well as in individuals diagnosed with schizophrenia and higher consumption of cannabis in their lifetime (>20 times/life) compared with those with no or sporadic use (<5 occasions in their lifetime).⁷ However, it should be noted that in the latter study, a subgroup of high-frequency cannabis users with schizophrenia tested positive for urinary Δ⁹-THC still displayed elevated AEA in the CSF, though statistical significance was not reached. Therefore, the overall lower AEA levels in high-frequency cannabis users diagnosed with schizophrenia may not be attributed to recent cannabis use.

Previous reports on alterations in circulating AEA due to Δ⁹-THC and cannabis consumption have shown considerable variability. The aforementioned studies found no effects on human sera,^{6 7} while others reported increases in AEA alongside mixed effects on OEA and PEA levels in participants undertaking³¹ or undergoing cessation³² from high-cannabis use. Lower AEA concentrations have been observed in the plasma of individuals with psychosis who self-reported high-cannabis use,³³ whereas higher AEA levels were found in individuals with cannabis use disorder, accompanied by concomitant increases in OEA.³⁴ Clinical evidence supports elevated eCB/NAE response following acute, oral administration of Δ⁹-THC at a dosage identical (20 mg) to a treatment arm used in this study, with results displaying higher levels of AEA alongside 2-AG and OEA in plasma 2–3 hours post in-take.³⁵ Conversely, a longer clinical trial reported a decline in plasma AEA over 28 days in participants undergoing cessation for their cannabis use disorder, with stable AEA levels in participants that received 800 mg of CBD, while this effect was absent in those receiving 400 mg of CBD.³⁶ The contradictory reports of AEA/NAE levels may be a consequence of their biphasic response towards Δ⁹-THC, which is postulated to be mediated by synthesis and/or degradation processes^{37 38} or through catecholaminergic and glucocorticoid signalling that, in turn, promote eCB/NAE concentrations.3741 Thieme et al reported that plasma concentrations of AEA and 2-AG undergo repeated changes (both upregulation and downregulation) at multiple time points up to 48 hours post intravenous injections of Δ⁹-THC.³⁸ This infers that those acute investigations, including this study, may have limitations in interpreting the effects of Δ⁹-THC on AEA and its congeners. Additionally, formulation and mode of administration could also impact the response to Δ⁹-THC^{16 42} and its combination with CBD,³⁷ as discussed below.

OEA and PEA concentrations appeared to plateau with CBD-alone treatment, suggesting that the maximal affinity for inhibiting their degradation may have been reached. This is postulated to be due to either a higher affinity of FAAH for AEA⁴³ or the metabolism of OEA and PEA via alternative pathways, such as N-acylethanolamine acid amidase (NAAA),⁴⁴ which is not primarily responsible for the hydrolytic deactivation of AEA,⁴⁵ nor inhibited by CBD.⁴⁶ Additionally, NAPE-PLD is not exclusively responsible for the formation of NAEs.⁴⁷ Alternative pathways involve phospholipase C, phospholipase A or α/β Hydrolase Domain-Containing Protein 4 catalytic hydrolysis, as reviewed by Simard et al.⁴⁸ While the importance of these multiple pathways remains unknown, they may facilitate the differential synthesis of AEA, OEA and PEA, potentially varying based on the specific tissue or physiological process involved.⁴⁹ Differential effects have also been noted across the overall pathway, with preclinical models recognising FAAH as a critical step for regulating NAEs in the brain but not heart tissue.⁵⁰ Hence, identifying the source(s) of our circulatory AEA/NAE alterations could elucidate relevant aspects or overlapping mechanisms of the metabolic pathway contributing to the dose-dependent changes observed with CBD and Δ⁹-THC. However, resolving the origin of circulating eCBs/NAEs presents a challenge, owing to their lipophilic nature, which means they are identifiable in almost every tissue and produced by nearly all cell types.⁵¹

In this study, NAEs continued to increase, alongside a greater upturn of AEA, from the combination of CBD+Δ⁹-THC, leading us to speculate a positive contribution towards CBD-derived mechanisms occurs with the inclusion of Δ⁹-THC (‘entourage effect’).⁵² Though beyond this project’s scope, prior evidence of CBD+Δ⁹-THC therapy supports analgesic efficacy⁵³; however, safety and tolerance of Δ⁹-THC remain debated.⁵⁴ Furthermore, single-dose administration of CBD+Δ⁹-THC through inhalation yielded no effect on eCB/NAE levels.³⁷ It is important to note that doses administered by Chester et al were designed to reflect recreational cannabis use, with CBD doses of only up to 30 mg. Additionally, comparisons between formulations are challenging due to the delayed absorption and metabolism from the first-pass effect of oral phytocannabinoid administration.⁵⁵ Refinement of Δ⁹-THC dosage and pharmacokinetics to support CBD efficacy for certain indications while minimising adverse effects warrants future investigation.

Our study has several limitations. Though endogenous effects were observed, our sample size remains relatively small. This may account for eCB/NAE changes higher in dosage (e.g., Δ⁹-THC 20|mg effects on AEA) or magnitude (CBD|800 mg+Δ⁹-THC|20 mg response on PEA at 65 min) did not reach statistical significance. The exploratory nature of combining randomised clinical trials, along with the increased likelihood of variation in results when working with small sample sizes, constrains our comparative measures to within-treatment arms. This underscores the necessity of addressing these limitations in expanded studies to authenticate our observations. In particular, clinical trials that allow for more controllable comparisons between CBD and Δ⁹-THC (including combined effects), increased sampling times that better encapsulate endogenous alterations or validate non-apparent responses observed herein (eg, CBD|600 mg) and, if permissible, higher concentrations or longer administrations to clarify whether endogenous effects are dose-dependent or reflect a pre-existing state. Influences on endogenous eCBs/NAEs, and the required dosage to achieve them, may also have been affected by food intake. Phytocannabinoid bioavailability increases with high-fat meals, likely due to their lipophilicity requiring intestinal lumen solubilisation before absorption.⁵⁶ Although participants in both studies received standardised meals, they were not identical. This may have slightly belated the absorption of CBD and Δ⁹-THC between the studies (GEI-TCP-II and LOGIN-TS4), delaying any subsequent effects on AEA and NAEs. However, this would not account for differences between CBD-alone versus combination (CBD+Δ⁹-THC), which were controlled in the same study and administered with an identical dosage of CBD. Finally, we also acknowledge that this study does not address the question of whether the observed peripheral changes are indicative of a cerebral response. Although our observations are consistent with the reported clinical benefits of CBD^{10 11} and Δ⁹-THC effects on the CSF,^{6 7} response in peripheral systems may differ from cerebral effects, as there is evidence of a lack of association between these two biofluids.^{57 58} To the best of our knowledge, no comparative investigations have yet to assess CBD’s effects on the eCBs and NAEs in the CSF alongside paired peripheral fluids.

In summary, our data supports both previous clinical and mechanistic evidence for AEA’s and NAEs’ response to phytocannabinoids. Our results could reflect a dose-dependent metabolic signature for CBD, with potential enhancement from Δ⁹-THC, encouraging future investigations into these endogenous lipid mediators as indicative markers of CBD efficacy, both as a stand-alone or co-administered cannabinoid therapy.