An occasionally irregular blog about orthodontics

Does Myofunctional orthodontics work? We have a trial to help us decide!

Does Myofunctional orthodontics work?  We have a trial to help us decide!

Does Myofunctional orthodontics work: We have a trial!

I think that one of the most interesting areas in orthodontics at the moment is myofunctional orthodontics and I’ve posted about this before. The main reason for my interest is that while it is a controversial area there may be something to the overall philosophy of correcting soft tissue function. I am also interested in the possibility of providing treatment with an appliance that comes “off the shelf” without the need for impressions. If this worked it would be great!  This blog post is about a recent trial of myofunctional orthodontics.

The main concept behind myofunctional orthodontics is  that we need to address the aetiology of malocclusion when we are providing treatment. As a result, we should consider the effect of soft tissues and breathing. Up to this point this seems logical to me. Proponents of this approach then  suggest that if the soft tissues and breathing are corrected with an appliance early in a child’s life any changes are stable and readily achieved. While I agree that we should be attempting to correct the aetiology of a malocclusion, I am not so sure that the soft tissue function etc can be corrected by an orthodontic appliance. This interesting trial in which a Myofunctional appliance was used may shed some light on this issue.

This paper was in the closed access journal the European Journal of Orthodontics. Maybe it is time for the European Orthodontic Society to consider whether it is right to restrict research findings?

Gothenburg, Sweden

Gothenburg, Sweden

The research was carried out by a team based in Gothenburg, Sweden.  This is a great city that I visited when I was 16 on my way to go walking inside the Arctic circle.





images-17Treatment of large overjet in Angle Class II: division 1 malocclusion with Andresen activators versus prefabricated functional appliances—a multicenter, randomized, controlled trial.

Emina Cirgic et al. European Journal of Orthodontics: Advanced access. DOI: 10.1093/ejo/cjv080

What did they ask?

In their introduction they pointed out that large overjets are seen in 15% of 10-year-old Scandinavian children. They also outlined that the main benefits of treating this problem early are the potential to reduce the incidence of trauma and improve self-concept. Traditionally treatments have been provided by activators and they raised the question on whether a preformed activator would be as effective as a modified Andresen Activator.

What did they do?

They carried out a randomised controlled trial in 12 general dental practices with the treatment being provided by the general practitioners.

They used the following inclusion criteria for the study:

  • 6 to 14-year-old children
  • Overjet of more than 6 mm or less if their lips were incompetent
  • Well aligned arches with no other features of malocclusion

The children were randomly allocated to receive treatment with a slightly modified Andresen activator or the prefabricated Myobrace appliance. Randomisation was done by opening a sealed envelope that contained the treatment allocation for each patient.

The Myobrace is an appliance that is made in seven sizes and the size of the appliance was chosen according to the manufacturer’s recommendations.  Details of this treatment can be found here.

The children were asked to use the appliance every night and for two hours during daytime.

Data was collected at

  • The start of treatment (T0)
  • When the over jet had been reduced to less than 3 mm followed by a six month retention period (T1)
  •  12 months post-treatment (T2)

They measured the following at the data collection stages

  • Overjet
  • Overbite
  • Molar relationship
  • Lip seal (yes or no)

This was the only data that they collected. Importantly, they did not collect cephalometric information.

The treatment was classified as unsuccessful if there was no reduction in the overjet over six months or they were unable to wear the appliance.

They carried out a sample size calculation that suggested that they needed to enrol 38 patients in each group. They based this on being able to detect a clinically significant difference of 1mm between the appliances.

They carried out an Intention to Treat analysis, which means that the data for all patients was included in the analysis, even if they dropped out of treatment. This approach, therefore, evaluates the average effect of a treatment on an “average” treatment regardless of the outcome.

What did they find?

When they did the ITT analysis they found that the overjet reduction was significantly greater in the Andresen group than the Myobrace group. However, this relapsed significantly in the Andresen group and at the end of the follow up period there were no differences between the appliances.

I have included this data in the table below.

Start8.4 (7.9-8.8)8.9 (8.5-9.3)0.15
End of treatment6.2 (5.6-6.8)5.6 (4.7-6.4)0.22
12 months post treatment6.5 (5.9-7.7)6.6 (5.7-7.4)0.77

This shows that most of the differences were small and the 95% confidence intervals were rather wide for the post treatment overjet measurements. This means that there is a lot of uncertainty in the data. While the authors pointed out that some of the differences were stated as being statistically significant, I am not sure that I think they were clinically significant. For example, at the end of treatment the mean overjet difference between the groups was 0.4mm and the change looked about 1mm but this was difficult to work out because of the graphical presentation of the data.

Similar information was found for the other dental measurements.

I was particularly interested in the effect of the appliances on lip seal, because this may have corresponded to change in soft tissue behaviour. Importantly, lip seal was achieved in 61% of the successful and 38% of the unsuccessful treatments. There was no difference between the groups.  However, we do not know if this would have occurred without treatment as there was not an untreated control group.  Nevertheless, this was interesting.

Successful treatment defined by an outcome of an overjet less than 3mm was found in only 37% of the patients. Again, there was no difference between the groups. Finally, 70% of the Myobrace and 53% of the Andreasen treatments were not successful.

What did I think?

The authors pointed out that the study had real world generalisability as it was set in general practitioner clinics and GDPs provide most of the functional appliance treatment in Sweden. This makes this study rather unique and very interesting.

They also suggested that they did not use a cephalometric analysis because other trials have shown limited differences in cephalometric measures. As a result, they only collected patient-centred outcomes and I totally agree with this approach. There is little to be gained by over analysis of cephalometric measurements in massive cephalometric festivals (see this post).

I was confused over their sample size calculation and the numbers they actually enrolled in the study. The sample size calculation suggested that they should enrol 38 in each group. However, they attempted to enrol 120 in each group to take predicted poor co-operation into account. Furthermore, as it took a long time to recruit the patients they stopped the trial early. This resulted in an unbalanced sample of 40 patients in the Andreasen group and 57 in the Myobrace group. While, this does not lead to a biased allocation because the randomisation was still carried out, I am fairly sure that this has some implications for the analysis of the data. They may have taken this into account, but it was not mentioned in the discussion.

If you refer back to my previous post on how to read a trial, I drew attention to the method of generation of the randomisation sequence and blinding of examiners. I could not find any indication of how they generated their randomisation. But more importantly, the method of data collection was not clear and I think that the data was collected by the operators. This means that the operators knew what treatment the patient received and bias could be introduced because of a lack of blinding.

When we consider all these factors this means that this study was at high risk of bias, but it is not possible for us to work out the direction of the bias. We need to bear this in mind when we consider the findings. I know that I have spent sometime on this aspect of the trial, but I think that this study is important and I wanted to look at it very closely.

So what can I conclude?

While I think that there are issues with this study, I feel that this provides us with the following information.

  • There is a greater level of co-operation with the custom made activator than the Myobrace.
  • The overall level of co-operation for both appliances was low.
  • The Activator was more costly than the Myobrace and also needed additional visits for impressions.
  • It appears that there may be no difference in the relative effectiveness of the two treatments. But this finding should be treated with caution because of the potential biases in the study.

Will this change my practice? While the use of the Myobrace is tempting I am concerned with the low co-operation rate and I think that I will still stick to custom made functional appliances, until I have more compelling evidence.

ResearchBlogging.orgČirgić, E., Kjellberg, H., & Hansen, K. (2015). Treatment of large overjet in Angle Class II: division 1 malocclusion with Andresen activators versus prefabricated functional appliances—a multicenter, randomized, controlled trial The European Journal of Orthodontics DOI: 10.1093/ejo/cjv080

Tags: , , , , , , ,

There Are 17 Comments

Trackback URL | Comments RSS Feed

  1. Sam Whittle says:

    Dear Kevin, Thank you for this blog discussing a prefabricated appliance that has gained some popularity amongst primary care ‘orthodentists’ here in Australia. There has been considerable concern amongst specialist orthodontists that myofunctional treatment has been recommended for children much younger than those in this study, [securing significant remuneration for the suppliers of such therapy], while little is understood or applied to the diagnosis and management of the growing dentofacial complex. I will look at this paper when my journal arrives but am most interested in changes claimed between different age groups. In light of the large AJODO class 2 study from North Carolina that suggests no justifiable difference amongst 3 treatment cohorts within the study of children with similar overjets, I am struggling to give this current study much attention. I would be interested in your interpretation of this study in light of the argument: ‘to treat or not to treat moderate class 2 malocclusions in the mixed dentition’.

    • Kevin O'Brien says:

      HI Sam, thanks for the comments. yes, you are correct that this study and the current drive for early intervention needs to be interpreted in the light of the findings from the UNC, Florida and UK based RCTS into the effects of early treatment. We know that these studies reported little advantage in early treatment but there still may be benefits from the reduction in trauma and socio-psychological well being. Nevertheless, my interpretation of this current study was that the interventions were not particularly effective because of the low compliance rate. If there is an effect of early treatment, this cannot be achieved if the patients cannot or will not wear the appliances. I hope that I have answered your question

      • Sam Whittle says:

        Thanks for the reply Kevin. It seems if I add your reply to me and the one to my colleague Mark Cordato, we end up with a very clear answer. Cooperation is such a huge issue, particularly with removable appliances; I wonder if we can truly design a study that eliminates this as a factor to give a clearer picture of appliance effects?
        Thanks for your blogs during the year. Wishing you a restful and enjoyable festive season from your friends down under.

        • Kevin O'Brien says:

          Thanks for the comment, I agree that co-operation is important, however, if we are to truly test a treatment method we still need to factor in co-operation. For example, someone could develop a treatment method that has great effects but only 10% of our patients could co-operate fully.This would mean that in the real world the appliance is not effective.

  2. Mark Cordato says:

    Dear Kevin, I do not have access to the EOJ. I have several points on this study. I had trouble with the statistics. It could be the way I have interpreted your table, it is a blog so a fully annotated table would be an issue, but, if the parenthesised numbers are the range at 1SD then the end of treatment figures show the better end of the 1SD are nowhere near the 3mm threshold of success mentioned elsewhere. I am guessing the tables you presented from the article were the millimetre (mm) values of the overjet of patients.
    End of treatment 6.2 (5.6-6.8) 5.6 (4.7-6.4)
    12 months post treatment 6.5 (5.9-7.7) 6.6 (5.7-7.4)
    The numbers are consistent with a 2 sided test so 1 SD would cover 68.27% and so only 14% (1 side, the lowest values) would be less than 5.6 or 4.4 respectively. It gets worse, to reach a figure of 3.0mm or less the Myobrace would be more than 4SDs and the Activator just shy of 4SDs. Again, it could be a very small segment of the data provided but the more I look at it the less ‘real’ they seem. The thought did cross my mind you were trying to check on us Kevin!

    With such poor compliance rates either those patients were excluded from the stats or large numbers improved 1+mm in the Myobrace group and 2+mm in the activator group compared to the untreated numbers. Start 8.4 (7.9-8.8) 8.9 (8.5-9.3) May I suggest we have a problem!

    If the numbers in the discussion are valid it still demonstrates poor efficacy. Even in the table the overall change from 8.5 to 6.6 still has a problem for most patients after a phase of treatment and as has been shown by Bjork and others, some Class II s spontaneously improve which does not seem to be accounted for.

    Unless the child is experiencing psycho-social discomfort from their dental appearance or they are rough and boisterous and more likely to have trauma, I can see only cost and no benefit from treatment. Airway considerations may be a part of the discussion but maxillary expansion seems to improve function as do respiratory physicians and ENTs, if there is still an issue then a protraction appliance could have a place. In general, I see only rare indications for functional appliances especially if there is a likelihood that fixed appliances will be required. Your Cochrane review lays it bare that a functional appliance provides no skeletal or dento-alveolar benefit so I only see costs in terms of clinic time, lab fees and patient compliance over gaining Class II correction in one phase. It has to cost somewhere, either higher fees to the patient or less profit to the orthodontist. We should be striving for efficacy and try to minimise inefficiency.

    I really do enjoy you letting us into your thought and logic processes and I really do enjoy the Blog!

    • Kevin O'Brien says:

      Hi Mark, thanks for the comments. The data that I presented were the means and CIs for the ITT sample and this, therefore, included the data for the patients who were deemed unsuccessful. I think that this results in the rather poor performance of both appliances. This is reflected in the low co-operation rates and we can conclude that there is poor efficacy of this treatment method and delivery.

      So this does go back to the original conclusions of the early Class II treatment studies and the best time to provide treatment for most children with Class II malocclusion is one phase in adolescence. I hope that this answers your question

  3. Kevin, As a recently new user, and lecturer for (disclaimer), of the Myobrace appliances, I have learned a couple of things that this study does not tell me:
    1) I don’t use the Myobrace as an orthodontic appliance. It may look like a positioner and the one you illustrated is supposed to be used to move teeth, but for me that is not the best use for the appliance. Think of it more like an exercise accessory….say, like a jump rope. Jump ropes require compliance, too, and low compliance would condemn a jump rope as ineffective. But if you use the jump rope with a personal trainer that MAKES you use the jump rope properly, I bet you can get great results with it, right? Well, Myobrace is best used when the child is encouraged to use it properly and to do specific exercises with it whose aim is to correct breathing, lip seal, tongue posture and properly swallowing. So my first question about the study is “what did they do to ensure good compliance?”. I know this question is anathema to most of us orthodontists who try to do anything we can to eliminate cooperation as a factor for reasons of predictability and efficiency. But if you are going to change habits – habits which are the etiology of most malocclusion – then WE have to be responsible for getting cooperation. So what did the investigators do to ensure compliance? What did the investigators do to TEACH good habits to these kids? (Giving dance shoes to a kid to see if they will dance does nothing to teach him to dance, does it?)

    2) Reduction of overjet, by itself, is not a satisfactory metric. If the maxilla starts out retrusive, as it is in most Class II cases (see McNamara, 1981….and it’s only gotten worse with more respiratory ailments and mouth breathing habits), then reduction of overjet by retracting the upper incisors can be considered detrimental to the outcome of the case (assuming the goals of treatment are to put the teeth where they belong in the face). If you can get FORWARD growth of the jaw structures (to provide more room for the teeth, room for the tongue on the palate, and reduce obstruction of the pharyngeal airway) AND reduce the overjet, THEN overjet reduction is valuable.

    3) If you don’t use a simple to apply, cost effective method for changing poor oral habits in your practice, what do you use? If you have a good myofunctional therapist to do the work and give proper attention to the case, well, fine. If your MFT has the time to treat all the children in your practice with soft tissue dysfunction (do you have two or three?) then you are right on target. On the other hand, if you are waiting until your orthodontics is in trouble before you make your referrals, you are way behind in handling the problems. It’s like waiting for the whole kitchen to be on fire before you pull out the extinguisher. I encourage you to look for soft tissue dysfunctions BEFORE they have their way with facial growth. And then consider using something that you can use to easily teach better habits.

    This study, unfortunately, tells you nothing about that.

    • Kevin O'Brien says:

      Thanks for the good question. I probably agree with you, although I am a sceptic. The theory tells me that there is a role for changing oral habits, but I have not seen any research evidence that shows that this can be achieved and then results in a change in the position of the teeth and skeletal basis. This research is overdue and I am wondering how to do this study. It certainly seems very worthwhile to me

  4. Helen Jones says:

    I wholeheartedly endorse Barry Raohael’s comment. I have just visited the Myobrace clinic in Bondi, Sydney, Australia and seen some impressive results. This system is not about just fitting an appliance but it is about educating the patient about tongue position, lip seal and nasal breathing and providing the exercises and motivation to make these changes happen..

  5. NOEL STIMSON - editor Cranio UK journal says:

    “Treatment of large overjet in Angle Class II: division 1 malocclusion with Andresen activators versus prefabricated functional appliances—a multicenter, randomized, controlled trial.”

    I was intrigued by your blog on this Swedish study. I am sure you are correct in your criticisms and comments about the biases and other features of the trial’s construction, but I can see some even more fundamental anomalies that were not commented on.

    I confess to being a fan of the Myobrace system (though I had retired before I learned about it) and I find it unusual to compare this relatively new (20 years) biologic muscle-based system with the totally mechanistic Andresen appliance that I first used as a student over 50 years ago!

    What troubles me is the high incidence (70%) of unsuccessful Myobrace cases in this trial when it generally regarded as a high-compliance system (80-90% compliance). This begs the question as to whether or not those patients received the intensive management programme of visits to the myofunctional therapist coupled with home-based activities that are all essential components of the Myobrace programme.

    My concern is compounded by the fact that, according to the Myofunctional Research website, there appear to be no Myobrace certified practitioners in Sweden, Norway or Denmark. So who was delivering and supervising the Myobrace treatment?

    I was also puzzled that the case selection criteria include an overjet of greater than 6mm and a normal arch relationship (your phrase is “well aligned arches with no other features of malocclusion”) in spite of the “Angle Class II division 1” description in the study’s title!

    Even more puzzling is the inclusion in the ‘data’ from all the non-compliant (i.e. unsuccessful) cases. I find it strange that you have not commented on the fact that a study’s results must inevitably be affected when they are based on data from patients who effectively failed to take part in the study, unless of course that is what the study was about! Surely a true comparison of two such disparate appliances can only include those in the ‘compliant’ groups.

    • Kevin O'Brien says:

      Thanks for your comment. I will only address the one about including children who were non compliant in the data analysis. This is standard practice in trials and is called Intention to Treat. This means that all participants who enrolled in the study are analysed. This is done because investigators are trying to measure the effect of attempting an intervention on a sample of patients regardless of outcome or co-operation. This is because co-operation cannot be guaranteed with some treatments i.e orthodontics or even taking medicines. If these people were excluded the study results would be biased to successful treatments. It is like carrying out a drug trial and only including those patients who got better! This really is basic trial design and is very relevant to orthodontic studies.

  6. noel STIMSON - editor Cranio UK journal says:

    Seems to me that including non-participants in the data of an orthodontic study would introduce a bias towards unsuccessful treatment. I understand the process of Intention to Treat and I am sure it works well in studies where the main question is “What is the completion or compliance rate for this treatment?”, but to allow non-data from subjects who did not actually take part in the study to affect the final result of the study (i.e. bias) is baffling to me.

    Intention to Treat may be part of many normal study protocols, but it is a point where science and common sense start to collide.

    • Kevin O'Brien says:

      Noel, I have replied to this in another of my replies to your questions. This is the correct way to analyse data in a trial because it is essential to include all the participants despite co-operation. If this is not done then the results are biased. I am sorry that you cannot see this but this may be your problem? It would also be helpful if you do not post multiple identical questions, as I cannot always answer them. I do this blog in my spare time

      • noel STIMSON - editor Cranio UK journal says:

        Kevin, I really have to refute your suggestion that “I have a problem” seeing that bias is the result of not including the data of non-participants in a study. I see it all too plainly, but I also can see the bias the other way when you do include it as normal. I did read your earlier reply and I was not satisfied by your saying “but this is the way we have always done it”. I am not posting multiple identical questions, simply attempting to extend the argument for a more satisfactory explanation.

        • Kevin O'Brien says:

          Thanks and to clarify, I did not state that this is the “way that we have always done it”, I stated that correctly carried out trials use this methods. I wonder if you should read the excellent book by Professor Trisha Greenhalgh on “how to read a paper”. It provides a great insight into the carrying out and interpretation of scientific research. You would learn a lot

        • CP says:

          If you are a practitioner, and you want to know how likely a treatment is going to be, you want to know the efficacy including non compliance. You don’t know ahead of time if the patient will be compliant. It’s more useful to know the success rate for *all* patients, not just a subset. However it is also useful to know the success rate for the compliant subset as well, for *if* the patient is compliant. Maybe it will motivate the patient. But, that’s a big if. You don’t want to base a general policy on the subset, but on the success rate for all patients, because you can’t look into the future and predict which patients will be compliant.

  7. Joel Hodges says:

    Thank you very much for posting about this study. I think that we’re all seeing the issue with this study trying to address two questions. Can the Myobrace work? And can the Myobrace work for an average patient? I think that the second answer is addressed by the study but it would be interesting to see the results for the most successful participants just to hint at what the appliance can do under ideal conditions.

Post a Comment

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


Pin It on Pinterest

Share This