A newly published federally funded study is providing a more detailed picture of why people with cancer use medical cannabis and how the benefits they perceive may differ depending on whether their products contain cannabidiol alone or a combination of CBD and delta-9-tetrahydrocannabinol. Researchers from the University of Wisconsin–Madison, University of Pennsylvania, Johns Hopkins University and the nonprofit Realm of Caring Foundation analyzed reports from 65 adults with cancer who had used cannabis therapeutically. Across the group, pain relief, better sleep or physical relaxation, improved emotional well-being and reduced reliance on other medications repeatedly emerged as perceived benefits. The study was published August 14, 2026, in Frontiers in Psychiatry.
The cannabinoid composition appeared to matter. Pain relief was particularly prominent among patients who reported using both CBD- and THC-containing products, while emotional regulation was the strongest theme among people whose reported exposure involved CBD-dominant products without THC-containing preparations. The research received support from the National Cancer Institute and National Institute on Drug Abuse, along with a research contract from Realm of Caring. Yet this was not a randomized clinical trial, and it did not objectively measure whether cannabis reduced pain or improved sleep. Instead, researchers applied an unusual network-analysis method to patients’ descriptions of their own experiences, making the study useful for identifying patterns that should now be tested experimentally.
Researchers Looked at How Cannabis Benefits Occur Together
The paper, titled A Comparative Network Analysis to Explore Cancer Patient Experiences with Cannabis, differs from the typical medical cannabis survey. Rather than simply counting how many patients said cannabis helped with pain, sleep or anxiety, researchers used Epistemic Network Analysis, or ENA, to examine which experiences appeared together in individual responses. Patients were asked how therapeutic cannabis had helped them, whether it caused problems, why they began using cannabinoids and, where applicable, why they stopped. Human researchers coded the responses into themes such as pain relief, emotional regulation, sleep or physical relaxation, appetite stimulation, medication reduction, adverse effects, stigma, high cost and lack of medical guidance.
The sample included 65 adults with an average age of 53; 75 percent were women and 75 percent were White. Thirty-three were categorized in the CBD-only group and 32 in the CBD+THC group. Importantly, all participants reported some CBD-dominant cannabis use, while the CBD+THC group additionally used THC-dominant products or products containing roughly balanced amounts of both cannabinoids. Products were diverse: CBD-dominant oils were the most common, but patients also reported THC oils, edibles, flower, topical CBD products, concentrates and other formulations. Ninety-two percent were still using cannabis when surveyed.
Pain Relief Was the Most Commonly Reported Benefit
Pain stood out more than any other individual theme. Researchers recorded 112 instances of pain relief, accounting for 21 percent of all coded experiences. Emotional regulation followed at 15 percent, medication reduction at 14 percent and sleep or physical relaxation at 10 percent. The prominence of pain was especially pronounced in the CBD+THC network, where it was strongly linked with better sleep or relaxation, emotional regulation and reduced medication use. Pain was also important among CBD-only users, but emotional regulation occupied the more central position in that group’s network.
That difference is intriguing because THC has a stronger established analgesic rationale than CBD alone, although neither cannabinoid has been conclusively demonstrated to provide broad relief from cancer pain. Among CBD+THC users, the strongest links involving pain were with sleep or physical relaxation and emotional regulation, followed by medication reduction. The investigators argue that looking at these connections may better reflect real cancer symptom management than examining pain as an isolated endpoint: someone who hurts less may sleep better, while better sleep and lower anxiety may also change how disruptive that pain feels. Still, the authors explicitly caution that their sample was self-selected and nonrandom, meaning the cannabinoid-group differences are hypothesis-generating rather than proof that adding THC produces superior analgesia.
Sleep, Relaxation and Emotional Well-Being Were Closely Connected
Sleep was another recurring component of patients’ experiences. Across both cannabinoid groups, reports of improved sleep or physical relaxation appeared alongside pain relief, emotional regulation, appetite benefits and medication reduction. The relationship between pain and sleep was particularly strong among CBD+THC users. That pattern has a plausible biological basis because acute THC can have sedating properties, although the effects of cannabinoids on sleep are complicated by dose, tolerance, duration of treatment and differences between THC and CBD. The investigators themselves note that evidence for CBD’s effect on sleep remains considerably more variable.
Among CBD-only users, emotional regulation was the most interconnected benefit. Patients described improvements involving anxiety, mood, relaxation and general well-being, and those reports frequently overlapped with sleep and pain. This should not be interpreted as evidence that CBD is an established treatment for anxiety in cancer patients. A systematic review of cannabis interventions for depression, anxiety and stress in people with cancer found that the evidence remains limited, while the American Society of Clinical Oncology has concluded that available data are insufficient to recommend cannabis for anxiety or depression in oncology. The new study instead shows that emotional effects are important to patients themselves and deserve stronger controlled research.
Some Patients Reported Using Less Conventional Medication
Medication reduction was the third-most common positive theme, appearing 75 times in the responses. It was most closely associated with pain relief in both cannabis groups. Some patients reported lowering prescription medication doses after beginning cannabis, while others described cannabinoids as an alternative to medications whose side effects they found difficult to tolerate. This finding adds to a broader body of observational research suggesting that some medical cannabis patients intentionally use cannabis to reduce consumption of opioids, sleep medications or other drugs. But the study did not independently verify prescription records, opioid dosages or medication discontinuation.
That distinction is particularly important in oncology. Opioids remain a core treatment for moderate-to-severe cancer pain, and observational reports of substitution do not establish that withdrawing opioids in favor of cannabis would provide equivalent pain control. Researchers funded by the National Cancer Institute are now conducting substantially larger longitudinal studies specifically designed to measure cannabis use, cancer symptoms and opioid consumption over time. One ongoing NCI-supported project plans to follow 600 ambulatory patients with cancer receiving opioid therapy, comparing cannabis users with nonusers while repeatedly measuring pain, sleep, anxiety, quality of life and objective and subjective opioid outcomes.
The Findings Echo Earlier Studies of Cannabis Use During Cancer Treatment
The new results are consistent with a 2023 study published in the American Cancer Society journal Cancer. Researchers surveyed 267 adults undergoing treatment at an NCI-designated cancer center and found that 26 percent had used cannabis during the previous month. Among cannabis users, 86 percent reported using it for pain and 70 percent for sleep problems. Patients perceived some of their greatest cannabis-related relief in sleep problems, nausea and vomiting, headaches, pain, muscle spasms and anxiety. However, cannabis users actually had more severe overall physical and psychological symptoms than nonusers, illustrating why cross-sectional studies cannot determine whether cannabis caused improvements. People experiencing greater symptom burdens may simply be more likely to seek cannabis in the first place.
A larger prospective Israeli study published in 2022 followed oncology patients after beginning prescribed medical cannabis. Of 324 patients who initiated treatment, 126 remained in the six-month assessment, and researchers reported a median 18 percent reduction in overall cancer symptom burden from baseline. Most measured outcomes improved over time, while reported adverse effects were generally non-serious. That prospective design provides stronger evidence than a one-time survey because symptoms were assessed before and after cannabis initiation, but without a randomized placebo group it remains impossible to separate pharmacological effects from changes in cancer treatment, expectations, natural symptom fluctuation or other factors.
Clinical Guidelines Remain Much More Cautious About Cancer Pain
The strongest reason not to interpret the new study as proof of efficacy comes from randomized evidence. ASCO’s 2024 clinical guideline evaluated 13 systematic reviews along with additional randomized and cohort studies and concluded that evidence remains insufficient to recommend for or against cannabis or cannabinoids for most cancer-related symptoms, including cancer pain. Randomized studies of nabiximols—a standardized THC/CBD oral spray—have generally failed to show a statistically significant overall reduction in cancer pain compared with placebo when results are pooled. For sleep among patients with chronic cancer pain, cannabinoids produced only a very small improvement in clinical trials.
There is one oncology indication with substantially better cannabinoid evidence: refractory chemotherapy-induced nausea and vomiting. ASCO says adults whose nausea or vomiting persists despite guideline-based antiemetic therapy may consider adding dronabinol, nabilone or a quality-controlled oral 1:1 THC:CBD extract. The guideline simultaneously recommends against using cannabis or cannabinoids as cancer treatment itself outside a clinical trial. That point matters because some participants in the new study expressed beliefs that cannabinoids might directly attack their cancer. Laboratory studies of cannabinoids and tumor biology remain an active research area, but patient reports of symptom relief provide no evidence that cannabis shrinks tumors, delays cancer progression or improves survival.
Cost and Cannabis Stigma Remain Significant Problems
The paper also highlights two problems that have little to do with pharmacology: price and stigma. High cost appeared 20 times in patient responses, while stigma appeared 24 times. Patients described struggling to afford cannabis on limited incomes and experiencing negative reactions from family or others because of THC use. Because most medical cannabis remains outside traditional prescription-drug reimbursement systems, patients often pay the entire price themselves. Researchers argue that their findings strengthen the case for considering insurance coverage where medicinal cannabis is legally available, particularly if future clinical trials identify formulations with reliable benefits.
Access is also complicated by the gap between patient use and clinician expertise. The paper notes earlier research finding that only about 30 percent of oncologists felt sufficiently informed to make cannabis recommendations. ASCO has since urged cancer clinicians to routinely and nonjudgmentally ask patients about cannabis use, discuss treatment goals and possible risks and guide patients toward reliable information rather than leaving those conversations entirely to dispensaries or internet sources. This is particularly relevant because cannabis can cause sedation, cognitive impairment, anxiety or unwanted intoxication, and CBD and THC may interact with medications through drug-metabolizing pathways.
What the New Study Does—and Does Not—Prove
Several limitations prevent strong causal conclusions. The 65-person sample was small, predominantly White and recruited through social media and word of mouth. Participants selected themselves into a study concerning therapeutic cannabis, creating obvious potential for favorable-selection bias. Cancer type, cancer stage and treatment regimen were not collected, even though a person receiving curative therapy for early-stage breast cancer may have very different symptoms from someone receiving palliative treatment for metastatic disease. Cannabis products were also identified entirely through patient reports without laboratory verification of THC, CBD or other cannabinoid concentrations.
There was no placebo group, no cannabis-free cancer comparison group, no objective sleep measurement and no clinical measurement of pain improvement before and after treatment. The authors also reported adverse experiences—including sleepiness, lethargy, intoxication, paranoia, anxiety and increased appetite or weight gain among CBD+THC users, along with stomach upset and bad dreams among some CBD users. One author, Johns Hopkins researcher Ryan Vandrey, disclosed consulting or advisory work for several cannabis or pharmaceutical companies; the study itself received NIH support and a research contract from Realm of Caring. These disclosures do not invalidate the results, but they are important context for evaluating the research.
Why the Study Still Matters
Despite those limitations, the study contributes something useful that randomized trials often miss: it shows how cancer patients themselves describe the interconnected nature of symptom relief. Pain, sleep, mood, appetite and medication use do not occur independently during cancer treatment. A therapy that changes one may influence several others, and cannabinoid composition may change which effects patients value most. The network analysis suggests that THC-containing regimens deserve particular investigation for pain-sleep symptom clusters, while CBD-dominant preparations may warrant more focused research on emotional well-being and relaxation. Those hypotheses can now be tested with standardized products, objective measurements and randomized designs.
For now, the most accurate conclusion is narrower than saying medical cannabis has been proven effective for cancer pain or insomnia. Cancer patients in this federally supported study frequently reported pain relief, improved sleep or relaxation, emotional benefits and reduced medication use, with pain playing an especially prominent role among users of THC-containing products. Those experiences agree with patterns seen in several previous oncology studies, but clinical guidelines continue to regard the evidence for most cancer symptoms as insufficient. The gap between widespread patient use and limited high-quality evidence is exactly why larger federally funded cannabis trials now underway may be especially consequential for cancer care.
