Is there a link between extracting premolars and breathing disorders?
I will start my blog posts this year with a post on breathing disorders. I suspect this will be the first of many, on this subject, this year. There appears to be an increasing number of research projects in this area. This new publication in a high-impact journal sets a standard for the quality of research. It provides us with a lot of useful information.
As regular readers will know, there is considerable controversy concerning the role of orthodontic treatment in causing or curing breathing disorders. Little evidence supports these concepts. Unfortunately, there appears to be an increasing trend toward promoting this tenuous association despite the absence of evidence.
One of the promoted theories is that orthodontic treatment involving the extraction of four first premolars causes breathing disorders. The source of this is an erroneous belief. People mistakenly think extraction treatment involves retraction of the teeth. They believe this retraction then leads to breathing disorders.
A study team from West Virginia did this interesting study. Community Dentistry and Oral Epidemiology published the paper.
Sleep disruption and premolar absence, NHANES, 2017–2020: A cross-sectional study
R. Constance Wiener, Michael Hnat, Peter Ngan
Community Dent Oral Epidemiol. 2024;52:809–816. DOI: 10.1111/cdoe.12983
What did they ask?
They conducted this study to
“Determine whether the absence of first premolars is related to the prevalence of sleep-disordered breathing in adults aged 18 to 65 years”.
What did they do?
They conducted a cross-sectional study using data from the National Health and Nutrition Examination Survey (NHANES), carried out from January 2017 to March 2020. This survey included approximately 5,000 US residents. Further information can be found here.
I have also pasted a description of the study directly from the website here.
Each year, about 5,000 adults and children in communities across the United States participate in NHANES. They use a random, scientific process to select the people they invite to participate. This process ensures that this group of people can accurately represent the health and nutritional status of everyone in our diverse nation. To collect data about a wide range of health and nutrition topics, NHANES includes—
- Interviews about health, diet, and personal, social, and economic characteristics
- Visits to a mobile exam centre for dental exams and health and body measurements
- Laboratory tests by highly trained medical professionals
All the participants were aged between 18 and 65.
The presence or absence of premolars was the primary explanatory variable.
They also gathered sleep variables from questionnaires. The most meaningful of these measures was the modified summative sleep disturbance score. A score of 1 had zero indicators of self-reported sleep disturbed breathing, a score of 2 had one indicator and a score of 3 had 2-4 indicators.
Finally, they collected basic demographic information on each participant.
They analysed the data with logistic regression and the main outcome variable was the summary score.
What did they find?
They included data on 4742 participants, 51% of whom were female. Of these, 5.4% had no first pre-molars. When they examined the sleep data, 28% reported trouble sleeping.
They provided a large amount of data and I am only going to outline the sleep disturbance score for premolar extractions. The analysis revealed that when 4 premolars were missing 33% of the participants had a score of 1, 40% had a score of 2 and 26.3% were in group 3.
When they analysed the data, they found no significant association between the sleep scale and the absence of first premolars. Yet, they identified an association with sex, age, body mass index, education, and chronic disease.
The final conclusion is
“Concerns about the impact of first premolar extractions on sleep disorder breathing were not supported in this study”.
What did I think?
I want to congratulate the authors on conducting such an excellent study. Community Dentistry and Oral Epidemiology may not be familiar to readers of this blog. However, the journal is a high-impact dental journal. Getting an orthodontic paper published in this journal is a significant accomplishment.
The authors used data from a very large, well-respected study. They examined data from a very large sample using established measures.
Interestingly, only a small proportion of the participants had four first molars removed. This may reflect the relatively low uptake of orthodontic treatment in the past compared to now.
Some might criticise the authors for assuming that the teeth have been extracted for orthodontic purposes. Nevertheless, I feel that this assumption is logical. It is difficult to consider other reasons for the loss of these teeth.
Another criticism is that the team did not measure SDB with any quantitative measure. Nevertheless, they evaluated the participant’s experiences of sleep and breathing problems. In my view, this is completely correct as it evaluates patient values, rather than laboratory tests, for example, polysomnography.
When I examined the results, I found them reassuring, and I recognise that they align with other studies. They also seem logical, as well-established factors, like BMI, gender, and chronic illness, are linked to sleep-disordered breathing. Furthermore, the findings correspond with the AAO white paper on sleep-disordered breathing.
Finally, this study is well-conducted. It supports the concept that there is no association between orthodontic extractions and breathing disorders. Airway-focused orthodontists should read this paper and join in the discussion?
Emeritus Professor of Orthodontics, University of Manchester, UK.
I know this orthodontic group but I’d hardly describe it as ‘such an excellent study’, respectfully. The primary reason for this is that no sleep studies were undertaken, which are regarded as the first option in the diagnosis of OSA by physicians – otherwise, many dentists would jump on the ‘airway wagon’.
I take your point. However, I cannot help feeling that self reported symptoms are as important as any lab tests. This is because we need to measure the effects of a “disorder” on people and not solely rely on quantiative measures. This is one of the great problems with orthodontic research and its reliance on numbers.
Understood, but the corollary of the conjecture doesn’t pass muster. For example, hypertension can have few symptoms till a catastrophic event, such as a heart attack or stroke, occurs –
Yes, I agree, but I’m not sure that this is a good example. While breathing disorders are serious, when you are looking for a cause and effect it is completely relevant to measure clear symptoms rather than “hidden signs” for practical reasons. It is not possible to do sleep studies on samples this large.
Approx. 50% of patients with OSA are diagnosed with hypertension, while approx. 80% with resistant hypertension have OSA. The study didn’t look at endotypes of OSA and exposomes weren’t taken into consideration. In addition, variables were self-reported, subject to recall bias, and social desirability bias – but it’s a good start in the right direction.
With all due respect, in a study of this size then I think self reporting is an adequate way of screening participants for a condition that has strong symptomatic characteristics. Sleep studies would not be done for non-symptomatic people anyway so that would be more akin to a ‘fishing expedition’ and would skew the study when compared to others.
Thank you for all of your work! Looking forward to more in 2025. I see you’ve previously posted about some good journals to seek research articles, but what makes a research journal high-impact or high-quality? Is it that the articles are peer-reviewed, a history of the journal, the quality of the studies themselves, the number of articles published, etc.? How can we reliably evaluate how valid journals are especially for journals outside of dentistry/orthodontics?
Thanks for the question. Research journals are measured by their impact on their field. This is mostly based around the number of citations of papers. Therefore, a journal that publishes high quality studies will get high citations and therefore a high impact factor. As a result, this predicates on the quality of the papers, the peer review process and reputation of the journal. The impact factor of the journal is often listed on the website and this can give an indication of the journal. For example, the New England Journal of Medicine which is a leading journal has an impact factor of 96. But this is also influenced by the field of research. Dental journals have a low impact facto; the angle orthodontist has an impact factor of 3.0. Community Dentistry Oral Epidemiology impact factor is 2.3. But these are high for the field. It is also really difficult to get into CDOE because their editorial standards are v high. So you can see it is not straightforward!
Good afternoon Sir,
Thank you for highlighting this study.
Anatomically, considering the loss of two premolars in an arch as having an impact on the airway is similar to thinking that Orthodontics can fix TMD or malocclusion is the sole cause of TMD.
Scientific temper and common sense needs to prevail whenever we are talking about complex development processes like breathing.
The next challenge for Orthodontics (with most of it being practiced in a primary care environment) should be refining our retention protocol to avoid retreatments. Sleep apnoea can be managed well in Secondary Care with Multi Disciplinary teams.
Thank you.
Yours sincerely,
Karun Sagar
Orthodontic Assistant
Thank you for this interesting study and your analysis of it. I find this topic to be fascinating, and as mentioned, neither viewpoint supported by solid peer-reviewed evidence in the public square.
You did discuss my greatest concern with this article: Self-reporting SBD signs and symptoms. That would seem to be the lowest bar possible for actual data. Again, no ax to grind here, but this doesn’t stand up as really anything more than “I think I don’t snore”.
With gratitude for all you do,
Dr. Brian Gray
Thanks and yes there may be issues with the outcome measures used here. However, they do have some validity as they reflect patient perceptions. In this respect they are valuable and realistic. This is better evidence than that obtained from a referred population that has a sleep study. I cannot think of a better method to answer this question.
The limitations of sleep questionnaires make them incapable of properly examining the relationship between orthodontic extractions and disordered breathing.
It doesn’t make any sense to keep repeating these studies with flawed methodology.
It makes even less sense to promote these studies as well done and providing meaningful answers.
As David Singh pointed out, this study is flawed. If you are looking at a link between these two issues, Sleep Disordered Breathing and extraction of Pre-molars, you need a diagnosis of each condition to look further. In the medical community, a diagnosis of Sleep Disordered Breathing is required in order to use the descriptor. 80% of Obstructive Sleep Apnea is undiagnosed. A patient’s lack of recognition of the condition doesn’t mean it isn’t there. That is an unreliable basis for the study.
I take your point. However, it is not possible to screen a large population with sleep studies. Therefore, it is perfectly reasonable to use self reports. Importantly, these may be more important to people than studies for a “disease” that we need to go looking for? You stated that 80% of OSA is undiagnosed. I see this figure stated commonly. But I cannot find a reference to the study that shows this. Can you or Dave Singh let me have this reference?
😔”Which came first, the chicken or the egg?
Breathing disorders (low tongue posture-> narrow palate->crowding-> extraction needed)?
Or opposite way?
Thanks or alternatively, the breathing disorders and malocclusion were not associated with each other?