Palatal Expansion for a Child's Airway: It Does Not Work Alone
Widening the upper jaw is widely marketed as an airway treatment for children. Pooled against control, rapid maxillary expansion on its own changed the apnea-hypopnea index by 0.02 events per hour. Combined with other treatment it works. That distinction is the whole article.
Rapid maxillary expansion is sold hard as an airway intervention for children. In the best pooled evidence, used on its own, it reduced the apnea-hypopnea index by 0.02 events per hour — which is to say, not at all. Used alongside other treatment, it works. The difference between those two sentences is what parents are rarely told.
What is palatal expansion?
An appliance fixed across the upper teeth, widened slightly each day, which separates the two halves of the upper jaw at the midline suture and holds them apart while new bone fills the gap.
The airway logic is real: the floor of the nose is the roof of the mouth. Widen the upper jaw and you widen the nasal floor, lowering nasal resistance. In a child with a narrow high-arched palate and a crossbite, that is a genuine anatomical change rather than a theory.
It works in children because the midline suture has not fused. In adults the equivalent needs miniscrew assistance or surgery, which is a different procedure with a different risk profile.
What did the evidence find?
Yu and colleagues, in Sleep Medicine Reviews in 2023, ran a network meta-analysis — the design that lets you compare treatments against each other rather than one at a time. Eleven studies, 595 patients under 18, with AHI as the primary outcome.
| Treatment | AHI change vs control | 95% CI | p |
|---|---|---|---|
| RME + adenotonsillectomy | −5.13/h | −7.50 to −2.76 | < 0.0001 |
| RME + mandibular advancement | −3.79/h | −5.21 to −2.37 | < 0.0001 |
| Myofunctional therapy | −2.45/h | −4.76 to −0.14 | 0.038 |
| Mandibular advancement alone | −2.18/h | −3.48 to −0.89 | 0.001 |
| Rapid maxillary expansion alone | +0.02/h | −1.72 to 1.75 | 0.985 |
Heterogeneity was modest at I² = 32.6%, which means these studies were reasonably consistent with each other — the flat result for RME alone is not an artefact of pooling wildly different trials.
A p-value of 0.985 with a confidence interval sitting symmetrically across zero is as close to "no detectable effect" as this kind of analysis produces.
Evidence tier 1 that expansion alone does not reduce AHI in children. Tier 1 that the combinations do. The uncomfortable part is that the marketing rarely distinguishes between them.
Why does it work in combination but not alone?
Because it addresses the wrong level to finish the job.
Expansion lowers nasal resistance. In most children with obstructive sleep apnea the obstruction is adenotonsillar — tissue crowding the pharynx behind the nose. Widening the nasal floor upstream of a pharyngeal blockage improves the easy part of the route and leaves the narrow part untouched.
Pair it with adenotonsillectomy, which removes the pharyngeal obstruction, and you get the largest effect in the table. Pair it with mandibular advancement, which addresses the tongue base, and you get the second largest. The pattern is consistent: expansion is a useful adjunct to something that treats the main obstruction, and it is not the main treatment.
That is the same principle this whole site is organised around. Locate the narrowing, then pick something that can physically reach it. Nose, mouth or throat.
Does that mean expansion is pointless?
No, and this is where the honest version has to be careful.
The trial outcome measured was AHI. Expansion has other legitimate indications — crossbite, dental crowding, a transverse discrepancy that needs correcting on orthodontic grounds regardless of breathing. Those reasons stand on their own and an orthodontist recommending it for them is not overreaching.
It may also improve nasal breathing and mouth breathing without moving the AHI, and for a child whose problem is nasal obstruction rather than apnea, that could matter. The analysis above cannot speak to that, because it was not the endpoint.
What is not supported is presenting expansion to a parent as a treatment for their child's sleep apnea, or as an alternative to adenotonsillectomy. On this evidence it is neither.
What should a parent ask?
- "Is this for my child's bite, or for their breathing?" Both are valid reasons. They justify different expectations and only one of them is supported by the AHI data.
- "Has the apnea been diagnosed with a sleep study?" Not estimated from snoring. Paediatric thresholds differ from adult ones.
- "What is treating the adenoids and tonsils?" If the answer is nothing, and they are enlarged, expansion is being asked to do a job it has been shown not to do. Adenotonsillectomy, and what the CHAT trial found
- "What happens if we do nothing for six months?" A fair question in mild cases — in the CHAT trial, 46% of untreated school-age children normalised within seven months.
- "What is the total cost, and over how long?" Expansion is a months-to-years commitment with retention afterwards.
The line
Nothing sold for adult snoring belongs near a child's airway, and no orthodontic appliance is a substitute for diagnosing what is actually obstructing it. A child who snores habitually, pauses, gasps or sleeps restlessly needs a paediatric assessment first. Snoring in children, and when to worry.