Background

Introduction

Amelogenesis imperfecta (AI) is a heterogeneous group of inherited enamel conditions, with an estimated prevalence between 1:700 and 1:14,000. Enamel formation is disturbed, resulting in defects which can be associated with significant clinical, functional, and psychosocial consequences (Crawford et al. 2007).

Identifying and classifying AI has historically been focused on clinical phenotyping. For non-syndromic, or “isolated” AI, the classification proposed by Witkop (1988) has traditionally been the most widely adopted. Increased knowledge of associated genes and molecular pathways has led to the proposal of new, more accurate classification systems (Bloch-Zupan et al. 2023; Wright 2023), allowing the clinical appearance to be matched to the molecular diagnosis (Gasse et al. 2013; Hany et al. 2024).

Defects in enamel formation have been linked to pathogenic defects in more than 100 genes (Bomfim et al. 2026). In addition, more than 115 genetic conditions have been associated with hereditary enamel defects (Chu et al. 2026; Wright 2023). Genetic testing for AI can be undertaken via a targeted gene panel, whole-exome sequencing, or whole-genome sequencing. An AI gene panel test was introduced in England in 2016, as part of the UK Genetic Testing Network, now the National Genomic Directory (NHS England 2021). The gene panel encompasses a defined set of genes known to cause AI and is offered by several Paediatric Dental Units. This panel test does not carry the risk of recognition of coincidental pathogenic gene variants.

Dental anomalies

Amelogenesis Imperfecta has been observed alongside a variety of other disturbances in dental development (Poulsen et al. 2008). The co-segregation of anomalies suggests that the pathogenic variants associated with AI may affect multiple odontogenic tissues, extending beyond ameloblastic dysfunction. The presence of dental anomalies could reflect the expression of AI-associated genes in multiple cell types involved in dental development. Conversely, some dental anomalies may arise secondarily to the altered enamel phenotype (Peters et al. 1992; Collins et al. 1999). The relative contribution of the pathogenic variant itself, and secondary consequences of the enamel defect on dental development, is likely to vary between anomalies, and remains incompletely understood. The most commonly described anomalies are discussed below.

Pre-eruptive crown resorption (PECR) may be described as irregularly distributed, “moth-eaten” coronal areas visible radiographically on unerupted teeth (Collins et al. 1999). A systematic review of dental anomalies associated with AI found PECR to be reported in six papers (Poulsen et al. 2008). A cross-sectional study reviewing radiographs found PECR occurred in 41% of unerupted teeth of individuals with AI, compared with none of their unaffected relatives (Collins et al. 1999). However, no sizable epidemiological studies of PECR exist (Besa-Witto et al. 2025). The mechanism by which coronal resorption occurs in AI remains unexplained. Most research on pathological external crown resorption concerns the cervical root area, initiated after traumatic periodontal ligament injuries (Besa-Witto et al. 2025). However, the crown differs from the root, as the dentine is covered by enamel rather than cementum. This is significant when considering the mechanism of PECR, as individuals with a hypoplastic phenotype will have very thin, or even absent enamel.

It has been theorised that a lack of reduced enamel epithelium in AI results in exposure of the underlying dentine to the surrounding connective tissue, triggering aggressive osteoclast-mediated resorption. This results in irregular cavities in the crown filled with connective tissue. An alternative explanation attributes resorption to mechanical trauma to the developing dental follicle, causing the fragile reduced enamel epithelium to be destroyed or damaged, exposing the underlying dentine, again triggering osteoclast-mediated resorption (Besa-Witto et al. 2025). Local factors have also been theorised as potential causes of PECR, including ectopic tooth position, and periapical involvement of the preceding primary tooth (Bhatia et al. 2015). These proposed mechanisms suggest that PECR may reflect both the consequences of abnormal enamel development and secondary effects of structural fragility or damage to the developing tooth and its surrounding tissues.

An additional frequent clinical finding for individuals with AI is delayed tooth eruption. Collins et al. (1999) found in their cross-sectional study that individuals with AI were more likely to have markedly delayed eruption (27%; 6/22), compared to unaffected family members without AI (8%; 1/13). A more recent cross-sectional study of panoramic radiographs found that 53% of those with AI had retained primary teeth, compared to 12% of unaffected controls. In addition, 78% had multiple impacted teeth, compared with 13% of controls (Kammoun et al. 2024). This was consistent with a previous study by Seow (1995), reporting a six-fold increase in the risk of permanent tooth impaction in individuals with AI.

A systematic review by Poulsen et al. (2008) found that those with AI may develop intra-pulpal calcifications (pulp stones). Two cross-sectional studies have shown a greater incidence in AI (14–48%) compared to healthy controls (0–18%). The aetiology remains unclear, with calcifications possibly resulting from tissue-specific expression in the dentine-pulp complex (Collins et al. 1999; Kammoun et al. 2024).

Although occurring most frequently as an isolated anomaly, taurodontism has also been associated with genetic disorders, including AI (Jayashankara et al. 2013). It has an estimated prevalence of 11.8% in the general population, but was found to be significantly more prevalent in AI, estimated at 40% (Adorno-Farias et al. 2019; Decaup et al. 2022). A Tunisian study of panoramic radiographs found increased prevalence of taurodontism in patients with AI at 18%, compared to 3% of healthy controls (Kammoun et al. 2024). Conversely, a study conducted in the USA found no significant difference between AI and healthy controls (Collins et al. 1999).

Another area of interest relates to the gingival condition of patients with AI, with case reports describing hyperplastic gingivae covering areas of PECR on affected teeth (Bhatia et al. 2015; Urzúa et al. 2021). It remains unclear, however, whether this localised gingival hyperplasia occurred as a primary feature, or secondary to resorption of dental hard tissues. Individuals with AI may also experience gingival overgrowth that is independent of plaque control or systemic disease (Macedo et al. 2005; O'Connell et al. 2014; Heyl et al. 2020).

O'Sullivan et al. (2011) identified homozygous loss-of-function variants in the FAM20A gene as the cause of AI and Gingival Fibromatosis Syndrome (AIGFS), a syndromic form of AI characterised clinically by gingival hyperplasia and dental anomalies including hypoplastic AI, intrapulpal calcifications and delayed eruption (Cherkaoui Jaouad et al. 2015). FAM20A is expressed in both ameloblasts and gingival fibroblasts, supporting its dual role in enamel formation and gingival tissue regulation. Clinically, dental features include severe enamel hypoplasia, delayed eruption and gingival enlargement. Radiographic signs include failure of eruption, pulp calcifications and hyperplastic dental follicles.

The phenotypic effects of pathogenic variants in FAM20A extend beyond the oral cavity, also being associated with renal calcifications as part of enamel-renal syndrome (OMIM #204,690). If this highly distinctive clinical phenotype is identified, subsequent investigation via renal ultrasound is indicated (Bloch-Zupan et al. 2023). Although AIGFS was initially regarded as a distinct condition from enamel-renal syndrome, they are now considered phenotypic variants of the same FAM20A-related disorder (de la Dure-Molla et al. 2014).

Individuals with AI may also present with anterior open bite (AOB). Indeed, it is the most commonly reported malocclusion affecting individuals with AI (Poulsen et al. 2008), with a reported prevalence of 24–60% (Messaoudi et al. 2023). The wide range in reported prevalence may reflect differences in study populations and AI phenotypes, as well as variation in methodology between studies. A recent scoping review reports the global prevalence in the general population at 16.8%, however this was primarily attributed to digit sucking, rather than being skeletal in origin (Gonzalez et al. 2025).

In their systematic review and meta-analysis of craniofacial characteristics in AI, Messaoudi et al. (2023) found that individuals with AI tended to have a more vertical craniofacial growth pattern with a smaller overbite, a larger intermaxillary angle and a class II skeletal pattern with a more retrognathic mandible. The pathophysiology of this vertical craniofacial growth, and the resulting propensity for AOB, remains unclear. Both locally acting environmental factors and genetically determined skeletal growth patterns have been hypothesised.

One proposed explanation for the development of AOB is that defective enamel may contribute to this vertical growth pattern, whereby resultant decreased bite forces in individuals with AI may lead to posterior tooth overeruption and the development of AOB. Oral habits, such as thumb sucking and tongue thrusting, have been proposed as potential local environmental contributors to AOB in individuals with AI, although their contribution remains uncertain (Ravassipour et al. 2005; Witkop and Sauk 1976). Indeed, they may act as confounding factors, thereby complicating the assessment of any causal relationships.

Although pathogenic variants responsible for the enamel phenotype may genetically predispose an individual to an AOB, it is also possible that enamel defects and open bite in individuals with AI may result from different mechanisms. Studies in this area have been described as having low methodological quality, carrying moderate to high risk of bias, lacking representative control groups, and failing to accurately describe occlusal diagnoses with validated tools (Broutin et al. 2023).

Dental maturation

Estimation of dental age, or the stage of dental maturation, has a variety of clinical, and forensic purposes. Existing observational studies have shown conflicting results regarding the potential impact of AI on dental maturation compared to chronological age, employing a variety of dental ageing estimation methods in their assessment (Aren et al. 2003; Seow 1995; Gabardo et al. 2020; Kirzioglu et al. 2019). It has been suggested that although delayed dental eruption is a common feature in AI, dental maturation per se does not appear to be affected.

Study rationale

The existing literature regarding the co-existence and aetiology of dental anomalies in children with AI is sparse. A better understanding of phenotype-genotype correlations is important. This helps us to understand the likelihood of individuals with a particular pathogenic variant having other co-existent dental anomalies, which may influence management. The primary aim of this descriptive cross-sectional study was, therefore, to assess for the presence of dental anomalies in individuals with AI, recording this alongside the molecular diagnosis where available.

Methods

Study design

This was a descriptive cross-sectional study investigating the presence of dental anomalies and dental maturation in children with a diagnosis of AI. This study involved secondary analysis of participant data from an existing study “Molecular Genetic Investigations of Amelogenesis Imperfecta”, which aimed to increase insight into the key genetic events in amelogenesis, and the impact of pathogenic variants on dental hard tissues (Hany et al. 2024, 2025).

The chief investigator (CI) for the primary study was Dr Alan Mighell. This is an ongoing study led by the Leeds Dental Institute & Leeds School of Dentistry. Local named leads were appointed in the other participating centres, including Leeds (Professor Richard Balmer), Bradford (Professor Peter Day), Sheffield (Professor Helen Rodd and Dr Joana Monteiro), and Birmingham (Miss Catriona Brown).

Data utilised from the primary study included radiographs, clinical photographs and for some participants, the molecular diagnosis responsible for AI.

Ethics

The primary study received full approval from the Research Ethics Committee (REC) and Health Research Authority (HRA) (REC reference: 13/YH/0028 for IRAS project ID: 82,448) on 31 January 2013. Approval for the secondary analysis of existing data from one arm of the primary study was subsequently granted by the local trust research governance coordinator, in agreement with the chief investigator, and local named leads. Confirmation that the secondary analysis could proceed without the need for ethical amendment was provided by the trust research governance coordinator on 22 February 2024. The chief investigator gave formal approval for the secondary analysis on 2 April 2024.

Participants/recruitment

Participants were recruited to the Sheffield arm of the primary study either via their attendance at the Paediatric Dentistry Department at Charles Clifford Dental Hospital or invited by letter following a review of their clinical record or their position on the departmental waiting list. Written consent was obtained by a member of the research team, including consent for future research. The participant list from the Sheffield arm of the primary study consisted of 149 individuals, including participants with AI, and unaffected family members.

The inclusion criteria for this secondary analysis included participants with a clinical/genetic diagnosis of AI and an available full panoramic radiograph taken prior to 18 years of age. Individuals with healthy enamel phenotype, or other enamel abnormality were excluded, along with those lacking the required radiograph. Following review of participants recruited to the primary study, 30 participants met these essential criteria and were included in this secondary analysis.

Variables

Data were recorded on a pre-piloted clinical record form (CRF), including the participants’ sex and age at the time of the panoramic radiograph exposure. Radiographs and clinical photographs were reviewed by the lead investigator (RB). Additional review was sought from a co-investigator (JM), where required, to reach consensus.

The enamel phenotype was assessed and recorded as either hypoplastic or hypomineralised. The presence of six dental anomalies was recorded from the available radiographs and, when available, clinical photographs. These included PECR, delayed/ectopic/failed eruption, pulp calcifications, taurodontism, gingival enlargement and anterior open bite (AOB). Eruption abnormalities were recorded where teeth remained unerupted beyond the expected time range as set out in the Atlas of Tooth Development and Eruption (AlQahtani et al. 2010). Gingival enlargement and AOB were only assessed for participants with available clinical photographs. Taurodontism was assessed using the Seow and Lai (1989) technique. This technique involves assessment of mandibular first permanent molars with mature apical development, where the occlusal height is visible. Finally, dental maturity was assessed using the Demirjian method and compared to the participants’ chronological age at the time of the radiograph.

The CI of the primary study securely shared the available molecular diagnosis, to which the identified dental anomalies were mapped.

Training and calibration

The lead investigator (RB) received training in the use of the Demirjian method for dental ageing. A pilot study of five randomly selected cases was conducted by the research team (RB, JM) to assess both intra-examiner and inter-examiner reliability for the identification and classification of dental anomalies, and dental ageing estimation. Disagreements were discussed until consensus was reached, and the data collection form was modified accordingly. Following changes, a new pilot study of five additional cases was undertaken, and reliability assessed again using Cohen’s kappa coefficient (Cohen 1960). The resulting kappa values are summarised in Tables 1 and 2, and reported further in the Results section.

Data collection

Anonymised data were collected by the lead investigator between 30 August 2024 and 19 October 2024. No further clinical examination was undertaken as part of this study. During data collection, any uncertainties or discrepancies were discussed by the two researchers until agreement was reached, in cases where clear guidance or evidence was lacking.

Statistical methods

Data were inputted into a standardised data collection sheet using Microsoft Excel (Microsoft 365, version 16.103.4, Microsoft), and into the SPSS electronic software programme (SPSS, version 29; IBM Corp., New York, USA) for the purpose of statistical analysis. Simple descriptive statistics were undertaken in relation to the criteria studied, in the absence of a comparison group.

Results

Intra-examiner and inter-examiner reliability

The pilot study included an assessment of both intra-examiner and inter-examiner reliability to evaluate the consistency of diagnostic judgements of dental anomalies, and dental ageing estimation using the Demirjian method. Reliability was quantified using Cohen’s kappa coefficient (Cohen 1960).

The resultant kappa values for intra-examiner reliability are presented in Table 1. The disagreement in one case of pre-eruptive crown resorption resulted in a kappa value of 0.545, which describes a moderate level of agreement. For all other criteria, agreement was observed in 100% of cases, with a resultant kappa value of 1.00.

Table 1 Result of intra-examiner reliability using kappa coefficient, for the assessed dental anomalies, and dental ageing estimation
Full size table

The kappa values for inter-examiner reliability are represented in Table 2. In the assessment of pre-eruptive crown resorption, JM recorded one case of pre-eruptive crown resorption that RB did not. A similar discrepancy was noted for pulp calcification, with JM identifying one case that was not noted by RB. As a result, the agreement for these two criteria was reached in four cases (80%), resulting in a Cohen’s kappa coefficient of 0.545, which represents a moderate level of agreement. All remaining criteria demonstrated complete agreement (100%), with a kappa score of 1.00.

Table 2 Result of inter-examiner reliability using kappa coefficient, for the assessed dental anomalies, and dental ageing estimation
Full size table

Demographics and enamel phenotype

A sample of 30 children was included from the participant list of the primary study, having met the essential inclusion criteria. Sixteen participants (53.3%) were male and 14 (46.7%) were female. Age at the time of the assessed panoramic radiograph ranged from 5 years 1 month (61 months) to 14 years 4 months (172 months), with a mean age of 9 years 8 months (116 months, SD = 23.3). Nine individuals (30%) were deemed to have hypoplastic AI, whilst 21 participants (70%) had hypomineralised AI. No individuals had a mixed phenotype recorded due to the binary classification of the enamel phenotype employed in this study.

Assessment of dental anomalies

Seventy-three percent of participants (22/30) were found to have at least one dental anomaly in addition to the enamel phenotype. The prevalence of the dental anomalies is summarised in Table 3. Co-existence counts for the assessed dental anomalies are summarised in Table 4.

Table 3 Prevalence of dental anomalies in the study group
Full size table
Table 4 Co-existence counts for dental anomalies, excluding missing values
Full size table

Three participants with PECR had hypoplastic AI and four had hypomineralised AI. Co-occurrence of PECR and delayed/ectopic/failed eruption was observed in four participants. Co-occurrence of PECR and gingival enlargement was observed in one participant (Table 4). Figure 1 shows a participant with suspected PECR affecting the unerupted maxillary canines.

Fig. 1
Fig. 1
Full size image

Panoramic radiograph showing coronal radiolucencies UR3 UL3 suggestive of PECR (Participant 3, aged 9 years 9 months)

Delayed, ectopic or failed eruption (including impacted teeth) was observed for 10 participants (33.3%), as seen in Fig. 2.

Fig. 2
Fig. 2
Full size image

Panoramic radiograph showing generalised hypoplastic AI, retained primary teeth and delayed eruption of teeth (Participant 2, aged 10 years 1 months)

Four participants had at least one permanent tooth with intra-pulpal calcifications, whereas in two participants this finding was limited to one primary tooth. Two participants had intra-pulpal calcifications affecting multiple permanent teeth.

Twenty participants had two mandibular first permanent molars that were suitable for assessment of taurodontism, whereas two participants had only one, resulting in the assessment of a total of 42 teeth, from 22 individuals. Two teeth from the same participant were classified as having mild taurodontism (hypodont), according to the classification system proposed by Shaw (1928).

Clinical photographs were unavailable for three participants; therefore, the presence of gingival enlargement and AOB was only assessed for 27 participants. Of these, seven participants (25.9%) had an AOB. Five participants (18.5%) had evidence of gingival enlargement in the absence of evidence of plaque accumulation. Co-occurrence of gingival enlargement and delayed/ectopic/failed eruption was observed for two participants (Table 4).

Assessment of dental ageing

Dental age could not be estimated for one participant due to radiographic superimposition of the mandibular lateral incisors and canines. Sixteen participants (55.2%) had advanced dental age, 12 (41.4%) had delayed dental age, one participant had concordance between the chronological and dental ages (3.4%). Finally, twenty-six participants (89.7%) had a dental age within 18 months of their chronological age (Fig. 3).

Fig. 3
Fig. 3
Full size image

Dental age compared to the chronological age. In addition to the results displayed, one participant had a concordance between the dental and chronological ages

Molecular diagnosis of AI

Thirteen participants had an available molecular diagnosis, which were mapped to the identified dental anomalies (as summarised in Table 5). Given the small number of participants with an available disease-causing variant, these results are descriptive only and should be interpreted with caution. The origin of the disease-causing variant responsible for the AI diagnosis is presented in Table 6. Where this information was unavailable, this was recorded as unknown.

Table 5 Prevalence of different dental anomalies according to the pathogenic variant responsible for the AI diagnosis
Full size table
Table 6 Origin of pathogenic variant
Full size table

Discussion

Discussion of key results

This study combined radiographic and clinical data with genetic information for a well-characterised cohort and was unique in utilising this data in the systematic assessment of dental anomalies in AI. Access to high-quality clinical photographs and radiographs meant that no additional recruitment was necessary. Longitudinal treatment records were available for many of the participants, providing an extended timeline of clinical and radiographic data, which facilitated the assessment of multiple anomalies over time.

Pre-eruptive crown resorption was noted in 23.3% of the sample, appearing higher than in non-AI populations, where a prevalence of 3.54% has been reported (Gurunathan, Pandiar and Krishnan, 2025). Interestingly, of the six participants with a pathogenic variant of FAM83H, four had evidence of PECR. In their study of PECR in individuals with Junctional Epidermolysis Bullosa (JEB), Besa-Witto and colleagues (2025) utilised a range of imaging modalities, including conventional radiographs, and Cone Beam Computed Tomography (CBCT), to find several cases in which no resorption was identified on panoramic radiographs, but multiple resorptive lesions were observed on CBCTs. The absence of CBCT in this study, therefore, may have led to an underestimation of the true prevalence of PECR.

Surprisingly, taurodontism was reported for only two teeth (4.8%) in a single participant (4.5%). This prevalence was considerably lower than that reported in AI in existing studies (Adorno-Farias et al. 2019; Decaup et al. 2022; Kammoun et al. 2024). The Seow and Lai (1989) technique for assessment of taurodontism is susceptible to change in the crown height from post-eruptive changes, such as attrition and post-eruptive breakdown. As this influences the crown:root ratio, it risks compromising accuracy, potentially leading to underdiagnosing taurodontism.

Da et al. (2024) proposed using the cementoenamel junction as a reference point to calculate the body to root (B/R) ratio, instead of the occlusal surface used to calculate the crown-body to root (CB/R) ratio described by Seow and Lai (1989). Their study compared the B/R ratio with existing commonly used assessment methods described by Seow and Lai (1989) and Shifman and Chanannel (1978), across different age groups. They found that whilst the Seow and Lai method (CB/R ratio) was less susceptible to age-related changes than the Shifman and Chanannel method, the proposed B/R ratio was least susceptible. The authors recommended using the B/R ratio for assessment of taurodontism to minimise the impact of age-related changes. This may also be relevant when assessing taurodontism in individuals with AI, where rapid post-eruptive change to the clinical crown height may occur. Greater standardisation in the assessment of taurodontism in individuals with AI in future research would facilitate meaningful comparison across studies and populations.

The ability to assess individuals clinically, for example by recording a plaque-free score, would facilitate differentiation between plaque-induced gingival inflammation and gingival enlargement of alternative aetiology. Furthermore, future research should consider distinguishing between localised and generalised gingival enlargement. This would allow localised gingival enlargement, such as that associated with PECR, to be differentiated from full thickness gingival enlargement.

An AOB was seen in seven cases (25.9%), which is higher than the 2–7% reported in the general population, but at the lower end the 24–60% range reported for individuals with AI (Kammoun et al. 2024). In the assessment of AOB, a clinical history of any habits would have been helpful, to determine whether the AOB was likely to be skeletal or habit-related, such as digit-sucking. In the absence of this, the prevalence of AOB associated with AI could be overestimated.

Participants with advanced (N = 16) and delayed (N = 12) development were both observed. However, most participants had a dental age within 18 months of their chronological age, which is of no clinical consequence. This result is comparable to existing studies of dental maturation in AI, which found dental age not to be significantly delayed in comparison to healthy populations (Kirzioglu et al. 2019; Gabardo et al. 2020).

Finally, it is recognised that the Demirjian method may overestimate dental maturity internationally, as the reference standards are based on a French-Canadian cohort (Celikoglu et al. 2011). Nevertheless, it remains the most used method for dental ageing estimation and allows comparison to existing research, which supported its selection in this study. In the absence of a measurement tool based on a UK population, the use of multiple methods would allow comparison, and enable assessment of any difference in precision.

Limitations

Several limitations should be considered when interpreting the findings of this study. As no primary data were collected, the investigators did not have access to an unaffected comparison group. As such, it cannot be concluded whether the observed dental anomalies are specific to AI or merely reflect their prevalence in the wider population. In the absence of a comparator group, an association between the co-existence of dental anomalies and AI could not be fully established. In addition, reliance on the primary study restricted the sample size and, therefore, the statistical power. Furthermore, the small number of participants with a confirmed genetic diagnosis limited the ability to establish phenotype-genotype correlations.

As phenotypical classification of AI was based on radiographic and photographic analysis this may have lacked accuracy due to the missing overall clinical context. The binary phenotype classification used in this study (hypoplastic vs hypomineralised) inherently limits recognition of mixed presentations and, therefore, may not fully reflect the phenotypic variability for some individuals.

Finally, the sample was representative only of the local population and, therefore, may not reflect the variability expected in more ethnically diverse cohorts. To better establish genotype–phenotype correlations, multicentre and international studies are needed to capture a broader range of pathogenic variants and clinical presentations of AI.

However, despite these limitations, this study provides a valuable foundation on which further research can build.

Implications for clinical practice and further research

The findings of dental anomalies within the AI population reinforce the importance of comprehensive dental assessment for this patient group, including referral to specialist paediatric dental units where appropriate. The common co-occurrence of dental anomalies highlights the need for thorough clinical and radiographic assessment to facilitate early identification and management of concomitant anomalies and associated pathologies. The clinical significance of the identified dental anomalies varies. Some anomalies, such as PECR or failed eruption, can result in adverse outcomes and may have direct implications for treatment planning. Others, such as taurodontism, may not necessarily result in adverse oral health outcomes, or require specific intervention. However, their identification can provide relevant information in clinical assessment and treatment planning in the developing dentition, which may, therefore, inform subsequent management.

In investigating dental anomalies in participants with several molecular diagnoses, this study builds on previous research combining clinical and radiographic assessment with the molecular diagnosis to investigate genotype/phenotype relationships in AI. Wright et al. (2011) assessed the distribution of pathogenic variants in a large AI population, evaluating genotype/phenotype relationships. A recent imaging-genomics study combined clinical examination, panoramic radiographic imaging, and whole-exome sequencing in a non-syndromic AI cohort, proposing a novel radiographic approach for evaluating developmental defects of enamel (Leban et al. 2025). Next-generation sequencing has been combined with clinical phenotypic assessment to propose a new more accurate classification system, which includes phenotypic observations, mode of inheritance, and the molecular diagnosis (Bloch-Zupan et al. 2023).

A larger, adequately powered sample in future studies would reduce the risk of Type II error and strengthen the validity of findings on the prevalence of dental anomalies associated with specific pathogenic variants. Given the rarity of AI, multicentre collaborations will be essential to achieve sufficient statistical power and improve the generalisability of results to more diverse populations. Expanding the scope of clinical, radiographic and genetic data would also enable more detailed investigation of phenotype–genotype correlations.

Conclusions

This study identified several dental anomalies occurring alongside AI across participants with a variety of genetic variants. Dental age did not appear to differ substantially from chronological age. By characterising co-existing dental anomalies within this population, this study provides insight into potential patterns that may support early recognition, more accurate diagnosis and improved patient management. It cannot be concluded presently whether the observed dental anomalies are associated with AI or simply reflect their presence in the general population. Future research with larger, adequately powered samples, and an unaffected comparison group, will be critical to assess for correlations between the observed dental anomalies and AI.