February 27, 2017

Another nail in the coffin of the non-extraction obsession: a new study

Do extractions influence facial aesthetics? A long-term study.

In some of my previous posts I have discussed the effect of extraction/non-extraction on facial profile. However, I managed to miss this paper that was published last year. I feel that it certainly provides us with useful information.

As we all know, the extraction/non-extraction debate continues to rage. There are strong views on either side. While some are extreme and even verge on scaremongering. Others are more reasonable and point out that it is not the extractions that may cause damage to facial profiles. It is much more likely to be a consequence of poor treatment mechanics.

This paper was published in the EJO in 2015

crowdInfluence of premolar extractions on long-term adult facial aesthetics and apparent age

Guilherme Janson et al

Eur J Orthod (2016) 38 (3): 272-280  doi:10.1093/ejo/cjv039



The authors did this study to find out if there is any effect of extractions on facial aesthetics and perceived age of patients.

What did they do?

They took a sample of 63 treated patients who had the following characteristics:

  • Class II division 1 malocclusion
  • Their age was greater than 21 years old
  • Availability of records
  • Frontal and lateral extra-oral photographs taken a minimum of eight years after their treatment.

They divided them up into three groups:

  • Non-extraction treatment
  • Two maxillary premolar extraction
  • Four premolar extraction.

They had treated the patients with  edgewise fixed appliances using a standard set of mechanics.

Interestingly, they used headgear to reinforce anchorage in the non-extraction and four premolar extraction groups. They used class II elastics when necessary.

They recorded the following data;

  • Post-treatment occlusal outcome with the HBO system
  • Cephalometric data (a large amount)
  • Post-treatment lateral and frontal extra-oral radiographs

They standardised the images and showed them to a panel of raters. This group used a 10 point scale to record facial attractiveness and their perception of the patient’s apparent age. 76 orthodontists and 83 lay people made up the rating panel.

They carried out an appropriate statistical analysis and sample size calculation.

What did they find?

They presented a large amount of data. I certainly did not look too closely at the cephalometric festival of multiple variable testing!

I thought that the most important and relevant finding was that there were no clinically significant difference in any of the morphological measurements between the groups at the post-treatment stage.

Importantly, when they looked at facial aesthetics and apparent age, 8 years after treatment, there were no differences.

What did I think?

I will start my summary by considering the factors that may lead to bias in this study. I felt that these were;

  • This was a retrospective study based on the availability of records. It must, therefore, have selection bias.
  • They did not randomly allocate the treatment decision.  This could lead to operator bias.
  • The raters recorded their impressions from still life photographs. This does not necessarily  represent real-life social interaction.

I have thought about the effect of these factors and I think that the most important is the selection bias because of the retrospective data collection.  This means that we need to assume that they may have collected the records of the “best” treatments.  As a result, this data may only be relevant to high quality treatment.   We can only conclude that they found no difference between extraction and non-extraction treatments when high quality care was provided.  I can also interpret this as meaning that it is not the extraction decision that may cause facial damage; it is more likely to be the choice of mechanics.  I cannot help thinking that this type of treatment is not done by many orthodontists with the increased use of functional appliances and improved methods of reinforcing anchorage.

Nevertheless, if we consider that it is not possible to randomise the extraction/non-extraction decision, then their methods are likely to be “as good as we can get”.

I have thought about their conclusion and we need to remember that “absence of evidence”  does not mean “evidence of absence”. It simply means that they did not find any evidence of harm from extraction treatment.

My conclusion

When I put all this information together, I cannot help feeling that this paper reinforces information from other studies that I’ve posted about before.

My overall conclusion from the literature and this paper is;

“We have not found any evidence that extracting teeth as part of the course of orthodontic treatment causes any harm to the facial aesthetics of our patients”.

I think that I will carry on amputating body parts (teeth) for patients in which I feel extractions are necessary.

Finally, I wonder if it is almost time to stop discussing orthodontic extractions?

Related Posts

Have your say!

  1. If was not a practicing orthodontist who had to deal with real world factors like patient desires and general dentist input then I might think the same way you do. However, if there is “no difference” between extraction and non- extraction outcomes” but patients are generally against tooth removal (as are referring dentists) then doesn’t this study put all the nails in the coffin of extraction mixed orthodontists and bury them? Are you suggesting that “quality treatment” involves headgear in extraction and non-extraction cases? It sure seems like it and I can tell you that won’t fly outside academia or a country where paternalism is still a thing.
    I find it dangerous when academics hold out cephalometic evaluations of 63 cherry picked patients as “proof”, especially when that orthodontists has as much visibility as you do. Keep up the good work but please don’t forget that there are people attached to those teeth and those cephs represent a human being with a multitude of desires and varying levels of compliance. The orthodontist in private pracrice can’t retrospectively pick our patients! 🙂

    • HI Ben, thanks for the comments. I really agree with what you say. But maybe I did not make myself clear. I was suggesting headgear was quality treatment, I was trying to say that it was not the extractions that has the potential to cause harm, it is more likely the choice of treatment mechanics or poor treatment mechanics. I maybe an academic but I did treat patients, until 3 weeks ago when I retired from clinical practice. I, therefore, agree with you that I would always try to avoid removing teeth, as this is something that neither I, my patients or referring practitioners would like. But on the occasions when it was necessary I was not worried about deleterious profile change, because I managed my mechanics.

      I suppose the message that I was trying to get over from this paper was that while it is good to avoid extractions, it is not good to try and treat all patients non-extraction by over expanding and proclining when this may also cause harm and spoil facial appearance. As usual, we need to seek a balance in our decisions etc..

      • Extraction/non-extraction is not a treatment goal, only a treatment modality. To satisfy the goals of facial balance, occlusion, TMJ and periodontal health, extractions are frequently indicated. Inadequate diagnosis and improper mechanics should not be conflated as a flaw of the treatment modality. Although I generally agree with the conclusion of this study (and Dr. Vaden’s paper) the selection criteria (full-cusp class II) for this study might be overly broad, and not sufficiently delineating of the several phenotypes one observes. As an example, if a class II malocclusion presents with protrusive incisors, extractions might be an optimal modality whereas if maxillary incisors are well-positioned, and the mandible is retrusive, a different modality might be called for. Thus, it would really depend on the totality of the diagnosis and not just the molar relationship.
        As the experts, patients look to us to recommend the best available option. If we as a specialty eschew that role and endorse unscientific positions based on specious reasoning, is it a wonder that we are now beset by DIY ortho, Teleortho etc.? What patient would knowingly desire lip incompetence (and drooling), protrusion, poor facial balance, uncoupled incisors, teeth positioned outside the alveolus, and the other litany of undesirable outcomes? I don’t know of any. Patients want to look better, and not worse than when they started. All too frequently (with a non-extraction at all costs philosophy) they end up looking worse! If patient desires are detrimental to occlusion, stability, periodontal health etc. then there’s always the option of refusing to provide such treatment.

    • I completely agree with Ben Burris. I had a orthodontics professor at NYU who used to preach ” when in doubt take four out” which I always thought was the most absurd thing to say to your graduate students!

  2. Kevin– I have read with interest your posts on almost everything in orthodontics. You are doing a great service to our specialty because you are level headed!! I have read , with great interest, your posts on extraction/nonextraction. Yes–it is important to me because so many patients are being harmed by this nonextraction mentality. Not everyone needs extraction, but if they do, we must do it.

    I furnished the long term treatment results for a study that was published in AJO/DO. You might be interested in it. We took a different tact in that we compared long term post extraction faces to a non treated sample. T2 faces had to match the nontreated faces at the outset of the study. Article is:Extraction vs no treatment:Long-term facial profile changes;AJO/DO;vol 145;#5;May2015;pp596-603.

    Thank you for all you do for our specialty. Sorry we will not have you with us in San Diego. It is our loss. Jimmy

  3. This subject should have been put to rest in 1947 with Hayes Nance.

  4. Thank you for another piece of very useful information. I enjoy your blog and find it to be most helpful and practical.

  5. Kevin there are many many studies , I’m sure you are aware that repeatedly state that the lip and mid face flattens as the upper incisors and canine retract.
    (Even the one you posted last a few weeks ago said so !)
    The canine eminence is also lost as the canine routinely falls back 3-4 mm. This is not picked up by papers as cephalomtric radiographs do not measure or assess this important facial landmark .
    It is impossible to remove two upper premolars 8mm each and not have some flattening of the lip as the canines retract. You may argue it is clinically insignificant and you feel it is an acceptable compromise , but I am struggling to agree with you headline title.
    I still think lower arch extractions are required at times , but still not as often as is currently practiced in the NHS.
    Less than 3 % of cases in the caucasion population require upper teeth extracting . Current NHS extraction rates are still at a staggeringly high 40%
    It is inappropriate to extract for severe crowding as it is easily resolved with expansion and PSL brackets giving the face increased lower volume, producing a far superior FACIAL result . It is routine at my practice . (And many others)
    To extract or not to extract should be based on the lower facial profile , NOT the severity of crowding . If the lower face is low in volume then expand expand expand , followed by psl brackets to unravel crowding .
    It works beauifully without moving teeth out of bone if not done too quickly . I have been doing it 25 years .
    The tmj improves as late clicks become early , and early clicks disappear as they recapture the disc , with the natural increase in vertical .
    Sorry to disagree with you but I see more attractive faces , better tmj results .
    To remove 16mm of teeth from an upper arch is incorrect in the vast majority of crowding cases . The upper teeth do not end up overly proclined if you successfully expand to create width.
    You are still to be commended for raising the issue for discussion , and nailing your colours to the mast , As I am happy to respond . I do think there needs to be a shift and more research so we reach consensus.
    The exsisting PSL research you have quoted is laugable as they extracted teeth !
    Best Regards

    • Most of the literature states that PSL doesn’t really make a difference except for a marginal increase in palatal width (I say this as a damon user!). Though I agree with you that extractions are too often used as a solution to crowding, I think underutilization is just as dangerous. Are there no situations for bilateral fullstep CL II patients to just get two upper premolars extracted? What about overly protracted mandibular teeth? The beauty of the profession is that there isn’t one solution and using expansion for everything seems devoid of critical thinking and lazy.

      I definitely disagree with some of Kevin’s conclusions but they are founded in research and evidence. Your evidence seems to be case based, especially some of this TMJ stuff like recapturing the disc. It makes it difficult to take your post seriously.

  6. Kevin,
    You say the following in conclusion…..
    “When I put all this information together, I cannot help feeling that this paper reinforces information from other studies that I’ve posted about before.
    My overall conclusion from the literature and this paper is;
    “We have not found any evidence that extracting teeth as part of the course of orthodontic treatment causes any harm to the facial aesthetics of our patients”.
    I think that I will carry on amputating body parts (teeth) for patients in which I feel extractions are necessary.
    Finally, I wonder if it is almost time to stop discussing orthodontic extractions?”

    The discussion you are referring to is centered around ‘facial aesthetics’.
    The HUGE conflict here is not all about the effect that extraction has on facial aesthetics but the effect that the ‘amputation of body parts’ and the forcible repositioning of the jaws has on airway and breathing integrity.
    Let there be two camps in the aesthetics field – no problem with that. The ‘evidence’ however that the removal of 4 bicuspids and the removal/impaction of 4 third molars – amounting to the compromise of 25% of the teeth – simply cannot be ignored.
    As an ‘outsider’ I can see orthodontics splitting into two disciplines.
    One is Aesthetic Orthodontics where the visual outcome is the most important and scant attention is paid to the down-the-line consequences of the mechanical intrusion and forcible retention.
    The other is Healthy Orthodontics where the focus is on optimal function and cranial balance with an aesthetic outcome which is the best possible given the starting position.

  7. In this article he has not mentioned anything regarding space requirements.. for example in a crowded cases entire space might be used by extraction spaces . We can observe the soft tissue changes only when we use space for retraction of incisor. This is my opinion regards the article published sir. Plz do correct me if I am wrong sir

    • This sort of study is easily misinterpreted. Given proper planning and execution, the results of the two approaches will converge. Both extraction and non-extraction, however, can either help or hurt depending on initial crowding and protrusion. Susceptibility bias must be considered whenever we consider studies in which the usual suspects are rounded up years after treatment

  8. Prof this study is not really applicable to the extraction non extraction debate as typically extractions in a IIdivI case is likely to enhance dentist facial aesthetics in a protrusive patients and where a patient requires treatment to mask skeletal discrepancy. But the debate is irrelevant as appropriate case selection and mechanics is more relevant to this outcome

  9. thanks for the article.

    I’m interested in a couple of points –
    problems are likely to be caused by choice of mechanics than extractions – what mechanics had you in mind?
    you will continue to prescribe extractions where you feel it necessary – presumably most of us will do things we think are necessary and not do things that we think are unnecessary. The question then becomes, do we have objective criteria for when extractions are necessary (or helpful, if the definition of necessary is particularly operator-dependent)

  10. I think one other thing that isn’t always clear in this extraction V non-extraction debate is: what were the extractions supposed to achieve? Make space to relieve crowding, correct a centreline or correct an overjet?

    If it’s meant to allow correction of crowding, then the amount of crowding could absorb some of the extraction space without much of an effect on lip support.

    If it’s meant to allow incisor retraction, then it’s hard to believe it won’t have some effect on lip support. It may be an amount that we would tolerate, but if it’s a dental solution to a skeletal problem, and that problem is a small/retruded mandible….I think there’s going to be trouble.

    Time for curry.

    • I would say that the extractions are mostly necessary for the relief of severe crowding. Certainly, in the UK most Class II malocclusions are treated with functional appliances and therefore reduce the need for extractions

  11. Certainly nice debate. Not all extractions are done for the same reasons. Hence careful selection of extraction or non-extraction cases is a cornerstone piece in this debate. If not then we will defuse any meaningful difference, if indeed exists. At the end how that space was used or not is the point of the mechanics.

  12. “Finally, I wonder if it is almost time to stop discussing orthodontic extractions?” Nice try. Won’t happen in our lifetimes.

    Kevin, I think this discussion will only stop when we’ve learned ways to predictably normalize rest oral posture and (thereby) optimize forward growth of the jaws. Then there won’t be a perceived need to force the dentition to fit into a box that is too small. Do you see that happening any time soon?

    As others have commented, health really will trump aesthetics in this “debate.” (Although the aesthetic question is far from settled. “We have not found any evidence…” means just that: they haven’t found it *yet*. You know that.) When patients are aware of the direct connection between tongue space and the health of the airway, they will never consent to treatments which potentially decrease that space.

    Brian Hockel

  13. Kevin,
    The comments clearly show that the debate will continue, but I would like to suggest that there is something bigger at play here.
    I believe that this debate is just the orthodontic manifestation of what behavioral psychologists and decision theorists call “Loss Aversion”. (Definition below) It turns out it is very hard for human beings to “lose” something, anything actually, once they possess it, teeth included. Anyone interested in looking into this in detail would be well served to read “Thinking Fast and Slow” by Daniel Kaneman. I think that Loss Aversion is an important concept to understand and teach in orthodontic graduate programs because it is a very powerful undercurrent in this debate and shapes the feeling of our patients and our referring dentists.
    I believe Loss Aversion is also the basis of the seeming need to constantly invent new reasons that extraction treatment is “bad” for people. The desire to avoid loss is a powerful intuitive “gut feeling” that makes some orthodontists (and dentists) believe that “tearing out 16mm of perfectly good tooth structure” must be bad so as soon as one myth is debunked, another is created. “Restricting the Airway” is the current culprit of the hour, even though there are several studies showing no connection between apnea and extractions.
    The debate will continue because orthodontists continue to choose gut feelings and anecdotal observations over objective data. It is up to the profession and those like Kevin here to elevate the debate to objective, high quality data to make our points knowing that our patients trust us to base our decisions on more than our gut feelings.

    Definition of Loss Aversion: “In economics and decision theory, loss aversion refers to people’s tendency to prefer avoiding losses to acquiring equivalent gains: it’s better to not lose $5 than to find $5. Some studies have suggested that losses are twice as powerful, psychologically, as gains.” (From Wikipeda)

  14. Dear Dr. Kevin O’brien, thanks for the useful information you provide strengthening evidence based Orthodontics. By this post I may understand that whenever the face demands extraction we should extract. Like in Bimax or severe crowding conditions. Please correct if my interpretation if needed.
    Thanks and Regards.

  15. These studies are of little use because, as pointed out in the blog, the treatment decisions are not, and cannot, be randomised. The only result is that competent orthodontic decisions result in good results; ie if patients need extractions and have extractions they get a satisfactory result and vice versa. What we need is a study of so called borderline cases,some of whom were treated with extractions and others not.This would show any real effects of unnecessary extractions.
    However, I consider this debate unnecessary if , where possible, borderline extraction cases, 40% of my practice, are treated with extraction of second molars which enables the crowding to be relieved without having an effect on the profile or lip anatomy.

  16. Hi Kevin:

    Keep stirring that pot 🙂 But seriously, one of the selection criteria was “proclined maxillary incisors”. When I looked at the stats (briefly), none of the groups had proclined maxillary incisors, according to the nasolabial angle at the outset. It seems that the midface was retruded in all three groups. Here is the obfuscation of facial recognition and facial perception. Long story – so I won’t go into it here. Suffice to say the authors state, ” Laypeople attributed significantly smaller scores for facial aesthetics and greater apparent ages than orthodontists”. Is this relevant finding camouflaged by (flawed) statistics? One could argue the merits of multivariate analysis here – but I won’t. Pertinently, you can’t really measure a football with a ruler. Size-change masks shape-change and without the application of robust geometric morphometrics, pseudo-2D parameters derived from 2D analyses of an entity that does not exist in either true 2D or 3D space (i.e. lateral photographs) are of limited value. Bottom line; Beauty is in the eye of the beholder. Keep up the good work!

  17. Hi Kevin.

    Thank you very much for the info and the article. I have some major issues with this article which I think limits its clinical application significantly. Most importantly, as mentioned by others and yourself, each single patient has their own diagnosis, needs and treatment plan and should not be treated as a single statistic. I think most of us agree these days that extractions are necessary when they’re absolutely necessary, but to simply say they do not affect facial profiles is a dangerous and incorrect statement IMO. I’ve seen plenty of cases where extractions have destroyed esthetics, not just in profile but also by poor angulation of U incisors, narrow arches and buccal corridors. Here are my issues with this article: 1. As mentioned in the article, we too often live in the profile view and lip protrusion, and evaluating esthetics from frontal view are more important IMO (speaking as some one who had U/L4s out, doesn’t have retrusive lips, but has buccal corridors the size of cucumbers). However, I don’t think anyone would suggest that extractions can affect the anterior esthetics when patient is not smiling, which is how they evaluated the frontal photographs. I understand that individual variations could cause issues with comparing smiling photos (where the real difference in arch width, buccal corridors and angulations would lie) but this on its own creates a huge question mark on the results of the study. What about buccal coridors? wider smiles? incisal display and anuglation? 2. The patients in the study were are all “phenotypically mediterranean”. These patients tend to have more protrusive lips to begin with, which don’t tend to retract as far with extractions compared to the Caucasian tight-lip retrusive lip patients that led to the non-extraction Damon movement. Again, limiting the results to that ethnic background if at all applicable. 3. The cases were retrospectively evaluated, which as you mentioned can have bias. More importantly, “The three groups had to present similar occlusal and cephalometric orthodontic treatment outcomes.” which means that likely they all started at vastly different situations but with the same treatment goal, at least AP-wise (no transverse evaluation). This basically excludes any mess-ups of cases that were dumped too far back due to extractions. As an overall, I don’t think this study adds a nail in any coffin really. A simple statement like “extraction doesn’t cause harm to facial esthetics” is a dangerous one because it takes out the real risk factor of causing esthetic harm due to extractions from the treatment planning process of a single individual which may be ways away from the “mean”, on the wrong side.

  18. How I see it.

    1. “Facial Aesthetics” in this context probably means the amount of lip support – protrusion or retrusion – and to lesser extent buccal support of cheeks; and increase or decrease of lower anterior face height.

    2. It is generally accepted that orthodontic treatment can deliberately or accidentally make some changes to “Facial Aesthetics” described above.

    3. Probably the major factor in these changes is the mechanics chosen – deliberately or accidentally ( very simple example of such differing mechanics, – retraction of upper labial segment on round wire or alternatively on rectangular wire).

    4. Extraction or non extraction per se has only minimal effect on “Facial Aesthetics”.

    5.In deciding on extraction or non extraction therefore, “Facial Aesthetics” should not be the overriding influence.

    6. The consideration of other factors; Dental Aesthetics; Dental Occlusion; Dental Health (caries situation); Patient wishes (avoiding extractions); Differing levels of stability of treatment result; Are probably more important factors to take into account when deciding to extract or not.

    Anybody agree ?

  19. just a question from a lay person here (sorry!) on the point about the choice of treatment mechanics or poor mechanics – these being the driver of facial changes, and extraction having no effect – what dies ‘choice’ and ‘poor’ mean here ? are there lots of options that constitute choice ? and it is easy to find evidence of poor technique ?

    are choice and poor technique reversable?

    Thanks – really enjoyed the evidence based reads


Leave a Reply

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