A Big Review Found Medical Cannabis May Barely Help Insomnia. Here Is What That Means

If you lean on weed to fall asleep, a Cochrane review out today has news you won't love. Medical cannabis may barely move insomnia, and the proof is very shaky.

I still use flower when I want the night to settle. My tolerance isn't a typical tolerance. There is not enough research yet on one person's endocannabinoid system, the body's own cannabis-like signaling. A trial average and a night that helped me can both be real.

What Did the Cochrane Review Find on Medical Cannabis for Insomnia? #

A Cochrane review of nine trials found little to no meaningful effect on sleep quality, time asleep, or daytime function. The trials covered 537 adults, and the proof behind those rows is very low. These were randomized trials, so chance, not a person, decided who got the product and who got the comparison.

The paper is Medical cannabis for the treatment of insomnia, out today in the Cochrane Database of Systematic Reviews. The record is DOI 10.1002/14651858.CD016216.pub2. A Cochrane review pools trials that clear a preset rule list. Lucia Varela at the University of Manitoba led it, and Luis Garegnani at Hospital Italiano in Buenos Aires is the senior author. Cochrane's news post today says those two centers ran the work.

They searched major medical databases through April 5, 2025. The plain-language page prints the authors' conclusion:

Medical cannabis may have little to no clinically meaningful effect on quality of sleep/satisfaction with sleep, total sleep time, or daytime functioning. It may lead to an improvement in quality of life, although the evidence is very uncertain. Medical cannabis may result in a slight increase in adverse events, but the certainty of this evidence is low.

Eight trials used a placebo, a dummy that looks like the real product. One used amitriptyline, an older antidepressant some doctors still use at night. Ages ran from 18 to 76. Women made up 72.66% of the people. Six studies were chronic insomnia, a long pattern of bad nights. Three were short-term. Most looked at insomnia on its own. One folded in insomnia tied to other illness.

Chronic, in Cochrane's own explainer, means trouble at least three nights a week for three months or more. Short-term means under three months.

What they measured	What the math showed	Trials and people	How sure
Sleep quality or satisfaction	Score -0.49. Range -1.06 to 0.08. Trials disagreed at 91%	7 trials, 374 people	Very low
Time asleep	7.68 minutes. Range -23.04 to 38.39 minutes. Disagreement 80%	6 trials, 194 people	Very low
Daytime function	Score -0.34. Range -0.83 to 0.16. Disagreement 65%	5 trials, 144 people	Very low
Quality of life	Score 0.84. Range -0.61 to 2.29. Disagreement 78%	3 trials, 95 people	Very low
Unwanted effects	Odds ratio 11.25. Range 4.46 to 28.35. Trials agreed	3 trials, 98 people	Low

Those figures are from the PubMed abstract of the review. The score is a resized average so different sleep quizzes can share one line. Zero means no difference. The range is where the true result might still sit. If it crosses zero, no change is still allowed. The disagreement percent shows how far the trials pulled apart. High means they do not tell one story. An odds ratio compares odds of an unwanted effect. A zero on the dummy side makes that ratio look huge.

The free abstract does not report minutes to fall asleep, waking in the night, or withdrawal. Those were on the plan, and they are not in the table above.

Varela told Cochrane's news desk: "This does not mean that medical cannabis is ineffective. It means that, based on the evidence we have, we cannot draw reliable conclusions about its effectiveness for insomnia in the general population."

This table does not reach into my body and erase a night that helped me.

What Does Little to No Meaningful Effect Mean? #

Little to no meaningful effect means the best estimate is a small gap, and the range still includes no real change. For time asleep, the average gap was 7.68 minutes. The range ran from 23 minutes less to 38 minutes more.

Eight minutes is about one press of the snooze button. The same range also allows a real loss and a much larger gain. A tiny help and no help both fit inside it.

Clinically meaningful means a change you would notice and care about. A blip on a chart can be real in the math and still too small to matter in bed. The authors used that bar for sleep quality, time asleep, and daytime function.

Sleep quality landed at -0.49, with a range from -1.06 to 0.08. The trials disagreed at 91%. The free pages do not show me whether a minus sign means better sleep or worse sleep on their quizzes. The authors' translation is little to no meaningful effect, so I am staying with their words.

Quality of life averaged 0.84, and the authors say it may improve. The range ran from a worsening (-0.61) to a large gain (2.29). Cochrane's news post says that spread fits anything from a slight worsening to a large improvement. The evidence is very uncertain.

This review is not "the biggest" just because it is Cochrane. A 2025 paper in Sleep Medicine Reviews pooled six trials and 1,077 adults who did or did not have insomnia. Subjective sleep quality rose (score 0.53, range 0.03 to 1.02) with heavy disagreement (88%). CBD-only arms in that paper showed no clear gain. Cochrane pooled more trials but stays on insomnia and calls its own proof very low. The other paper pooled more people but mixed in adults without an insomnia diagnosis.

• Time asleep: about eight minutes, range crosses zero.
• Sleep quality: trials disagree hard, range crosses zero.
• Life quality: average looks better, range still includes a loss.
• None of those sleep rows are high-certainty proof.

What Does Very Low Certainty Mean? #

Very low certainty means the true effect could sit far from the number in the table. Low certainty is one step firmer, and it still leaves a wide door. Cochrane prints these grades so a shaky average is not mistaken for a settled fact.

The abstract names three reasons the sleep rows were marked down:

• Bias: the way a trial was run could tilt the result.
• Inconsistency: the trials disagreed. That is the high percent on the sleep rows.
• Imprecision: too few people, so the estimate is loose and the range stays wide.

The news post calls the whole body of evidence very low to low. Side effects are the low row. Sleep quality, time asleep, daytime function, and life quality are very low.

Very low certainty leaves the question open for a single person. Varela said the team cannot draw a reliable conclusion for the general population. Garegnani, on the same page, said clinicians and patients should recognize the uncertainty and consider individual circumstances while better evidence is generated. He said not to read the paper as either an endorsement or a rejection.

Most studies had fewer than 50 people. The longest followed people for 12 weeks. Both points are on Cochrane's public pages. Twelve weeks cannot speak for a year of nightly use. The group also skewed younger and female, at 72.66% women. A man in his sixties with a bad back is barely in that pool.

Which Cannabis Products Were in the Trials? #

The free Cochrane pages say the trials used THC, CBD, or both, given as oils, capsules, or gel. THC, tetrahydrocannabinol, is the main compound that intoxicates. CBD, cannabidiol, is a cannabis compound that does not bring that same high. The plain-language page is the source for that product line.

The abstract adds a split on how the products were supplied:

Kind of product	How many studies	What you are looking at
Licensed plant-based cannabinoid	1	A cannabis compound with a formal drug approval
Licensed synthetic cannabinoid	1	A lab-made cousin with a formal drug approval
Magistral preparation	7	A pharmacy mix made for one person

Nabiximols is a mouth spray with THC and CBD, sold as Sativex in some countries. It does not appear by name on these free pages, so I will not assign it to the nine trials. The full paper sits behind the Cochrane Library. Until that table is public, I am not tying the spray to any trial.

CBN, cannabinol, is a breakdown product of THC. Shops market it for sleep. The plain-language product line does not list CBN. The search closed April 5, 2025, so a later CBN paper is outside this review. Lavender and colleagues tested one night in 20 adults. They compared 30 mg of CBN, 300 mg of CBN, and a dummy. Time awake after falling asleep did not move in a clear way. At 300 mg they saw more stage-2 sleep (a lighter sleep stage) and a better self-rating. Time to fall asleep got shorter, and brief arousals (short wake-ups) dropped. They logged 247 mild or moderate unwanted effects across the arms. One night in 20 people is a pilot. The shopper comparison of gummies and flower is in CBN gummies vs indica flower.

A May 2026 nurse-practitioner review was kinder to CBD and CBN and cooler on THC. The abstract skips the trial count. It also skips a certainty grade. When two 2026 papers split, I stay with the one that prints its doubt.

The news post says clinical medical cannabis usually comes as tablets, sprays, or oils. The "what we found" line for these nine trials says oils, capsules, or gel. I am not merging those into one shopping list. A jar of Michigan flower is a third thing entirely, and it was not the assigned product here.

Did Side Effects Show Up Clearer Than the Sleep Change? #

Unwanted effects were the clearest signal, and they still come from three trials and 98 people. The odds ratio was 11.25, with a range from 4.46 to 28.35. An odds ratio compares odds, not headcounts. Nobody had an unwanted effect on placebo alone, so the ratio jumps.

When the unwanted effect showed up	People
Only during the cannabis period	37
Only during the placebo period	0
During both periods	14
During neither period	47

That count is in the abstract and on the plain-language page. The authors call it a slight increase, at low certainty. Cochrane's news post says people were substantially more likely to have mild effects, such as dry mouth and dizziness. Same review, two wordings. The headcount is the part you can check.

The plain-language background lists dizziness, dry mouth, memory trouble, and hunger. It says those can show up in up to 80% of users. That 80% line is general background on the same page. The nine-trial pool is the 37-person row.

Low certainty on harms means the size of the risk could still move. This row is steadier than the sleep rows because the three trials agreed. Agreement inside a small pool is not a precise risk for every shopper in Detroit.

Dry mouth and dizziness matter if you drive home after a midnight shift. They also matter if you already take other night medicine. The review does not give a drug-by-drug interaction table. That conversation belongs with your prescriber.

Does a Shaky Review Mean Cannabis Cannot Help Your Night? #

The review cannot tell one person that a helpful night was imaginary. The trials assigned a medical product to adults who met insomnia rules. Then they averaged the groups. Your jar, your hour, and your other medicines were not in that average.

I have used flower for years, and my tolerance isn't a typical tolerance. There is not enough research yet on how one person's endocannabinoid system, the body's own cannabis-like signaling, answers a product. A night that helped me is still a night that helped me.

I keep one limit. If cannabis is standing in for care of depression, breathing pauses in sleep, or a new illness, see a doctor. The review is not a pass to skip that visit. Snoring, gasping, and morning headaches belong in a clinic. A metro Detroit night shift can make them easy to blame on being tired.

What I take from the review:

• A group result can be tiny while one person still feels a difference.
• A calm morning can still hide a choppy measured night. That split is the October 9 post.
• A weak average does not predict one person's night.

Garegnani said to weigh the individual case while better evidence is built. When someone asks me if the plant treats insomnia, I say the trials did not show a solid treatment effect. I also say I still use flower in my own evening.

What Do Sleep Doctors Use Before Cannabis? #

Major sleep groups tell clinicians to start chronic insomnia with CBT-I. CBT-I is cognitive behavioral therapy for insomnia, a short course on habits and the thoughts that keep you up. The work is practice. A sleep medication, if it comes, comes after.

The American College of Physicians made this a strong recommendation in 2016, on moderate-quality evidence. They say all adults with chronic insomnia should get CBT-I as the first treatment. If that course alone fails, they say to talk through a short run of medication. That second step is a weak recommendation, on low-quality evidence.

The American Academy of Sleep Medicine said the same thing in 2021, and marked multicomponent CBT-I as strong. Multicomponent means several pieces in one course.

A typical course includes:

• A fixed wake time, even after a bad night.
• Less time in bed, so the bed pairs with sleep.
• Out of bed when you are awake and frustrated.
• Less clock-watching in the dark.

A 2026 AASM guideline looked at starting CBT-I and a sleep medication at the same time. It suggests that pair over medication alone. It suggests against the pair when the other choice is CBT-I alone. Both calls are conditional, on low certainty. People who want more time asleep early, and who care less about daytime symptoms, may still pick the pair. Cannabis is not the medication named in that abstract.

A fixed wake time is a hard rule on a Detroit rotating hospital shift. It is also hard when a metro Detroit auto plant starts in the dark. CBT-I has to bend around that clock, so find a Michigan clinician who knows shift work. A mail-order jar does not replace the course. The long map of cannabinoids and terpenes, if you still want it, is the cannabis and sleep guide.

What Happens to Sleep If Nightly Cannabis Stops? #

Sleep is often broken in the stretch after regular cannabis use stops. Peter Gates and colleagues reviewed 36 human studies and found sleep often interrupted in withdrawal. Withdrawal means the rough patch after a regular habit stops. The authors could not name which part of sleep breaks. Most of those studies were too messy for a final map.

That pattern is why a sudden stop can feel like proof you need the plant to sleep. The rough nights can be withdrawal. They can also be the insomnia that was there first. Those 36 studies do not sort that for you.

If the same flower has gotten weaker over months, the reset and the timeline live in the tolerance-break piece. I will not redo that how-to here. Tell a prescriber before you stop cold if you take other night medicine. A bad week of sleep is a work problem when you drive at dawn or run a crew.

What I watch in my own weeks:

• The first nights off can feel louder.
• A weaker effect after daily use is tolerance. The flower in the jar can be unchanged.
• Going back only to escape the rough patch can lock the loop in.

There is not enough research yet to say how long that rough patch lasts in one body. The Gates review gives no single timetable, and I will not make one up.

What Does This Mean If You Shop in Michigan? #

Insomnia is absent from Michigan's medical-card list. A Detroit flower buy is still a flower buy. This review does not turn it into a sleep prescription. The Cannabis Regulatory Agency list includes cancer, PTSD, chronic pain, severe nausea, seizures, and severe ongoing muscle spasms. A card for chronic pain is a pain card, not a sleep result.

Adults 21 and older can buy at a licensed Michigan shop with no card. The CRA consumer page is the state's frame for that adult-use lane. Cochrane lists oils, capsules, and gel for these trials, and the supply split was two licensed cannabinoid medicines and seven pharmacy mixes. Cured flower in a Michigan jar is a different object.

The trials skewed younger, and women were 72.66% of the pool. Cochrane's news post says the findings may fit men, older adults, and people with other illnesses less well. A Detroit night-shift worker should ask a clinician how those trials fit.

Question a Michigan shopper actually has	What this review can say
Will a medical card for insomnia clear?	Insomnia is not on the CRA list
Is shop flower the product they tested?	The public record says oils, capsules, or gel, plus licensed cannabinoid medicines
Does a calm night prove the trials wrong?	No single night overturns a group average, and the average is very uncertain
What should I ask a clinician?	Ask about CBT-I first, then about other night medicines and about stopping

Divine Toke is a Michigan mail-order shop for sun-grown flower. There is no storefront and no public counter. We do not sell oils, capsules, gels, or CBD products. CBT-I and a doctor visit come before any jar.

How Is This Different From the Middle-Age Sleep Study? #

The October 9 post follows years of use and a wrist monitor. This page is about assigned medical products in insomnia trials. Read felt rest versus measured sleep for the middle-age cohort. I am not restating that paper's odds here.

A calm morning and a measured night can come apart. That is the cohort's story, built on people watched across years. Cochrane assigned a product and graded the trials as very uncertain.

Use each page for its own question:

• Years of use, and a watch that disagrees with the morning: the October 9 post.
• "Does a medical cannabis product treat insomnia in trials?": this page.
• Gummies versus flower, and what CBN claims outrun the trials: CBN gummies vs indica flower.
• Why the same amount feels weaker after daily use: cannabis tolerance breaks.

The wider sleep map stays in the cannabis and sleep guide. That is the long version. This page is one review, dated today.

Frequently Asked Questions #

Nine trials leave medical cannabis as a maybe, with very shaky proof.

Does medical cannabis treat insomnia? #

The Cochrane review says it may have little to no meaningful effect, and the proof is very uncertain. Nine trials and 537 adults, ages 18 to 76, compared medical cannabis mostly with a placebo. Eight trials used a dummy. One used amitriptyline. The authors' conclusion is on the plain-language page. Time asleep differed by 7.68 minutes, and the range crossed zero.

How many trials were in the Cochrane insomnia review? #

Nine studies, 537 people. Women were 72.66% of that pool. Six studies looked at chronic insomnia. Three looked at short-term insomnia. Most studies had fewer than 50 people, per Cochrane's summary. The search for trials stopped on April 5, 2025. The review itself came out October 13, 2026.

Did the review test CBN or nabiximols? #

The free pages name THC, CBD, or both. They do not name CBN or nabiximols. Products were oils, capsules, or gel. Two studies used licensed cannabinoid medicines, one from the plant and one synthetic. Seven were pharmacy mixes. A later one-night CBN trial in 20 adults, PMID 41698831, falls after the April 2025 search close.

What does very low certainty mean? #

The true result could sit far from the published number. Cochrane marked sleep quality, time asleep, daytime function, and life quality as very low. Side effects are low certainty, which is a bit firmer and still limited. The abstract blames bias, disagreement between trials, and small samples. Very low means the proof is weak. One person's night can still land somewhere else.

Does little effect mean it fails for every person? #

No. A group average can miss one person's night. Lead author Lucia Varela said the paper does not mean medical cannabis is ineffective. It means the team cannot draw a reliable conclusion for the general population. That quote is in Cochrane's news post.

What is CBT-I? #

CBT-I is a short course of habit and thought changes, and it is the first treatment sleep groups recommend. The American College of Physicians gave it a strong recommendation in 2016. The American Academy of Sleep Medicine marked multicomponent CBT-I strong in 2021. A 2026 AASM guideline suggests against adding a sleep medication when CBT-I alone is the other choice.

Can stopping cannabis make sleep worse? #

Yes. Sleep is often interrupted after regular use stops. A review of 36 human studies found that pattern and refused a single map of which sleep stage breaks. The rough nights can be withdrawal, the insomnia you already had, or both. How to take a break without guessing a timetable is in the tolerance-break piece.

Is insomnia a Michigan medical card condition? #

No. Insomnia is not on the state list. The CRA qualifying-conditions page names cancer, PTSD, chronic pain, severe nausea, seizures, and severe muscle spasms, among others. Adults 21 and older can still buy adult-use flower in Michigan without a card. A pain card is not a sleep-trial result.

Does this review cover flower from a shop? #

The public record lists oils, capsules, and gel, with two licensed cannabinoid medicines in the mix. It does not describe cured flower from a Detroit counter. Seven products were pharmacy mixes made for one person. Divine Toke ships sun-grown flower by mail in Michigan and does not sell those trial oils. Shop flower and a Cochrane capsule are different objects.

How long did the trials follow people? #

Most ran a few weeks. The longest ran 12 weeks. Cochrane's news post is the source for the 12-week cap. That window cannot answer what a year of nightly use does. It also cannot answer a Michigan winter of 5 a.m. starts.

Should I change what I use tonight? #

Do not rip up a nightly habit tonight because a review hit the web today. Stopping can interrupt sleep, as the withdrawal review above describes. If you want to change, talk with a clinician, especially if you take other night medicine or you drive before dawn. Ask about CBT-I. The review's own senior author said to weigh the individual case.

How is this different from the October 9 sleep study? #

October 9 is years of use and a wrist monitor. Today is assigned medical cannabis in insomnia trials. The cohort page is felt rest versus measured sleep. This page does not repeat that paper's numbers. Use October 9 for "I have used this for years." Use today for "did the trials show a real treatment effect?"

Where This Leaves a Night You Already Know #

I am keeping flower for my own nights. I will not call the plant an insomnia treatment.

If bad nights hit three times a week for months, start with CBT-I. That is the first step named by the American College of Physicians. Tell the clinician your real schedule if you work nights at a metro Detroit hospital or start in a dark Michigan morning. A 10 p.m. bedtime plan will not fit you.

If nightly flower has gotten weaker, read how a tolerance break works before you stop cold. Gummy claims versus flower sit in CBN gummies vs indica flower. A calm morning that a tracker disputes is the October 9 sleep study. The long version remains the cannabis and sleep guide.

Divine Toke ships sun-grown Michigan flower by mail and has no storefront. A treatment plan starts with a clinician.

This article is for educational purposes only and is not medical advice. Always consult your healthcare provider before starting any new wellness routine.