Dental maturity has been considered a suitable indicator of chronological age and regarded as superior to the stature and skeletal methods for evaluating an individual’s somatic maturity. So, it is considered a valuable tool to estimate biological age in forensic or archaeological studies. Yet, for accuracy improvement, other factors have been considered important when applying dental age estimation techniques, and agenesis is thought to be one of those factors.

Our study points to delayed dental maturation of the second mandibular left molar in individuals with at least one 3M agenesis, being this effect more pronounced as the number of 3M agenesis increases. This result agrees with the one obtained by Lebbe et al. [23] pointing out that the total number of agenesis needs to be considered when estimating age, as having more teeth agenesis is negatively correlated with dental age.

The effect of each tooth is, however, not clear. The maxillary 3M agenesis was associated with 37 mineralization stages in the 13- and 14-year-old males, and in females, only with the 11-year-old. As for mandibular agenesis, there was a significant association in 14-year-old males and 13-year-old females. Thus, maxillary agenesis may have an effect earlier than mandibular agenesis, and the effects of 3M agenesis are perceived in girls first.

This influence is difficult to explain. It may be that, at some ages, mineralization is more susceptible to congenitally absent teeth, making some stages more prone to this influence. The results comparing mean age in both groups (S1 and S2) support this hypothesis as stage G was the only one to depict statistically significant differences. In our case, the more susceptible ages, based on the greater number of associations, are 11 years old for females and 14 years old for males, which, in the Portuguese population, coincides with the pre-puberty ages [24, 25].

As for other studies, to date, to the best of our knowledge, this is the first study analyzing the third molar agenesis impact on second mandibular molar mineralization. Other investigators addressed delayed mineralization of permanent canines, premolars, and second molars in individuals with missing teeth, not including third molars [13, 26]. Daugaard et al. found that the second premolar unilateral absence is associated with a tooth mineralization delay on the same and the contralateral side [13]. Our research did not study the second lower right molar mineralization stages, so we cannot assess the described effect. The same authors also reported a relationship between missing second premolars and delayed second molars in girls but not boys. Our investigation does not confirm their findings, as the effect was seen in both sexes; Ruiz-Mealin et al. [26] did not also report sex differences in dental maturation delay in patients with agenesis. Yet, our results point to more robust differences between groups in females (p = 0.028 and p = 0.049, for females and males, respectively). This latter finding of differential effects in females and males may be of particular interest concerning sex differences in dental development, and the fact that, in our sample, the effects seem to be more pronounced in females suggests that this is most likely a substantial finding and not merely a consequence of multiple statistical tests.

Further studies also account for this effect of delayed teeth mineralization. Tunç et al. [27] reported a mean difference of the estimated dental age of 0.3 years between the agenesis and the control group. Similar results were obtained by other authors [28,29,30], who also reported the differences being more expressive in females.

The relevance of performing this study concerning specifically the second molar mineralization relates to the fact that, in some countries, 14 is the limit of criminal liability. For instance, in Italy, the subject may be liable from 14 to 18 years, depending on the degree of maturity reached [1]. Other reasons may relate to age estimation of victims of sex crimes. In Portugal, the 172nd article of the Portuguese Penal Code determines different sentences for abuse of minors if they are over or under 14 years of age [20, 21]. Regardless, the underestimation or overestimation of age in this age group caused by technical errors may conflict with ethical issues since the penalties applied will be inadequate. The error of the estimated age being different from the real age becomes even more relevant in child abuse or child pornography scenarios since the punishments are more severe in case the victim is younger than 14 years old [20].

Another point we would like to address is the methodology of 37 staging. Being a two-rooted tooth, and because the mineralization process does not happen simultaneously in both roots, we have staged 37 according to the root that displayed the more advanced mineralization stage (i.e., the one that would give the individual the higher age). We recognize that usually, it is the other way around. Yet, in this case, a higher age estimation, i.e., being older, leads to a better (more favorable) penal measure for the offender. In this way, we believe that by doing so, we favored technical and not ethical errors.

As for limitations, we cannot tell if our sample with no agenesis depicts a “normal development rate of the second molar,” but as no other factors were identified, we have assumed it does. Still, other factors can modify stage attainment [7, 9]. Nevertheless, when compared with a study in a similar population [20], our non-agenesis subsample depicted similar Demirjian’s stage attainment ages.

We also recognize that by using the Demirjian stages (or any other), we introduce large uncertainties since the assigned stage is attained, and suddenly, it goes from one stage to the next. Still being so widely used, it allows for comparison with other studies; additionally, stages (particularly Demirjian stages) are frequently used as markers or thresholds for given ages [31,32,33], and in this sense, this study can benefit those practices.

Finally, we recognize that the effect is discrete, and the differences in the development of tooth 37 are minor. If a p-value 0.01 instead of 0.05 was adopted, not so much would be significant.