A 2025 narrative review finds that cannabinoids can offer modest pain relief for patients with neuropathic pain and MS-related spasticity, but average improvements rarely exceed one point on a ten-point pain scale. Side effects, including dizziness in one in four patients and hepatotoxicity risk with high-dose CBD, weigh meaningfully against those gains.

Chronic pain affects approximately one in five adults globally, and the fallout from the opioid crisis has left clinicians and patients searching for safer alternatives. Cannabinoids occupy a unique space in this conversation, widely accessible in many jurisdictions but still lacking the large-scale, high-quality trial evidence that would support firm clinical guidelines. This 2025 review attempts to map the current state of play across efficacy, safety, and regulation, making it directly relevant to clinicians fielding daily patient questions about cannabis-based treatments and to policymakers shaping access frameworks.

Chronic pain management remains an area of intense clinical need, with conventional analgesics offering their own burdens of tolerance, organ toxicity, and dependence. Cannabinoids, acting through CB1 and CB2 receptor systems, TRPV1 channels, descending inhibitory circuits, and glial neuroinflammatory pathways, have a plausible biological rationale as adjunctive analgesics. This 2025 narrative review from investigators at Advocate Illinois Masonic Medical Center and the University of Illinois synthesizes the existing clinical literature across multiple pain conditions, safety data, and international regulatory landscapes, aiming to identify where the evidence is strongest and where critical gaps remain.

The review finds that the most robust data support cannabinoid use in neuropathic pain, where cited trials show pain score reductions of 6 to 9 points on a 0 to 100 scale, and in MS-related spasticity, where nabiximols provides approximately one point of improvement on a 0 to 10 scale. For fibromyalgia, osteoarthritis, and musculoskeletal pain, results are described as inconsistent. Average pain reductions across conditions generally fall between 0.5 and 1.0 points on a 10-point scale, placing them at or below standard minimal clinically important difference thresholds. One cited observational study reported a 64% reduction in opioid consumption among cannabis users, and an inhaled cannabis study found 67.2% of migraine patients experienced pain relief within two hours. Adverse events are clinically meaningful: dizziness occurs in 25% of nabiximols users, somnolence in 8%, and treatment discontinuation in 12%. High-dose CBD carries a measurable hepatotoxicity risk. The authors conclude that cannabinoids should be considered adjunctive rather than first-line, reserved for patients unresponsive to conventional therapy, and call for standardized formulations, harmonized regulations, and large-scale randomized controlled trials.

Cannabinoids in Chronic Pain: Modest Effects, Real Risks, and the Evidence We Still Need

One in five adults lives with chronic pain, and the tools we have to help them all carry their own burdens. When cannabis entered the conversation as a potential alternative, the hope was understandable. A 2025 narrative review by Sic and colleagues now asks: does the evidence justify that hope, or are we again reaching for a solution before the science is ready? What this paper actually does is synthesize existing clinical trials and mechanistic literature into a condition-by-condition overview. It is not itself a clinical trial, and it does not employ systematic search methods. Think of it like reading only the books on the top shelf of a library and concluding you have surveyed the whole collection. You may have found the most prominent works, but you cannot know what sits in the stacks below. That distinction matters, because when the authors cite a 64% reduction in opioid use or 67% migraine relief from inhaled cannabis, those figures come from individual studies, some observational, whose designs cannot exclude confounding. The review presents these numbers without always weighting them against their inferential fragility. What the paper does get right, and what I genuinely respect about it, is its refusal to oversell. The authors explicitly describe cannabinoids as adjunctive rather than first-line, and they quantify adverse events with specificity: 25% dizziness with nabiximols, 12% discontinuation, hepatotoxicity risk at high CBD doses. That kind of honesty is uncommon and clinically valuable.

The central methodological issue is the gap between statistical detection and clinical meaning. The average pain reduction cited across conditions is 0.5 to 1.0 points on a 10-point scale. A scale that can detect a difference of half a point is technically working, but if you are trying to decide whether a therapy is clinically meaningful, you need to know how much improvement actually changes a patient’s life, not just that the needle moved. Standard thresholds for minimal clinically important difference in pain research are generally 1.5 to 2.0 points, and the review does not systematically apply those thresholds. This omission matters enormously for clinicians, because it means the published averages might describe real but functionally irrelevant relief for many patients, even as a subset experiences more substantial benefit that the averages conceal. The review also does not engage with publication bias, which is an important blind spot: positive trials are more likely to be published, and in a narrative synthesis without systematic search, the effect estimates could be inflated in ways we cannot quantify. Nor does it address vulnerable populations, such as adolescents, pregnant patients, or those with psychiatric comorbidities, where the risk calculus shifts substantially.

What would I tell a patient who asks about this evidence? For someone with nerve pain or MS-related spasticity who has not responded to standard treatments, I would say that cannabinoids may offer modest additional relief, that the average improvement is small, that side effects like dizziness are common, and that we do not yet have strong long-term safety data. This is a carefully considered adjunct, not a replacement for a current regimen. To a colleague, I would frame this review as a useful map of the terrain but not a quantitative evidence base for protocol development, and I would encourage direct examination of the underlying primary studies before making prescribing decisions. To a policymaker, I would argue that the most urgent need is not simply expanding access but removing barriers to large-scale, well-designed clinical trials that could finally resolve the questions this review can only pose. Effect size honesty is a form of clinical respect. A therapy that offers 0.5 to 1.0 points of pain relief on a 10-point scale to a patient suffering daily is not nothing, but it is also not a solution, and the gap between those two truths is where good clinical judgment lives.

This review sits at a useful but intermediate point in the research arc for cannabinoids in pain management. It arrives after several systematic reviews and meta-analyses, most notably the 2015 Whiting et al. JAMA review, have established the same broad pattern: modest efficacy concentrated in neuropathic pain and MS spasticity, with inconsistent results elsewhere. The IASP position statement remains cautious. This narrative synthesis adds updated mechanistic context and regulatory discussion but does not advance the quantitative evidence base, and its lack of systematic methods means it cannot challenge or refine the pooled estimates from prior meta-analyses.

From a pharmacological and safety standpoint, clinicians should note the 25% dizziness rate and 12% discontinuation rate with nabiximols, which are clinically relevant in patients who may already be managing polypharmacy. The hepatotoxicity signal with high-dose CBD warrants liver function monitoring, particularly in patients taking concurrent hepatically metabolized medications including certain anticonvulsants, antidepressants, and statins. Drug-drug interactions via CYP3A4 and CYP2C19 inhibition by CBD deserve careful attention. For clinicians considering cannabinoid therapy, the most actionable recommendation is to restrict its use to treatment-refractory neuropathic pain and MS spasticity patients, initiate at the lowest available dose, monitor for dizziness, sedation, and hepatic function, and set explicit response thresholds with the patient at the outset to determine whether continued use is justified.

This is a narrative review, which occupies a lower tier in the evidence hierarchy than systematic reviews or meta-analyses. No search strategy, inclusion or exclusion criteria, or formal quality assessment of cited studies is reported. Its conclusions therefore reflect the authors’ editorial judgment about the existing literature rather than a reproducible, auditable synthesis. The single most important inference constraint is that the completeness of the evidence surveyed cannot be verified, meaning publication bias and selection bias in literature coverage may shape the conclusions in ways that are invisible to the reader.

This review broadly confirms the findings of the 2015 Whiting et al. JAMA systematic review, which identified moderate-quality evidence supporting cannabinoids for chronic neuropathic pain and MS spasticity, with low-quality or insufficient evidence for other pain conditions. The effect size range cited here (0.5 to 1.0 points on a 10-point scale) is consistent with the pooled estimates from prior meta-analyses, reinforcing the pattern of statistically detectable but clinically modest effects. Where this review extends prior work is in its integration of regulatory context and comparative NNT data for conventional analgesics, offering clinicians a more complete decision-making framework. However, it does not challenge or refine the quantitative conclusions of earlier systematic reviews and does not introduce new primary data or novel analytic methods.

The most consequential analytic choice in this review is the decision to conduct a narrative rather than systematic synthesis. Had the authors employed a reproducible search strategy, applied formal inclusion and exclusion criteria, and assessed risk of bias in cited studies, the resulting conclusions might have been more conservative. Systematic methods would have required explicit engagement with null and negative trials that may have been underrepresented in the narrative approach. Additionally, a formal application of minimal clinically important difference thresholds to each condition would likely have reclassified some of the cited analgesic effects from “modest but real” to “statistically detectable but clinically uncertain,” materially changing the practical implications of the review for prescribing clinicians.

The most likely overinterpretation is treating the 64% opioid reduction figure as a causal, RCT-level finding. This figure derives from observational data cited by the authors, and confounding by indication, self-selection, and unmeasured variables cannot be excluded. It is a hypothesis-generating observation, not evidence sufficient to recommend cannabinoids as an opioid-sparing strategy. Similarly, the 67.2% migraine relief rate from inhaled cannabis appears to draw on a single study whose design details are not fully characterized in the review, making it premature to generalize this finding. Readers should also be careful not to mistake this narrative review for a systematic review or meta-analysis; its conclusions carry qualitatively different inferential weight.

This review contributes a clinically conservative, mechanistically grounded overview of the 2025 cannabinoid-pain evidence landscape. It does not establish definitive efficacy, causal opioid-sparing effects, or generalizable effect sizes. For now, cannabinoids remain a reasonable adjunctive consideration for treatment-refractory neuropathic pain and MS spasticity under medical supervision, with meaningful adverse effects that require monitoring. The field urgently needs standardized formulations, long-term safety data, and large-scale randomized controlled trials before any broader clinical recommendations are warranted.

Does this review prove that cannabis works for chronic pain?

No. This is a narrative review that summarizes existing studies rather than generating new data. It finds modest evidence of benefit in neuropathic pain and MS spasticity specifically, but average pain reductions are small and may not reach the threshold that most pain researchers consider clinically meaningful. The evidence is not strong enough to support cannabis as a first-line treatment for any chronic pain condition.

Can cannabis replace my opioid pain medication?

The review cites one observational study reporting a 64% reduction in opioid use among cannabis users, but this is not causal evidence from a controlled trial. Many factors could explain this association. You should never change or stop opioid medications without close medical supervision, as doing so can be dangerous. Any decision to incorporate cannabinoids should be made with your physician as part of a comprehensive pain management plan.

What are the main side effects of medical cannabinoids?

According to the data cited in this review, dizziness affects about 25% of patients using nabiximols (a THC and CBD spray), somnolence affects about 8%, and roughly 12% discontinue treatment due to side effects. High doses of CBD carry a risk of liver injury. These are not rare or trivial effects and should be discussed with your doctor before starting any cannabinoid therapy.

Is this a systematic review?

No. This is a narrative review, meaning the authors selected and summarized studies based on their own judgment rather than following a structured, reproducible search protocol. Narrative reviews are useful for providing an overview of a field but carry less inferential weight than systematic reviews or meta-analyses because they may inadvertently omit relevant studies, particularly those with negative or null findings.

Have thoughts on this? Share it: