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.
The value of a treatment is not captured by a symptom score alone.
A person may still live with pain, anxiety or poor sleep yet find it easier to work, cook, exercise, see friends or take part in family life. Another may report that a medicine lowers a number on a questionnaire without making daily life feel meaningfully better.
That distinction is becoming more important in medical-cannabis research. A 2026 systematic review found an overall signal of improved health-related quality of life among patients receiving medicinal cannabis. Recent interviews have also described something less easily condensed into a scale: greater agency, steadier relationships and renewed participation in ordinary activities.
The findings are encouraging. They are not a licence to declare that cannabis improves everybody's life, or that every reported change was caused by treatment. Much of the evidence remains observational, products differ substantially and people who continue treatment are not always representative of those who start it.
Still, this literature asks a better question than “did the symptom disappear?” It asks whether treatment helped a person live.
What quality of life actually measures
Health-related quality of life is broader than happiness and narrower than a general judgement about whether life is good. It usually covers several connected domains: mobility and physical function, pain, sleep, mood, energy, ability to perform usual activities, relationships and a person's overall perception of their health.
Researchers measure those domains using questionnaires such as the EQ-5D, the SF-36 or condition-specific tools. These are patient-reported outcomes. They do not replace laboratory results or clinician assessments, but they capture effects that a blood test cannot.
For someone with a long-term condition, complete symptom removal may be unrealistic. A worthwhile change might instead mean making breakfast without assistance, returning to a weekly class, concentrating through a work meeting or having enough energy to spend time with children.
That is why quality of life should not be treated as a soft extra. It can reveal whether a statistically significant change has translated into something a patient can feel and use.
It also creates a challenge. Quality of life is influenced by treatment, expectations, relationships, income, housing, other medicines, access to care and events outside healthcare. A person reporting improvement after starting medical cannabis may be describing a real and important change without proving that cannabis alone produced it.
What the 2026 systematic review found
Researchers from the University of Sydney reviewed studies published between January 2015 and April 2025 that measured quality of life before and after medicinal-cannabis treatment using validated instruments.
From 16,674 records, they retained 64 studies: 12 randomised controlled trials, 38 cohort studies, 13 case series and one non-randomised experimental study. The studies covered different conditions, formulations, countries and follow-up periods.
Randomised trials produced a small short-term improvement in health-related quality of life, with a standardised effect size of 0.30. The observational studies showed larger improvements across short-, medium- and long-term follow-up, with effect sizes ranging from 0.43 to 0.74.
That gap between trial and observational results matters.
Randomisation is designed to separate treatment effects from expectation, natural symptom fluctuation and differences between people who choose treatment and those who do not. Observational studies are better at showing what happens in ordinary care, across longer periods and more varied populations, but they are more vulnerable to those alternative explanations.
The combined message is not that one method is right and the other is wrong. It is that quality-of-life improvement appears repeatedly, while the best-controlled estimate is more modest than the change reported in real-world cohorts.
The review also identified a basic reporting problem. Only five of the 64 studies defined what they meant by health-related quality of life, and about a third did not explain why they had measured it. Different questionnaires can emphasise different domains, so an “improvement in quality of life” is less informative when the underlying concept is left vague.
A correction published in March replaced an incorrect version of the review's randomised-trial figure. The original article has been updated. This does not erase the overall conclusion, but it is another reason to work from the corrected record rather than repeating a chart without checking it.
What patients describe beyond the numbers
A recent Australian study adds detail to the quantitative picture. Researchers followed nine adults who had been prescribed medicinal cannabis for an anxiety disorder, collecting diary entries at three and six months and interviewing participants within the first year of treatment.
The participants did not generally describe anxiety as cured. They described it as more manageable.
Across the interviews, people reported greater emotional regulation, less reactivity and more perceived control. Some felt better able to use breathing techniques or cognitive coping strategies. Others described changes in sleep, motivation, relationships, social engagement and the ability to complete everyday tasks.
This idea of agency is important. A treatment can be useful not because it replaces every other form of care, but because it creates enough stability for a person to use therapy skills, rebuild routines or re-enter situations that illness had made difficult.
The study also shows why positive patient stories should not be converted into an efficacy claim. Nine people are not a representative sample of everyone prescribed cannabis for anxiety. There was no untreated comparison group or objective quality-of-life measure in this qualitative component. Expectations, recall, concurrent therapy and life events could all influence the accounts.
Participants willing to spend months documenting their experience may also have had more favourable views than people who stopped early or felt no benefit. The researchers were explicit that the study explored how people understood their experiences; it did not establish causation.
That does not make the interviews disposable. Clinical trials can identify an average difference. Qualitative research can show what a difference looks like in real life and which outcomes deserve to be tested next.
A larger cohort found improvement—and substantial attrition
The Australian QUEST initiative provides a broader observational view. It recruited adults with chronic health conditions who were newly prescribed one of four medicinal-cannabis oils and asked them to complete validated questionnaires before treatment and during follow-up.
Of 2,744 people who completed baseline measures, 2,353 provided at least one follow-up and were included in the analysis. Chronic pain was the most common reason for treatment, followed by insomnia and anxiety. More than half were being treated for more than one condition.
Average quality-of-life scores improved after treatment began and the gains were broadly maintained among respondents over 12 months. On the EQ-5D index, the adjusted average change from baseline to mean follow-up was 0.114, above commonly used thresholds for a meaningful difference. Sleep, fatigue, pain, depression and anxiety measures also improved across the cohort.
The study's size, pre-treatment baseline and repeated validated measures make it valuable. It also had no randomised control group, and follow-up participation fell sharply. Only 778 people completed the 12-month assessment. The paper reports this as 38%, although 778 divided by the stated analysis sample of 2,353 is approximately 33%; we retain the counts rather than repeat the inconsistent percentage.
Among enrolled participants, 322 formally withdrew. Reported reasons included lack of therapeutic benefit, choosing another treatment, unwanted side effects and cost. Those people belong to the treatment story just as much as those who remained and reported improvement.
Attrition does not automatically invalidate a cohort, but it can make the remaining group look more positive. People who benefit or tolerate treatment may be more motivated to continue questionnaires than people who do not. Regression to the mean is another concern: patients often seek a new treatment when symptoms are unusually bad, after which some improvement may have happened regardless.
The strongest conclusion is therefore that many continuing patients reported sustained, meaningful improvement—not that the oils were proven to cause every change.
The newest cross-sectional study is more limited
A July study from Brazil assessed quality of life and sleep among 71 people already receiving supervised full-spectrum cannabis oil. Responses tended towards satisfaction, and better sleep was associated with psychological wellbeing and energy for daily activities.
The study used recognised questionnaires and documented the oils being used. But it assessed participants at one point in time, without a pre-treatment baseline or control group. Only 71 of 689 invited patients supplied complete responses.
It therefore cannot show whether quality of life improved after treatment, how respondents differed from non-respondents or what would have happened without cannabis. The authors appropriately describe the result as favourable patient perception rather than evidence of efficacy.
This is a useful example of how headlines can outrun a paper. “Patients on medical cannabis report generally positive quality of life” is supported. “Cannabis improved quality of life” requires a design capable of measuring change and testing competing explanations.
Why the positive signal still matters
It would be easy to respond to these limitations by dismissing all patient-reported improvement as bias. That would be a mistake.
Consistent positive signals across interviews, registries, cohort studies and controlled trials indicate that quality of life is an outcome worth taking seriously. Even the systematic review's smaller randomised-trial effect suggests the subject is not explained entirely by enthusiasm from selected patients.
The evidence also challenges a narrow model of success. Cannabinoid medicines may produce different effects across pain, sleep, anxiety, appetite and medication burden. A modest change in several connected domains could feel more important than a larger change in a single score.
At the same time, “the whole person” should never become a slogan used to avoid precise measurement. If researchers claim that a treatment improves life broadly, they should define the domains in advance, use validated tools and publish results for people who stop as well as those who remain.
Positive reporting is credible when it includes non-response, adverse effects, cost and uncertainty. It does not need to place every patient in a visual or written language of suffering to prove that their condition is serious.
What this means in the UK
These studies do not change NHS treatment guidance or establish medical cannabis as a routine treatment for anxiety, chronic pain or poor sleep.
For generalised anxiety disorder, NHS care continues to centre talking therapies—usually cognitive behavioural therapy—and established medicines such as selective serotonin reuptake inhibitors when appropriate. People should not stop psychological treatment or prescribed medicine because observational cannabis research sounds hopeful.
For UK patients already receiving a legal private prescription, however, the research supports a more useful review conversation. Follow-up should ask not only whether pain or anxiety fell, but whether the person is sleeping, functioning, connecting and participating more successfully. It should also ask about cognition, daytime impairment, side effects, cost and any activities made harder by treatment.
Goals should be individual and observable. “Improve quality of life” is too broad to monitor. “Attend a weekly exercise class,” “sleep through four nights a week,” “return to part-time work” or “manage the school run without overwhelming anxiety” gives patient and clinician something concrete to review.
Where a treatment produces no functional benefit, causes impairment or creates an unaffordable burden, a positive population average should not be used to pressure someone to continue. Patient-centred care includes recognising non-response.
Our guide to discussing medical cannabis with a GP explains how to describe previous treatments, current symptoms and practical goals without presenting cannabis as a predetermined answer.
What better research should do next
Future randomised trials should treat quality of life as a planned outcome, define it clearly and explain which change would matter to patients. They should be long enough to distinguish an early expectation effect from durable participation in daily life.
Studies also need product-level precision. “Medicinal cannabis” can mean purified CBD, THC-dominant flower, balanced oils or licensed medicines with very different doses and routes. A quality-of-life result cannot be transferred automatically from one product to another.
Retention deserves equal attention. Researchers should report why people stop, continue following them where possible and test how different assumptions about missing data change the result.
Finally, patients should help choose the outcomes. Researchers may prioritise symptom intensity while patients prioritise independence, relationships, sleep or the ability to leave home. Both sets of information are needed to understand whether treatment is genuinely useful.
The editorial view
Medical-cannabis coverage too often swings between two images: miraculous recovery and visible suffering. Neither represents most people living with a long-term condition.
The quality-of-life literature offers a more honest frame. Patients are not merely collections of symptoms, and improvement does not have to look like a cure. It can look like gardening, working, laughing, parenting, exercising or feeling able to make plans again.
The evidence is not yet strong enough to promise those outcomes. The controlled signal is small, much of the larger improvement comes from observational research and people who discontinue treatment are easily lost from view.
But caution should refine a hopeful result, not erase it. The emerging picture is that some prescribed-cannabis patients report more than symptom relief: they report greater capacity to take part in life. The next generation of research should test that possibility with the same seriousness applied to pain scales and laboratory measures.
Key takeaways
- A 2026 systematic review included 64 studies measuring health-related quality of life before and after medicinal-cannabis treatment.
- Randomised trials showed a small short-term improvement; observational studies reported larger improvements over short-, medium- and long-term follow-up.
- Interviews with nine adults prescribed medicinal cannabis for anxiety described greater agency, emotional regulation, sleep, relationships and daily participation, but could not establish causation.
- In the 2,353-person QUEST cohort, average quality-of-life improvements were maintained among respondents, while follow-up fell to 778 people at 12 months.
- A July cross-sectional study found favourable quality-of-life and sleep perceptions among 71 existing patients but had no pre-treatment baseline.
- Patient-reported outcomes are clinically valuable, but expectations, selection, attrition, other care and natural symptom fluctuation can influence them.
- UK patients should judge treatment against specific functional goals as well as symptoms, tolerability, impairment and cost.
Sources
- Health-related quality of life in patients receiving medicinal cannabis: systematic review and meta-analysis — Quality of Life Research
- Correction to the systematic review's randomised-trial figure — Quality of Life Research
- Beyond symptom change: quality-of-life experiences among adults prescribed medicinal cannabis for anxiety disorders — International Journal of Clinical Pharmacy
- QUEST Initiative 12-month follow-up study — PLOS One
- Real-world quality of life and sleep outcomes with THC- and CBD-rich cannabis oil — Frontiers in Pharmacology
- Generalised anxiety disorder: treatment and support — NHS
- Medical cannabis — NHS
