| Audience | Parents and caregivers of children with developmental and epileptic encephalopathies, adults with treatment-resistant epilepsy, pediatric epileptologists, neurologists, and cannabis-science readers |
| Primary Topic | CBD nonseizure outcomes in developmental and epileptic encephalopathies |
| Source | Read the full source |
CBD and Nonseizure Outcomes in Epilepsy: What a New Systematic Review Shows
A PRISMA-guided systematic review published July 14, 2026 in Epilepsy Research pooled 32 studies of highly purified CBD in patients with developmental and epileptic encephalopathies and complex treatment-resistant epilepsies beyond Lennox-Gastaut, Dravet, and tuberous sclerosis complex. Thirty-one of 32 studies reported improvement in at least one nonseizure outcome, most consistently in neuropsychiatric function and cognition.
| Study Type | PRISMA-guided systematic literature review; search conducted March 2024 |
| Population | 1,343 patients across 32 studies with developmental and epileptic encephalopathies (DEEs) and complex treatment-resistant epilepsies (TREs), excluding Lennox-Gastaut syndrome, Dravet syndrome, and tuberous sclerosis complex |
| Intervention | Highly purified plant-derived CBD oral solution (Epidiolex in the US / Epidyolex in the EU) |
| Comparator | No uniform comparator; narrative synthesis of mostly observational and open-label studies |
| Databases Searched | Embase, Medline, and Cochrane Central Register of Controlled Trials |
| Domains Assessed | Neuropsychiatric function, cognitive function, concomitant antiseizure medication use, communication, behavior, motor function, healthcare utilization, quality of life, global change, sleep, and development |
| Improvement Reported | 31 of 32 studies reported improvement in at least 1 nonseizure outcome; neuropsychiatric function (9/9 studies) and cognitive function (9/9 studies) showed the most consistent reporting |
| Most Common Adverse Events | Diarrhea (17-34%), vomiting (5-50%), and decreased appetite (7-20%), consistent with CBD’s known safety profile |
| Journal | Epilepsy Research |
| Published | July 14, 2026 |
| Authors | Debopam Samanta, Lisa Moore-Ramdin, Marco Navetta, and Antonietta Coppola |
| PMID / DOI | 42520581 / 10.1016/j.eplepsyres.2026.107873 |
| Major Limitation | Mostly observational, open-label, or retrospective studies with moderate to high risk of bias, and no uniform comparator or blinding |
The authors searched Embase, Medline, and Cochrane Central in March 2024 for studies reporting nonseizure outcomes after CBD initiation in patients with developmental and epileptic encephalopathies and complex treatment-resistant epilepsies other than Lennox-Gastaut syndrome, Dravet syndrome, and tuberous sclerosis complex, the three syndromes CBD is currently FDA-approved to treat.
Thirty-two studies covering 1,343 patients met inclusion criteria. Results were narratively summarized rather than statistically pooled, because study designs and outcome measures varied widely.
Every one of the 9 studies that assessed neuropsychiatric function and every one of the 9 studies that assessed cognitive function reported improvement in at least some patients. Communication (7 of 7 studies) and behavior (7 of 7 studies) showed the same pattern.
Motor function (6 of 6), healthcare utilization (5 of 5), quality of life (5 of 5), global change (4 of 4), sleep (2 of 2), and development (2 of 2) also showed improvement in every study that measured them, and 7 of 8 studies reported reduced use of concomitant antiseizure medications.
Reported adverse events were consistent with CBD’s established safety profile and were predominantly gastrointestinal: diarrhea in 17 to 34 percent of patients, vomiting in 5 to 50 percent, and decreased appetite in 7 to 20 percent, depending on the study.
Wide ranges across studies reflect differences in dosing, population, concomitant medications, and how adverse events were tracked and reported.
The authors themselves describe the included evidence as mainly observational studies with moderate to high risk of bias. Most were open-label, uncontrolled, or retrospective, without blinding or a placebo comparator.
When every included study reports improvement in a domain, that consistency can reflect a genuine effect, but it can also reflect publication bias, expectation effects, or the tendency for open-label studies to report favorably.
CBD’s FDA approval is specific to seizures in Lennox-Gastaut syndrome, Dravet syndrome, and tuberous sclerosis complex. This review looked at a broader and more heterogeneous population of developmental and epileptic encephalopathies and treatment-resistant epilepsies.
The nonseizure findings do not extend the approved indication and should not be read as evidence that CBD is proven effective for cognition or behavior in these other syndromes; they identify a research signal worth confirming in controlled trials.
CBD’s regulatory approval rests on randomized controlled trials that measured seizure frequency in three specific syndromes. This review shifts the lens to nonseizure outcomes and to a broader population of developmental and epileptic encephalopathies and treatment-resistant epilepsies.
The next useful step is prospective, controlled research with standardized, validated cognitive and behavioral measures, a defined comparator, and prespecified nonseizure endpoints, so the consistent observational signal identified here can be tested rigorously.
Families living with developmental and epileptic encephalopathies frequently tell me that alertness, communication, and mood matter as much as seizure counts. This review takes that priority seriously and gathers a genuinely broad set of nonseizure signals, which is valuable groundwork.
I would not present this as proof that CBD improves cognition or behavior. It is a well-organized summary of a mostly observational literature with real bias risk. It supports a careful, individualized conversation and continued monitoring, not a general claim of nonseizure benefit.
How to Interpret the Nonseizure Outcomes Review
A systematic search process improves confidence that the literature was gathered comprehensively.
A narrative summary of observational studies is still a lower tier of evidence than a randomized controlled trial.
A Four-Step Reading Frame
Note the study design
This is a systematic review of mostly observational and open-label studies, not a randomized controlled trial or meta-analysis with pooled effect sizes.
Respect the consistency, cautiously
Improvement was reported in nearly every study and domain, which is notable, but consistent reporting across open-label studies can also reflect bias rather than a uniform true effect.
Separate seizure control from nonseizure effects
Some reported cognitive or behavioral improvement may follow indirectly from fewer seizures or reduced antiseizure medication burden, not necessarily a direct effect of CBD.
Watch for the adverse event pattern
Gastrointestinal effects, including diarrhea, vomiting, and decreased appetite, were common and should be part of any monitoring plan.
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 data differently. The perspectives below explore how this study looks through several evidence-based lenses.
A Broad Signal, Not a Guarantee
Most studies in this review reported some improvement in alertness, communication, behavior, or another nonseizure domain.
Track specific, measurable goals with your care team rather than expecting improvement across every domain at once.
Useful Hypothesis-Generating Synthesis
The review organizes a scattered observational literature into a clear picture of which nonseizure domains have been studied and how often improvement was reported.
It should inform counseling and monitoring priorities, not stand in for controlled trial evidence when discussing expected benefit.
Near-Universal Improvement Is a Red Flag as Much as a Signal
When 31 of 32 studies report improvement in an open-label literature, publication bias and selective reporting are plausible contributors.
The absence of pooled effect sizes or confidence intervals limits how much weight any single domain’s finding should carry.
Narrative Synthesis Cannot Replace Meta-Analysis
Without a common comparator or standardized outcome measures, the review cannot quantify how large any nonseizure effect might be.
Study quality across the 32 included papers was not uniform, and the review’s own assessment flagged moderate to high bias risk.
Extends Beyond the Three Approved Syndromes
Randomized trial evidence for CBD has focused on seizure counts in Lennox-Gastaut syndrome, Dravet syndrome, and tuberous sclerosis complex.
This review gathers evidence from a wider population of developmental and epileptic encephalopathies and treatment-resistant epilepsies, and from a wider set of outcomes than seizures alone.
Watch for Gastrointestinal Effects
Diarrhea, vomiting, and decreased appetite were the most commonly reported adverse events, consistent with CBD’s known tolerability profile.
Wide reported ranges suggest these effects vary meaningfully by dose, population, and concomitant medications, so individualized monitoring matters.
Controlled Trials With Standardized Measures Are the Next Step
Prospective, controlled studies with validated cognitive and behavioral instruments, a defined comparator, and prespecified nonseizure endpoints would let researchers test this signal directly.
Disentangling a direct CBD effect from improvement driven by fewer seizures or lower antiseizure medication burden is an important open question.
Evidence Tier Should Match Communication
A systematic review of observational studies is a meaningful contribution, but it should be communicated at the evidence tier it represents.
Regulatory and clinical guidance should continue to distinguish the approved seizure indication from unproven nonseizure claims.
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?
It reviewed 32 studies of 1,343 patients with developmental and epileptic encephalopathies and complex treatment-resistant epilepsies, looking at nonseizure outcomes after starting highly purified CBD.
Which patients were included?
Patients with developmental and epileptic encephalopathies and complex treatment-resistant epilepsies other than Lennox-Gastaut syndrome, Dravet syndrome, and tuberous sclerosis complex, the three syndromes CBD is FDA-approved to treat.
What nonseizure outcomes were assessed?
Neuropsychiatric function, cognitive function, concomitant antiseizure medication use, communication, behavior, motor function, healthcare utilization, quality of life, global change, sleep, and development.
How many studies reported improvement?
31 of 32 included studies reported improvement in at least one nonseizure outcome in at least some patients.
What were the most common side effects?
Gastrointestinal effects were most common: diarrhea in 17 to 34 percent of patients, vomiting in 5 to 50 percent, and decreased appetite in 7 to 20 percent.
Is this the same evidence that led to CBD's FDA approval?
No. FDA approval is based on randomized controlled trials measuring seizures in Lennox-Gastaut syndrome, Dravet syndrome, and tuberous sclerosis complex. This review looks at nonseizure outcomes in other, broader epilepsy populations.
Does this review prove CBD improves cognition or behavior?
No. Most included studies were observational, open-label, or retrospective with moderate to high bias risk, so the finding is a research signal rather than proof of benefit.
Could the reported improvement come from something other than CBD directly?
Yes. Some nonseizure improvement may reflect fewer seizures or reduced antiseizure medication burden rather than a direct effect of CBD on cognition or behavior.
Does the review provide an effect size?
No. Because study designs and outcome measures varied widely, the authors used narrative synthesis rather than statistical pooling, so no combined effect size is reported.
What research is needed next?
Prospective, controlled studies with standardized cognitive and behavioral measures, a defined comparator, and prespecified nonseizure endpoints are needed to test this signal rigorously.
