December 05, 2016

Another trial suggests that AcceleDent may not be effective…

Another trial suggests that AcceleDent may not be effective…

Regular readers of this blog will know that I have posted about AcceleDent before. In these posts I discussed that the claims made by the manufacturer and the Key Opinion Leaders may not supported by research. This new study adds to the evidence about this intervention.

Peter Miles, an Australian orthodontist, carried out this trial in his own practice. I knew Peter when he was a resident at the University of Pittsburgh and he was a student who always asked the difficult questions. He has carried out several randomised in his own office. The AJO-DDO published his most recent paper on AcceleDent.

Assessment of the changes in arch perimeter and irregularity in the mandibular arch during initial alignment with the AcceleDent Aura appliance vs no appliance in adolescents: A single-blind randomized clinical trial

Peter Miles and Elizabeth Fisher

Am J Orthod Dentofacial Orthop 2016;150:928-36)


This is a well written paper. They wrote a very nice review of the current knowledge behind AcceleDent and pointed out that retrospective studies supported the use of the device.   Paradoxically,  randomised trials showed that the device was not effective.

They set out to discover whether the use of AcceleDent increased the rate of initial tooth movement as measured by an irregularity index or anterior arch perimeter.

What did they do?

They did a single centre randomised trial with a 1:1 allocation of intervention vs control.

The PICO was:

Participants: Children up to age 16 years old having fixed appliance orthodontics for a Class II malocclusion requiring extraction of two first premolars.

Intervention: AcceleDent Aura used for 20 minutes per day

Control: Treatment as usual, no AcceleDent

Outcome: Change in mandibular arch perimeter and Irregularity Index in a 10 week period measured from study casts. The secondary outcome was discomfort and analgesic intake.

They carried out a sample size calculation based on a previous trial run by AcceleDent. Randomisation and allocation concealment and were good. They blinded the operator to the group assignment. Similarly, the model assessor did not know the treatment allocation and the time point of the models.

The statistical analysis was relevant to the data.

They analysed the data for all the patients and then repeated this for the patients who co-operated well, according to the timers in the AcceleDent devices.

What did they find?

They showed that there were no differences between the groups at the start of treatment.

After the 10 week period there was no difference in arch perimeter and irregularity index between the groups. In other words, in this study the AcceleDent did not work.  I have included some of the relevant data in this table.

AcceleDentAcceleDent ShamFixed onlyP
Mean and 95% CI (mm)1.09 (0.84-1.35)1.16 (0.75-1.58)1.00 (0.61-1.38)

They also found that there was no difference between the groups when they only included the good co-operators.

Finally, they did not find any differences in discomfort between the groups.  However, the group who had AcceleDent took less analgesics at 24 hours. 19 patients took analgesics in the control group compared to 12 in the AcceleDent group. As a result, the use of AcceleDent resulted in a reduction of 7 patients taking analgesics at 24 hours. I calculated the numbers needed to treat for this data. This showed that you would need to treat 3 patients with AcceleDent to avoid one patient needing to take analgesics at 24 hours. The confidence interval was large going from 2 to 9. This is not a great effect and my academic opinion is that this cannot justify the use of AcceleDent.

They did a really nice discussion of their results and drew parallels with the introduction and promotion of self-ligating brackets. They pointed out that the first studies that were carried out were retrospective and there was widespread adoption of this new technology. Yet, when trials were done we found that these benefits were not supported. This is not unusual with the introduction of new technology. Unfortunately, this appears to be the pattern with AcceleDent.

What did I think?

I thought that this was a very well carried out small trial in a ‘real world” setting. I think that it is very interesting that the results of this trial agree with other well carried out trials. It is also relevant to consider that these results do not support the results of a trial run by an AcceleDent consultant and published in a lower quality journal.

I also thought that this was a great example of an interesting and relevant study being done in an orthodontic clinic and not in a dental school. Peter proves that good research can be done in this way.

I am concerned that practitioners are promoting this technology in the absence of good evidence. This is because the AcceleDent appliance must cost the patient money. There is now increasing evidence that may cast doubt on the effectiveness of AcceleDent. In this respect, I wonder if orthodontists can keep promoting this product?  Finally, it would be great if some of the  AcceleDent Key Opinion Leaders could comment on this paper.

Related Posts

Have your say!

  1. It appears that no system can increase the speed of tooth movement beyond 1mm per day. However tooth movement in adults can be half that rate. Have there been any trials into increasing speed of movement in adults to match those of adolescents?

  2. A couple of years ago, Proffit at the AAO meeting talked about accelerating tooth movement using different things. I remember him saying that Acceledent did not even have a hypothesis as to why it worked but they still claimed it did. He was doubtful.

    Another disappointment. We all want these things to work.

  3. Hi Kevin:

    As mentioned before, I believe that the protocol used “for 20 minutes per day” is ineffective. Try 8hrs per day. Proffit’s book notes that any device needs to be in the mouth for that length of time to have a clinical effect. In the basic research done on rabbits, the application of the stimulus was well beyond 20 mins, if I’m not mistaken. Let’s have a discussion on the hypothesis of “sutural homeostasis” if we want to have some idea of how this “epigenetic orthodontic” procedure might work in children and in adults .

  4. It would be interesting to see a study where measurements were taken at shorter intervals….perhaps if the measurements had been at 3-4 weeks instead of 10, maybe the Acceledent group would be more aligned by that time (assuming the only measurements were done at baseline and 10 weeks, none in between). Like could patients progress through their first 3-4 archwires at a faster rate than normal? I don’t use or promote Acceledent, but I can see the argument that MAYBE you could progress archwires faster, although this would of course require seeing patients on shortened intervals during treatment. Hard to blind the operator and other issues I suppose as to if/when to progress to the next wire, etc, but I can see this being the push-back argument from promoters of the product. We just need a good study to find out…..

    • Hi Greg; In this study I also did impressions of the teeth at 5 weeks and 8 weeks as well as at 10 weeks and there was no difference at any time point. On the AJODO website there is a video I did accompanying the article which has graphs (not in the published version) which demonstrate this visually and the tables in the article state the measurements. I was blinded to the grouping of subjects which is actually fairly simple to do in this study design. I feel the question about a shorter time interval is therefore already answered in the paper. Vibration has an anabolic (osteoblastic) effect but no obvious hypothesis for a catabolic (osteoclastic) effect. Even if it did, to then precisely turn each of these on and off in the exact required locations in a tooth’s supporting tissues for accelerated dental movements seems unlikely. More significant ‘trauma’ of the bone/tissue would be required but this then shuts down after ~3 months of repair. While a few shiny chalices may appear in the marketplace, I feel we are still searching for the holy grail of accelerated movement.

  5. Article: Physical properties of root cementum: Part 25. Extent of root resorption after the application of light and heavy buccopalatal jiggling forces for 12 weeks: A microcomputed tomography study
    Edina Eross · Tamer Turk · Selma Elekdag-Turk · […] · M. Ali Darendeliler
    Full-text available · Article · Jun 2015 · American journal of orthodontics and dentofacial orthopedics: official publication of the American Association of Orthodontists, its constituent societies, and the American Board of Orthodontics

  6. Hi Peter: I agree that vibration has an anabolic (osteoblastic) effect. Please see the article below (1) altho’ we used a different vibrational protocol. The hypothesis for a catabolic (osteoclastic) effect could be osseous remodeling, according to temporo-spatial patterning (2). I agree that further studies are needed in this respect.
    1. Singh GD, Heit T, Preble D. Changes in 3D midfacial parameters after biomimetic oral appliance therapy in adults. J Ind Orthod Soc. 48(2), 104-108, 2014.
    2. Chaplain MAJ, Singh GD and McLachlan JC (Eds) “On Growth and Form: Spatio-Temporal Patterning in Biology”. John Wiley and Sons Publishers, England. ISBN: 0-471-98451-5 (1999).

  7. After reviewing the methods in this study, I honestly would not have expected the results to be any different than reported. I don’t think any acceleration device would have shown any clinical difference, since the biomechanics and arch wire sequence used was exactly the same in both groups. To see Accelerated tooth movement, the patient needs to refrain from the use of any NSAIDs, which was recommended but not measured or enforced. Example: All Pt’s who would have taken any NSAID’s would have been excluded from the group. In my clinical experience with accelerated treatment, it is important to adjust biomechanics, whether you are changing wires more frequently or switching aligners in a shorter time interval to see a reduction in treatment time or increased rate of tooth movement (which was conveniently not looked at in this study). In order to take advantage of the enhanced biological response (osseous modulation), one must change the treatment sequence to see any significant difference, which was not done. By maintaining a standard protocol without shortening the follow-up interval or changing wires within the evaluation period, one would not expect an increased expression of the prescribed mechanics. A small diameter wire like 014 Niti will not fully express the prescription of the bracket nor will it affect the arch width, thus what was measured did not account for the changes in vertical position of the teeth. The body responded to the biomechanics faster in the vibration group but was limited by its expression based on the limited mechanics. Thus both groups showed about the same change in Littles irregularity Index. Creating a clinical scenario that the only variable is the use of pulsatile forces would not by itself show an accelerated result. That is the misunderstanding of the study and for all those who think just adding a device will increase the rate of tooth movement or decrease treatment time. Osseous modulation in conjunction with smart biomechanics is where we see a reduction in treatment time. Even more important to me is increased predictability of the intended outcome in less time.

    Furthermore, I question why the investigators evaluated patients for only 10 weeks. A short time creates statistically significant results, but it’s not clinically significant, since orthodontics is never a 10 week process and we don’t end with a 014 Niti wire. Aside from their limitations in assessing orthodontic tooth movement, clinically significant differences in arch expansion and crowding are not to be expected between groups in such a short period, especially with a light 014 Niti wire. It may have been more appropriate to at least evaluate the effect of pulsatile forces on total treatment time vs. a 10 week time frame. Because of these limitations, I can’t say that this study disproves the effect of AcceleDent. I think this study was not set up to show the effect, but to show that there is no effect. The bias is inherent in the way the study was set up.

    • Hi David and thanks for your comments. In the study we did record all analgaesic use as well as type although not reported. Only 3 subjects used an NSAID and 2 of these were in the no appliance group. The one in the AcceleDent group used it at baseline and 5 weeks but thereafter used paracetamol. Such a small number of cases would not influence the outcome. There are a few choices to measure movement and the Texas study felt the most important was arch perimeter which is what the study was powered for. Irregularity is a common index and felt to be universally accepted and understood by both academics and clinicians and so it was used also. The protocol was to remove any subjective bias in changing wires faster which is what likely occurred in the original retrospective SL bracket studies demonstrating ‘faster’ treatment whereas later RCT’s found no difference. I disagree with your bias statement as not only were we blinded but the purpose of this design in having impressions at 5, 8 and 10 weeks would still demonstrate a difference earlier if there truly was one. An 0.014″ NiTi wire in an 0.018″ slot bracket is a routine wire and common clinical practice and even if movement was accelerated it would be noted on the models taken at 5 weeks which it wasn’t. The study has 3 parts and the 2nd part is on extraction space closure in the upper arch. I am currently collating the data for that and soon to commence writing that up.

    • “I can’t say that this study disproves the effect of AcceleDent.”

      I find this statement a little troubling since you cannot disprove something that has not been proven. I think that we would all admit that science is empirical dealing with nature as it exists. The number one criterion and standard of evaluation of scientific theory is evidence and ideally that evidence can be replicated. Hence, scientists prefer theories for which there is more and better evidence to theories for which there is less and worse evidence. In this particular case and at this time, is there evidence that Acceledent works as claimed? It appears that it does not, however, upon further investigation that could change.

      We now come to the ethics of the situation and that is left to each individual practitioners moral compass.

  8. Can we grow mandibles? Are extraction cases more stable long-term than non-extraction cases? Are lower third molars responsible for late lower anterior crowding? These are some of the questions that have been bouncing around for almost a century. I truly believe that a healthy debate within the scientific community does our beloved specialty a great service, and open the clinicians’ eyes and minds to ultimately better treat our patients. Almost half of the research in the literature would answer “Yes” to the above 3 questions, while the other half would give a resounding “No”.

    I welcome Kevin’s blog and enthusiasm, but I must respectfully admit that I see the article written by Miles & Fisher from a completely different prospective. Although the authors mention this study to be reflective of real-world practice, I don’t see its clinical applicability based on the choices made to evaluate AcceleDent. Baseline arch perimeter and mandibular arch irregularity do not really provide insight into the severity of the movement that may be involved in treatment, especially at only 10 weeks. It may have been more real-world clinically relevant if the investigators at least followed these patients for a longer period and assessed the overall time to alignment with AcceleDent.

    Additionally, I agree with Greg on taking measurements at shorter intervals as periods of increased orthodontic tooth movement during the follow-up could have been missed, especially since the study was only 10 weeks long. Simply put, the orthodontic tooth movement may have been “accelerated” and teeth moved sooner to their intended position, only to “sit and wait” for the next few weeks to be measured, thereby masking the acceleration. In addition, the shorter interval could have allowed clinicians to modify mechanics in both groups as needed since they were blinded anyway. In my experience, accelerated interventions maximize our ability to adjust standard mechanics protocol, and not allowing such an opportunity limits the clinical impact of the study.

    • Hi Sam. We took impressions at 5, 8 and 10 weeks and there was no difference at ANY time interval in either the irregularity or arch perimeter. A previous criticism of the study by Woodhouse et al. JDR 2015 by proponents of vibration was that they found no difference as they did not measure arch perimeter which is why I included this measurement. If there was a true acceleration in movement then both arch perimeter would increase faster and/or irregularity would reduce faster and this would have shown up in the 5 week models and it didn’t – there was no difference. I have also responded to David’s comments which may cover some of your concerns.

  9. Hi Sam and David, I am going to reply to your useful comments in one post. I am going to address these from the point of view of trials methodology with reference to orthodontic treatment. I am sure that Peter Miles, will be back to answer your specific questions about the study.

    Firstly, you are correct in that the most useful outcome from a orthodontic trial is whether total treatment time was influenced by AcceleDent and I guess that we will have to wait for these results. In the meantime, we only have short term outcomes. However, when we look at the results from the high quality studies, they are currently showing no effect. Even when we look at the results from the rather poor study that was carried out by Pavlin, the difference between AcceleDent and treatment as usual was very small. In fact, the statistical analysis showed that there was no difference, when you look at the 95% CIs. So this is where we are with the research. In summary, it may appear that AcceleDent does not have an effect.

    You both make the point that because AcceleDent allows us to change mechanics and work up through arch wires at shorter intervals and this should have been taken into account in the study. However, this would lead to a problem in that it would not be possible to split the effect of AcceleDent from the effect of different treatment mechanics. For example, I cannot help thinking that it is likely that I can speed up the treatment of my patients if I adjusted my aligning wires every two weeks. Or are you saying that to use AcceleDent effectively you need to see the patients at more frequent intervals? I cannot help feeling that we are then going to be asking our patients to pay for a device, use it once a day and then attend more frequently. This adds considerably to the burden and cost of care.

    Furthermore, in all the information that I have seen on AcceleDent it does not mention that we need to use different mechanics.

    You make the point that the study was set up to show that there was no difference in interventions. This is how most trials are set up and is standard trials methodology.

    I think that Peter is correct in drawing attention to the findings of the studies into self-ligation that showed that there was no advantages to this new technology.

    I hope that I have addressed your comments..

Leave a Reply

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