Different cannabis-based medicines have been studied for different conditions. Read what the studies found, who took part and where the findings fall short. Follow them through the years, then look at the questions still open.
Sources checked 3 October 2026 · 18 records · 7 subjects · 4 open questions. This is a snapshot of the research. It does not update automatically or guide prescribing.
Hemp seed oil is not CBD oil. These studies do not show that Mount Herbs seed oil or protein powder can treat a medical condition.
8 July 2022 online Double-blind comparison: 12 weeks
What was studiedTransdermal cannabidiol alongside existing antiseizure medicines.
Who was studied188 adults with drug-resistant focal epilepsy.
In 188 adults with focal epilepsy, neither studied transdermal CBD level significantly improved seizure frequency compared with placebo.
Where the evidence stopsAn uncontrolled extension cannot overturn the blinded result. Findings do not negate evidence for a different oral formulation in particular childhood syndromes.
How much can this evidence tell us?
Randomized, double-blind trial. Adverse events: 50.4% with CBD versus 41.3% with placebo; the difference was uncertain.
20 June 2024 online · September 2024 issue RELEASE MSS1 · 21-day treatment periods
What was studiedNabiximols THC:CBD oromucosal spray versus placebo.
Who was studied68 people with MS spasticity inadequately managed with existing medicines.
The clinician-rated lower-limb muscle-tone endpoint did not improve significantly with nabiximols versus placebo. On MAS LLMT-6 (Modified Ashworth Scale lower-limb muscle-tone score), changes were −0.23 versus −0.26; adjusted difference 0.04 (P=.7152).
Where the evidence stopsA negative objective endpoint can coexist with patient-reported symptom benefit in other trials. Neither establishes slowed MS progression.
How much can this evidence tell us?
Double-blind randomized crossover trial; short treatment periods.
16 August 2024 online · December 2024 issue Supportive care; not cancer-directed treatment
What was studiedQuality-controlled oral 1:1 THC:CBD added to guideline-consistent antiemetics, compared with added placebo.
Who was studied147 evaluable adults with refractory chemotherapy-induced nausea or vomiting.
In adults with chemotherapy nausea or vomiting despite standard antiemetics, an added oral 1:1 THC:CBD extract increased complete response from 8% to 24%. Response meant no vomiting or retching and no rescue medication during the first 120 hours after chemotherapy in the first cycle assessed on study. It was not a tumour response.
Complete response during 0–120 hours
THC:CBD + antiemetics24%
24%
Placebo + antiemetics8%
8%
No emesis and no rescue medication; 147 evaluable participants. Absolute difference 16 percentage points (95% CI 4–28). Not tumour response.
Adverse effectsModerate/severe sedation: 18% versus 7%. Dizziness: 10% versus 0%.
Where the evidence stopsSedation, dizziness and transient anxiety were more frequent. The trial enrolled fewer people than planned, and only 10% received olanzapine. Results do not establish superiority to every current antiemetic regimen.
How much can this evidence tell us?
Absolute response difference: 16 percentage points (95% CI 4–28). ASCO advises against using cannabis to replace cancer-directed treatment; adding it as anticancer therapy belongs in clinical trials.
14 February 2025 Historical trials are not modern standard care
What was studiedDifferent cannabinoid medicines and comparators; adult and paediatric studies.
Who was studied26 randomized trials of chemotherapy-induced nausea and vomiting.
A synthesis of 26 randomized trials found better control of chemotherapy nausea and vomiting with cannabinoids than placebo, but no significant overall advantage over active antiemetic medicines.
Where the evidence stopsTwenty-three trials were published before 2000. Many comparisons predate today’s multidrug prevention regimens. The review overlaps with contemporary trial evidence and is not wholly independent confirmation.
How much can this evidence tell us?
Evidence must be read by comparator and treatment era. It supports investigation of symptom relief, not a claim that cannabinoids treat cancer.
7 July 2025 28 days of exposure; laboratory follow-up through day 35
What was studiedOral CBD solution compared with placebo. The study exposure is context, not a dosing recommendation.
Who was studied201 adults randomized: 151 CBD, 50 placebo. The published rate is a Kaplan–Meier estimate, not 8 divided by 151.
In a trial of healthy adults, eight CBD recipients developed liver-enzyme elevations above three times the upper limit of normal; none did with placebo. The estimated event rate was 5.6%. Enzymes returned to normal after CBD was stopped.
Where the evidence stopsThese were laboratory safety signals, not eight cases of liver failure. The tested exposure was 5 mg/kg/day; the result does not quantify risk at every dose. Four weeks cannot establish long-term safety.
How much can this evidence tell us?
A controlled safety signal at the studied exposure. Prescription CBD also needs medication review and liver monitoring.
22 August 2025 online · March 2026 issue Trial duration: 24 weeks
What was studiedPlant-derived CBD tablets versus placebo.
Who was studied200 adults with fibromyalgia · single-centre trial.
Average pain changed by −0.4/10 with CBD and −1.1/10 with placebo. The 0.7-point difference favoured placebo (95% CI 0.25–1.2).
Mean decrease in pain at 24 weeks
CBD0.4/10
0.4/10
Placebo1.1/10
1.1/10
Pain decrease on the original 0–10 scale; larger means greater reduction. 200 participants randomized. Placebo performed better.
Adverse effectsAdverse events were generally mild and similarly distributed.
Where the evidence stopsThe result challenges routine CBD pain claims for fibromyalgia. It does not answer every cannabinoid question in every pain condition.
How much can this evidence tell us?
Randomized and double-blind. Adverse events were generally mild and similarly distributed.
29 August 2025 online · December 2025 issue Interpret by formulation and outcome
What was studiedCBD-only, THC-containing and other cannabinoid formulations.
Who was studiedSix trials · 1,077 adults with or without insomnia or poor sleep, without comorbidities.
Six trials found a pooled improvement in subjective sleep quality with some cannabinoid formulations. CBD-only preparations did not show a significant benefit, and improvements measured by actigraphy were not established.
Where the evidence stopsResults varied substantially across studies. Populations and preparations differed, and perceived sleep quality is not the same outcome as objectively measured sleep. A pooled result cannot be applied to every product.
How much can this evidence tell us?
High heterogeneity (I² 88%) limits confidence in the pooled headline. This is not evidence for a hemp seed sleep remedy.
A small average benefit, with more adverse effects.
29 September 2025 online · December 2025 issue 3-week titration, then 12-week blinded treatment; later extension/withdrawal phases
What was studiedStandardized oral THC-dominant full-spectrum extract, VER-01.
Who was studied820 adults with chronic low-back pain randomized.
VER-01 reduced pain by an adjusted 0.6/10 more than placebo (95% CI 0.3–0.9). A ≥30% pain reduction occurred in 54.1% versus 39.5%.
Participants with ≥30% pain reduction at 12 weeks
VER-0154.1%
54.1%
Placebo39.5%
39.5%
Responder outcome from this trial only; it is not the mean pain-score change. 820 participants randomized.
Adverse effectsStopped for adverse events: 17.3% versus 3.5%. Overall adverse events: 83.3% versus 67.3%.
Where the evidence stopsAdverse events: 83.3% versus 67.3%; discontinuation due to adverse events: 17.3% versus 3.5%. The later randomized-withdrawal primary endpoint was not significant.
How much can this evidence tell us?
Large industry-sponsored trial. Results cannot be generalized to other extracts, seed foods or certain long-term benefit.
30 September 2025 online · December 2025 issue Comparison through 27 weeks
What was studiedVER-01 oral extract versus individualized opioid treatment.
Who was studied384 adults with chronic low-back pain.
The primary constipation outcome favoured VER-01 over individualized opioid therapy (risk ratio 0.25; 95% CI 0.09–0.69).
Where the evidence stopsKey week-27 pain and sleep comparisons were not significant. Adverse-event rates were 73.0% versus 73.7%. Lack of masking and industry sponsorship require caution.
How much can this evidence tell us?
Open-label design. This is not blanket proof of superiority across outcomes or cannabinoid preparations.
23 December 2025 online · February 2026 issue Search through 28 July 2025
What was studiedExtracted comparable THC:CBD; synthetic or purified THC; lower-THC preparations analysed separately.
Who was studied25 placebo-controlled trials · 2,303 participants · 64% of trials concerned neuropathic pain.
Across 25 trials, some THC-containing medicines produced small average reductions in chronic pain. Extracted products with comparable THC and CBD, delivered through the mouth lining, reduced pain by about 0.54 points more than placebo on a 0–10 scale. Dizziness, sedation and nausea increased.
Where the evidence stopsMost evidence concerned nerve pain over 1–6 months. Preparations differed, and low-THC/CBD-dominant products may not improve outcomes. Long-term benefit and harm remain uncertain.
How much can this evidence tell us?
Comparable-ratio preparations probably slightly reduce pain. This does not establish the same result for every cannabis product.
Partial relief is different from substantial relief.
19 January 2026 Search through 29 January 2025
What was studiedTHC-dominant, balanced THC:CBD and CBD-dominant medicines assessed separately.
Who was studied21 studies · 2,187 adults · treatment lasting 2–26 weeks.
For chronic nerve pain, the updated review found no clear evidence that THC-dominant, balanced or CBD-dominant medicines produce substantial pain relief of at least 50%. Balanced THC:CBD may help some people achieve a smaller, 30% reduction.
Where the evidence stopsCertainty was low to very low. Adverse effects can cause people to stop treatment. These findings concern selected trial preparations, not all cannabis use.
How much can this evidence tell us?
For balanced THC:CBD, about 7 extra people per 100 achieved ≥30% relief, while 5 extra per 100 stopped because of adverse effects. Authors judged these effects not clinically relevant.
16 February 2026 online Single-night laboratory study
What was studiedOral CBN isolate at two active levels, compared with placebo. CBN is distinct from CBD and THC.
Who was studied20 adults · randomized, double-blind, three-condition crossover trial.
In 20 adults with diagnosed insomnia, isolated cannabinol (CBN) did not significantly improve the primary outcome: time awake after first falling asleep. One active condition improved some secondary measures, including sleep-onset time and perceived sleep quality.
Where the evidence stopsThe study was small and tested single nights. Secondary findings cannot override a null primary outcome or establish long-term effectiveness and safety.
How much can this evidence tell us?
Early clinical research. Larger, longer trials are needed before drawing conclusions about routine insomnia care.
16 March 2026 online · April 2026 issue Search through 13 May 2025
What was studiedSeveral cannabinoid formulations, studied for mental or substance-use disorders as the primary indication.
Who was studied54 randomized trials · 2,477 participants; typically small studies.
A review of 54 randomized trials found no significant treatment benefit for anxiety disorders, PTSD or psychotic disorders. No eligible depression-treatment trial was found. Adverse events were more common with cannabinoids.
Where the evidence stopsMost evidence was low certainty and 44% of trials had high risk of bias. Some other outcomes showed preliminary signals, but the review did not support routine cannabinoid treatment of mental disorders.
How much can this evidence tell us?
Absence of a depression trial is an evidence gap, not proof of no effect. Adverse-event odds ratio 1.75 (95% CI 1.25–2.46); an odds ratio is not an absolute percentage-point risk increase.
Pivotal trials 2017–2020 · US label revised May 2026 Established evidence, retained for clinical relevance
What was studiedPurified cannabidiol oral solution, used alongside existing antiseizure medicines.
Who was studiedSeparate randomized trials: Dravet, 120 participants; Lennox–Gastaut, 225; tuberous sclerosis complex, 224.
Pharmaceutical purified CBD, added to existing antiseizure treatment, reduced specific seizure types in trials of Dravet syndrome, Lennox–Gastaut syndrome and tuberous sclerosis complex. In the United States, prescription Epidiolex is approved for seizures associated with these conditions from age 1.
Where the evidence stopsThis is evidence for a standardized prescription oral solution. It does not establish equivalent effects from cannabis, shop-bought CBD or hemp seed oil. Liver effects, sleepiness and drug interactions require clinical monitoring.
How much can this evidence tell us?
Defined trial populations and prescription indication. US approval does not establish Indian approval or availability. Do not change antiseizure treatment without the treating clinician.
28 May 2026 online · June 2026 collection Two six-week treatment periods
What was studiedOral cannabidiol versus placebo, with washout between periods.
Who was studied40 adults randomized; 38 in the primary analysis · spinal-cord-injury neuropathic pain.
Oral CBD reduced average pain by 0.54/10 more than placebo (95% CI 0.21–0.88 improvement). Adverse events affected 68.4% during CBD and 52.6% during placebo, mostly minor.
Where the evidence stopsSmall sample, short follow-up and only six women limit generalization. This does not resolve the negative fibromyalgia result in another population.
How much can this evidence tell us?
Early condition-specific signal. Larger and longer trials are needed.
2 August 2026 online · October 2026 issue Search through 2 December 2025
What was studiedTHC:CBD extracts, including medicines delivered through the mouth lining; other cannabinoids assessed separately.
Who was studied27 randomized-trial publications · over 3,000 participants overall; seven studies in the direct THC:CBD analysis.
The selected 2026 review found modest improvements in patient-reported MS spasticity, most consistently with standardized THC:CBD extracts. A direct comparison across seven studies favoured these extracts by 0.82 points on a spasticity rating scale.
Where the evidence stopsCertainty was low, studies varied, and objective functional benefits were limited. Adverse effects were common. Symptom relief does not mean that the medicine slows MS progression.
How much can this evidence tell us?
Low certainty in the 2026 review. Some studies selected previous responders, limiting generalisation. The established 2022 Cochrane review supports short-term perceived-spasticity benefit for nabiximols.
Sources checked 3 October 2026 US guidance; not an Indian approval statement
What was studiedTHC: tetrahydrocannabinol. CBD: cannabidiol. CBN: cannabinol. Hemp seed oil and protein are separate seed ingredients.
Who was studiedGeneral education. Pregnancy, breastfeeding, adolescence and a history of psychosis warrant particular caution and professional advice.
Hemp seed oil is pressed from seeds; it is not CBD oil. FDA explains that seeds do not naturally contain CBD or THC, although trace amounts can enter during processing. Cannabinoid medicines use specified compounds and formulations. Findings cannot be transferred between them.
Where the evidence stopsTHC can impair attention, coordination and driving. Cannabis use can lead to dependence. CBD can interact with medicines and cause sleepiness and liver effects. Discuss use with a qualified clinician, especially alongside other treatment.
How much can this evidence tell us?
Mount Herbs’ supplied range is 100 ml hemp seed oil and 500 g hemp protein powder. No therapeutic claim or cannabinoid content is established for these products here.
Explore four areas of research: what we know so far, what is being tested and what remains unanswered.
CLINICAL QUESTION
Which formulation, for whom?
Looking ahead · checked 3 October 2026
Can a reproducible benefit be identified for a defined condition, formulation and outcome?
What we know so far
Recent pain trials return different answers: modest benefit for one THC-dominant extract, a negative CBD fibromyalgia result, and a small CBD spinal-cord-injury signal.
The next question
Replication, longer follow-up, better masking, and meaningful daily-function outcomes matter more than a universal cannabis claim.
What we still don’t knowThis is an editorial synthesis of the linked trials, not a prediction of approval or a treatment recommendation.
ARISTOCRAT asks whether nabiximols added to temozolomide improves overall survival in recurrent MGMT-methylated glioblastoma.
What we know so far
NCT05629702 · first posted 29 November 2022; last update 5 May 2026. The record estimates 120 participants and has no posted results.
The next question
The record last reported recruiting, verified April 2026. Its estimated primary completion was September 2026; actual recruitment or completion is not confirmed by that estimate.
What we still don’t knowNo cancer-treatment benefit can be inferred. Survival research is distinct from nausea relief. Registry snapshot checked 3 October 2026.
We chose these sources to help you make sense of cannabinoid research. We checked them on 3 October 2026; this page does not update itself, and it is not a new systematic review. We give priority to systematic reviews, randomized trials in people and official medicine information. Older findings sit alongside newer work when they still matter. Animal studies, personal stories and marketing claims cannot, on their own, show that a treatment helps patients.
Where we could verify them, we use online publication dates. Medicine-label changes and guidance are marked separately. A publication date tells you when a paper appeared; a review’s search cutoff tells you how recent its included evidence is. A review published in 2026 might only include trials found through 2025. Two reviews may also use the same trials, so they are not necessarily independent confirmations. Pain scores, response rates and seizure reductions measure different things and cannot be used to rank medicines across conditions.
When we describe evidence as strong or uncertain, we follow the cited review’s wording or explain the study design. We have not created our own clinical rating system. Where a review was read at abstract level, we identify its original indexed publication and link the full text when available. A clinician has not independently reviewed this summary. It does not give prescribing instructions or establish whether a medicine is approved or available in India.
A newer paper is not always a more reliable one. We found a 2026 sleep meta-analysis, but we could not fully appraise it or reconcile its reported search dates. We therefore do not use it here as evidence of benefit. Read the paper · Asim et al., 2026 ↗
