February 21, 2022

Maxillary expansion improves the quality of life for Sleep Apnoea patients?

This is another post on the effects of implant retained maxillary expansion. In previous postings, I pointed out that there are few trials into this treatment.  This new paper addressed some of my concerns.  As a result, it may provide us with useful information.

Currently, clinicians are promoting Mini Implant-assisted rapid palatal expansion (MARPE). However,  when we consider the effects of this appliance, it appears that the evidence base is relatively weak. Most investigations are retrospective and others analyse the same cases in different papers. As a result, it is difficult to come to clear conclusions about this form of maxillary expansion.

As a result, I thought it was great to see this new trial that looked at MARPE and quality of life in a sample of patients with Obstructive Sleep Apnoea.

A multi-national team wrote this paper. Progress in Orthodontics published the paper

Mini-implant assisted rapid palatal expansion (MARPE) effects on adult obstructive sleep apnea (OSA) and quality of life:a multi-centre prospective controlled trial.

Daniel Paludo Brunetto , Christoph E. Moschik , Ramon Dominguez‑Mompell , Eliza Jaria , Eduardo Franzotti Sant’Anna and Won Moon

Progress in Orthodontics: Advanced access. Doi: https://doi.org/10.1186/s40510-021-00397-x


What did they ask?

They did this study to:

“Assess the clinical outcomes of adult non-obese OSA patients who underwent MARPE treatment, using home sleep testing and OSA related quality of life measures”.

What did they do?.

A sample size calculation showed that they needed 8 participants in each arm of their trial to detect a 9.5 variation on the Apnea/hypopnea index (AHI).

The PICO was


The authors identified patients from three institutions. The inclusion criteria were above 18 years old and had a skeletal maxillary transverse discrepancy. In addition, they had OSA that was identified by a home sleep test.


MARPE using a Maxillary Skeletal Expander (MSE) appliance for maxillary expansion. This was done until “the centre of the upper alveolar ridge was positioned 2-3mm towards buccal compared to the center of the lower aleveolar ridge”. Treatment was successful if the CBCT detected radio translucency at the mid palatal suture.


No treatment


Primary outcome Apnea/Hypopnea Index. Secondary outcomes: Epworth Sleepiness scale, Quebec Sleep test, CBCT imaging.

They confirmed the subject’s OSA with a home sleep test and offered the participants treatment. Those who declined this treatment were allocated to the control group. The treatment was not randomly assigned. As a result, this was not a randomised trial. It is a quasi-randomised or controlled clinical trial. This is important when we consider the quality of the evidence.

They collected data at the start of treatment and 6 months after the maxillary expansion was complete.

They did a simple univariate statistical analysis.

What did they find?

They enrolled 20 participants in the intervention group from Jan 2016 to Jan 2019. Three were lost to follow-up, and three were not responsive to treatment. This left 14 patents for data analysis. They did not do an Intention to Treat analysis.

They recruited 12 participants to the untreated group.

At the start of the study, there were no differences in the outcome measures between the two groups.

At the end of treatment, there were several important statistically significant differences between the groups.

The AHI for the untreated group was 24.18 (SD=11.37), and for the treated group, this was 11.45 (SD=6.16). Furthermore, of the participants who had a successful expansion, 11 out of the 14 had a 50% reduction in AHI.

When they looked at the Epworth Sleepiness Scale, this was 12.16 (SD=3.9) for the control and 7.32 (3.6) for the treated group.

Finally, the Quebec Sleepiness Questionnaire score for the control was 4.92 (0.71) and 5.52(0.5) for the treatment group.

Their overall conclusions were;

“MSE had a 85% success rate for morphological change. It also improved daytime sleepiness, OSA related quality of life and AHI”.

“MSE may be considered as an auxiliary treatment for OSA in non-obese young adults with a maxillary transverse discrepancy”.

What did I think?

This was an ambitious study, and it was great to see a team of investigators use outcome measures relevant to our patients. This is much more useful than a series of cephalometric tables that report on minor morphological differences.

It was also important to see that they detected statistically significant differences between the groups. However, I was unsure if these were clinically significant, and the authors did not provide any information on this matter.

Study design

As with all studies, we need to critically look at the study design. This was defined as a controlled clinical trial because the participants were allocated on their preferences for not having treatment. We, therefore, need to assume that there were some differences between the groups because they took different decisions. However, we do not know the effect of these differences. Therefore, we can only assume that the study is at risk of bias.

I think that it is useful at this point to consider whether it was possible to carry out a randomised trial to reduce the risk of bias. The authors stated that it was “unethical to not treat participants that knowingly needed expansion for occlusal purposes”. I am afraid that I do not agree.  They could have done a trial in which they allocated participants to the intervention or delayed treatment. This would be ethical.

Furthermore, the sample size was very small. This means that the data is subject to a large amount of variation and the effect of outliers.

We also need to consider that 6 of the treated patient group dropped out or were excluded because the treatment was unsuccessful. Notably, the authors did not do an Intention to Treat analysis that included these patients. We, therefore, need to wonder if the results would have been different if this data had been included. Again, this is a significant deficiency of this study.

I contacted the lead author about the absence of an ITT analysis.  They kindly got back to me and let me know that they took this decision because this was a preliminary study and they wanted to make sure that there was not too much heterogeneity in the data. They are doing a larger study that will include an ITT analysis.

Final comments.

I have thought carefully about this study. While on brief reading, the findings appear to be important. I am not sure that I can be so confident about their conclusions because of the issues that I have previously described.

Nevertheless, this small study is a good step forwards in researching the effects of MARPE. Therefore, we can consider this to be a good pilot. It does provide great information to plan a more extensive study. But it has a high degree of uncertainty and should not change practice.  I really look forward to seeing the results of their larger study.

Related Posts

Have your say!

  1. I’m not sure that the statement that it’s unethical not to treat if there are occlusal reasons for treatment really holds water. They were all over 18 so were presumably seeking treatment for OSA only. In addition there are no occlusal reasons for treating posterior cross bite, probably in most cases. This statement implies that all those in the treatment group were different, occlusally (ie they had x bites) from those in the control group (no cross bites), that doesn’t strike me as a good control group.
    Yet again, as with a lot of ortho stuff, the question isn’t really “does intervention x do something” but “does the intervention x effect last for a sufficient time”. I’m not sure any studies like this should be published until followed up for at least a year post treatment.

  2. Dear Kevin

    Thanks for posting this paper.

    Given that many malocclusion phenotypes, such as maxillary transverse discrepancy/deficiency(MTD), Class II mandibular retrusion and vertical excess(‘adenoid face’, ‘Long-face syndrome’), are often initially detectable before the age of 7, and will most often persist and worsen beyond, do you think these 26 young adults with Dx:MTD and OSA co-morbidity might have experienced an earlier improved OHRQOL(per Brazilian study you’d previously reviewed here) had their MTD been corrected by/before age 7?

    • Thanks for your comments. In answer to your question, I do not know as we do not have any real strong evidence that treating children with expansion can prevent OSA. There maybe something to treatment for these problems, but currently the evidence is lacking to support this treatment without many caveats about its effectiveness. Best wishes: Kevin

      • Thank you Kevin

        My actual question (below) though centered around improved OHRQOL as a possible Tx response to maxillary expansion with RPE if done by/before age 7 rather than in early adulthood, and not at all about whether or not RPE might prevent OSA.

        ‘…do you think these 26 young adults with Dx:MTD and OSA co-morbidity might have experienced an earlier improved OHRQOL(per Brazilian study you’d previously reviewed here) had their MTD been correctedby/before age 7?’

        • Sorry to be unclear. If we consider the study on quality of life. We could say if they had transverse maxillary deficiency, it appears that treatment between 8-10 years is likely to improve their oral health related quality of life. This is the conclusion of the study that I posted about in November last year. However, you cannot suggest that early treatment is better than adolescent treatment because this study did not look at comparing early with adolescent treatment. In order to conclude this someone would need to do a similar study design to the early Class II treatment studies.

  3. Another interesting adult study. One thing we have to clearly differentiate here is pediatric maxillary orthodontic appliance expansion vs. adult maxillary surgical expansion. Let’s not ‘muddy the water’. In this study, the mean AHI was reduced to 11.45. By current standards, this failed to reach the definition of medical success (>50% reduction in AHI and <10/events per hr). However, some believe that AHI is a rather arbitrary parameter in that an apneic event lasting 10 seconds or an event lasting 60 seconds are equally weighted in the current AHI definition. Moreover, an SpO2 nadir of say 90% for 10 seconds is treated the same way as a nadir of say 70% lasting 60 seconds. For this reason, some believe it might be better to calculate the integral to gauge the patho-physiologic and hence QoL impact. Some argue that two patients with the same AHI may have very different risks of stroke or heart attack since the perceived improvement in AHI can mask the increased incidence of hypopnea allied with profound desaturations.
    p.s. I really like Dr Pearson's comments on the (lack of) need to treat asymptomatic posterior crossbite, which is worthy of further investigation with respect to OSA.

  4. a thought – would any means of expansion be successful ? IIRC there is no significant difference in the outcomes/success between the various expansion methods (excluding SARME)

Leave a Reply

Your email address will not be published. Required fields are marked *