Archive restoration — 26 September 2026: This article has been restored from our original source files. Its publication date and reporting context remain unchanged; it is not a new update on current policy, availability or clinical guidance.
A statistically significant result is not automatically a life-changing one. A new trial of high-dose cannabidiol for neuropathic pain after spinal-cord injury demonstrates exactly why that distinction matters.
Published in EClinicalMedicine, the Australian study found that purified oral CBD reduced pain more than placebo during six weeks of treatment. The trial was randomised, double-blind and placebo-controlled, and each participant received both treatments in a crossover design.
Those are meaningful strengths. So is the finding that 37.8% of participants achieved a reduction in pain of at least 30% with CBD, compared with 11.1% on placebo.
The average difference between treatments, however, was 0.54 points on a 0-to-10 pain scale. Participants took up to 800 mg of CBD a day—far more than the amount found in most high-street oils. Most measures of sleep, mood, pain interference and quality of life did not improve specifically with CBD.
The most accurate conclusion is therefore narrower than either side of the cannabis debate may prefer. This small trial identified a credible signal that high-dose, pharmaceutical-grade CBD may help a subset of people with a specific form of neuropathic pain. It did not establish CBD as an effective general treatment for chronic pain.
What the trial tested
Researchers in Sydney recruited adults with a complete or incomplete spinal-cord injury and neuropathic pain lasting at least three months. Forty people entered the study and 38 were included in the primary analysis. Their average age was about 55, 32 were men and six were women.
The participants were not treatment-naive. They were taking an average of 2.4 pain medicines, and more than three-quarters used a gabapentinoid. Other treatments included paracetamol, baclofen, tricyclic antidepressants and opioids. Their existing medicines remained stable during the trial.
Each participant completed two six-week treatment periods: one with 99% pure oral CBD and one with a matching placebo. A four-week washout separated the periods. The order was randomised, and participants, clinicians, researchers and the dispensing pharmacist were masked to it.
CBD was supplied in 200 mg capsules. The daily dose increased from 200 mg on days one to four, to 400 mg, then 600 mg, reaching 800 mg from day 13 to day 42. Participants took the capsules with food, which can increase CBD absorption.
Pain was recorded on a 0-to-10 visual scale three times a day on alternate weekdays. This repeated measurement is useful because neuropathic pain can change across a day. In this study it was lowest in the morning and highest in the evening.
The trial excluded people with recent cannabis use, unstable pain medication, relevant medicine interactions, substance dependence, major psychiatric illness or suicidality. Those safeguards made the experiment cleaner, but they also mean the result cannot be assumed to apply to every person with spinal-cord injury or chronic pain.
The main result was real but modest
Average pain intensity during active CBD treatment was 3.82, compared with 4.36 during active placebo treatment. The adjusted difference was 0.54 points in favour of CBD, with a 95% confidence interval from 0.21 to 0.88 points.
The difference was statistically significant under the study’s model. That is evidence against a no-difference hypothesis; it is not the probability that the result arose by chance. It does not tell us on its own whether the average patient would consider the change worthwhile.
Between the baseline week and the final treatment week, pain fell by an average of 14% with CBD and 6.5% with placebo. The researchers describe the CBD change as modest and below the 30% reduction commonly used as a marker of moderately important improvement in chronic-pain trials.
That average also hides different individual experiences. Fourteen of 37 participants with the required data—37.8%—reached a reduction of at least 30% with CBD. Four participants, or 11.1%, reached the same threshold with placebo. The difference was statistically significant.
Responder rates did not differ significantly at the study's 10%, 20% or 50% thresholds. Because this was a post-hoc analysis rather than the pre-specified primary endpoint, it should be treated as a promising clue to confirm in a larger trial, not a guaranteed response rate for patients.
The practical reading is that the average effect was small, while a minority appeared to experience a more substantial benefit. Future research needs to establish whether that subgroup is reproducible and whether clinicians can identify likely responders before exposing everyone to a high dose and its cost.
Most secondary outcomes did not improve
Pain intensity is important, but chronic neuropathic pain affects more than a number on a scale. The trial also measured the character of neuropathic pain, pain interference, catastrophising, sleep, depression, anxiety, stress and aspects of quality of life.
Most of those outcomes did not show a CBD-specific benefit. There was no significant treatment effect on sleep quality, pain interference, stress, pain catastrophising or the main anxiety measure. Measures of depression and quality of life did not improve with CBD.
This does not cancel the primary result. Trials should not be declared negative simply because every secondary outcome fails to move. It does limit what can be said about the wider value of the treatment.
A medicine that lowers pain intensity by half a point but does not improve sleep, daily function or wellbeing may still matter to an individual. It is not the same as a treatment that reliably restores participation or makes life broadly easier. A larger study should define both outcomes before recruitment and measure them for longer.
The researchers also reported that people with longer-standing injuries and pain tended to show greater pain reduction. That association was exploratory, came from a small sample and does not yet provide a basis for selecting patients.
This was not retail CBD oil
The study's dose is central to the story.
Participants took purified CBD capsules at up to 800 mg a day under research supervision. Many shop-bought products contain only a small fraction of that amount in an entire bottle. Their composition and absorption can vary, and a label saying “CBD” does not make two products clinically equivalent.
The NHS warns that online or retail CBD products are not guaranteed to be of good quality or to provide a health benefit. Some may contain unexpected THC or other ingredients. The trial cannot be used to validate those products, their doses or their marketing claims.
Nor is 800 mg a sensible self-experiment. CBD can interact with other medicines and affect liver function. This population was already using multiple drugs, including medicines acting on the nervous system. The researchers screened for relevant interactions and kept treatment stable; a customer buying oil does not receive that trial infrastructure.
The study also tested CBD alone. It says nothing directly about THC-rich cannabis flower, balanced THC/CBD products or vaporised medicines. Findings from one formulation and route should not be transferred to another as though “cannabis” were a single intervention.
What the safety data showed
Twenty-six participants reported adverse events during CBD treatment and 20 did so during placebo. There were 67 events with CBD and 51 with placebo, although the event rate per affected participant was similar between treatments.
The most frequently reported issues during CBD treatment were tiredness or sleepiness, nausea, diarrhoea, feeling unwell, loss of appetite and stomach discomfort. One person withdrew because of mild nausea while taking CBD. Another withdrew during placebo after several symptoms.
Three more serious events occurred across the trial: a urinary-tract infection and a femur fracture during CBD treatment, and a stroke during the washout after CBD. Investigators judged none to be caused by the treatment, and all three participants recovered.
Overall, the six-week CBD period appeared reasonably tolerable within this small, screened group. It was not large or long enough to define uncommon harms, sustained adherence, liver effects or the consequences of combining high-dose CBD with the many medicines used in routine care.
Another limitation is masking. Just over half of participants correctly guessed when they were receiving CBD, close to chance. But 78% correctly identified placebo. If participants expected less improvement when they believed they had placebo, that could have influenced self-reported pain.
Residual CBD and metabolites were also detectable in some participants after the four-week washout. The researchers adjusted for treatment order and argued that the remaining concentrations were too low to be therapeutic, but crossover studies are most straightforward when the first treatment has cleared completely.
How it fits with the wider evidence
Evidence for CBD alone in neuropathic pain has been limited and inconsistent. Earlier randomised trials using substantially lower oral doses did not show a clear reduction in pain. Trials of THC, THC/CBD combinations and whole-plant products answer different questions and should not be pooled casually with pure CBD.
A 2022 meta-analysis focused on cannabinoids after spinal-cord injury concluded that the available randomised evidence did not show a significant reduction in pain and called for larger, better-designed studies. The new trial adds a more rigorous test of high-dose CBD, but 38 participants are not enough to settle the question.
Its crossover design is efficient because each person acts as their own control. That can reduce the noise caused by differences between patients. The trade-off is the possibility of carryover, unmasking and a study experience that differs from long-term treatment in ordinary practice.
The next trial should be larger, multicentre and long enough to examine function, sleep, medicine reduction and sustained safety. It should pre-specify responder analyses, include a more representative balance of women and men, and test whether benefit survives after treatment stops.
It should also address value. Pharmaceutical-grade CBD at 800 mg a day can be expensive. A modest average benefit may look different once cost, monitoring, medicine interactions and treatment burden are included.
What it means for patients in the UK
The study does not change UK guidance.
NICE currently says CBD should not be offered for chronic pain in adults unless it is part of a clinical trial. For neuropathic pain in non-specialist settings, its recommended initial options remain amitriptyline, duloxetine, gabapentin or pregabalin, with switching when a treatment is ineffective or not tolerated.
That guidance can be frustrating for people whose pain remains severe despite trying several medicines. This trial is relevant precisely because current treatment is often inadequate. But a research signal is not the same as a new standard of care.
Someone already receiving prescribed cannabis-based treatment should not increase CBD to match the study dose. The formulation, interaction profile and clinical circumstances may be different. Any change should be discussed with the prescriber, particularly where medicines for epilepsy, blood thinning, mood, sleep or pain are involved.
People using a high-street CBD product should not interpret the paper as proof that their oil treats neuropathic pain. The trial tested a controlled medicine at a pharmaceutical dose in a narrowly defined group.
Our broader review of THC and chronic-pain evidence explains why cannabinoid, formulation, dose and outcome all need to be named before a result can be applied to care.
The editorial view
This is the kind of result cannabinoid medicine needs: controlled, transparent and specific enough to be tested again.
It is also the kind most vulnerable to being flattened into a misleading headline. “CBD works for pain” discards the small sample, the 800 mg dose, the narrow diagnosis and the modest average effect. “CBD failed because the difference was small” discards the significant responder signal and the burden faced by people with spinal-cord-injury pain.
The responsible position sits between those claims. High-dose CBD produced a measurable reduction in pain and may have delivered a clinically meaningful benefit to a subset of participants. Most broader outcomes did not improve, the average change was limited and the trial needs replication.
Progress does not require pretending that an early positive study is definitive. It requires protecting the exact finding long enough for the next study to discover whether it is real, durable and useful in practice.
Key takeaways
- A randomised, double-blind crossover trial compared high-dose oral CBD with placebo in 38 adults with neuropathic pain after spinal-cord injury.
- Participants took up to 800 mg of purified CBD a day for six weeks—far more than typical retail CBD products.
- Average pain was 0.54 points lower with CBD than placebo on a 0-to-10 scale; the researchers characterised the effect as modest.
- A 30% or greater reduction in pain occurred in 37.8% of participants with CBD and 11.1% with placebo, but this responder analysis was post hoc.
- Most secondary measures, including sleep, pain interference, stress and quality of life, did not show a CBD-specific improvement.
- Adverse events were mostly mild, with tiredness, nausea and digestive symptoms among the most common during CBD treatment.
- NICE does not currently recommend CBD for chronic pain outside a clinical trial. The study does not justify self-dosing or validate high-street CBD oil.
Sources
- High-dose cannabidiol for chronic neuropathic pain associated with spinal cord injury: a randomised clinical trial — EClinicalMedicine
- Full open-access trial report — PubMed Central
- Trial registration ACTRN12622000634774 — Australian New Zealand Clinical Trials Registry
- Cannabis-based medicinal products: recommendations — NICE guideline NG144
- Neuropathic pain in adults: pharmacological management — NICE guideline CG173
- Medical cannabis — NHS
- Cannabinoid use for pain reduction in spinal cord injuries: meta-analysis — Frontiers in Pharmacology
