Teeth develop from the dental lamina, then commence interactions with the epithelia and underlying mesenchyme. As tooth development advances, enamel knots mediate crown size and cusp formation . Dental anomalies can occur due to disturbance of these processes by genetic factors, environmental factors, or both [10,11,12,13,14,15,16]. When unfavorable factors affect the initiation of tooth development, the tooth may not develop at all (tooth agenesis), whereas when they affect later development, they may result in changes in tooth morphology. Because the permanent teeth develop at different time-points, the influences of unfavorable factors can differ between tooth types. Later-developing tooth types in the same tooth family may have more of a chance of being exposed to unfavorable factors earlier in development. If the unfavorable factors exert influence early in the development of odontogenesis, such as at the time of initiation of tooth formation, there may be more of a possibility of severe effects such as tooth agenesis. If they exert influence at a later stage in development, the effects may be reduced and may only influence aspects of tooth morphology such as tooth size and cusp formation. Juuri and Balic  proposed that a gradual reduction in the odontogenic potential of dental lamina may explain why tooth agenesis most frequently affects the last tooth to develop within a tooth family. If the unfavorable effects on odontogenesis are weak, this gradual reduction in odontogenic potential may only affect the last tooth to form, as is apparent in some hypodontia patients. If the unfavorable effects are strong, however, more of the teeth may be affected by tooth size reduction and CMT. Concordant with these two ideas, later-forming teeth are much more susceptible to tooth agenesis and size reduction than earlier-forming teeth.
In this study, in the hypodontia group, mandibular second premolars were the most frequently CMT, followed by mandibular and maxillary lateral incisors, which is consistent with results reported by Endo et al. , who investigated 3358 Japanese orthodontic patients. Many researchers have reported that mandibular second premolars were the most frequently missing tooth type in tooth agenesis patients; however, in some reports, the maxillary and mandibular lateral incisors were the most frequently missing tooth type in Japanese [31, 32] and in other races [3, 9]. Regardless of the change in order, these teeth are the last teeth to form in their tooth family, which supports the contention that the last forming teeth are more susceptible to tooth agenesis. Although the mandibular second premolars were the most frequently missing tooth type in both the oligodontia group and the hypodontia group, there was a dramatic difference in frequency—25.9% (hypodontia) vs. 88.2% (oligodontia), and in the oligodontia group, the second, third, and fourth most frequently missing teeth were the maxillary second premolars (87.3%) and first premolars (63.7%) and mandibular first premolars (44.1%). These results are consistent with those of Endo et al.  and Ogaard et al. , which may indicate that an increased number of missing premolars is a typical feature of oligodontia patients.
Notably, in the current study, there was no significant difference in the reduction of mesiodistal tooth width in the hypodontia group except for maxillary lateral incisors and mandibular first molars in female, whereas other researchers reported a significant difference in mild tooth agenesis patients [17, 22, 34]. This may relate specifically to the fact that our sample population was Japanese, or it may relate to some other difference in sample collection. In the oligodontia group in the current study, however, as in the study of Gungor and Turkkahraman , in both sexes, reduction of mesiodistal tooth width was greater in cases of severe agenesis. Brook et al.  further concluded that as the number of CMT increase, the degree of tooth size reduction increases. The results of the current study are concordant with this in both sexes, with a higher correlation in male patients, particularly with regard to maxillary teeth. The details of CMT and reduction of mesiodistal tooth width pertaining to sex differences remain unclear at present, and further investigations are necessary in this respect.
Maxillary lateral incisors exhibited a higher frequency of CMT in both the hypodontia group and the oligodontia group in the current study. In a few reports, the frequency of tooth agenesis of maxillary lateral incisors was reduced in severe tooth agenesis patients compared to mild tooth agenesis patients [4, 5, 31]. In the current study, however, the frequency was still high. Moreover, maxillary lateral incisors exhibited the greatest difference in the oligodontia group, as has been reported by other researchers [20, 22]. Thus, it is possible that these teeth may be more susceptible to disturbances during development than other teeth because maxillary lateral incisors have simpler morphology, which may be more easily affected than that of any other teeth.
In contrast to the finding of high frequency of CMT in maxillary and mandibular second premolars in the oligodontia group, the mesiodistal tooth widths of these teeth, even in the oligodontia group, were not significantly differ from the sizes in the controls. These results differ from the findings of another study that maxillary and mandibular second premolars were also smaller in tooth agenesis patients [20, 22]. The discrepancy between the results of our study and their study may indicate that this phenomenon is specific to Japanese patients with tooth agenesis. However, our oligodontia patients included only four maxillary second premolars and three mandibular second premolars that were measured in both sexes, which suggests that larger numbers of teeth are needed to clarify this issue statistically. Further collection of samples is necessary to resolve this question.
The current study focused on the mesiodistal tooth width since it is believed that proper mesiodistal tooth width ratio of the maxillary and mandibular arches is necessary to have proper interdigitation, overbite, and overjet . It is therefore important to know the characteristics of mesiodistal tooth width in tooth agenesis patients for making a treatment plan. However, because buccolingual dimensions and tooth shape were also affected in tooth agenesis patients [20, 22, 29], further analysis including three-dimensional analysis is essential to clarify the relation of tooth size/shape and tooth agenesis in Japanese patients.
All the samples, including the controls, were collected from Japanese orthodontic patients which limits the generalizability of our study. Richardson and Malhotra  and Johe et al.  reported tooth size discrepancies greater than ± 1 SD in 33.7 and 41% of their orthodontic patients, respectively. Our control samples also include disproportions in tooth size. However, the mean dimensions for the controls were similar to those reported by Brook et al. . Nevertheless, because the samples were taken only from a Japanese population, the results should not be extrapolated to other racial population.