Thank you for your letter and your interest in our article. The first question raised, regarding the amount of crowding present before treatment, might not add a significant value to the discussion. The reason is that the subjects were selected from a pool of nonextraction patients. We outlined in the methods that, of the 858 screened patients, 309 were treated without extractions. The sample group of 60 subjects was selected from those who had mild to moderate crowding that required no extractions. Therefore, the amount of crowding—a transversal plane assessment—is nothing but another statistical mean when evaluating the long-term changes in the deepbite malocclusion of nonextraction patients. Speculatively, the severity of crowding could have an impact on anterior alignment. However, Erdinc et al argued that, with the exception of the interincisal angle, there is no statistically significant differences between extraction and nonextraction patients who had clearly distinguishable initial crowding values. A systematic review confirmed these findings. After all, posttreatment mandibular rotation, which might actually affect the degree of deepbite relapse as speculated in our article, was not associated with the relapse of mandibular incisor alignment.

I believe your second question has also found an answer, since you referred to a study that relates to the quality of finished occlusions with 4 premolar extractions. However, I certainly agree that posttreatment occlusion, regardless of the treatment modality, is a potent factor in maintaining the treatment outcome. This is why patients with good-quality treatment results and records were included, and the 309 available subjects were reduced to 60 in the study.

As for your third question, the minimum period of retention for any patient in the study was 2 years, as stated in the article. Therefore, our results are limited to the outcome assessments for at least 2 years of retention time, which is a common practice. On a side note and for further reference, Bondemark et al concluded that the lengths of treatment times and retention periods could not be used to predict stability changes because of large individual variations.