Introduction

Molar-Incisor Hypomineralization (MIH) is designated as a qualitative defect of unknown etiology in the enamel development1,2. Since 2003, the European Academy of Pediatric Dentistry (EAPD) has proposed its first nomenclature to define a pathology of unknown etiology that affects one or more permanent molars and may include permanent incisors1.

As a potential oral public health concern, the prevalence of MIH became imperative to determine as a measure of interest in oral health programs. The prevalence of MIH was reported to range 2.8 to 40.2%, yet this inconsistency leads to a challenging interpretation and is mainly caused by the lack of standardization among clinicians/researchers3. As a result, the EAPD introduced a diagnostic and classification system for MIH, with the purpose of improving epidemiological assessments3,4.

Two systematic reviews have estimated the prevalence of MIH between 13.1% and 14.2, with significant variances amid regions5,6. Moreover, Schwendicke et al.5 estimated MIH prevalence on country scale via the Global Burden of Disease database, which may explain the variances between those regions. Additionally, both searches were conducted until mid 2017, and ever since, a number of new epidemiological studies have been published. However, other relevant information remains to be elucidated, namely the prevalence of moderate/severe cases, prevalence of molars and incisors affected and the prevalence of hypomineralization of the second primary molars (HSPM). For these reasons, conducting a new systematic review on the topic would be convenient and relevant globally.

In this sense, and given the increase research on the prevalence of MIH, we aimed to comprehensively investigate the global prevalence of MIH, as well as its associated characteristics.

Methods

Protocol and registration

The protocol for this systematic review was defined by all authors and registered at the National Institute for Health Research PROSPERO, International Prospective Register of Systematic Review (http://www.crd.york.ac.uk/PROSPERO, ID Number: CRD42021229435). We based our review design following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guideline7.

Focused question and eligibility criteria

We aimed to answer the following PECO question: “What is the global prevalence of MIH?”. The respective statements were as follows: Clinical/Epidemiologic studies in humans (P, Population); Diagnosis of MIH (E, Exposure); No MIH (C, Comparison); Prevalence of MIH (O, Outcome).

The primary aim was the prevalence of MIH. The secondary aims were the prevalence of moderate/severe MIH cases, prevalence of molars and incisors affected and the prevalence of HSPM.

Studies were eligible for inclusion based on the following criteria: (1) Observational studies reporting the prevalence of MIH; (2) Studies with clear reporting of MIH definition; (3) Subjects with no systemic disorders; (4) Studies including both genders.

In contrast, studies based on specific population, for example, children born preterm, studies which only reported on primary molars, and studies which focused on non-representative samples (e.g., institutionalized populations, particular professions, those with specific dental outcomes like high caries experience, among others) were excluded.

Search strategy

Identification of studies for this systematic review was performed through detailed search strategies developed for each database (Pubmed, MEDLINE, CENTRAL [The Cochrane Central Register of Controlled Trials], Web of Science, SciELO [Scientific Electronic Library Online], EMBASE [The Excerpta Medica Database], LILACS [Latin-American scientific literature in health sciences], and TRIP [Turning Research Into Practise]) up to July 2021. Our search strategy was based on the following algorithm: "(hypomineralization OR hypomineralisation OR hypomineralized OR hypomineralized OR hypoplasia OR demarcated OR opacities OR MIH OR cheese molars) AND (survey OR questionnaire OR cross-sectional OR prevalence OR frequency OR population OR sample OR sampling) AND (molar OR molars OR incisors)".

Study selection

Study selection was assessed independently by two investigators (LBL and JB), who performed the assessment of titles and/or abstracts of retrieved studies. For measurement reproducibility purposes, inter-examiner reliability following full-text assessment was calculated via kappa statistics. Any disagreements were resolved by discussion with a third author (VM).

Data extraction process and data items

Data extraction was performed by two reviewers independently and in duplicate (LBL and JB). The agreement between the reviewers was assessed by Kappa statistics. Any paper classified as potentially eligible by either reviewer was ordered as a full text and independently screened by the reviewers. All disagreements were resolved through discussion with a third reviewer (VM). The following information was gathered in general description, research characteristics, methodology, and outcome measurements. The following standard information was extracted from each eligible study: first author’s name, year of publication, year of study conduct, country and place (region, city) of sampling, setting of sampling, sampling strategy, case definition, setting, observation setting, sample size, age of participants, total sample size, prevalence estimation, sex-specific sample size and prevalence (if available), the diagnostic criteria of MIH, mean number of affected teeth, and funding. Also, severity of MIH was registered whenever studies reported it. We considered studies that have defined or used classifications that considered severe cases of MIH as having demarcated enamel opacities with breakdowns, caries, persistent/spontaneous hypersensitivity affecting function and strong aesthetic concerns.

We applied no publication year nor language restrictions. Grey literature was searched via http://www.opengrey.eu/. If not reported, corresponding authors were contacted to obtain baseline data.

Risk of bias (RoB) assessment

The Newcastle–Ottawa (NOS) Scale for case–control studies was used by two calibrated reviewers (LBL and JB). For calibration purposes, a random sample of 10 studies was assessed and reassessed 2 days later (to calculate Cohen's kappa). We have categorized studies as of low RoB (with 7–9 stars), moderate RoB (studies with 5–6 stars), and high RoB (with less than 5 stars) (as previously performed8,9). If any doubt occurred, they were resolved by discussion with a third author (VM).

Summary measures and synthesis of results

We began by conducting a prior sensitivity analysis to understand if studies reporting MIH with the 2003 case definition would differ from other alternative case definitions. Predefined tables were prepared to collect continuous data, mean values and standard deviations (SD). Random-effects meta-analysis and forest plots of prevalence were calculated in R version 3.4.1 (R Studio Team 2018) using ‘meta’ package10, through DerSimonian-Laird random-effects meta-analysis. Subgroup meta-analysis was conducted for two reasons: (a) comparing the EAPD case definition with other alternative methods; (b) comparing continental prevalence of MIH. Also, a meta-analysis of binary outcome data comparing females and males prevalence was performed. Heterogeneity and publication bias were computed via I2 test statistics (p < 0.1) and Egger’s significance test, respectively11. Substantial heterogeneity was considered when I2 statistics exceeded 50%11. In meta-analysis with 10 or more studies included, we analyzed publication bias11. Meta-regressions were conducted using continuous variables to appraise potential sources of heterogeneity, such as sample size, female/male ratio (FMR), geographic location (latitude and longitude) and year of publication. The regression approach also allowed to quantify the percentage of heterogeneity that could be explained by that variable. All tests were two-tailed with alpha set at 0.05 except for the homogeneity test whose significance level cutoff was 0.10 due to the low power of the χ2 test with a limited number of studies. Estimates were described with 95% confidence interval (CI).

Additional analyses

We employed the Strength of Recommendation Taxonomy (SORT) to appraise the strength and quality of the evidence12.

Results

Study selection

The online search retrieved strategy 2290 possibly relevant publications. After duplicates removal, 357 papers were judged against the eligibility criteria, and 1576 were excluded after titles and/or abstracts review. Among 138 articles assessed for full paper review eligibility, 22 articles were excluded with the respective reasons for exclusion detailed in the Supplementary S2. As a result, a final number of 116 observational studies were included for qualitative synthesis (Fig. 1). Inter-examiner reliability at the full-text screening was considered very substantial (kappa score = 0.915, 95% CI 0.895–0.925).

Figure 1
figure 1

PRISMA flowchart.

Studies characteristics

The characteristics of the included studies are presented in Table 1. We identified 116 different cohorts13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111,112,113,114,115,116,117,118,119,120,121,122,123,124,125,126,127,128 from fifty different countries, across five continents. The year of publication of the included studies ranged between 2003 and 2021. Overall, a total of 135,181 participants were included in this review, being 52,876 girls and 52,872 boys, even though 18 manuscripts did not report on sex distribution. Thirty-four papers did not report the prevalence of MIH according to sex. Seven studies reported data on HSPM. Most studies recorded the MIH-related hypomineralization according to the diagnostic criteria of the EAPD case definition1, and others indices were also used such as the modified Developmental Defects of Enamel (mDDE) index129, the Kemoli88, Mathu-Muju and Wright130 criteria, and the diagnostic criteria of Cho et al.17,92.

Table 1 Characteristics of the included studies.

Three cohorts had their data reported in more than one article (Petrou et al.78 and Petrou et al.109; Balmer et al.13 and Balmer et al.14; Negre-Barber et al.110 and Negre-Barber et al.111); thus, these papers were grouped under a single name study as follows: Petrou et al.78,109; Balmer et al.13,14; and, Negre-Barber et al.110,111. Also, three studies reported in the same study two cohorts: Biondi et al.16 reported data for Buenos Aires (Argentina) and Montevideo (Uruguay); López Jordi et al.108 reported data for Buenos Aires (Argentina) and Montevideo (Uruguay); and Irigoyen-Camacho et al.62 reported data for both 2008 and 2017 cohorts.

Assessment of RoB within studies

Inter-examiner reliability at RoB analysis was considered very substantial (kappa score = 0.885, 95% CI 0.865–0.905). The RoB for observational studies, with the NOS, ranged from 6 to 9 stars (Supplementary S3). After the assessment, forty-eight had the maximum score (9/9). Additionally, fifty-three and six articles scored 8/9 and 7/9, respectively. Only one paper was of moderate RoB (score = 6/9). The main sources of inconsistencies arose from the representativeness of the cases. While all articles succeed to apply an adequate MIH case definition, selection of control, ascertainment of exposure, equal method of assessment of cases and controls and non-response rate (100.0%, n = 113), studies failed to provide adequate representativeness of the cases (48.7%, n = 55), two studies failed the definition of controls (1.8%) and 8.8% only provided information regarding MIH and not any other variables (n = 10).

Prevalence of MIH

A first subgroup meta-analysis confirmed that estimates from studies using the EAPD 2003 classification were significantly different from studies with alternative classifications (categorized as ‘others’) (p = 0.0061) (Supplementary S4). This initial analysis comprised 133,734 participants. Thus, we proceeded with the analyses using only studies reporting prevalence through the 2003 MIH case definition.

Global prevalence (primary outcome)

The overall prevalence of MIH for a total of 113,089 participants was estimated at 13.5% (95% CI 12.1–15.1, p < 0.001) (Table 2), with high heterogeneity (I2 = 98.0%) (Supplementary S5). Cumulative meta-analysis confirmed the overall estimate was not influenced by a particular study or group of studies (Supplementary S6A). We further confirmed the non-existence of influential studies through leave-one-out meta-analysis (Supplementary S6B).

Table 2 Meta-analysis on the prevalence of MIH cases, severity of cases, number of affected molars, cases with affected incisors and HSPM.

The prevalence of moderate to severe cases of MIH was estimated at 36.3% (95% CI 29.9–43.2, I2 = 95.2%) (Table 2, Supplementary S7). Detailed information on the definition of severity in each study was collectively presented in Table 3. Regarding the number of affected molars, estimates point to 24.3% of cases with one molar (95% CI 18.9–30.7, I2 = 94.2%), 26.7% of cases with two molars (95% CI 23.9–29.7, I2 = 65.0%), 18.1% of cases with three molars (95% CI 13.8–23.3, I2 = 90.0%) and 26.8% of cases with four molars (95% CI 21.1–34.7, I2 = 94.0%) (Supplementary S8-S11). The cases with affected incisors were estimated at 38.7% (95% CI 32.1–45.8, I2 = 93.2%) (Supplementary S12), while cases with both molars and incisors were estimated at 42.1% (95% CI 34.9–50.0, I2 = 95.5%) (Supplementary S13). Lastly, the prevalence of HSPM cases was estimated at 3.6% (95% CI 1.9–6.8, I2 = 96.3%) (Supplementary S14). All the latter results had high heterogeneity.

Table 3 Detailed case definition of MIH severity for each study with the respective reported prevalence.

Sex and geographic location (secondary outcomes)

We further analyzed whether the prevalence results were influenced by study sample size, female/male ratio, geographic location (latitude and longitude) and year of publication (Table 4).

Table 4 Meta-regression analyses on the effect of female/male ratio (FMR), latitude, longitude and year. Values are provided as estimate (Standard Error) [Variance explained (%)].

Overall, MIH was influenced by the study sample size explaining 7.7% of the accounted heterogeneity, respectively. The year of publication (estimate = − 0.09, p = 0.023) demonstrated a slight influence on the prevalence of MIH cases with one molar affected (explained 11.6% of heterogeneity).

We then explored whether the prevalence between males and females would differ regarding MIH. Meta-analysis confirmed the latter result from meta-regression that MIH is not sex-related and females and males present a non-significant difference on the prevalence of MIH (0.986, 95% CI 0.940–1.035, I2 = 32.6%, p = 0.564) (Fig. 2).

Figure 2
figure 2

Forest plot of meta-analysis comparing MIH prevalence of female versus male participants.

We further explored the prevalence of MIH per continent (Table 5). Among the five continents analyzed, America was the continent with highest prevalence (15.3, 95% CI 12.8–18.3, p < 0.001, I2 = 96.3%) and Asia had the lowest prevalence (10.7, 95% CI 78.5–13.5, p < 0.001, I2 = 98.7%). The differences between continents (test for subgroup differences) were not significant (p = 0.1643).

Table 5 Meta-analysis on the prevalence of MIH per continent.

Additional analyses

No publication bias was detected in the overall analysis (Table 2), except for the prevalence of cases with one molar affected (p = 0.004).

Using the SORT recommendation, we concluded the estimates obtained are classified as SORT A, that means, the results provide high level of confidence.

Discussion

Summary of main findings

The results of the present systematic review estimated a pooled prevalence of MIH at 13.5%. The moderate to severe cases of MIH were estimated at 36.3% of all cases. Having three molars affected with MIH is the least probable situation and affected incisors were seen in 36.6% of the cases. The prevalence of HSPM in MIH cases was estimated at 3.6%. The sample size was a significant source of heterogeneity for the overall MIH prevalence and the year of publication for the prevalence of one molar affected. Sex, year of publication and geographic location were not deemed influential factors in almost all the results. Continents showed no different prevalence on MIH, with the American continent displaying the highest prevalence and the Asian continent the lowest.

Quality of the evidence and potential biases in the review process

Overall, these results were categorized with a SORT A recommendation, which means that all studies found coherent conclusions regarding the prevalence of MIH and that these results are consistent and good-quality patient-oriented evidence. Furthermore, this is the first systematic review providing pooled estimates on molars and incisors affected with MIH and HSPM cases.

As previously presented, two previous systematic reviews have focused on the prevalence of MIH. Overall, our results provided similar prevalence to the one reported by Schwendicke et al.5 (13.1%) and slightly above from Zhao et al.6 (14.2%). However, comparing with the latter, the present systematic review expanded the number of countries (49), confirmed the downgrading of alternative case definition of MIH to the overall pooled estimate (while the previous reviews combined classifications), and present new prevalence estimates concerning clinical characteristics of MIH (molas and incisors affected, severity and HSPM).

Regarding the comparison between sexes, our result fully align with those by Schwendicke et al.5 (OR 0.92; 0.81–1.04) and Zhao et al.6 (regression estimate = 0.005, p-value = 0.938), which means that both girls and boys present similar distribution of MIH lesions.

When analyzing the prevalence among continents, the comparison with literature is not reasonable as we only accounted for the EAPD classification, and this explains why Oceania had no studies available (despite two publication by Mahoney et al.118,119). Also, in Zhao et al.6, Africa was the continent with lowest prevalence, yet in our review Asia had the lowest prevalence. The American continent includes for the first time studies from the United States of America and Mexico which may explain a decrease in MIH prevalence from the two previous studies, however remains as the continent (super-region) with highest prevalence.

Regarding the methodological aspects, by comparing the EAPD with alternative diagnostic methods as a subgroup analysis we confirmed the downgrading potential of alternative methods to the overall estimates. Thus, this step methodological assortment into the analyses despite the substantial heterogeneity from the meta-analytical estimations. Also, our analyses on the severity, teeth affected and HSPM were severely reduced because this sort of data is still scarce. Future studies shall provide extensive information on these characteristics to confirm these results. Also, we were unable to explore hypothetical MIH-related factors (both medical, sociodemographic and environmental) once again because of the lack of relevant information, and this should be taken into account in future epidemiological studies.

All in all, readers must bear in mind that although the overall prevalence seems to be constant over the time, new prevalence data has been pooled that contribute to understand the clinical characteristics of this enamel defect entity.

Strengths and potential limitations

This systematic review was conducted following PRISMA a strict guideline for data reporting, a comprehensive literature search and a meticulous predefined protocol. Furthermore, prior to any analysis, we compared the EAPD case definition with other classifications than the EAPD, and we confirmed substantial differences with a downgrading in prevalence when alternative methods were applied. We have attempted to explore ways to mitigate heterogeneity, and all studies used to compute estimates (and that employed the EAPD case definition) were of high methodological quality. Another advantage of this study is that we have expanded the search for potential sources of heterogeneity with the addition of geographic measures and the further assessment into the new prevalence estimates. Also, the number of included participants has increase, which is logical given the increase in studies included, yet this is a point to keep in mind.

Nevertheless, there are a number of limitations important discussing. Almost half of the studies had not fulfilled the criterion of representativeness and this is a point where future studies shall be careful. These results should be prudently analyzed because of the elevated heterogeneity observed in some of the reported estimates, though from our analyses the heterogeneity mostly derives from the variability between regions already discussed in a previous study assessing meta-analysis of prevalence131, rather than the sources of heterogeneity considered as proven through meta-regression.

Also, a number of studies have not employed the EAPD case definition for MIH and after the subgroup analysis aforementioned they were not accounted for the analyses. It is essential that there is a standardization of the classification used, which is a topic already widely discussed in the literature132. Ergo, and given the results of the present systematic review, several challenges may emerge. First, a global partnership between all geographic representative associations shall be attained, to ultimately ensure a standardization of MIH reporting and, certainly, will encourage new and updated epidemiological and clinical data. Second, this suggested consensus will clarify the terminologies and guidelines towards a global alliance that will benefit all people affected by MIH. All in all, these may contribute to overcoming the lack of epidemiological data and a still methodologically unsettled reporting approach.

Only a percentage of the overall included studies reported data on the teeth affected with MIH, the severity of cases or HSPM cases. Several classifications for the severity of MIH have been proposed133,134, and some date before the EAPD 2003, such as Leppäniemi135 or the Wetzel & Reckel scale117,136. Moreover, the MIH Treatment Need Index (MIH-TNI) was recently presented, which is137 part of the Wuerzburg MIH concept. Nevertheless, the lack of a homogeneous definition may have contributed to the heterogeneity of results, making it urgent to establish a consensual severity classification.

Hence, future studies should focus on data on these prevalence characteristics to deepen our knowledge regarding the specifics of MIH. These information are of the utmost relevance for clinicians and may aid the development and implementation of future oral health programs.

Conclusion

The estimated prevalence of MIH was estimated at 13.5%. Moderate to severe cases of MIH were estimated at 36.3%. Affected incisors were seen in 36.6% of the cases. The prevalence of hypomineralization of the second primary molars in MIH cases was estimated at 3.6%. Overall, these results were categorized with a SORT A recommendation.