The compounds that give cannabis its distinctive aroma may offer a path to pain relief without the psychoactive side effects, but the evidence so far comes from mice, not humans.
Researchers at the University of Arizona College of Medicine have shown that four terpenes, aromatic compounds found in cannabis and many other plants, produce measurable pain relief in mouse models of post-operative pain and fibromyalgia. Importantly, the effect works through adenosine A2A receptors, not cannabinoid receptors, meaning there is no “high” and no potential for abuse.
The study, published in Pharmacological Reports, tested geraniol, linalool, β-caryophyllene, and α-humulene, all terpenes that are FDA-listed as Generally Recognized as Safe (GRAS) food additives.
The researchers induced pain in mice using two established models: a paw incision (simulating post-operative pain) and reserpine injection (a pharmacological model of fibromyalgia). They then injected the terpenes (200 mg/kg, intraperitoneal) and measured mechanical sensitivity using von Frey filaments.
All four terpenes reversed mechanical allodynia, the pain response to normally non-painful touch, with peak effect at 60,90 minutes post-injection.
The pain relief was completely abolished when the mice were pre-treated with istradefylline (3.2 mg/kg), a selective adenosine A2A receptor antagonist. This proves the effect is specifically A2A-receptor-dependent, not mediated by sedation, opioid receptors, or cannabinoid receptors (CB1/CB2).
“This is the mechanism,” says John Streicher, the study’s senior author. “These terpenes are not cannabinoids. They don’t produce a high. They target a different pathway entirely.”
The hot plate test (measuring latency to react to heat) showed no change, confirming that the reduced pain sensitivity was not due to motor impairment or general sedation.
Which terpene worked best
Geraniol, found in roses, citronella, and lemongrass, had the strongest effect, followed by linalool (lavender, basil) and α-humulene (hops, sage). β-caryophyllene (black pepper, cloves) was the weakest but still significantly above baseline.
The four terpenes are found in many common plants beyond cannabis. Geraniol, for instance, is a major component of rose and citronella oils and has a long history of use in food and cosmetics.
Why it matters
Chronic pain affects approximately one in five adults worldwide, and opioid addiction remains a public health crisis. Terpenes, naturally occurring, widely available, and already approved for food use, represent a potential new class of analgesics that could bypass the addiction liability of opioids and the psychoactive effects of cannabinoids.
“The fact that these compounds are already GRAS-listed means they could potentially move into human trials more quickly than a new chemical entity,” notes Caleb Seekins, the study’s first author.
Streicher’s lab has since tested sublingual administration in mice with positive results, suggesting a possible route for human use.
The limits
This is preclinical mouse data, no human trials have been conducted. The terpene dose used (200 mg/kg, injected) is high, and the human-equivalent dose is unknown.
The study tested only mechanical allodynia (pain from touch), not thermal pain, spontaneous pain, or affective (emotional) aspects of pain. Only a single dose was tested, with no dose-response curve or pharmacokinetic data. There are no data on chronic use, tolerance development, or oral bioavailability.
Streicher holds equity in Teleport Pharmaceuticals and Botanical Results and serves as a consultant for Black Rock Nutraceuticals, though none of these entities were involved in the study.
What’s next
The University of Arizona team is working toward human trials, with sublingual and oral formulations under investigation. The GRAS designation of these terpenes could shorten the regulatory pathway, but clinical testing for pain as a medical indication still requires full FDA approval.
Source
Seekins CA, Welborn AM, Schwarz AM, Streicher JM. “Select terpenes from Cannabis sativa are antinociceptive in mouse models of post-operative pain and fibromyalgia via adenosine A2a receptors.” Pharmacological Reports, 77(1), 172,181 (2025). DOI: 10.1007/s43440-024-00687-1
Disclosure: Based on a preclinical (animal) study. No human trials have been conducted.
Funding: NIH award R01AT011517.