Answers are grounded in this paper and cite their sources.
Preclinical and clinical studies suggest that cannabidiol (CBD) holds therapeutic promise across a range of indications. However, a major barrier to realizing CBD's full therapeutic potential is its inherently low oral bioavailability, partly due to its poor aqueous solubility. To address this, various oral delivery technologies have been explored to enhance the kinetic and apparent solubility of CBD. However, these 'solubilizing' strategies, may inadvertently increase the risk of acid-catalyzed conversion of CBD to tetrahydrocannabinol (THC) in the gastric environment. The possibility of CBD converting to THC in the acidic gastric environment has been debated for many years. Notably, the study by Merrick et al., which reported significant conversion in sodium dodecyl sulfate (SDS)-modified simulated gastric fluid (SGF), raised concerns that oral CBD products might lead to THC intoxication. Yet, other SGF-based studies and human data consistently indicate minimal conversion. We critically revisit the CBD-to-THC literature and discuss the methodological pitfalls associated with using SDS-modified SGF. We propose three conditions required for CBD-to-THC conversion: low pH, CBD accessibility to the acidic environment, and stabilization of the protonated carbocation intermediate essential for ring closure converting CBD to THC. We recommend evaluating micellar CBD formulations containing anionic surfactants like dodecyl sulfate or similar that meet these conditions in acidic SGF. We also discuss the differences between simulated and physiological gastric conditions, including pH, retention time, composition, and dynamic processes, that are likely to limit conversion risk in vivo, although detected in SGF. Overall, our literature review, comparison of human gastric conditions with in vitro SGF models, and our contextualizing of the Merrick et al. study collectively support that concerns about CBD-to-THC conversion in humans remain, at present, a speculative scenario.
No takes yet. Share an insight, caveat, or question.
(1953121) et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: