This is, to our knowledge, the first study investigating the prevalence of DDE since 1990 in a nationally representative sample of 18-year-olds living in all sixteen provinces of Poland. The study was conducted as part of the Ministry of Health national programme– the only nationwide programme assessing oral health. Even though the programme was launched in 1999, it was in 2017 that it was extended with an additional examination of DDE. Reliable data or studies on larger populations are not available.
Retrospective determination of the aetiology of enamel defects is difficult. The presence of demarcated opacity and hypoplasia in the form of isolated, sporadically located lesions indicates local causes. Diffuse opacities are usually found on the teeth with simultaneous enamel secretion and maturation, pointing to environmental aetiology and are related to systemic causes. Traumatic damage or periapical inflammatory lesions in the primary tooth or its early extraction may disrupt normal matrix deposition or enamel mineralisation and, consequently, lead to enamel defect (a demarcated spot and/or hypoplasia) in the permanent tooth.. The distribution of defects among Polish adolescents aged 18 years old was purpose of our previous paper . The data from previous monitoring studies allow to conclude on the consequences of dental caries as a possible causal relationship with DDE. The 18-year-old adolescents in the study were aged 3 years in 2002, when the prevalence of caries among Polish children at that age was 56.2%, with a mean of 2.9 teeth affected . In the same period (2002), caries was found in 86.9% of 6-year-olds, with 5.9 primary teeth involved .
A number of studies have shown a positive relationship between DDE and the severity of caries [9–20, 33]. The results of our study are consistent with previous studies that suggested that DDE increased the risk of dental caries, since the influence of enamel defects in the development of caries was observed [14, 33–35]. Fotedar et al. also demonstrated a significant association between caries and enamel opacity among 12- and 15-year-olds from India. The relationship between DDE and the severity of dental caries was also confirmed in our study, which demonstrated significantly increased caries severity (expressed as DMFT and DMFS) in patients with DDE (7.66 ± 4.68 vs 6.35 ± 4.12 and 11.59 ± 9.15 vs 9.51 ± 8.07, p < 0.001). However, the severity of caries was increased in individuals presenting with qualitative enamel defects (demarcated and diffuse opacity) rather than those with quantitative defects (hypoplasia). Nevertheless, no significantly increased risk of caries among Polish adolescents in the presence of developmental enamel defects was shown, which may be due to the high prevalence of caries and a several-year residence time of teeth in the oral cavity. Some of these subjects will have their original DDE obliterated by caries, restoration and extraction.
In relation to enamel defects, all the types of defects can be associated with dental caries. Enamel hypoplasia is more susceptible to dental caries . In opposite, individuals with diffuse opacities decurrently of greater fluoride exposure are less prone to exhibit dental caries . The term ‘diffuse opacities’ is used interchangeably with dental fluorosis when is caused by an excessive intake of fluoride. The amount of fluoride contained in drinking water should be considered, especially if 0.70 mg/l or more of fluoride is present in the water [16, 23]. The studied sample of adolescents was not exposed to fluorine in the public water. The fluoride level in the public water was below 0.5 ppm at the time when the participants of the present study were born. The majority of the participants had municipal drinking water with a low level of fluoride (< 0.5 mg/ml). The low prevalence of diffuse opacities among participants in this study might be due to consumption of low level of fluoride in the public water. Another explanation might be the effect of remineralization. Compared with demarcated opacity, it is easier for diffuse opacities to be remineralized, similarly to assumption of Wong et al.  study.
The results of our study, however, demonstrated that the type of DDE could not explain the heterogeneity, probably because the aetiology of these defects share some similar influencing factors. It is possible that the longitudinal observation of DDE differ among populations with different prevalence of DDE. When analysis was performed for the different types of DDE, it was found higher prevalence of DEO compared to DIO. Reduction of frequency of diffuse opacities has been showed, similarly to Wong et al.  study. Demarcated opacities have distinct boundaries separating them from normal enamel which are thus more unlikely to disappear from mechanical and chemical reasons.
A question should be asked whether these are demarcated defects that promote the development of caries or whether dental caries develops due to the presence of causative factors of this disease in individuals with dental caries in primary dentition. A positive association between enamel defects and dental caries was identified in meta-analysis of Vargas-Ferreira et al. . Individuals with DDE had higher pooled odds of having dental caries experience [OR 2.21 (95%CI 1.3; 3.54)]. A higher chance of dental caries should be expected among individuals with enamel defects.
It is beyond doubt that the risk of differential misdiagnosis due to the lack of precise data on different fluorine sources in childhood, is an important limitation of epidemiological studies assessing the incidence of enamel opacities classified as dental fluorosis.
It’s important to emphasize that DDE and carious lesions were distinguished, diagnosed and recorded based on locations and surface features, followed diagnostic criteria and recommendations [25–30]. Training and calibration of all examiners in the present study resulted a good intra- and inter-examiner reliability. Dini et al.  demonstrated a two times lower risk of dental caries in children with diffuse enamel opacities compared to children with no or demarcated opacities.
The half of adolescents worry about their teeth. These findings are not in accordance with Sujak et al. who suggested that very few subjects were concerned about the appearance of their teeth, or were not aware of their teeth being different. In the Vargas-Ferreira et al.  study children with DDE did not indicate any decrease in self-perception. However, this condition was associated with an impact on the functional limitation domain.
The strength of this study was large number of 18- year-old adolescents. A population-based sample was used, contrary to a clinical convenience sample, in some studies. The presence of controlling for confounding such as socioeconomic factors to find its influence on findings was taken into account and analysed, in contrast to some other studies. When considering the methodology of the present study, it should be mentioned that the recording of carious lesions and DDE followed the most recently published recommendations [26, 28–30]. The trained group of dental examiners showed good intra- and inter-examiner reliability values and good capacity to identify DDE and to discern the different types of DDE. Furthermore, restorations subsequent to carious lesions were delineated from DDE-related restorations, i.e. atypical restorations due to DDE were not scored as caries-associated restorations and, hence, were not part of the F-component and the DMF index. Therefore it was possible to distinguish between DMF and DDE-related lesions correctly. Restorations subsequent to carious lesions were delineated from DDE-related restorations, i.e. atypical restorations due to DDE were not scored as caries-associated restorations and, hence, were not part of the F-component and the DMF index. Therefore it was possible to distinguish between DMF and DDE-related lesions correctly.
It is important to mention that the cross-sectional observational epidemiological design of the study does not allow to establish the temporal causal relationship between DDE and caries. Further investigations using longitudinal design are needed to confirm these findings.
This study has several limitations that need to be taken into account for an adequate interpretation of the results, acknowledging the age of the subjects. Adolescents were recruited from the public high schools only. Only those who signed consent for participation were included to the study, what may cause selection bias. The examinations regarding dental caries should be separated from DDE to avoid observational bias, however, this was not possible in the present study. Some of the subjects will have their original DDE obliterated by caries, restoration and extraction. Due to the difficulty to differentiate between molars and incisors hypomineralisation and caries, misclassification bias could be taken into account. On the other hand, the tooth- and surface-related recording of MIH-related defects and restorations—which has been used in caries epidemiological trials for decades—is another step forward helping to determine the extent and severity of MIH precisely . In addition, another limitation of the study refers to the lack of analysis according access to oral health services, dietary and oral hygiene habits aspects, data on sources of fluoride exposure other than water fluoridation. These factors may be considered as potential effect modifiers that may lead to a weak association between dental caries and enamel defects. Also, the reasons for missing teeth were not recorded; hence, some teeth that were missing due to caries were designated as non-carious. Diagnosis was based on visual and tactile examinations under artificial lighting conditions only. Radiographs were not taken, thus small caries lesions might have been underestimated nor recorded, what is a common problem in cohort studies. Some studies have investigated teeth with natural lighting, others with a flashlight for illumination and have used sterile gauze to remove debris. It is important to mention that in the present study all permanent teeth have been evaluated, while most studies assessed only index teeth. In addition - DMFT was used instead of more detailed index like ICDAS II. Different indices and criteria, examination variability, methods of recording, and varying age groups in various DDE studies, may limited the comparisons of the results. Thus, a standardised index should be used in future studies. Finally, because dental caries and DDE may share key risk factors, such as a disadvantaged background, a common risk approach should be more rational .