We read with interest the article by Surendran and Thomas (Tooth mineralization stages as a diagnostic tool for assessment of skeletal maturity. Am J Orthod Dentofacial Orthop 2014;145:7-14). First and foremost, we want to express our appreciation to the authors for their efforts in exploring whether dental calcification can be used as a first-level diagnostic tool for assessment of skeletal maturity by using an accurate, simple, practical, and economical method. Their conclusions give directions on the clinical usefulness of the determination of dental maturity for the assessment of treatment timing for a skeletal malocclusion. However, we have 2 main concerns and 1 question about this research.

The first concern is about the selection criteria of the subjects. Since the authors used a cross-sectional design, choosing appropriate subjects for the removal of interfering factors has a significant impact. Assessment of dental maturity was carried out through the calcification stages from panoramic radiographs, so some factors affecting tooth mineralization but having no impact on skeletal maturity should also be controlled. For example, subjects should not come from high-fluoride areas or have a history of caries, periodontal disease, central cusp, enamel hypoplasia, or serious diseases of the deciduous teeth. In addition, socioeconomic differences are much more pronounced in skeletal growth than in dental development. So, if the selection criteria included some terms that reflect socioeconomic conditions, this research would be stronger.

The second concern is about the discussion of the results. The authors concluded that first premolar stage E alone has a better predictive value than the combination of FEED for identification of the MP3-F stage through the likelihood ratio, and combinations of the dental maturation stages had similar predictive values as individual teeth in this study, but we wondered whether this was contrary to people’s general knowledge. The combination of FEED represents a shorter period of time than the first premolar stage E alone, which contains the combination of EEDD, EEED, FEDD, FEED, and FEEE in the aspect of time range. It is common sense that a shorter time would predict more accurately. In addition, we think this study can extend to compare with the combinations of 2 or 3 teeth. For example, combinations of the dental maturation stages of the canine and the first premolar might have a great predictive value for identification of the MP3-F stage.

Apart from the concerns above, we have a question about the resultant analysis. In Table I of the article, there were 3 100% values of the skeletal maturation stages, but in Table II, all of them appeared to be null values. But in a real-world situation, a subject whose canine is stage E, first premolar is stage D, or second premolar is stage D can be in MP3-FG stage or even MP3-G stage. Then the value discarded is significant. Should the values be abandoned? Will this affect the conclusion?