Medicine and risk at the same time.
Both are true.
What cannabis can do as therapy. Where the addiction risk sits biologically. Why fatigue after cannabis could be more than character, what animal data suggest and what is still open in humans. And which other paths exist to reduce pain, sleep better and become clear again.
I am a doctor. I prescribe medical cannabis only for clearly defined indications, and only when the expected benefit outweighs the known risks in my judgement. And at the same time I see what high-potency consumption may do to young people.
Both are true. Cannabis can be medicine, for a small, clearly defined group. Cannabis is at the same time a risk, for a large, often invisible group. Anyone who describes only one of the two sides gives an incomplete picture.
This article is the honest dual view. From the perspective of clinical psychoneuroimmunology. From functional medicine. From anthroposophic medicine. With studies. And with what you can do today to reduce pain, sleep better and become calmer, without cannabis.
Why this article is needed right now
Before we go into the biology, a picture has to land that many patients do not yet know in this form. Since 1 April 2024, cannabis has been removed from the German narcotics law and can be prescribed on a normal eRecipe. That was regulatory progress. What has come of it in practice needs explaining.
How large the increase is exactly, I cannot state responsibly here. Figures from market reports and industry estimates circulate, but no official statistic is available to me. So I deliberately do not name them. That the order of magnitude is considerable is shown by the reaction of the legislator. Part of the increase runs through telemedicine platforms that issue cannabis prescriptions after an online questionnaire. The German Federal Ministry of Health has put forward a draft law requiring an in-person initial contact before a first prescription, plus a ban on mail-order sales of cannabis flower. I am not judging individual providers or colleagues here. I am only describing what the legislator considers in need of regulation.
Why does this matter for you? Because a prescription alone is not yet therapy. I am not describing individual providers here and I am not judging colleagues; I cannot assess someone else's course of treatment from the outside. What I can describe is my own standard: I prescribe only after a personal consultation, with a clear diagnosis, and with follow-up over time. The fact that it is fast and easy today to obtain a cannabis prescription does not mean it is the right therapy for you. It only means that the market has a very low entry barrier.
Easy access to a substance is no indicator of its appropriateness. The question is not whether you can get cannabis. The question is whether cannabis serves your specific problem, or whether it is a tool that displaces other tools that would be more sustainable.
What cannabis does in your body
Before we talk about benefit and risk, let us briefly look at the biology. Your body has its own cannabis system. It is called the endocannabinoid system, ECS for short. It regulates mood, pain, appetite, sleep, immune function and stress response. You have it long before you have ever thought about cannabis.
Two receptor families carry the system. CB1 sits mainly in the brain and nervous system. CB2 sits mainly on immune cells. Your body produces its own cannabinoids, especially anandamide, often called the "bliss molecule" because its name comes from the Sanskrit word for bliss.
Cannabis brings two main molecules into play. THC binds strongly to CB1, is psychoactive, and can influence pain perception and appetite. How strongly differs a great deal by complaint; further down I look at the numbers more closely. Cannabidiol, CBD for short, acts indirectly on the ECS, can dampen THC's effect, and is not intoxicating. There are signals for anxiolytic, anti-inflammatory and anticonvulsant effects of its own; how well they are documented differs a great deal by indication.
One point that often gets lost with cannabidiol: it can interfere with the breakdown of other medicines in the liver. Best studied is the interaction with clobazam, a prescription-only antiseizure and sedative medicine. The combination can raise the concentration of its active metabolite and with it the side effects. When given together with valproate, elevated liver values were seen more often, and for coumarin-type blood thinners there are case reports of altered clotting values. If you take medication regularly, discuss cannabidiol with your doctor beforehand, even if you buy it over the counter.
Patsalos PN et al. Clinical implications of trials investigating drug-drug interactions between cannabidiol and enzyme inducers or inhibitors or common antiseizure drugs. Epilepsia. 2020;61(9):1854 to 1868. DOI: 10.1111/epi.16674
Cannabis is not a foreign substance. It is a body-internal system that you trigger from outside with an extremely strong dose. The question is not whether it works, but what that dose does to the system over time.
The mitochondria hypothesis: a possible lead, not yet demonstrated in humans
Here comes a lead that rarely appears in consultations. It comes from basic research in animals and cells. In humans it has not been tested.
For a long time, science assumed CB1 receptors sit only on the cell membrane of nerve cells. In 2012, Bénard's team found something else. CB1 receptors also sit directly on the mitochondria, the power plants of your cells. They are called mtCB1.
In mouse neurons and cell cultures, Bénard and colleagues were the first to demonstrate functional CB1 receptors on the membranes of neuronal mitochondria. Activation of mtCB1 lowered cAMP, the activity of complex I in the respiratory chain and mitochondrial respiration. This is a finding from animal and cell experiments. It makes it plausible that cannabis can also reach into cellular energy production in humans. It does not demonstrate it in humans.
Bénard G et al. Mitochondrial CB1 receptors regulate neuronal energy metabolism. Nat Neurosci. 2012;15(4):558 to 564. DOI: 10.1038/nn.3053
Four years later Hebert-Chatelain showed in mice: cannabinoid-induced memory impairment did not occur when the mitochondrial CB1 receptors in the hippocampus were genetically removed. That suggests the cognitive dampening could work not only through synaptic effects, but also through the energy supply of the brain cells. Transfer to humans is not yet demonstrated.
Hebert-Chatelain E et al. A cannabinoid link between mitochondria and memory. Nature. 2016;539(7630):555 to 559. DOI: 10.1038/nature20127
In isolated mitochondria from rat brain, THC lowered maximal oxidative capacity by 71 percent and succinate-driven respiration by 65 percent. At the same time, hydrogen peroxide production rose sharply. This is a test-tube finding in animal tissue at high concentrations. How strong the effect is in a person who consumes cannabis, and how long it lasts, was not examined in this work.
Important for context: the authors had a different aim than the fatigue question. They were looking for a mechanism behind something observed in neurology, namely strokes in young cannabis users. That link is not proven. But it belongs in an honest picture, especially since a section for 18 to 28 year olds follows further down.
Wolff V et al. Tetrahydrocannabinol induces brain mitochondrial respiratory chain dysfunction and increases oxidative stress: a potential mechanism involved in cannabis-related stroke. Biomed Res Int. 2015;323706. DOI: 10.1155/2015/323706
What could that mean in practice? The well-known cannabis aftermath of heaviness, sluggishness, low drive the next day need not be purely psychological. It need not mean "I have just become lazier". From the animal and cell data above you can derive the thought that the energy supply of nerve cells may play a part. That is a physiological consideration without a human study to back it. I am deliberately giving it to you as a hypothesis, because it opens a different question than "pull yourself together".
On top of that comes a second observation. In a small imaging study, a reduced striatal dopamine synthesis capacity was measured in regular users who had experienced psychosis-like symptoms under cannabis. Whether the loss of drive follows from it is not settled. It is a snapshot in a very small group, and it says nothing about what is cause and what is consequence. It does take the moral edge off the topic: what is measured here is something physical, not a character trait.
Using PET imaging, Bloomfield and colleagues compared 19 regular users who had experienced psychosis-like symptoms under cannabis with 19 non-users. That selection matters: the group is not representative of everyone who uses regularly. On average, the users had lower striatal dopamine synthesis capacity, most clearly those who met abuse or dependence criteria. For context: 38 people is a very small sample, the study is a snapshot without follow-up over time, and the authors specifically did not find a link with psychosis-like symptoms. They themselves question the simple dopamine explanation.
Bloomfield MAP et al. Dopaminergic Function in Cannabis Users. Biol Psychiatry. 2014;75(6):470 to 478. DOI: 10.1016/j.biopsych.2013.05.027
If you feel more tired, slower, less motivated after extended cannabis use, that is not a character flaw. It could also be biology. This perspective takes blame out of the topic. It says nothing about how fast or how completely anything changes for you, that is individual and not predictable.
Where cannabis as medicine has solid data
Medical cannabis is a prescription-only medicine. On individual preparations, on indications, and on whether a prescription is an option for you, I am not permitted to make public statements. That is not caution on my part, German medicines advertising law requires it. What I can describe here is the general evidence. It does not apply to everything, not to every form, not to every person. Everything beyond that belongs in a personal medical consultation.
79 trials, 6462 participants, only four of them at low risk of bias. Moderate-quality evidence was found for chronic pain and for spasticity. For chemotherapy-induced nausea, weight gain in HIV, sleep disorders and Tourette, the evidence was of low quality. The authors also report an increased risk of short-term adverse events, including serious ones: dizziness, dry mouth, nausea, fatigue, euphoria, vomiting, disorientation, confusion, loss of balance and hallucination. The effects are real, but smaller than public discussion often suggests.
Whiting PF et al. Cannabinoids for Medical Use: A Systematic Review and Meta-analysis. JAMA. 2015;313(24):2456 to 2473. DOI: 10.1001/jama.2015.6358
Nabiximols, an oromucosal THC and CBD spray, is approved in Germany as a prescription-only medicine for treatment-resistant spasticity in multiple sclerosis. A pooled analysis of three randomised trials with 666 patients shows: on a 0 to 10 scale, spasticity fell by 1.30 points under nabiximols and by 0.97 points under placebo. The difference from placebo was therefore 0.32 points, statistically significant but small. Slightly more of those treated reached at least a 30 percent improvement. Adverse events occurred in a high proportion of participants, mostly mild to moderate. Part of the context: two of the four authors of this pooled analysis were employed by the manufacturer of the product.
Wade DT et al. Meta-analysis of the efficacy and safety of Sativex (nabiximols), on spasticity in people with multiple sclerosis. Mult Scler. 2010;16(6):707 to 714. DOI: 10.1177/1352458510367462
In children and young adults with Dravet syndrome, a severe drug-resistant epilepsy, monthly convulsive seizures fell under cannabidiol from a median of 12.4 to 5.9. Under placebo they fell from 14.9 to 14.1, so the placebo-adjusted difference was around 23 percentage points. A parallel 2018 trial showed similar effects in Lennox-Gastaut syndrome. Cannabidiol is approved as a finished medicinal product for these indications. The effect comes at a price: diarrhoea, vomiting, fatigue, fever, somnolence and abnormal liver function tests occurred more often under cannabidiol than under placebo, as did study withdrawals. A central safety point of these trials is the interaction with the clobazam given alongside, see the note further above. Cannabidiol in this form is a prescription-only medicine and belongs in neurological or paediatric hands.
Devinsky O et al. Trial of Cannabidiol for Drug-Resistant Seizures in the Dravet Syndrome. N Engl J Med. 2017;376(21):2011 to 2020. DOI: 10.1056/NEJMoa1611618
16 RCTs, 1750 patients with chronic neuropathic pain. Pain relief of at least 50 percent was reached by 21 percent under cannabis-based medicine versus 17 percent under placebo, on low-quality evidence. Against that stand clear adverse effects: nervous system events in 61 versus 29 percent, psychiatric adverse events in 17 versus 5 percent, and twice as many withdrawals due to adverse events. The authors conclude that the potential benefit may be outweighed by the potential harms. On long-term risks they found no data.
Mücke M et al. Cannabis-based medicines for chronic neuropathic pain in adults. Cochrane Database Syst Rev. 2018;3(3):CD012182. DOI: 10.1002/14651858.CD012182.pub2
Beyond that, there are smaller positive signals for cannabidiol in social phobia (Bergamaschi 2011, Crippa 2011). The evidence for sleep disorders and general anxiety is considerably thinner than public presentation often suggests.
Where the evidence has gaps you should know about
Cannabis is often presented as a universal solution. For sleep, for stress, for anxiety, for "inner restlessness". Honestly, this is a marketing phenomenon, not a research finding.
For sleep disorders, cannabis has acute sedative effects; with chronic use a tolerance can develop. On withdrawal, sleep can get markedly worse, which may keep the consumption going without the underlying cause being addressed. For anxiety, cannabidiol can act anxiolytically in low doses, while THC can trigger anxiety, especially at higher doses or in predisposed individuals. For depression and PTSD, the evidence is mixed, with signals that chronic use may worsen outcomes.
If cannabis acutely helps you sleep or calm down, that is real. But subjective relief is not proof of efficacy. And it is certainly not proof that cannabis is the best thing you can do for your sleep or your anxiety.
Addiction risk is biology, not weakness
Here it gets uncomfortable. Cannabis has a well-documented dependence potential. That belongs in any honest account.
36,309 US adults were systematically interviewed. Result: 6.3 percent of adults met DSM-5 criteria for a Cannabis Use Disorder at some point in life, 2.5 percent within the past twelve months. Young adults, men and people on low incomes were affected considerably more often. Only around 13 percent of those affected ever took up treatment or a self-help programme. How high the risk is for one individual cannot be derived from a population survey.
Hasin DS et al. Prevalence and Correlates of DSM-5 Cannabis Use Disorder. Am J Psychiatry. 2016;173(6):588 to 599. DOI: 10.1176/appi.ajp.2015.15070907
On top comes the Cannabis Withdrawal Syndrome, officially recognized in DSM-5. A meta-analysis by Bahji and colleagues (2020) of 47 studies with 23,518 participants found it in around 47 percent of people with regular or dependent use. In population samples the figure was 17 percent, in outpatient groups 54 percent, in inpatient groups 87 percent. Variation between studies was very high, which limits how much the pooled figure can carry. Typical on cessation are irritability, anxiety, sleep disturbance, depressed mood, appetite loss, sometimes sweating and tremor. The classic "I can quit anytime" is often relativised within the first week of quitting. This is not weak willpower, this is pharmacology.
Two thresholds you must know: adolescence and high potency
There are two constellations where the evidence is so clear that minimisation no longer holds.
1037 New Zealanders, prospectively followed from birth to age 38. IQ tested at age 13 (before any cannabis exposure) and at age 38. In this cohort, participants with adolescent onset and persistent heavy use showed the clearest cognitive decline, around 8 IQ points in the most affected group. With adult onset the decline was not detectable, and stopping did not fully restore functioning in the early-onset group. Whether the decline is due to cannabis alone or also to social factors is still professionally debated. This is an observational study, not proof.
Meier MH et al. Persistent cannabis users show neuropsychological decline from childhood to midlife. Proc Natl Acad Sci USA. 2012;109(40):E2657 to E2664. DOI: 10.1073/pnas.1206820109
Systematic review of 66,816 individuals across 18 studies, of which 10 entered the meta-analysis itself. The heaviest-using group had an odds ratio of 3.90 for schizophrenia and psychosis spectrum diagnoses, compared to non-users. Clear dose-response relationship. Causality is not finally proven, but this is the strongest risk constellation in cannabis research.
Marconi A et al. Cannabis Use and Risk of Psychosis Meta-analysis. Schizophr Bull. 2016;42(5):1262 to 1269. DOI: 10.1093/schbul/sbw003
Multicentre study, 901 patients with a first episode of psychosis plus 1237 controls from eleven sites in Europe and Brazil. Daily cannabis users had roughly threefold elevated odds of a first psychosis (odds ratio 3.2), and daily users of high-potency cannabis with over 10 percent THC had 4.8-fold elevated odds. Assuming the association is causal, around 30 percent of new psychosis cases in London and around 50 percent in Amsterdam could be attributed to high-potency use, and around 12 percent across all eleven sites. This is a case-control study; it cannot prove a cause.
Di Forti M et al. Cannabis and Incidence of Psychotic Disorder EU-GEI. Lancet Psychiatry. 2019;6(5):427 to 436. DOI: 10.1016/S2215-0366(19)30048-3
If several of these points apply to you, a family history of psychosis, being under 25, daily high-potency use, then statistically you are in the group for which the data are clearest. That says nothing about what will happen to you. It is a reason to look more closely, not a reason to panic.
The anthroposophic view: cannabis and the awake "I"
From the perspective of anthroposophic medicine, the human being is threefold. There is the nerve-sense system (consciousness, clarity, alertness, differentiation). There is the rhythmic system (breathing, heart, feeling, the centre). And there is the metabolic-limb system (will, digestion, warmth, movement). Health is balance between these three poles.
Cannabis shifts this balance characteristically. It dampens the awake nerve-sense system and amplifies the dreaming, dissolving pole. In Steiner's language: it pushes the human being towards the luciferic pole, away from sharp differentiation. That is exactly what makes the subjective appeal: the "becoming softer", the "letting go", the "sliding into images".
For occasional use, this tradition describes the shift as unproblematic. With chronic use, something else is described. The awake "I", the middle force that balances between dream and hardening, can become quieter. Differentiation can get harder, decisions softer, perception less precise. This is a description from a tradition more than a hundred years old, not a measurement. I find it useful as an image. It is not scientific proof.
Anthroposophic medicine does not say "cannabis is evil". It says: honour the clarity of your "I". Ask yourself: when do I need this substance because life right now demands something I cannot otherwise carry, and when do I use it to avoid the clarity that life is currently asking of me?
If the answer is consistently the second, then cannabis works against what the human being is meant to do: stay awake, present, and able to discern.
PNI and functional medicine: root work instead of symptom dampening
Clinical psychoneuroimmunology does not look at the symptom but at four lenses: nervous system, immune system, metabolism, hormones. From that perspective I see it this way: cannabis can dampen symptoms across all four lenses. Whether the cause is also addressed depends on the individual case. With the themes most people bring to me, my impression is that it stays unworked. That is a stance from practice, not a study result.
If someone uses cannabis because they cannot sleep, the more useful question, in my view, is not "does cannabis work for sleep". It is: why is your system not sleeping? Disrupted cortisol rhythm, blue light late at night, magnesium deficiency, unresolved conflict, caffeine-driven day, no morning daylight, mitochondrial weakness, inflammation, late sugar? That is PNI work. Cannabis can help with falling asleep short term. In my experience the underlying causes do not change through it.
If someone uses cannabis for chronic pain, the question is: which of the four PNI lenses drives this pain? Nervous system on permanent alarm? Silent inflammation? Tissue mitochondrial weakness? Hormonal shifts? Here too, cannabis can dampen. Whether anything happens at the roots is not settled by that.
From a functional medicine perspective the guiding principle is: symptom dampening without root work can cost more long term than it saves short term. That is a stance, not a study result.
Where I stand as a doctor on cannabis prescriptions
Whether medical cannabis is an option for an individual person is decided in a medical examination, not on a website. What I can describe here is the general evidence. And my own standard: never without a clear diagnosis, never without a personal consultation, never without longer-term follow-up, and rarely as a first lever.
1. Severe treatment-resistant spasticity in multiple sclerosis
For this indication there is an approved medicine in Germany and there are randomised trials. The measured difference from placebo is statistically significant but small, and adverse events are common. Whether it holds up in an individual case is decided in neurological care, not on a website.
2. Treatment-resistant chronic neuropathic pain
Here we have the Cochrane review by Mücke 2018, with a small added benefit over placebo and relevant adverse effects, above all nervous system and psychiatric ones. This evidence carries no promise of freedom from symptoms. It is about quality of life once the alternatives are exhausted.
3. Oncology-related symptoms
On cannabinoids in the context of cancer treatment I am not permitted to make public statements outside professional circles, neither on benefit nor on limits. German medicines advertising law forbids it. This question belongs in your oncology care, where it is best placed.
What I have to say just as clearly: in the vast majority of cases where patients come to me asking for cannabis, the indication is not unambiguous. It is about sleep problems, stress, anxiety, diffuse pain, "I do not know how to come down anymore". For these themes, cannabis is in my view rarely the first choice. Not because it does nothing, but because data collected over longer periods exist for other paths, and no dependence potential is described for them.
A prescription follows an indication, not a wish. That is not a judgement about individual colleagues, it is the standard I hold my own decisions to. If something was left unclear about your prescription, the right address for that question is the doctor who wrote it.
Other paths by indication
If you feel cannabis is the only thing that helps you, let us go through the evidence indication by indication. For some of these concerns there are data on other paths, collected over longer periods. Whether one of them works better for you than cannabis has never been directly compared. I present them so you know they exist, not as a ranking.
Pain
A Cochrane overview of 21 reviews covering 381 studies found favourable but small to moderate effects of exercise on pain severity and physical function, on low-quality evidence overall and with very few adverse events. That overview did not calculate a direct comparison with pain medication, so I am not making one. There is a second thought: after acute exercise, endogenous endocannabinoids rose in most human studies. According to the authors of that review, reliable proof that this is what produces the good feeling after sport is still outstanding.
Geneen LJ et al. Physical activity and exercise for chronic pain in adults: an overview of Cochrane Reviews. Cochrane Database Syst Rev. 2017;4(4):CD011279. DOI: 10.1002/14651858.CD011279.pub3
On omega-3 fatty acids there is a recent review by Xie and colleagues covering 41 randomised trials with 3759 patients, looking at chronic pain. The authors report a moderate reduction in pain intensity (SMD minus 0.55), though with very high variation between studies. The result was clearer in rheumatoid arthritis and migraine; in osteoarthritis and mastalgia there was none. Discussed mechanisms are reduced prostaglandin formation, pro-resolving mediators such as resolvins, and calmer microglia. That is an explanatory model, not proof measured in patients.
Two points of context. First, legally: as a food supplement, omega-3 may not be advertised in Europe with an effect on pain; only narrowly defined claims on heart function, blood pressure, triglycerides and, for DHA, on vision and brain function are authorised. What this study means for you depends on your findings and belongs in a medical consultation. Second, on safety: higher amounts can affect blood clotting. If you take anticoagulants or platelet inhibitors, or if surgery is coming up, speak to your doctor beforehand.
Xie L et al. Effects of omega-3 fatty acids on chronic pain: a systematic review and meta-analysis. Front Med (Lausanne). 2025;12:1654661. DOI: 10.3389/fmed.2025.1654661
One substance that often comes up in this context is naltrexone, an opioid receptor blocker. It is a prescription-only medicine and in Germany it is not approved for fibromyalgia. Use for this condition would be off-label, with separate informed consent, a particular duty of medical justification, and as a rule no reimbursement. I am reporting the evidence only. This is not a recommendation, and no offer of my practice is attached to it.
In four randomised trials with 222 fibromyalgia patients in total, low-dose naltrexone reduced pain scores significantly, though the difference from placebo was small at 0.86 points. On function and pain catastrophising there was no measurable effect. Vivid dreams and nausea occurred more often than under placebo; serious adverse events were not increased. The discussed mechanism is a transient opioid receptor blockade followed by elevated endorphin tone. That is an explanatory model, not proof.
What nobody may overlook here: if you take opioids, naltrexone is not an option. It can trigger an acute withdrawal syndrome and cancel the effect of your pain medication. Caution also applies in liver disease. This is one of the reasons it belongs in medical hands and never in a self-experiment. On dosages I deliberately say nothing here.
Vatvani AD et al. Efficacy and safety of low-dose naltrexone for the management of fibromyalgia: a systematic review and meta-analysis of randomized controlled trials. Korean J Pain. 2024;37(4):367 to 378. DOI: 10.3344/kjp.24202
Sleep
Cognitive behavioural therapy for insomnia is the first-line therapy for chronic insomnia in nearly all guidelines. In a meta-analysis of 20 studies with 1162 participants it shortened sleep onset by around 19 minutes and wake time after sleep onset by around 26 minutes. The effects appeared to be sustained at later time points, though the authors themselves call these long-term estimates less certain. No adverse outcomes were reported in this work. That does not mean there are none: in the first weeks, sleep restriction can make you tired during the day, and that matters for driving and at work. This work contains no direct comparison with cannabis.
Trauer JM et al. Cognitive Behavioral Therapy for Chronic Insomnia: A Systematic Review and Meta-analysis. Ann Intern Med. 2015;163(3):191 to 204. DOI: 10.7326/M14-2841
On top come the basics that many patients underestimate: dark, cool bedroom, no screens in the last hour before sleep, regular sleep times, magnesium, which contributes to normal function of the nervous system and should be taken only after medical advice if kidney function is impaired, no caffeine after 2 pm, daylight in the first hour of the day for the cortisol axis. PNI-aligned additions: check the cortisol day profile, avoid evening blood sugar swings, surface unresolved themes instead of deckling them with a substance.
Stress and chronic overload
Stress is the most common self-justification that brings people to a joint in the evening. There are data on this: three methods performed similarly well in a direct comparison, and no dependence potential is described for any of them.
A randomised trial with 76 participants directly compared physical activity, mindfulness meditation, and HRV biofeedback as five-week self-help exercises. Stress, anxiety and depressive symptoms fell in all three groups, with no significant difference between the methods. For context: there was no untreated control group and no comparison with cannabis. The study shows that these three paths perform similarly, not how large their effect is against doing nothing.
van der Zwan JE et al. Physical Activity, Mindfulness Meditation, or HRV Biofeedback for Stress Reduction RCT. Appl Psychophysiol Biofeedback. 2015;40(4):257 to 268. DOI: 10.1007/s10484-015-9293-x
Meta-analysis of 24 studies with 484 participants. HRV biofeedback was associated with a marked reduction in self-reported stress and anxiety (Hedges g 0.81 pre-post, 0.83 against control conditions). The authors explicitly note that more well-controlled studies are needed. The participant numbers are small, which qualifies the size of the effect.
Goessl VC et al. HRV biofeedback on stress and anxiety meta-analysis. Psychol Med. 2017;47(15):2578 to 2586. DOI: 10.1017/S0033291717001003
On top come the slower levers: anthroposophic compresses (lavender abdominal compress in the evening), warmth applications, time in nature, relational clarity, honest communication. For the anthroposophic applications there is clinical tradition but no strong study base; I name them as experience, not as evidence. For time spent in nature there are signals of favourable effects on stress markers, though the studies are small and methodologically inconsistent. Whoever uses cannabis against stress dampens a symptom. Whoever works on these points can change something in the system itself.
Anxiety and inner restlessness
Non-pharmacological levers with data behind them: slow, coherent breathing (four seconds in, six seconds out, five minutes daily) can raise HRV; a mediating influence via the vagus nerve is discussed. Mindfulness and MBSR: a meta-analysis of 38 randomised trials found a small effect on pain intensity in chronic pain, on low-quality evidence, plus effects on depressive symptoms and quality of life. Exercise reduced anxiety symptoms with a moderate effect size in a meta-analysis of six randomised trials with 262 adults; endocannabinoids and BDNF are discussed, not proven, explanations. For targeted psychotherapy at the roots (trauma, conflicts, relationships) there are data collected over longer periods. How robust they are depends on the method and on the condition. No direct randomised comparison of these paths with cannabis exists.
For generalised anxiety, the PNI lenses are worth checking: low magnesium may increase neuronal excitability, low vitamin D is associated with higher anxiety prevalence, thyroid and iron status can co-shape the inner tone. If one of these is off, even the most relaxing joint can do little about it.
Trauma and post-traumatic stress
This is where cannabis is particularly often used on personal initiative, with partly understandable subjective effects. The evidence shows a clear gap between subjective experience and what is documented as durable trauma processing.
A systematic review found exactly one randomised trial and ten observational studies with 4672 people in total on cannabis in PTSD. The evidence comes mostly from low-quality studies at high risk of bias, usually without a comparison group. In those studies, cannabis was associated with fewer PTSD symptoms and better quality of life. Common adverse effects were dry mouth, headache and psychoactive effects such as agitation and euphoria; in a small proportion, PTSD symptoms got worse. What does not follow from this evidence: that cannabis processes the trauma itself. That would require controlled trials, which so far do not exist.
Rehman Y et al. Cannabis in the management of PTSD: a systematic review. AIMS Neuroscience. 2021;8(3):414 to 434. DOI: 10.3934/Neuroscience.2021022
What is used in trauma with guideline support: trauma-focused cognitive behavioural therapy, including prolonged exposure, as first-line therapy. EMDR (Eye Movement Desensitization and Reprocessing), also recommended in the guidelines. Somatic Experiencing as a body-oriented approach, with a considerably thinner study base. EAET (Emotional Awareness and Expression Therapy) for somatised trauma sequelae with chronic pain, likewise still little studied. These methods are meant to work on the trauma itself, not only on its consequences. Whether they work better than a substance in an individual case has not been directly compared.
If you are young and use cannabis against life-feel themes
A particular constellation, because it is especially common today. Younger adults between 18 and 28 who use cannabis against "inner emptiness", "meaninglessness", "lack of depth", "too much overstimulation". Three points are important here: first, the adolescent and early-adult brain is in a critical maturation phase until the mid 20s. Second, high-potency material is in this constellation the largest risk factor for a first psychotic episode. Third, what often feels like "inner emptiness" is very often an unanswered hunger for meaning plus a measurable biological depletion (sleep, micronutrients, lack of movement, screen time, missing relational depth).
The most honest therapeutic answer here is a mixture of relational therapy, PNI diagnostics (thyroid, cortisol profile, vitamin D, iron, microbiome), physical activity, meaning work, and where appropriate anthroposophic support. Cannabis can soften this experiential gap acutely. It does not fill it.
Cannabis as a tool
Acutely effective, fast sedation, fast pain dampening, fast calming.
Tolerance and withdrawal symptoms are documented. Mitochondrial throttling has so far been shown mainly in animal and cell experiments, reduced dopamine synthesis in one small imaging study in a selected group. Whether root causes get worked on depends on the individual case.
Other paths
Exercise, breathing, CBT-I, mindfulness, root-level psychotherapy, targeted micronutrients according to findings. Medical routes outside a licence are not self-experiments and belong in medical hands.
Slower build-up, but for some of these paths data collected over longer periods exist, and no dependence potential is described for them. A direct comparison with cannabis is missing. Effect size and course differ from person to person.
If you want to cut down or stop
Anyone who uses regularly and wants to stop knows the pattern: a few days go well, then come sleep problems, irritability, vivid dreams, restlessness, the gnawing craving. That is not weakness, those are withdrawal symptoms. They usually fade within two to four weeks, and the first days are typically the hardest. How it goes for you cannot be predicted in advance.
One note that matters more here than anything else: the described withdrawal symptoms include marked low mood. If thoughts of not wanting to live any more come up with it, get help immediately. In Germany, Telefonseelsorge is reachable around the clock, free and anonymous, on 0800 111 0 111 or 0800 111 0 222. In an acute situation, call 112.
Stopping is rarely a fight against a substance. Usually the question behind it is what the system expected from cannabis: calm, sleep, reward, meaning.
Cannabis dependence is legally a substance use disorder. On treatment routes for a substance use disorder I am not permitted to make public statements outside professional circles, not on the evidence and not on medication either. German medicines advertising law forbids it. So there is deliberately no sequence, no method and no offer of my practice here.
If you want to cut down or stop, the addiction counselling centres in your city are the right address. They are free of charge and anonymous on request. In Germany, the Federal Centre for Health Education maintains a nationwide directory. Your general practice or a psychotherapy practice are a good first step too.
And a safety note: do not take anything on your own initiative to cushion withdrawal, neither medication nor high-dose supplements. That belongs in medical care.
In a secondary analysis of a medication trial in 62 treatment-seekers with Cannabis Use Disorder, reaching abstinence by the end of treatment was associated with higher quality of life. The same held for a reduced proportion of using days. The group is small, so no causal claim can be drawn from it. Still, the finding suggests that quality of life can change measurably over the course of treatment. I expressly derive no treatment offer from it.
Brezing CA et al. Abstinence and reduced frequency of use are associated with improvements in quality of life among treatment-seekers with cannabis use disorder. Am J Addict. 2018;27(2):101 to 107. DOI: 10.1111/ajad.12660
Three honest levers for your self-check
First lever. Ask yourself the honest question: if I stop today, would I feel free energy, or a loss? If the answer is "loss", that is no proof that cannabis benefits you. It may also mean that your system has adapted to the substance. A diagnosis does not follow from this question; it is only a reason to look more closely.
Second lever. Breathe slowly and evenly for five minutes a day (four seconds in, six seconds out). For slow breathing and HRV biofeedback there are data on improved heart rate variability and less self-reported stress. That this raises your endogenous endocannabinoid tone has not been studied, so I will not sell you that as a reason. The reason is simpler: it costs nothing and you can start today. Slow breathing is well tolerated by most people. If it makes you dizzy or restless, make it shorter and less deep. Anyone who suffers from panic attacks is best introducing such exercises with therapeutic support.
Third lever. Know the lines where risk rises most clearly in the studies: daily use, high-potency material, onset in adolescence, a family history of psychosis, cannabis as a permanent solution for sleep or stress. These constellations carry the clearest risks. This is expressly not a recommendation to use; I am only describing where the data are most unambiguous.
Clarity is not less pleasure. Clarity can be the precondition for enjoying anything deeply. An awake nervous system may allow more sensory depth than a dampened one. That is my stance, not a measurement. Whether this path works for you is something you can only find out by walking it.
In an emergency: call 112. If someone is confused after use, hears voices, feels persecuted, is panicking or unresponsive, that is a medical emergency. Then you call the emergency number 112, not a practice.
If you have thoughts of not wanting to live any more: in Germany, Telefonseelsorge on 0800 111 0 111 or 0800 111 0 222, around the clock, free and anonymous. In an acute situation, 112 as well.
Never stop prescribed medication on your own, including prescribed cannabis. That always belongs in a conversation with the doctor who prescribed it.
Driving and work: THC can impair fitness to drive for hours. In a review of 80 publications, most driving-related skills recovered only over roughly five to seven hours after inhalation, and oral intake can take longer (McCartney 2021). In German road traffic a statutory THC limit of 3.5 nanograms per millilitre in blood serum applies. For drivers in their probationary period and for everyone under 21, an absolute ban on cannabis at the wheel applies on top. So impairment can still be present when you have long felt sober.
Combinations: the sedating effect of THC can add up with alcohol, with benzodiazepine-type sedatives and with opioids. With opioids a risk of respiratory depression comes on top. Such combinations are dangerous.
Recurring nausea and vomiting: with regular use a distinct clinical picture is described, cannabinoid hyperemesis syndrome. Typical are cyclically recurring nausea, vomiting and abdominal pain, often with an urge to shower or bathe hot, and resolution after stopping use. It can lead to fluid and electrolyte loss. If you notice this in yourself, it needs medical assessment (Sorensen 2017).
Cannabidiol and other medicines: cannabidiol can interfere with the breakdown of other medicines in the liver, among them certain antiseizure drugs and coumarin-type blood thinners. If you take medication regularly, discuss it medically beforehand, including for over-the-counter products.
Particular caution in pregnancy and breastfeeding, before the age of 25, with psychosis or bipolar disorder in your own or your family history, and with cardiovascular disease.
This article is information. It is not medical advice, not a diagnosis, and not a treatment recommendation for your case.
If you would like to talk about your own situation, you will find the option to book an appointment below this article. What is possible in a consultation and what is not, we clarify there in person. For an existing dependence, addiction counselling and addiction medicine are the responsible services, and that is where I refer people too.
Sources
- Bénard G et al. Mitochondrial CB1 receptors regulate neuronal energy metabolism. Nat Neurosci. 2012;15(4):558 to 564. DOI: 10.1038/nn.3053
- Hebert-Chatelain E et al. A cannabinoid link between mitochondria and memory. Nature. 2016;539(7630):555 to 559. DOI: 10.1038/nature20127
- Wolff V et al. Tetrahydrocannabinol induces brain mitochondrial respiratory chain dysfunction and increases oxidative stress: a potential mechanism involved in cannabis-related stroke. Biomed Res Int. 2015;323706. DOI: 10.1155/2015/323706
- Bloomfield MAP et al. Dopaminergic Function in Cannabis Users. Biol Psychiatry. 2014;75(6):470 to 478. DOI: 10.1016/j.biopsych.2013.05.027
- Whiting PF et al. Cannabinoids for Medical Use Meta-analysis. JAMA. 2015;313(24):2456 to 2473. DOI: 10.1001/jama.2015.6358
- Mücke M et al. Cannabis-based medicines for chronic neuropathic pain. Cochrane Database Syst Rev. 2018;3(3):CD012182. DOI: 10.1002/14651858.CD012182.pub2
- Wade DT et al. Meta-analysis of the efficacy and safety of Sativex (nabiximols), on spasticity in people with multiple sclerosis. Mult Scler. 2010;16(6):707 to 714. DOI: 10.1177/1352458510367462
- Devinsky O et al. Cannabidiol for Drug-Resistant Seizures in Dravet. NEJM. 2017;376(21):2011 to 2020. DOI: 10.1056/NEJMoa1611618
- Devinsky O et al. Cannabidiol on Drop Seizures in Lennox-Gastaut. NEJM. 2018;378(20):1888 to 1897. DOI: 10.1056/NEJMoa1714631
- Bergamaschi MM et al. Cannabidiol Reduces Anxiety in Social Phobia. Neuropsychopharmacology. 2011;36(6):1219 to 1226. DOI: 10.1038/npp.2011.6
- Crippa JA et al. Neural basis of anxiolytic effects of cannabidiol (CBD) in generalized social anxiety disorder: a preliminary report. J Psychopharmacol. 2011;25(1):121 to 130. DOI: 10.1177/0269881110379283
- Patsalos PN et al. Clinical implications of trials investigating drug-drug interactions between cannabidiol and enzyme inducers or inhibitors or common antiseizure drugs. Epilepsia. 2020;61(9):1854 to 1868. DOI: 10.1111/epi.16674
- Hasin DS et al. DSM-5 Cannabis Use Disorder. Am J Psychiatry. 2016;173(6):588 to 599. DOI: 10.1176/appi.ajp.2015.15070907
- Bahji A et al. Cannabis Withdrawal Syndrome Prevalence Meta-Analysis. JAMA Netw Open. 2020;3(4):e202370. DOI: 10.1001/jamanetworkopen.2020.2370
- Marconi A et al. Cannabis Use and Risk of Psychosis Meta-analysis. Schizophr Bull. 2016;42(5):1262 to 1269. DOI: 10.1093/schbul/sbw003
- Di Forti M et al. Cannabis and Incidence of Psychotic Disorder EU-GEI. Lancet Psychiatry. 2019;6(5):427 to 436. DOI: 10.1016/S2215-0366(19)30048-3
- Meier MH et al. Persistent cannabis users show neuropsychological decline. PNAS. 2012;109(40):E2657 to E2664. DOI: 10.1073/pnas.1206820109
- Sorensen CJ et al. Cannabinoid Hyperemesis Syndrome: Diagnosis, Pathophysiology, and Treatment, a Systematic Review. J Med Toxicol. 2017;13(1):71 to 87. DOI: 10.1007/s13181-016-0595-z
- McCartney D et al. Determining the magnitude and duration of acute THC-induced driving and cognitive impairment: a systematic and meta-analytic review. Neurosci Biobehav Rev. 2021;126:175 to 193. DOI: 10.1016/j.neubiorev.2021.01.003
- Trauer JM et al. CBT for Chronic Insomnia Meta-analysis. Ann Intern Med. 2015;163(3):191 to 204. DOI: 10.7326/M14-2841
- Geneen LJ et al. Physical activity and exercise for chronic pain. Cochrane. 2017;4:CD011279. DOI: 10.1002/14651858.CD011279.pub3
- Hilton L et al. Mindfulness Meditation for Chronic Pain: Systematic Review and Meta-analysis. Ann Behav Med. 2017;51(2):199 to 213. DOI: 10.1007/s12160-016-9844-2
- Stubbs B et al. An examination of the anxiolytic effects of exercise for people with anxiety and stress-related disorders: A meta-analysis. Psychiatry Res. 2017;249:102 to 108. DOI: 10.1016/j.psychres.2016.12.020
- Vatvani AD et al. Efficacy and safety of low-dose naltrexone for the management of fibromyalgia: a systematic review and meta-analysis of randomized controlled trials. Korean J Pain. 2024;37(4):367 to 378. DOI: 10.3344/kjp.24202
- Siebers M et al. Endocannabinoids and the Runner's High. Neuroscientist. 2023;29(3):352 to 369. DOI: 10.1177/10738584211069981
- Brezing CA et al. Abstinence and reduced frequency of use are associated with improvements in quality of life among treatment-seekers with cannabis use disorder. Am J Addict. 2018;27(2):101 to 107. DOI: 10.1111/ajad.12660
- Xie L et al. Effects of omega-3 fatty acids on chronic pain: a systematic review and meta-analysis. Front Med (Lausanne). 2025;12:1654661. DOI: 10.3389/fmed.2025.1654661
- van der Zwan JE et al. Physical Activity, Mindfulness, HRV Biofeedback for Stress Reduction RCT. Appl Psychophysiol Biofeedback. 2015;40(4):257 to 268. DOI: 10.1007/s10484-015-9293-x
- Goessl VC et al. HRV Biofeedback on Stress and Anxiety Meta-analysis. Psychol Med. 2017;47(15):2578 to 2586. DOI: 10.1017/S0033291717001003
- Rehman Y et al. Cannabis in PTSD systematic review. AIMS Neuroscience. 2021;8(3):414 to 434. DOI: 10.3934/Neuroscience.2021022
- German Federal Ministry of Health. Medical Cannabis Act (MedCanG), in force since 1.4.2024. Planned amendment 2025/2026: mandatory in-person initial consultation before any cannabis prescription, ban on mail-order sales of cannabis flower.
- On the rise in prescriptions. The figures circulating on the increase in medical cannabis prescriptions and on patient numbers in Germany come from industry and market reports, not from official statistics. Since no reliable official source is available to me, I deliberately name no percentages in the text.