Sleep Disorders in Children With Cerebral Palsy: What a New Systematic Review Found About Melatonin, Cannabis-Based Medicines, and Other Options
| Audience | Pediatric neurologists, developmental pediatricians, physiatrists, primary care clinicians caring for children with cerebral palsy, and parents or caregivers weighing sleep treatment options |
| Primary Topic | A systematic review of 12 randomized controlled trials evaluating pharmacological and non-pharmacological interventions for sleep disorders in children with cerebral palsy |
| Source | Read the study via DOI |
Sleep Disorders in Children With Cerebral Palsy: What a New Systematic Review Found About Melatonin, Cannabis-Based Medicines, and Other Options
A systematic review in Developmental Medicine & Child Neurology included 12 randomized controlled trials examining nine pharmacological and non-pharmacological interventions for sleep disorders in children with cerebral palsy. Melatonin produced the clearest positive signal, with modest improvements in sleep latency and duration compared with placebo. Evidence for cannabis-based medicines, baclofen, neural stem cell therapy, and the non-drug approaches studied was inconsistent or negligible, while the overall evidence base remained limited and low certainty.
| Study Type | Systematic review of randomized controlled trials |
| Journal | Developmental Medicine & Child Neurology |
| Published Online | August 21, 2026 |
| Population | Children and adolescents with cerebral palsy, ages 5 to 17 years |
| Trials Included | 12 randomized controlled trials |
| Trial Size Range | 19 to 142 participants per trial |
| Searches | Major bibliographic databases and clinical-trial registries searched through May 2026 |
| Pharmacological Interventions | Melatonin, cannabis-based medicines, baclofen, and neural stem cell therapy |
| Non-Pharmacological Interventions | Massage, cranial osteopathy, music therapy, acupuncture, and postural support |
| Total Intervention Categories | Nine |
| Key Finding | Melatonin produced the clearest positive signal, modestly improving sleep latency and duration compared with placebo; findings for the other interventions were inconsistent or negligible |
| Safety | Reported adverse events in the included studies were generally mild, although the small and heterogeneous evidence base limits strong comparative safety conclusions |
| Economic Evaluations | No completed economic evaluations were identified |
| Authors | Nishant Jaiswal, Lizzie Fisher, Giorgio Ciminata, Ryan Mulholland, Anna Noel-Storr, Lesley Nutton, Valerie Orr, Paul Eunson, Nicola J Cooper, Alex Sutton, Olivia Wu, and Terry Quinn |
| DOI | 10.1111/dmcn.70435 |
The reviewers searched major bibliographic databases and clinical-trial registries through May 2026 for randomized controlled trials and economic evaluations of interventions intended to improve sleep in children with cerebral palsy. Twelve randomized controlled trials met the review’s inclusion criteria, with individual studies enrolling between 19 and 142 participants and including children and adolescents ages 5 to 17.
The review also incorporated input from clinicians and people with lived experience of cerebral palsy when considering outcomes and interpretation. That is especially relevant in cerebral palsy, where disrupted sleep may arise from several overlapping contributors rather than from a single isolated sleep disorder.
The pharmacological interventions represented in the review included melatonin, cannabis-based medicines, baclofen, and neural stem cell therapy. Among them, melatonin produced the clearest positive signal, with modest improvements in sleep latency and sleep duration compared with placebo.
Cannabis-based medicines were also represented in the review, but the evidence summarized did not demonstrate a consistent benefit for sleep. That is a narrower claim than saying cannabis-based medicines are ineffective. The clinical meaning depends heavily on which cannabinoid formulation, dose, route, population, and outcome were studied, as well as how much evidence was available for that particular intervention.
The review also identified trials involving massage, cranial osteopathy, music therapy, acupuncture, and postural support. These approaches did not produce a similarly clear or consistent positive signal across the available evidence.
This does not establish that each intervention is ineffective. It means that the randomized evidence currently available is too inconsistent or limited to support confident claims of sleep benefit comparable with the signal seen for melatonin.
Reported adverse events in the included studies were generally mild, with acceptable tolerability reported across the available evidence. That is useful, but it should not be overread. Twelve small trials distributed across several very different interventions cannot establish that all of those approaches have equivalent or well-characterized safety profiles.
Safety and efficacy are also separate questions. An intervention may be reasonably tolerated in a small trial while still failing to produce a measurable improvement in the outcome being studied. No completed economic evaluations were identified, leaving cost-effectiveness unanswered.
The review used formal risk-of-bias and certainty-of-evidence methods, including the Cochrane Risk of Bias 2 framework and GRADE. The resulting evidence base was limited and heterogeneous, with low certainty surrounding many intervention-specific conclusions.
Low-certainty evidence is not meaningless evidence. It tells clinicians where signals exist, where they do not yet reproduce reliably, and where another well-designed trial could materially change the conclusion.
Sleep disturbances in children with cerebral palsy are often intertwined with pain, spasticity, seizures, reflux, positioning difficulties, medication effects, respiratory problems, and behavioral or circadian factors. A therapy that does not demonstrate a primary sleep effect in a randomized trial may therefore still interact with sleep indirectly when it successfully treats another symptom.
This review also sits within a broader pediatric cannabinoid literature in which the strongest randomized evidence has historically concerned treatment-resistant epilepsy rather than sleep as a primary outcome. That difference in indication matters. Evidence should be interpreted according to what was actually tested, not according to the reputation of the intervention more generally.
I think the most useful way to read this review is to resist making it say more than it does. Melatonin produced the clearest positive signal for sleep in children with cerebral palsy. Cannabis-based medicines did not demonstrate a consistent sleep benefit in the evidence identified here. That is useful information, including for clinicians who are otherwise comfortable using cannabinoid medicines.
It is also not the same as concluding that cannabis-based medicines have no role for children with cerebral palsy. A child may have seizures, pain, spasticity, anxiety, or other symptoms entangled with sleep, and those questions need to be considered separately. For sleep itself, however, I would not present cannabis-based medicines to a family as though the trial evidence is comparable with the melatonin signal identified in this review. What I want next is more granular evidence: which cannabinoids, at what doses and ratios, by which route, in which children, for which sleep outcomes.
How to Read a Systematic Review Covering Many Different Interventions
Twelve trials spread across nine intervention categories can be tempting to compress into a simple headline. That would be a mistake. A systematic review can identify where evidence is strongest without establishing that every intervention with uncertain evidence has failed.
Four checks help keep the interpretation proportional to the data.
A Four-Step Reading Frame
Separate the clearest positive signal from the rest
Melatonin modestly improved sleep latency and duration compared with placebo. The evidence for the other interventions was less consistent. That tells us where the clearest signal currently lies, not that every alternative has been conclusively disproven.
Ask what was actually tested
“Cannabis-based medicine” is not a single pharmacological exposure. Cannabinoid composition, THC-to-CBD ratio, dose, route, timing, duration, and indication can fundamentally change the meaning of a trial result.
Weigh the size of the evidence base
Twelve trials, with individual sample sizes ranging from 19 to 142 participants, are spread across several distinct interventions in a heterogeneous population. Intervention-specific certainty can therefore be much thinner than the overall review title suggests.
Hold onto the certainty rating
Low-certainty evidence means future studies may materially change the estimated effect. That is a reason for careful interpretation, not for ignoring the evidence that already exists.
The Same Study Can Mean Different Things Depending on the Question Being Asked
Scientific papers rarely answer a single question. Patients, clinicians, researchers, policymakers, and critics often read the same evidence from different angles. These perspectives help keep a small, multi-intervention systematic review from being oversimplified in either direction.
Melatonin Has the Clearest Positive Signal Right Now
If your child has cerebral palsy and struggles with sleep, this review’s clearest finding is that melatonin modestly improved how quickly children fell asleep and how long they slept compared with placebo.
Cannabis-based medicines were represented in the evidence base too, but the available studies did not establish a consistent sleep benefit. That does not mean an individual child could never benefit. It means the trial evidence is not currently strong enough to promise that result.
A Useful Evidence Map, Not a Head-to-Head Trial
For pediatric neurologists, the review is valuable because it applies a systematic evidence framework across several pharmacological and non-pharmacological strategies used or considered in this population.
What it does not provide is a randomized head-to-head comparison of melatonin against cannabis-based medicines or the other interventions. Apparent differences in evidence strength may reflect differences in the quantity and quality of studies as much as differences in pharmacological efficacy.
Uncertain Evidence Is Not the Same as Proven Failure
When several different interventions are summarized as having inconsistent or negligible effects, readers should ask how many trials informed each intervention, how large they were, what outcomes they measured, and how similar those trials actually were.
That is especially important for a category such as cannabis-based medicines, where pharmacological exposures can differ substantially according to cannabinoid composition, dose, route, and treatment target.
A Better Starting Point for Family Conversations
Primary care clinicians can use this review to distinguish between what has a clearer trial signal and what remains uncertain. Melatonin has the clearer evidence for sleep itself. Cannabis-based medicines should not be presented as though a comparable sleep benefit has been established.
At the same time, sleep disturbance in cerebral palsy is often secondary to pain, seizures, spasticity, reflux, positioning, respiratory disease, or other contributors. Treating those contributors may be as important as choosing a “sleep treatment.”
Small Trials, Many Interventions, Limited Certainty
Twelve randomized trials, ranging from 19 to 142 participants, are distributed across nine intervention categories. That is a relatively thin evidence base from which to derive strong intervention-specific conclusions.
A rigorous systematic review cannot repair limitations in the studies it inherits. Risk-of-bias assessment and GRADE make those limitations visible, but they do not make sparse evidence more definitive.
An Unflattering Result That Should Be Taken Seriously, Not Overinterpreted
For clinicians who work extensively with cannabis-based medicines, the useful finding is straightforward: this review does not establish a consistent sleep benefit for cannabis-based medicines in children with cerebral palsy.
That result deserves to be taken seriously. It also needs pharmacological context. A cannabinoid intervention is defined by formulation, THC and CBD content, minor cannabinoids, dose, route, timing, duration, indication, and patient characteristics. More granular intervention-specific evidence is needed before a broad category can yield precise clinical guidance.
Sleep Is Often a Symptom of Something Else
Families living with cerebral palsy already know that sleep cannot always be separated neatly from pain, positioning, muscle tone, seizures, gastrointestinal symptoms, respiratory problems, medication effects, anxiety, and caregiver routines.
The involvement of people with lived experience in the review is therefore particularly valuable. An intervention that looks unimpressive when sleep is treated as an isolated outcome may raise a different clinical question when the child’s actual sleep disruption is driven by another treatable symptom.
A Direct Case for Larger, Better-Reported Trials
The review highlights a familiar problem in pediatric neurodisability research: clinically important questions are often being answered by small trials spread across heterogeneous interventions and outcome measures.
Future research needs adequately powered studies, clearer intervention reporting, standardized sleep outcomes, meaningful safety follow-up, attention to equity and heterogeneity, and economic evaluation. For cannabis-based medicines specifically, cannabinoid composition, dose, route, timing, and coexisting indications should be explicit.
Join the Conversation
Have a question about how this applies to your situation? Ask Dr. Caplan
Want to discuss this topic with other patients and caregivers? Join the forum discussion
Frequently Asked Questions
What did this systematic review study?
The review included 12 randomized controlled trials examining nine intervention categories for sleep disorders in children with cerebral palsy: four pharmacological approaches and five non-pharmacological approaches.
Which intervention showed the clearest benefit?
Melatonin produced the clearest positive signal, with modest improvements in sleep latency and sleep duration compared with placebo.
Did cannabis-based medicines help with sleep?
The available evidence identified in this review did not establish a consistent sleep benefit from cannabis-based medicines. That is not the same as proving that no child can benefit, nor does it address cannabinoid use for separate indications such as epilepsy, pain, or spasticity.
Did the review directly compare cannabis-based medicines with melatonin?
No. The interventions were evaluated within the same systematic review, but that should not be confused with a head-to-head randomized trial of melatonin versus cannabis-based medicines.
Why does the specific cannabis formulation matter?
Cannabis-based medicine is a broad category. THC and CBD content, cannabinoid ratio, dose, route of administration, timing, treatment duration, and the symptoms being targeted can all affect clinical outcomes. A broad category-level result therefore requires careful interpretation.
Were the treatments studied safe?
Reported adverse events in the included studies were generally mild and tolerability was acceptable. However, the trials were small and covered very different interventions, so the review should not be interpreted as establishing equivalent or comprehensive safety across all of them.
How strong is the overall evidence?
The evidence base was limited and heterogeneous, and certainty was low for many intervention-specific conclusions. Twelve relatively small trials were distributed across several different treatments and clinical circumstances.
Where was this review published?
The review was published online in Developmental Medicine & Child Neurology on August 21, 2026. Its DOI is 10.1111/dmcn.70435.
Does this mean cannabis-based medicines do not work for children with cerebral palsy?
No. The review addresses sleep outcomes. It does not establish that cannabis-based medicines lack benefit for other symptoms or indications that may occur in children with cerebral palsy. Those questions require their own evidence.
Were cost or economic factors studied?
No completed economic evaluations were identified, leaving cost-effectiveness unanswered.
What should future studies do differently?
Future trials need larger samples, clearer intervention descriptions, standardized sleep outcomes, stronger safety reporting, and better characterization of clinical heterogeneity. Cannabis-based medicine studies should report cannabinoid composition, dose, route, timing, duration, and coexisting treatment indications explicitly.
