Last verified: April 2026

The NASEM Rating — What “Conclusive/Substantial” Actually Means

Chronic pain received the highest evidence rating in the 2017 NASEM report — conclusive or substantial evidence that cannabis is an effective treatment. This rating was not based on a single landmark trial but on a convergence of data across multiple study designs, cannabinoid formulations, and pain subtypes. It remains the strongest therapeutic claim for any cannabinoid in any medical condition outside of CBD for epilepsy.

However, the rating comes with critical nuance that is routinely lost in translation. The evidence is strongest for neuropathic pain — pain arising from nerve damage or dysfunction — and weaker for nociceptive pain (tissue damage), inflammatory pain, and the diffuse pain syndromes (fibromyalgia, complex regional pain syndrome) that make up a large proportion of chronic pain clinic referrals. When a dispensary claims that “cannabis is proven for chronic pain,” they are extrapolating a neuropathic-specific finding to a category that encompasses dozens of distinct pathophysiological mechanisms.

Whiting 2015 — The Definitive Systematic Review

The most cited evidence synthesis is the Whiting et al. systematic review published in the Journal of the American Medical Association in 2015. The review encompassed 28 randomized controlled trials with a combined enrollment of 2,454 patients across various chronic pain conditions. The primary finding: patients randomized to cannabinoids were approximately 40% more likely to experience a 30% reduction in pain compared to placebo (OR 1.41, 95% CI 1.29–1.58).

The effect size warrants careful interpretation. A 40% greater likelihood of achieving a 30% pain reduction is statistically significant but clinically modest. It means that for every group of patients treated with cannabinoids, roughly 4 in 10 additional patients will experience meaningful improvement compared to what placebo alone provides. This is a genuine therapeutic signal — but it is not the dramatic efficacy that “conclusive evidence” might imply to a lay audience.

Whiting’s team also noted that the quality of included trials was generally moderate to low, with common methodological issues including small sample sizes, short durations, inadequate blinding (cannabis’s psychoactive effects make true blinding difficult), and heterogeneous outcome measures. The overall evidence supported cannabinoids for chronic pain, but the authors called for larger, longer, better-designed trials to determine optimal formulations and patient selection criteria.

There was moderate-quality evidence to support the use of cannabinoids for the treatment of chronic pain. The average number of patients who reported a reduction in pain of at least 30% was greater with cannabinoids than with placebo.

Whiting et al., JAMA 2015

Neuropathic Pain — The Strongest Subtype

Within the broader chronic pain category, neuropathic pain has the most robust cannabinoid evidence. The landmark trial was conducted by Donald Abrams and colleagues at the University of California San Francisco, published in Neurology in 2007. The study enrolled patients with HIV-associated sensory neuropathy — a condition that responds poorly to conventional analgesics — and demonstrated a number needed to treat (NNT) of 3.6 for smoked cannabis.

An NNT of 3.6 means that for every 3.6 patients treated, one will experience clinically meaningful pain relief that would not have occurred with placebo. This figure is directly comparable to first-line neuropathic pain agents: gabapentin has an NNT of 3.8, and is superior to amitriptyline, which has an NNT in the range of 4.0–6.0 depending on the neuropathy subtype. In other words, for HIV neuropathy specifically, smoked cannabis performs as well as or better than standard pharmaceutical options.

Additional neuropathic pain trials have demonstrated efficacy in diabetic peripheral neuropathy, post-traumatic neuropathy, and central pain syndromes. The mechanistic basis is sound: cannabinoid receptors — particularly CB1 and CB2 — are present on dorsal root ganglion neurons, spinal cord interneurons, and supraspinal pain-processing regions, providing multiple anatomical sites for pain signal modulation. The descending pain inhibitory system from the periaqueductal gray also involves endocannabinoid signaling, offering a top-down analgesic mechanism.

The Opioid-Sparing Hypothesis

One of the most important — and most overstated — claims in cannabis therapeutics is that medical cannabis reduces opioid use. The data are suggestive and politically compelling, but the evidence base falls well short of the standard required to establish a causal relationship.

The most-cited study is Boehnke et al. (2016), published in the Journal of Pain, which surveyed 244 medical cannabis patients at a Michigan dispensary. Respondents reported a 64% decrease in opioid use, along with improved quality of life and fewer medication side effects. The finding is striking but carries the inherent limitations of self-report data from a self-selected population at a cannabis dispensary — a group with obvious motivation and recall biases.

At the population level, Bradford and Bradford (2018) published a Medicare claims analysis in JAMA Internal Medicine demonstrating that states with medical cannabis laws showed an 8.5% reduction in daily opioid doses filled under Medicare Part D, along with reductions in benzodiazepine prescriptions. This ecological analysis suggests that medical cannabis availability is associated with reduced opioid prescribing at the population level.

However, ecological and survey designs cannot establish causation. Patients who seek cannabis may be simultaneously reducing opioids for other reasons (changing pain management philosophy, concern about opioid stigma, provider-driven tapering). And a critical 2019 re-analysis by Shover et al. in the Proceedings of the National Academy of Sciences found that the previously reported association between state cannabis laws and reduced opioid overdose mortality reversed direction when data through 2017 were included — states with cannabis laws now showed higher opioid mortality, suggesting the original finding was driven by confounders.

We found that the previously reported association between state medical cannabis laws and reductions in opioid overdose mortality reversed direction from 2010 to 2017.

Shover et al., PNAS 2019

Limitations — What the Data Cannot Tell Us

Several fundamental methodological challenges limit the chronic pain evidence base. Blinding is the most persistent: THC produces unmistakable psychoactive effects that make true double-blinding extremely difficult. Patients frequently guess their allocation correctly, introducing expectation bias. Some trials have attempted active placebos (low-dose THC or other mild psychoactive agents), but no universally accepted blinding solution exists.

Duration is another limitation. Most cannabis pain trials last 2 to 15 weeks. Chronic pain, by definition, persists for months to years. Whether cannabinoid efficacy is maintained over long-term use — or whether tolerance development erodes the analgesic effect — is largely unknown. The few long-term observational studies suggest that most patients maintain stable doses over 1–2 years, but these are not controlled data.

Formulation heterogeneity makes direct comparison between trials difficult. The Whiting review included studies of smoked cannabis, vaporized cannabis, oral THC (dronabinol), oromucosal THC:CBD spray (nabiximols), oral nabilone, and various CBD preparations. Whether the findings generalize to the specific product a patient purchases at a dispensary — with its particular cannabinoid ratio, terpene profile, dose, and route — is uncertain.

Finally, the effect sizes are modest. Cannabis is not a powerful analgesic in the way that opioids or regional nerve blocks are powerful analgesics. It appears to provide a moderate, clinically meaningful reduction in pain scores for a subset of patients — most consistently those with neuropathic pain. The honest framing is that cannabis is a reasonable option when conventional treatments fail or produce intolerable side effects, not a first-line therapy or a miracle cure for all pain.

Clinical Bottom Line

The evidence supports cannabis as an effective analgesic for chronic neuropathic pain, with an NNT comparable to established pharmaceutical agents. The broader chronic pain evidence is positive but less specific. The opioid-sparing data are suggestive and important but remain observational and ecologically confounded. Patients with neuropathic pain who have failed or cannot tolerate gabapentinoids, tricyclics, or SNRIs have the strongest evidence-based rationale for cannabinoid therapy. Patients with other chronic pain subtypes may benefit, but should understand that the evidence for their specific condition is weaker than the headline “conclusive evidence for chronic pain” implies.