The most important finding of this study is that a significant difference in the stable and unstable groups was noted in the horizontal distance between the maxillary and mandibular molars from mounted casts. The unstable cases occurred when the horizontal distance was more than 0.5 mm, and the percentage of stable cases when this distance exceeded 3.5 mm was 0. This result means that the occasional case that present with a class III molar relation due to hypo-growth of the maxilla must become stable. Measuring the horizontal distance between the maxillary and mandibular molars in CR may be helpful for predicting the results of class III early treatment in daily clinical practice, even at the chairside.
Although the number of cases in the stable group with anteroposterior functional shift was greater than the number of cases in the unstable group without functional shift, there was no significant difference. Anteroposterior functional shift has been thought to be a positive factor in early class III treatment . Most orthodontists believe that pseudo-class III cases are easier to treat than class III cases without anteroposterior functional shift; however, our findings call this belief into question.
In most previous studies [3,4,5,6,7,8], cephalometric factors alone were evaluated to predict the stability of early class III treatment. Possible reasons for this may be that most orthodontists assume that skeletal factors are associated with the outcome of class III cases. Additionally, the distance between the maxillary and mandibular molars of class III cases might be difficult to evaluate in the presence of a premature contact. Using mounted casts of class III cases enables the horizontal distance between the maxillary and mandibular molars to be measured, regardless of premature contacts. Hägg et al.  used cephalometric analysis to measure changes in the molar position at three stages (start of treatment, end of treatment, 8-year follow-up) and found significant differences between them; however, they were not evaluated as an associated pretreatment factor.
Regarding general factors, most orthodontists hypothesize that girls are more likely to achieve stability than boys. This may be due in part that Wolfe et al.  found that the anteroposterior discrepancies between the upper and lower jaws were larger in males than in females, and Alexander et al.  reported that the growth spurt in males was much greater than that in females. However, our findings indicated that there is no significant difference between the sexes. This result is consistent with those of a previous report that found no significant differences between the sexes for early class III treatment outcomes .
Most orthodontists also hypothesize that an early pretreatment age is more likely to have a favorable outcome than treatment at an older age. Indeed, maxillary protraction headgear has been reported to be more effective in younger patients [14, 15]. However, there were no significant differences in the pretreatment age between the two groups in the present study. This result is consistent with the findings of a meta-analysis showing no significant differences in the response to early treatment for class III malocclusion between starting at 7–10 years and 11–14 years of age .
Regarding cephalometric factors, mandibular length (Ar-Me) was associated with the stability outcome and was consistent with the results of previous studies [3,4,5,6]. The SN-Rm was also detected as a significant factor, a finding similar to those of previous studies concerning the association of the condylar axis inclination  and SN-Rm  with the outcome. Fudalej et al.  reported that the gonial angle was identified most frequently, and this factor was also indicated to be a significant predictor in the present study. Interestingly, significant differences were observed between the two groups in the ramus plane inclination and the gonial angle, but not in the mandibular plane angle. This may be because a small SN-Rm and an obtuse Gonial angle contribute to a normal mandibular plane angle. One of the most important points for clinicians to note is that it may be easy to overlook when the mandibular plane angle is normal, although cases with a small SN-Rm and large gonial angle may be unstable.
Mandibular incisors have been reported to be significantly retroclined in class III cases, except in the youngest patients ; therefore, the inclination of the mandibular incisors might be an important factor in predicting the outcome after the early mixed dentition stage. Our results indicated that there was a significant difference between the two groups in the IMPA and FMIA.
To summarize our findings relating to cephalometric factors, most of the data associated with stability were not related to maxillary measurements, contrary to what was reported by Ghiz et al. . Because the mandible and maxilla have different types of ossification, cartilage, and sutural components, we suggest that the maxilla is more likely to respond to orthopedic force than the mandible. A previous study indicated that 25.0% of class III cases were related to retrusion of the maxilla, and 22.2% of the cases were related to a combination of maxillary retrusion and mandibular protrusion . These findings therefore suggest that the remaining 52.8% of cases with mandibular protrusion might result in an unstable outcome.
Strength and limitations
Although the long-term outcome of class III treatment was evaluated after comprehensive orthodontic treatment in most studies [4, 5, 7, 8], we evaluated stability before the comprehensive treatment. Therefore, our findings were able to show the exclusive effect of early class III treatment.
However, given this feature of our study, the mean age of 15 years for evaluation might be criticized as being too young for male patients. Because the flexion point of growth in Japanese boys is 13 years , which is younger than that in Caucasian boys (15 years) , the mean ages of the participants were thought to be acceptable for evaluating stability in Japanese boys. Growth can be expected to continue at a decreasing rate, and any remaining growth might influence the prediction of stability. Another interesting factor we would like to consider is information from patient’s family. We do often ask the patient and/or parents whether any family member had orthognathic surgery or protruded mandible, but this is just a questionnaire and not reliable quantitative information. The general public even describes it as protruding teeth when they see the compensated maxillary anterior teeth. Thus, if possible, if the patient’s parents permit us to take their cephalometric radiograph, it would be valuable information to predict the stability and possibility of the surgery. Another limitation of our study is that, even though our sample size was statistically sufficient, adding additional subjects with a wider range of severity may provide further information. Thus, future studies should observe the participants until growth is completed, after the completion of phase 2 therapy, and/or during retention with increased sample size to show the longitudinal results.