November 13, 2023

Does Twin Block treatment cure pediatric obstructive sleep apnea?

The role of orthodontic treatment in treating airway problems in children is becoming very controversial. Over the past few years, proponents of airway-focused orthodontics have promoted orthodontic intervention for obstructive sleep apnea (OSA). Some have also published roadmaps for treatment, suggesting that orthodontic growth modification may have a role to play in this treatment. Unfortunately, while this treatment may be compelling, there is a lack of evidence for its effectiveness. Nevertheless, airway-focused orthodontists have built significant momentum and are promoting orthodontic treatment for paediatric OSA without evidence. As a result, they may be practicing fringe medicine and potentially causing harm both to patients and their parents’ wallets. There is, therefore, an urgent need for research in this area.

This new paper may help us shed some light on this controversy.

A team from Kota Bharu, Malaysia, wrote this paper. The European Journal of Pediatrics published it.


They presented a very clear and well-written discussion on the role of orthodontics in managing paediatric OSA. Importantly, they drew attention to the problems with using CBCT imaging as an outcome measure. I will return to this when I discuss this paper.

What did they ask?

They did this study to ask this question.

“What is the effect of the Twin Block therapy on upper airway parameters/dimensions in OSA children with mandibular retrognathia”?

What did they do?

They carried out a prospective observational cohort study based in one dental school clinic. The PICO was


Growing children in the age range of 8-12 years with a CVM of stage 2 or 3 with skeletal Class II malocclusion.


Twin Block appliance treatment for children who had polysomnographic proven OSA (Apnea-hypopnea index greater than 1.0 per hour)


Similar children who had an AHI less than 1.0/hr. This group did not receive Twin Block treatment.


Measurement of airway parameters from CBCT images taken at the start of treatment and 9 months later.

The operator asked the patients to wear their Twin Blocks full-time. The patients also completed an appliance wear diary.

The patients in the control group did not receive any treatment to correct the AP relationship. However, they received a phase of prefunctional treatment to correct occlusal interferences. It was not clear how or why these treatments were allocated.

The CBCT images were all taken following a clear protocol. Notably, the patients were seated in an upright position. The lips and tongue were in a resting position without swallowing while scanning.

The study team calculated the following parameters from the scans for the nasopharynx, oropharynx, and hypopharynx.

  • Airway volume of each region
  • Minimum Cross-sectional area (MCA)
  • AP and lateral dimensions of the smallest axial cross-section
  • Length of the upper airway.

Finally, they repeated a sleeping PSG for the study group 30 days after treatment was stopped. They did not obtain this data for the control group.

What did they find?

They enrolled 34 participants in each group. The TB treatment resulted in a reduction of 2.980 in ANB. They did not report this data for the “control” group.

They presented a large amount of data derived from the CBCT images. As with many orthodontic studies, this created much white noise with multiple comparisons of minor effects. So, I decided to try to keep things simple in my interpretation. As with other OSA papers, I thought of the airway as a simple tube and considered that the most crucial variable for airway obstruction would be the minimal cross-sectional area.

The study team also presented change data and not simple pre and post-treatment information. The latter is far more meaningful, and I approached the corresponding author for this information. Unfortunately, I did not receive a reply. I have put the change in MCA (mm2) in the table below. I calculated the 95% Confidence Intervals as this is more meaningful data than standard deviations.


Region Twin Block Control p
Nasopharynx 5.75 (5.1-6.3) 4.11 (3.57-4.6) 0.096
Oropharynx 40.47 (36.17-44.77) 4.45 (3.85-5.07) 0.001
Hypopharnx 19.91 (16.7-21.53) 3.86 (3.2-4.45) 0.003


When they looked at the change in AHI (event/hr). For the treatment group, the mean was 14.9 (5.5), and for the control this was 0.4 (0.3).

The team’s final conclusions were.

“The correction of Class II mandibular retrognathism  with a Twin Block appliance resulted in a significant increase in the upper airway volume and MCA for the oropharynx and hypopharynx”.

“There was also a significant reduction in AHI”.

What did I think?

This was an ambitious study that, however, has some flaws. We need to consider these.

Firstly, although the groups were matched for morphological features, the control group did not have OSA. This means that the groups must be different and are not comparable. This makes interpretation of the data very difficult.

Furthermore, in the introduction, the authors pointed out that CBCT imaging is inaccurate for measuring the airway as the views are taken with the patients sitting upright and not supine. Nevertheless, this is how they took their images.

It is a shame that I could not obtain the before and after values for the groups. This information is far more meaningful than treatment changes. This is because viewing treatment change only does not allow us to appreciate the actual effect size. More importantly, we have no idea whether the groups were different regarding the pre and post-treatment values. This is important when we consider that the control group did not have OSA.

However, it is interesting to look at the AHI values for the treatment group, as this data looks compelling. However, before we get too excited, we must remember that these changes may be due to normal airway growth and regression of lymphoid tissue. We also need to consider that the control group would also have been growing, and yet the change in AHI was minimal. This also illustrates that we need to see the pre and post-treatment AHI for the control group.

Final comments

This paper was interesting, and it certainly raised some questions about the effect of functional appliances on the airway and AHI. Unfortunately, the incomplete reporting of data makes it impossible to agree with the conclusions of the investigators. However, this study points towards the value of an RCT in which children with OSA could be randomised to either functional appliance treatment or a period of observation. This study would go a long way towards addressing the controversy on the role of orthodontics in treating paediatric OSA.

Related Posts

Have your say!

  1. The team’s final conclusions were.

    “The correction of Class II mandibular prognathism with a Twin Block appliance resulted in a significant increase in the upper airway volume and MCA for the oropharynx and hypopharynx”.

    “There was also a significant reduction in AHI”.

    Dearest prof.
    I think you are referring to mandibular retrognathism, not prognathism. A simple typo to correct

  2. It seems that too many of these studies obfuscate what is really important in evidence-based diagnosis and treatment for OSA. This research is better than much that is out there, though still does not deliver the information that we need.

    Polysomnography is still the gold standard in the diagnosis of OSA. So what is necessary in determining the validity of any treatment for OSA, is whether the apnea hypoxia index, as measured by polysomnography, is significantly improved after a treatment. In order to determine this accurately a polysomnography from time points both before and after treatment seems a necessary condition.

    In order for the research to be of evidence based clinical benefit to clinical practice, the control group would need to be similar to the experimental group and have polysomnography taken at the same time points as the experimental group, before and after treatment.

    I find it interesting that so many other factors, such as CBCT, cephalometric, and other evidence, known to be insufficient in providing definitive diagnosis for OSA still enter into the discussion. This is where the obfuscation appears to be taking place.

    Yes, it is interesting to see what cephalometric, CBCT, and clinical measurement values did and did not change. However, what we really truly want and need to know is if polysomnography, taken before and after treatment with both the experimental and control groups, showed a significant difference. That this is lacking from so much research on the subject almost seems suspect.

    • In keeping with your emphasis on polysomnography both before and after an intervention, I believe you are alluding to the control population as needing to be a similar AHI pre-intervention verses a control population that does not exhibit AHI as was used in this cohort study.

      The other thing that I was wondering is whether AHI is cyclical with allergies. Intuitively if would seem that inflamed lymphoid tissue might give a transient apnea or severity of apnea that would cycle regardless of the intervention and lead to belief in the intervention.


  3. I think you guys are being a little too harsh on this study. The authors clearly state “The aim of this study was to evaluate the effect of twin-block appliance therapy on upper airway parameters/dimensions in OSA children with class II mandibular retrognathic skeletal malocclusion using cone-beam computed tomography”. They clearly met that aim of looking at upper airway changes. I’m not saying I entirely agree with the narrative. For example, “CBCT imaging has been shown to be an effective and precise diagnostic tool for analyzing the upper airways and craniofacial structures” is a bit vague, since diagnosis of OSA cannot be done with CBCT. Also “Minimal cross-sectional area of upper ways may be the most relevant potential parameter when explaining how the upper airway anatomy plays role of in the pathogenesis of pediatric OSA” is a bit overreaching since tonicity of the upper airway as well as tonsillar tissues might also play significant roles in pOSA. We can also venture that PSG “used to be the gold standard” since newer technologies are proving to be more practical, especially in the pediatric population.

Leave a Reply

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