ResearchPod Summary
Melatonin is increasingly prescribed to treat chronic insomnia in children and adolescents. Despite this widespread use, the long-term safety profile of the hormone remains unclear, particularly regarding potential impacts on pubertal development and bone metabolism. This systematic review aimed to provide clinicians and guideline developers with a robust assessment of both short-term adverse events and long-term developmental outcomes.
The researchers conducted a systematic search of databases, including PubMed, Embase, Cinahl, and PsycINFO, for studies covering children and adolescents (aged 5–20 years) with chronic insomnia. The study design followed the GRADE (Grades of Recommendation, Assessment, Development and Evaluation) framework to assess the certainty of evidence. The analysis included 22 randomized controlled trials (1,350 participants) to evaluate serious and non-serious adverse events and four observational studies (105 participants) to evaluate long-term pubertal development.
The meta-analysis indicated that melatonin does not significantly increase the risk of serious adverse events compared to placebo. However, the data revealed an increased risk of non-serious adverse events (e.g., headache, nausea, fatigue, and mood changes). Regarding long-term development, the studies provided very low certainty evidence; three studies found no significant influence on puberty after 2–4 years of treatment, whereas one study suggested a potential delay in pubertal timing following longer durations of treatment (>7 years). Crucially, no studies were identified that investigated the relationship between melatonin treatment and bone health in this population.
Given the rising prevalence of off-label and self-prescribed melatonin use in young populations, the identified knowledge gap is significant. The findings suggest that while melatonin is generally safe from severe clinical outcomes in the short term, clinicians should not be complacent. The uncertainty surrounding long-term impacts on the developing endocrine system and potential non-serious side effects underscores the urgent need for more rigorous, long-term studies to inform clinical practice and pediatric health guidelines.
Sam: Melatonin’s long-term safety in children remains uncertain, despite its widespread use. A systematic review found more non-serious adverse events, but too little reliable evidence about puberty or bone health.
Alex: Those are different kinds of uncertainty. Did the review find evidence of harm, or find that researchers haven’t looked carefully enough?
Sam: Both, depending on the outcome. This is a systematic review by Mina Nicole Händel and colleagues, published in eClinicalMedicine in 2023. It brings together safety research on melatonin for children and adolescents with chronic insomnia, meaning persistent difficulty sleeping.
Alex: For researchers and clinicians interpreting those studies, the distinction matters. A treatment can look reassuring during a short trial while leaving developmental questions unanswered.
Sam: That distinction is the thread here. We’ll separate the short-term event data from the long-term developmental evidence, and ask what each can actually support.
Alex: Before we get into the evidence, why was a review needed? Was widespread use running ahead of safety research?
Sam: The authors describe substantial growth in paediatric melatonin use, including off-label use and self-prescribing. But safety reviews were scarce and mostly narrative. That made it difficult to judge the overall extent of unwanted events across studies.
Alex: So the contribution isn’t discovering an entirely new concern. It’s putting scattered evidence into a framework that distinguishes stronger answers from weaker ones.
Sam: Yes. They systematically searched the literature, combined results where that was defensible, and rated their confidence in the evidence. They also kept separate questions separate: adverse events during treatment, development through puberty, and bone health.
Alex: Why include puberty and bones alongside symptoms such as headache or drowsiness? Those seem like very different safety questions.
Sam: The puberty concern comes from a proposed hormonal mechanism, not an established treatment effect. Naturally occurring melatonin levels decline before puberty, raising concern that supplementation might disturb that pattern. Puberty also involves growth and bone development, making bone outcomes another question worth investigating.
AI-generated third-party summary by ResearchPod. Not official content or an endorsement by the paper authors or affiliated organizations.
Alex: A plausible mechanism can justify research without telling us what happens in patients. How did the review move from that concern to evidence?
Sam: It used different study designs for different timescales. Randomised studies supplied the comparisons for serious and non-serious adverse events. Observational follow-up studies supplied the longer-term developmental information, because that was the evidence available.
Alex: Let’s take the randomised evidence first. How large was the pool, and who was actually represented?
Sam: The review included twenty-two randomised studies reporting adverse events. Together, they involved about thirteen hundred and fifty participants. Most studies involved young people with underlying conditions, including autism, attention difficulties, or other neurological and developmental disorders.
Alex: That population matters if someone wants to apply the results to otherwise healthy children. Were the treatments themselves fairly consistent?
Sam: No. Doses, treatment lengths, and formulations varied. The trials were short, ranging from a week to a few months, so they cannot establish what happens after years of use.
Alex: What did combining those trials show about serious events? That seems the first place someone might look for reassurance.
Sam: The pooled comparison did not show an increase with melatonin versus placebo. Serious events include outcomes such as death, hospitalisation, or significant disability. The authors rated that evidence as moderate certainty, rather than high certainty, because the estimate was imprecise.
Alex: So “no increase detected” is fair, but “serious harm ruled out” would go beyond the evidence.
Sam: It would. The uncertainty around the estimate leaves room for differences the studies could not resolve. And this is reassurance within short trials, not a statement about lifelong safety.
Alex: What changed when they looked at non-serious adverse events? And “non-serious” doesn’t necessarily mean irrelevant to daily life.
Sam: The pooled results showed an increase with melatonin. On the absolute scale, that was about six additional participants experiencing an event per hundred treated, compared with placebo. Reported events included headache, nausea, fatigue or drowsiness, dizziness, and gastrointestinal problems.
Alex: That is easier to interpret than a relative increase alone. But does the pooled result establish that melatonin caused every symptom on that list?
Sam: No. The comparison concerns the overall number of participants reporting events, not proof that each listed symptom was caused by treatment. Confidence in the overall increase was moderate, while the exact size remained uncertain.
Alex: How consistently did the trials look for these events? A symptom checklist and a passing question at follow-up could produce different records.
Sam: That is a central weakness. The authors say systematic adverse-event assessment was rare, and complete reporting of event types and frequencies was rarer still. Some studies reported many events; others reported none in either group.
Alex: So part of the variation may concern how safety was investigated, not simply how participants responded. What about the longer-term evidence?
Sam: Here the evidence base shrinks to a handful of observational studies. After roughly two to four years of treatment, studies reported little or no apparent influence on puberty. A later follow-up suggested delayed timing after an average of about seven years.
Alex: That sounds like a possible duration effect. But a later signal could also reflect who remained in follow-up or how puberty was measured.
Sam: The authors call the duration interpretation speculative. These were small studies with serious risks of bias, including confounding and missing follow-up data. Some reports also followed the same population over time, rather than providing independent replication.
Alex: What did measuring puberty involve? That seems especially important when the proposed effect is a shift in timing.
Sam: Only one study used a physician-rated Tanner score, a rating of physical stages of puberty. Other assessments relied on self-report, questionnaires, or interviews with participants or parents. Comparisons sometimes came from previously published population data rather than a concurrently followed group.
Alex: Those methods might flag a question, but they make a confident developmental conclusion difficult. Were the authors able to combine them statistically?
Sam: They were not. The measurements and reporting differed too much for a meaningful pooled estimate, so the review described the studies separately. Its certainty rating for puberty was very low.
Alex: And for bones, was the conclusion similarly uncertain, or was there no eligible evidence at all?
Sam: No eligible studies reported bone mineral density or fracture risk. That means the review cannot establish either harm or reassurance for those outcomes. An empty evidence base is not evidence that nothing happens.
Alex: Given those gaps, does the review argue against melatonin treatment altogether, or against treating its safety as settled?
Sam: Against complacent use. The document identifies non-drug approaches as first-line treatment for childhood insomnia, with medication considered after those strategies have been tried. But this review assesses safety; it does not itself supply a complete comparison of treatment benefits and harms.
Alex: Who should read the full document, then, and where should they start?
Sam: Clinicians, guideline developers, and researchers studying paediatric sleep or development should read it. Start with the Results synthesis and the certainty assessment: they show where pooled comparisons are possible and where the evidence becomes fragile. Then read the Discussion’s account of puberty measurement and incomplete adverse-event reporting.
Alex: And for someone who only needs the conclusion to carry into a conversation with colleagues?
Sam: This review found more non-serious events with melatonin, without a detected increase in serious events during short trials. Long-term developmental safety remains unresolved. The line to remember is: short-term reassurance is not long-term safety.
Alex: Keep that distinction in mind when reading the next safety claim.