A cannabidiol study in rats with spinal cord injury reported better motor scores within a week. People often quote that result as though it pointed toward repairing human spinal cords. It does not. What human trials in spinal cord injury have actually studied, and in a few small cases shown, is something more modest and more useful: spasticity.
Nothing here is medical advice. No treatment, cannabinoid or otherwise, restores function to a damaged human spinal cord. In Thailand cannabis flower is a controlled herb dispensed only on prescription form PT 33.
1. The Animal Study
The study that cannabis literature cites is real. It is Kwiatkoski et al., Neurotoxicity Research (2012; published online 2011, DOI 10.1007/s12640-011-9273-8), from the University of São Paulo in Brazil. The researchers created a cryogenic spinal cord lesion in rats by applying liquid nitrogen to the exposed cord at the T10 level for five seconds. The animals received vehicle or CBD 20 mg/kg immediately before surgery, three hours afterwards, and daily for six days. On day 7 the CBD group scored about 7 on the BBB locomotor scale and the vehicle group about 4. The CBD group also had less tissue injury. The authors suggested CBD could be useful in the treatment of spinal cord lesions.
That is a legitimate preclinical neuroprotection result, and it belongs to a broader literature: CBD and THC have both shown neuroprotective effects in laboratory models of acute central nervous system injury, including stroke models.
2. Why It Does Not Translate as Written
Three features of the design limit what it can support. All three are usually dropped when the result is quoted.
Timing. The first dose was given immediately before the injury was created and the second three hours after. This is a neuroprotection protocol, meant to reduce the extent of damage as it occurs. It is not a treatment for an established injury, and it cannot tell us what happens to a person injured last year or last decade.
Species and injury. A freezing lesion made under controlled conditions in a rat is not a human traumatic spinal cord injury. Motor recovery after cord injury in rodents also differs substantially from recovery in humans.
The general translation problem. Doses and exposures used in animal studies often do not match what is achievable or tolerable in people. Preclinical neuroprotection is also a well-known area of translational failure: many compounds have protected rodent brains and spinal cords and then failed in human trials.
The honest description of the São Paulo result: a preclinical finding worth pursuing, with no human evidence behind it.
3. What Human Trials Have Actually Tested
Not repair, but spasticity. The evidence is small, but it comes from people with spinal cord injury.
Spasticity after spinal cord injury is involuntary muscle stiffness and spasm below the level of injury. It interferes with transfers, positioning, sleep and skin care. In a substantial share of patients, existing drugs control it only partly.
The SCIRE Project (Spinal Cord Injury Research Evidence) has online evidence modules on spasticity and on cannabinoids that summarise this literature. The four studies most often cited are:
| Study | Design (n) | Intervention | Result |
|---|---|---|---|
| Kogel 1995 | Open-label pre-post, n = 5 men with tetraplegia | Oral dronabinol, escalated from 5 mg twice daily to 20 mg three times daily | Mixed: two improved markedly, one fluctuated, one unchanged, one worsened |
| Hagenbach 2007 | Open-label phase (n = 22), then double-blind placebo-controlled phase (n = 15) | Oral dronabinol, rectal THC | Reduced spasticity scores during active treatment; SCIRE rates the controlled phase as methodologically compromised |
| Pooyania 2010 | Double-blind, placebo-controlled crossover, n = 12 enrolled, 11 completed | Nabilone 0.5 mg once daily, optional increase to 0.5 mg twice daily (0.5–1.0 mg/day), 4 weeks | Significant decrease in Ashworth spasticity score (most involved muscle and total); no significant difference on other measures; mild side effects |
| Wilsey 2016 | Crossover: placebo, 2.9% and 6.7% THC vapour, n = 42. A neuropathic pain trial; spasticity was a secondary outcome | Vaporised cannabis (four puffs per session) | 2.9%: spasticity reduction at some timepoints only. 6.7%: did not show a significant spasticity effect |
SCIRE uses its own grading system, in which “Level 1” and “Level 2” refer to the type of study design, not to how large or certain an effect is. SCIRE grades nabilone as Level 1b evidence from a single RCT (Pooyania, 11 completers) and dronabinol as Level 2 (one compromised RCT plus one five-person pre-post study). SCIRE’s cannabinoid module describes nabilone and 2.9% THC vapour as shown to be effective for spasticity. Each of those rests on one small crossover trial, and in the vapour trial spasticity was not the main question. SCIRE itself calls the results preliminary and says more research is needed.
Two details in that table repay attention.
One patient in the 1995 series got worse. Of five participants, one deteriorated. A group average can hide individual responses that go in different directions.
Only the lower THC concentration showed a spasticity signal. At 6.7% THC there was no significant spasticity effect. The trial was designed and powered for pain, not spasticity, so this does not show that the higher dose was worse or that a particular dose-response curve exists. It does mean a stronger concentration should not be assumed to work better.
4. A Mechanism Claim to Treat with Caution
Some older literature suggests that cannabis eases spasticity in spinal cord injury through an antispasmodic action rather than general muscle relaxation. This is an observation and a hypothesis. The small trials above were not designed to test it, and they do not establish it.
If it were true, it would matter in practice. A general muscle relaxant weakens everything, including muscles that a person with an incomplete injury still relies on for transfers, and including the muscles used for breathing. A more selective effect might reduce spasm without that global weakness.
There is no guarantee of that. Dizziness, sedation and a feeling of weakness are common side effects of THC and nabilone, and any of them can make transfers less safe. Whether a cannabinoid reduces spasm without reducing useful strength has to be judged for each person, with the treating team.
Spasticity is one of the areas where cannabinoid evidence in neurology is stronger, mainly in multiple sclerosis. The same measurement problems apply in spinal cord injury: patient-rated improvement is often larger than objective measures show, and blinding is hard with a drug whose effects people can feel.
5. Where This Leaves Things
| Question | Answer |
|---|---|
| Does a cannabinoid repair spinal cord injury? | No. The animal work is neuroprotection given around the moment of injury |
| Does it restore lost function? | No human evidence |
| Does it reduce spasticity? | Limited evidence: one small RCT of nabilone (11 completers) and a secondary outcome at limited timepoints for 2.9% THC vapour in one pain trial (n = 42); more research needed |
| Does it help neuropathic pain after SCI? | The 2016 vapour trial, designed for pain, reported an analgesic response; this fits the broader, still limited, neuropathic pain evidence |
| Is response uniform? | No. One documented patient worsened, and the higher THC concentration showed no significant spasticity effect |
6. Practical Notes
If spasticity after spinal cord injury is the target, take it to the rehabilitation physician or neurologist managing the case. Cannabinoids would sit alongside the existing options (baclofen, tizanidine, botulinum toxin, intrathecal baclofen), not replace them.
Two cautions specific to this population. Orthostatic hypotension is already common after spinal cord injury, particularly with higher-level lesions, and THC can lower blood pressure on standing. Together they are a real concern for anyone doing transfers or using a standing frame, especially combined with dizziness or sedation. Respiratory function is also reduced with higher lesions, which is a further reason to avoid smoking as a route.
7. In Thailand
Cannabis flower is a controlled herb under the Notification of the Ministry of Public Health Re: Controlled Herbs (Cannabis), B.E. 2568, issued 26 June 2025. It can be dispensed only against a prescription on form PT 33 from an authorised practitioner, for no more than 30 days, single use with no refills, from a GACP-certified source. Online sale and advertising of flower are not allowed, and recreational use is not permitted.
The tested medicines. Nabilone and dronabinol are the products with controlled trial data in spinal cord injury. We could not confirm a Thai FDA registration for either, so do not assume either can be prescribed or obtained here. Ask the treating physician what is actually available.
The vapour result is not a recommendation to vape. The 2.9% THC finding comes from a laboratory trial with a standardised device and dose. It is not advice to vaporise flower. Thailand bans the import and sale of e-cigarettes, and vaping devices are restricted. Any cannabis use for spasticity should go through a physician and a PT 33 prescription, with the route and dose the physician decides.
If you live with a spinal cord injury and have used a prescribed cannabinoid for spasticity: did it reduce the spasms specifically, or did it reduce everything, and did that affect your transfers? Trials have barely measured this difference, and it is something patients can describe.
Main sources: SCIRE Project, Cannabinoids for reducing spasticity after SCI and Cannabinoid Use in SCI (scireproject.com). Kwiatkoski et al., Neurotoxicity Research 2012, DOI 10.1007/s12640-011-9273-8. Pooyania et al., Archives of Physical Medicine and Rehabilitation 2010;91(5):703–707.
This is general educational information, not medical advice. No treatment restores function to a damaged spinal cord.