January 10, 2019

Headgear influences dental arch dimensions. Or does it?

I am under the impression that less headgear is being used as orthodontic treatment techniques evolve.  But are we discarding an effective treatment? This trial provides us with useful information.

I have not used headgear for ages. This is mostly because the children of inner-city Manchester were not “keen” on it.  Nevertheless, some operators in the UK and other countries use headgear as an essential part of their treatment.  This is particularly true of the Scandinavian countries (I think?). This is relevant in interceptive treatment. A team from Finland did this study.  The EJO published their paper.

Dental arch effects after early and later timed cervical headgear treatment—a randomised controlled trial

Johanna Julku et al

European Journal of Orthodontics, Advanced access. DOI 10.1093/ejo/cjy083

They have published several papers on this study and I have written about them previously when they reported on the effect of HG on the airway.

What did they ask?

They did the study to find out whether the timing of cervical pull headgear had an effect on dental arch dimensions. They also wanted to look at any influence of the gender of the participants.

What did they do?

They did an interesting randomised trial. They screened healthy 7-year-old children in school. To be included in the study the children had to have a Class II malocclusion based on molar relationship and an overjet of 6mm or more. They found that 67 children were eligible for the study and they all agreed to take part.

The PICO for the study was:

Participants: 7-year-old children with Class II malocclusion

Intervention: Cervical pull headgear with 500g force used for 8-10 hours per day.

Control: No treatment and then Headgear when they were older.

Outcome: Upper and lower arch measurements.

In effect, this study was similar to the other Class II studies where an intervention (HG) was either provided early or at a later age. They randomly allocated the children to early headgear treatment at aged 7 or HG 18 months later.

They collected data from study casts at the following stages of treatment.

T0:       At the start of the study

T1:       After 18 months of HG treatment or observation

T2:       After 24 months of treatment for the later treatment group.

I was not sure if the “earlier” treatment group had any treatment between T1 and T2?

They made several measures of dental dimensions on the dental casts.

Their power calculation was not too explicit as they did not state what outcome measure that they based this on. This suggested that the sample size should be 11 in each group. However, they enrolled three times this number of participants. This could result in a study being overpowered and lead to statistically significant findings when the effect size is small (see later).

What did they find?

67 participants started the study. After five years, 56 completed the study.  The groups remained balanced, and this should not lead to bias.

So that I can be brief, I have decided only to report and discuss two of the measures that they made from the casts. These are the distance from the canine tip to the distobuccal cusp of the first molar and the distance between the lingual/palatal surfaces of the first molars. If you want to look at the paper, there are many more measurements for you to look at.

This table contains the relevant data at the end of the study for the dental measurements in mm with 95% confidence intervals.


Maxilla Mandible
Canine to Molar Intermolar Canine to Molar Intermolar
Early Group 27.7 (27.2-28.2) 53.6 (52.5-54.7) 27.8 (27.2-28.3) 33.1 (32.2-33.9)
Change from start 1.3 2.4 -1.2 1.1
Later Group 27.5 (26.8-28.2) 52.3 (51.2-53.3) 25.9 (25.1-28.3) 31.5 (30.4-32.5)
Change from start 0.7 1.8 -1.7 0.3


This data is tricky to interpret. But the best way is to look at the effect sizes and decide if you feel that they are clinically significant. Then look at the 95% confidence intervals. There are two things to look for here. Firstly, if they overlap there is unlikely to be a statistically significant difference. Then look at the range of values if these are wide then there is a degree of uncertainty in the data.  I did this, and I felt that the effect of the treatment was small and not clinically significant. The CIs were tight, and so there is some certainty in the data.

When I looked at all the data, there were several instances of small effect sizes being statistically significant and this may be a result of the high power in this study.

Their overall conclusions were:

  1. Cervical Headgear treatment was effective in widening the maxillary dental arch
  2. There was a natural expansion of the lower arch.
What did I think?

This is an excellent trial that and was carried out well. However, when I look at the effect sizes, I am concerned that these are rather small. As a result, I am not sure about their conclusions. But I will leave this for you to decide.

I have reviewed this trial before, and it was difficult to build a total picture of the effects of the headgear treatment because they have reported this study the in several papers.  I have looked back at these.  The most comprehensive report they published was on the impact of the headgear on the skeletal pattern and the airway. I felt that this showed that there were minimal effects of the headgear on the skeletal pattern, with an effect size of 1.2 degrees on SNA. Furthermore, they also showed that there was no effect on the airway.

I cannot help thinking that when we consider all the results of this study, then we can conclude that cervical headgear therapy does not do too much? This is the real value of this study.

However, this may be the cynic in me and my bias against headgear? Maybe the inner city children of Manchester are right?










Related Posts

Have your say!

  1. I have two issues with the methodology -correct me if I am wrong-;
    1st, using the canine cusp to first molar DB cusp measurement for an outcome measure of arch length, I am not so much with it. Most of the times when you distalize either using extra oral or skeletal distalizers you don’t end with sole movement of the 1st molar but full distalisation of the Vic always segment in addition to the incisors and OJ reduction, most probably due to the trasdental fibers effect lighting the teeth together. So, I believe that using this measurement while neglecting the more anterior effect may underestimate the effects on the arch length.
    2nd, fixing the outcome measurement times at 12 months for the EG and 24 months for the LG, is there any evidence behind this? Did they report that all participants completed the trial did fullfilled molar correction into Cl I? Does this mean that they may have distalised beyond or less than Cl I molar relationship?!! Wouldn’t it have been better to do the measurements based on the molar relationship correction and report the average treatment times in EG and LG as a secondary outcome? I do believe that this would have been more logic and more patient-centered and clinically-relevant approach?

  2. I imagine your inner city Mancunian kids are no different to most kids world wide !Hard to believe that anyone still uses headgear.There are just so many means of achieving the same or better results with much greater compliance eg.Crossbow in major class two cases .

  3. The old joke, “I use headgear but my patients don’t”.

  4. Is there any published unbiased RCT that compares compliance with a removeable appliance such as a twin block with a headgear over a period of say 18months?
    In the not too distant past.

    It must be possible to accurately measure compliance with digital systems that are available.

Leave a Reply

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