Abstract

Introduction. Although numerous amounts of high-level evidence were present, they solely emphasized the tooth-level prevalence of three-rooted permanent mandibular first molar. Global patient-level prevalence and bilateral symmetrical distribution of this type of teeth were needed to be tackled across the world. The research question was “What is the global prevalence of three-rooted permanent mandibular first molars?” Materials and Methods. In vivo epidemiological studies undergone with Cone Beam Computed Tomography (CBCT) were eligible. The proportions of the prevalence of three-rooted permanent mandibular first molars were presented in the forest plots by random effect model. The calculation was performed with MetaXL version 5.3. Subgroup analysis, sensitivity analysis, and publication bias method were also calculated. Results. Seventy-two studies from 31 countries were selected for both qualitative and quantitative analyses. 26302 patients and 37994 permanent mandibular first molars were included in the analysis. 9% of permanent mandibular first molars all over the world demonstrated 3 roots. These three-rooted teeth were found in 10% of the world population, more than 45% of which revealed bilateral symmetry of that anatomy. Right-side dominance and no sexual dimorphism were seen in the distribution of three-rooted permanent mandibular first molars. Global tooth-level prevalence of Radix Entomolaris and Radix Paramolaris was 12% and 0.1%, respectively. Conclusion. The prevalence of three-rooted permanent mandibular first molars (PMFMs) was influenced by different geographical locations across the world and also by widespread habitation of the Mongoloid descent. The authors postulate that globalization, together with blending among ethnicities, may have a great impact on the reduction or accentuation of the anatomical significance in some populations.

1. Introduction

The term “three-rooted permanent mandibular first molar (PMFM)” is more generalized and less specific than others, based on the counting number of roots. So, most of the researchers used the more famous one, “Radix Entomolaris (RE),” instead of the former. “Radix” means “root” [1], and “Ento” denotes “Inside” [1], both of which in turn can be understood as “the root merging from the lingual side.” Comparatively, Radix Entomolaris can be stated as the accessory root originating from the lingual root trunk of the molar. Another word “Para” means “Beside” [1], which can be recognized as “buccal side of the mouth.” As a result, Radix Paramolaris (RP) can be defined as “a root originating from the buccal side of the molar.” Both of the two terms are constituted as three-rooted PMFMs. Recently, some evidence found that there could be many positions of an accessory root along the root trunk of PMFM, frequently distolingual or centrolingual and rarely distobuccal third roots [24] (Figure 1). However, the American Association of Endodontists (AAE) [5] recorded RE as a distolingual root and RP as a mesiobuccal one. Although definitions of the conditions should further be modified and more meaningful, we used the more generalized one, “three-rooted PMFM,” in our present meta-analysis.

From a clinical perspective, radix mandibular first molars display some significant features. Up to 32% of these teeth showed an additional tubercle or sixth cusp in its occlusal anatomy [6] in contrast to normal five-cusped crown morphology. They had averaged 0.3 mm wider in buccolingual measurement at the distal surface of the crown [7] than that of its two-rooted counterparts. Additionally, intercuspal distances between the distolingual cusp and the adjacent distobuccal and mesiobuccal cusps were slightly wider in three-rooted PMFM than in two-rooted one [7].

Buccolingual curvature of the third root of three-rooted PMFM comprised more than 30 degrees measured by Schneider method [8]. Nearly 60% of radix roots displayed this severe curve [8]. Mesiodistal curvature of these roots was less prominent than the buccolingual curve [8]. Some investigators found that an excessive degree of curvature was the origin of the separation of the rotary endodontic Ni–Ti files [9]. One systematic review figured out that the thermoplasticized method was superior in adaptation between root canal wall and gutta-percha than the lateral condensation method [10]. Most of the investigators of primary studies in this review postulated that the method may be suitable for complex anatomy such as high curvature.

Straight extra root could be more readily overlapped by distobuccal root than the curve one in periapical X-ray (PA) [11]. Consequently, the curvature of the distolingual root canal was more prominent in the proximal view than in the clinical view (PA view) [11]. In proximal view, buccolingual curvature was classified into straight, coronal curve, and apical curve in the buccolingual plane [12] (Figure 2).

The size of the distolingual roots may vary from short conical structure to normal root length up to 8 mm from cervix to apex [12] (Figure 3). Sometimes due to its tiny dimension, there was more apical structure from distobuccal and mesiobuccal roots that needed to be resected to access distolingual root during endodontic microsurgery [13].

Trifurcation, resulting from the presence of the extra root, was located approximately 1 mm apical to mesiodistal furcation [14]. This may worsen severe periodontitis in comparison with its two-rooted counterpart [15]. Surprisingly, the point of separation of Radix Paramolaris (RP) was commonly detected at the apical third of the root in an Israeli population [2] (Figure 1(f)).

Anatomy of the pulp chamber floor was switched from a normal triangular pattern in two-rooted PMFMs to trapezoidal in three-rooted PMFMs [4]. The orifice of the extra root was located approximately or more than 3 mm lingual to distobuccal orifice [8]. However, excessive deposition of secondary dentine could unite the roof and floor of the pulp chamber, thereby interfering with the orifice of the distolingual root. Intricate root canal anatomy was seen up to 26% of radix root [4]. RE usually has one canal with a nearly circular cross section [4].

Bilateral occurrence of PMFMs with distolingual roots was positively associated with the increase in the prevalence of the complicated root canal in mandibular lateral incisors [16] and the presence of an additional root and C-shaped root canal anatomy in mandibular first premolars in the Taiwanese population [17]. These studies were done by the CBCT in vivo method.

The empirical in vivo study applied the extraction method in the assessment of RE [18]. Nowadays, the method is not suitable for current clinical practice. And patient-level prevalence and bilateral symmetry of three-rooted PMFM cannot be identified by the extraction method. Then, this type of research was replaced with an in vivo X-ray methodology. Some investigators [19] suggested that two radiolucent lines, corresponding to root canal and periodontal ligament of an additional root, were diagnostically detected in PA view. This may be seen crossing the distobuccal root in the view. However, it needs 25–30-degree mesial horizontal angulation of the X-ray beam to capture RE [11]. Although some investigators used the X-ray method correctly, others confirmed RE only in one sagittal plane X-ray. One study argued that almost 60% of RE was sensitively detected in PA view [11]. Nowadays, micro-CT is a gold standard, which needs a priori extraction. But, the surgery cannot guarantee the preservation of intact root structure. So, this can lead to the liberation of underestimated and nongeneralizable findings. Finally, in vivo CBCT research methodology is a convenient solution to tackle the incidence of three-rooted PMFM in both research and clinical diagnosis. The axial slice of CBCT can ease the reliability to count the number of roots in PMFM. The data from such methodology could be readily recorded, available for research, and retrospectively extracted from both databases of public and private sectors.

The growing evidence demonstrated the incidence of the additional root in different types of teeth. The accessory root revealed approximately 0.8% at the palatal root of permanent maxillary first molars in the Greece-Turkish population [20], around 8% in permanent mandibular second premolars of the Indian population [21], up to 12% in permanent mandibular canines of the Iranian population [22], and almost 16% in permanent mandibular first premolars of the African-American population, respectively [23]. For PMFM, an earliest systematic review of the scope investigated that the pooled prevalence of three-rooted PMFM revealed 13% with the highest value of 22% in the Chinese population [24].

The recent systematic reviews of the specific populations pointed out that the incidence of three-rooted PMFM demonstrated up to 3% in Brazil [25] and almost 6% in Saudi Arabia [26]. One systematic review [27] investigated that the pooled prevalence of 3-rooted PMFMs was 8.9% resulting from the meta-analysis of 35 studies across the world. Although a huge amount of high-level evidence is currently being published in the scope of endodontic anatomy, they solely emphasized the tooth-level prevalence of three-rooted PMFMs. And some researchers used the studies of different research methodologies, pooling both in vivo and in vitro results. This may harm the consistency of the pooled estimate.

As a consequence, patient-level prevalence and bilateral symmetrical distribution of three-rooted PMFMs were needed to investigate across the world. Additionally, different frequency of the prevalence across Asia’s regions was still on demand, although a single systematic review for Saudi Arabia intervened recently. In addition, RE is more clinically significant than RP. Curiously, researchers and clinicians usually mark the prevalence of RE. For that reason, there was a lack of evidence pooling the prevalence of RP, which also has anatomical significance. Both RE and RP can be postulated as the term “three-rooted PMFM.”

The research question of the present meta-analysis was “What is the global prevalence of three-rooted PMFMs?”

The objectives were as follows:(i)To investigate the global tooth-level prevalence of three-rooted PMFMs.(ii)To estimate the global patient-level prevalence of three-rooted PMFMs.(iii)To detect global patient-level bilateral symmetry of three-rooted PMFMs.

2. Materials and Methods

2.1. Selection Criteria
2.1.1. Type of Included Studies

The studies eligible for inclusion were as follows:(1)In vivo.(2)Undergone by means of Cone Beam Computed Tomography (CBCT) or Spiral Computed Tomography (SCT) methodology.(3)Cross-sectional.(4)Prospective or retrospective.(5)Analytical or descriptive.(6)Primary or secondary objectives, including “number of roots” or the term inferred from the variations in the number of roots.

2.1.2. Characteristics of Excluded Studies

The features of excluded studies were as follows:(1)Studies that did not report the age of the patients or developmental condition of the root apex(2)In vitro(3)Methods other than CBCT or Spiral CT(4)Primary or secondary objectives not including “three roots” or “number of roots”(5)Studies that did not present “the number of roots”(6)Studies used more than one research methodology (e.g., CBCT + Periapical X-ray)(7)Secondary data analysis, book chapter, case reports, narrative reviews, editorials, opinions, letters to the editor, animal studies, and personnel communication

2.1.3. Target Conditions (Numerator Variables)

(1)Three-rooted PMFM (objective Ι, subgroup analyses ΙV and V, and sensitivity analysis VΙ)(2)Patient with three-rooted PMFM (objective ΙΙ)(3)Patient with the bilateral presence of three-rooted PMFMs (objective ΙΙΙ)

2.1.4. Population

Patient with permanent mandibular first molars (PMFMs) that had no external (or) internal resorption, no other root anomalies, fully formed root apex and the age, which was compactible to completed root development, was included.

The denominator variables were as follows:(1)Total number of teeth (PMFM) for objective Ι, subgroup analyses ΙV and V, and sensitivity analysis VΙ (tooth-level prevalence)(2)Total number of patients for objective ΙΙ and the total number of patients with three-rooted PMFMs for objective ΙΙΙ (patient-level prevalence)

2.2. Search Strategies

The literature was searched in the frame of condition, context, and population. The search words were as follows:(1)Condition. Three roots, three-rooted, 3-rooted, 3 roots, third root, three separate roots, distolingual root, two distal roots, extra distal root, extra distolingual root, DL root, DLR, DL, extra DL, extra root, additional root, supernumerary root, 2R2C, Radix Entomolaris, Radix Molaris, Radix Paramolaris, radix first molars, 3RM1, and radix molars(2)Context. CBCT, Cone Beam Computed Tomography, Cone Beam CT, Spiral Computed Tomography, and Computed Tomography(3)Population. Permanent mandibular first molars, permanent lower first molars, mandibular first molar, lower first molars, lower molar, mandibular molars, lower posterior teeth, mandibular posterior teeth, mandibular 1st molar, lower 6, and PMFM

The search strategies were decorated through 1 AND 2 AND 3 through the selected databases. PubMed, Goggle Scholar, Research Gate, ProQuest, and LILACS were included. There was no language and time restriction. Searching was undertaken until October 30, 2021. Magnifying from reference lists of the articles and Goggle Search was also done.

2.3. Data Collection

We accessed the formerly stated data for each study: sample characteristics (total number of PMFMs in the study sample, sample size (patients), number of three-rooted PMFMs, patients with three-rooted PMFM, bilateral symmetrical distribution of three-rooted PMFM in patients with PMFMs on both sides of the mandible, total number of teeth in male and females patients, total number of teeth on left and right sides of the mandible, total number of Radix Entomolaris and Radix Paramolaris, gender, age, geographical locations, country, and ethnicity), study characteristics (sampling frame, randomization, sample size calculation, and statistical analysis), and setting (CBCT scan, voxel size, field of view (FOV), mA, slice thickness, kVp, type of viewer software, type of examiners, and reliability test).

Manuscripts of some studies were translated from native languages to English by using the Mobile Application of Goggle Translate.

2.4. Assessment of Methodological Quality

We assessed the methodology of the selected studies by the Joanna Briggs Institute (JBI) Critical Appraisal tool for systematic reviews of prevalence studies. The appraisal tool contains nine questionnaires. Of these, the 9th question, which is routinely used to assess the response of the participants, was irrelevant for this meta-analysis and so was excluded. As a result, we attempted 8 JBI questionnaires. The selected articles were assessed and subgrouped as high risk of bias (JBI score ≤49%), moderate risk of bias (JBI score ranging from 50% to 69%), and low risk of bias (JBI score >70%) [28].

Together with the methodology quality assessment, the following domains of research methodology of included studies were subjectively analyzed, then listed, and graphed: sampling frame, randomization, sample size calculation, complete reporting of context and settings, reasonable population coverage, validity of measurement, reliability of measurement, complete outcome reporting, and appropriate statistical analysis. This was done by transforming the original data to proportion and then graphing.

2.5. Statistical Analysis

For the descriptive and qualitative purpose, the JBI scores were added together by the influence of the 8 questions mentioned above and transformed into proportions. The three formulas were as follows: for objective Ι. for objective ΙΙ. for objective ΙΙΙ.

Occasionally, patient-level prevalence data, which was not reported in primary studies, was back-calculated from unilateral and bilateral distributions of three-rooted PMFMs.

The proportions of the prevalence of the individual study were calculated and presented in the forest plots by random effect model. The estimation was calculated in an Excel spreadsheet and MetaXL version 5.3. Data extraction and back-calculation were undertaken whenever direct usage of the data was not possible from primary studies. P-value was agreed upon as 5%.

Tests for heterogeneity were Q and I2 statistics. I2 is the calculation of between-studies heterogeneity. The final results were pooled as global tooth-level prevalence of three-rooted PMFM, global patient-level prevalence of three-rooted PMFM, and global patient-level bilateral symmetry of three-rooted PMFM (objectives Ι, ΙΙ, and ΙΙΙ).

To explore heterogeneity, subgroup analyses were done through the prevalence of three-rooted PMFM according to different geographical locations across the world in addition to the pooled estimates of objectives Ι, ΙΙ, and ΙΙΙ. Additionally, subgroup analyses were undergone with regard to left and right and gender distributions (subgroup analyses ΙV and V).

To assess the pooled estimates of the tooth-level prevalence of RE and RP, sensitivity analysis (VΙ) was undergone by the exclusion from the selected studies which did not report “Radix Entomolaris,” “Radix Paramolaris,” and the terms matched with “RE” and “RP.”

2.6. Publication Bias Methods

Visual inspection of funnel plot asymmetry was the test for publication bias in the review. The x-axis of the funnel plot was set as double arcsine prevalence. Arcsine transformation was needed with the data of extreme values such as 0 or 1. Otherwise, values, out of the range of 0 and 1, which mean 0% and 100%, could be included in the confidence interval of the proportion [29]. Precision resulting from the inverse of Standard Error (SE) was set at the y-axis of the plot.

In MetaXL, funnel plot asymmetry was confirmed by Doi plot and LFK index for publication bias. Doi plot indicates “no asymmetry” (no publication bias), “minor asymmetry” (minor publication bias), and “major asymmetry” (major publication bias). Beyond ±1 of the LFK index describes the presence of publication bias [29].

3. Results

The proposal of the present meta-analysis was registered in PROSPERO and available at https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42022302195. The registration number is CRD42022302195.

Selection and exclusion of the records were demonstrated in PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flow diagram (Figure 4). A total of 72 studies were selected for both qualitative and quantitative analyses.

Qualitatively, 10 studies presented a JBI score of 5/8, 27 studies a score of 4/8, 24 studies 3/8, 7 studies 2/8, and 4 studies 1/8. As a result, we divided these into two categories: JBI scores 4 + 5 and ≤3. Thirty-seven studies obtained an average JBI score of 53.38%, indicating a moderate risk of bias, and 35 studies obtained the average score of 32.14% comprising high risk of bias.

The reported research methods of all included studies were categorized in Figure 5.

Table 1 presents the global prevalence of three-rooted permanent mandibular first molars with population, country, number of patients, number of teeth, geographic location, the condition termed in primary studies, settings, and study design.

26302 patients were included in the systematic review, of which 10003 were males and 11242 were females. Seventeen studies enclosed the gender status of study participants.

37994 permanent mandibular first molars were involved in this review. Four studies did not present the number of teeth. The number of patients with three-rooted PMFMs was included in the objectives of these studies.

A cross-sectional descriptive retrospective design was applied in 58 studies, 13 cross-sectional analytical retrospective studies, and only 1 cross-sectional analytical prospective study.

Nineteen studies presented that their outcome of interest was similar to the primary objectives (three-rooted PMFM) of the current meta-analysis. The remaining 53 studies showed “the number of roots” as a secondary objective.

Nine studies were conducted in South Korea, 8 in China, 6 in Turkey and Saudi Arabia, respectively, 5 in Taiwan and Iran individually, 4 in Brazil, 3 in Malaysia, and 2 in India, Portugal, and Chile particularly. Only one study was individually selected from Hong Kong, Japan, Egypt, South Africa, United States, Italy, Belgium, Serbia, Spain, France, Russia, Greece, Pakistan, Nepal, Thailand, Vietnam, Iraq, UAE, Yemen, and Israel.

Sixty-eight studies were published in English. Out of these, 4 studies were translated from native languages to English. Of these, 1 study from China was translated from Chinese to English, 1 Iranian study from Kurdish to English, 1 Saudi Arabia study from Arabic to English, and 1 from Japanese to English.

One study compared Portugal and Chinese populations. One study was conducted with both Belgium and Chilean populations. One study used both Saudi Arabian and Indian samples for comparative purposes.

Three-rooted PMFMs were not found in 5 studies presenting zero prevalence of this morphology.

3.1. Global Tooth-Level Prevalence of Three-Rooted PMFM (Figure 6)

Sixty-eight studies estimated tooth-level prevalence of three-rooted PMFM. The total number of teeth in the meta-analysis was 37994, in which 5503 three-rooted PMFMs were found.

Global tooth-level prevalence of three-rooted PMFM was 8.85% (95%CI: 6.60%−11.39%) (Q = 4706.52, , and I2 = 99%) by means of a random effect model. The occurrence ranged from 0% to 29% across the world.

By the subgroup analysis in accordance with the geographical locations, East Asian population revealed 24.1% (95% CI: 23%–25.2%) (range 15%–29%) (Q = 70.38, , and I2 = 67%), Southeast Asia 13% (95% CI: 9%–17.7%) (range 8%–22%) (Q = 22.34, , and I2 = 82%), South Asia 4.7% (95% CI: 2.5%–7.6%) (range 1%–9%) (Q = 28.16, , and I2 = 86%), West Asia 4.4% (95% CI: 3.1%–5.8%) (range 2%–13%) (Q = 55.969, , and I2 = 80%), Europe 2% (95% CI: 1.3%–2.9%) (range 0%–8%) (Q = 64.246, , and I2 = 78%), America 1.8% (95% CI: 0.2%–4.6%) (range 0%–6%) (Q = 74.096, , and I2 = 92%), and Africa 0.9% (95%CI: 0.3%–1.9%) (range 0.5%–1.1%) (Q = 0.451, p = 0.5, and I2 = 0%).

3.2. Global Patient-Level Prevalence of Three-Rooted PMFM (Figure 7)

Forty-four studies reported patient-level prevalence of three-rooted PMFM. The total number of patients in the meta-analysis was 16836, of which 2535 patients had three-rooted PMFM.

Global patient-level prevalence of three-rooted PMFM was 10.3% (95%CI: 6.9%–14.4%) ( Q = 2874.974, , and I2 = 98%) by random effect model. The prevalence ranged between 0% and 35% around the world.

The patient-level prevalence of three-rooted PMFM was then presented by the subgroup analysis with respect to the different geographical regions. In such case, East Asian population revealed 28.8% (95%CI: 27.3%–30.4%) (range 22%–35%) (Q = 37.983, , and I2 = 55%), South Asia 5.6% (95%CI: 2.2%–10.2%) (range 1%–11%) (Q = 21.774, , and I2 = 86%), West Asia 4.4% (95%CI: 2.9%–6.1%) (range 2%–13%) (Q = 56.114, , and I2 = 84%), America 2.1% (95%CI: 0.0%–5.9%) (range 0%–9%) (Q = 50.82, , and I2 = 92%), and Europe 1% (95%CI: 0.1%–2.4%) (range 0%–4%) (Q = 42.03, , and I2 = 86%).

There was not enough data to calculate the combined estimates for both Southeast Asia and Africa.

3.3. Global Patient-Level Prevalence of Bilateral Symmetry of Three-Rooted PMFM (Figure 8)

The data from 40 studies allowed us to calculate the patient-level prevalence of bilateral symmetrical distribution of three-rooted PMFM. A total of 2326 patients revealed three-rooted PMFMs. Of these, 1311 had three-rooted PMFMs on both sides of the mandible.

Global patient-level prevalence of bilateral symmetry of three-rooted PMFM was 46.22% (95%CI: 39.13%–53.39%) (Q = 385.13, , and I2 = 90%) by random effect model. The prevalence ranged from 0% to 100%.

The patient-level prevalence of bilateral symmetry of three-rooted PMFMs was then explored by the subgroup analysis with regard to the different geographical locations. In this scenario, East Asian population revealed 60.3% (95%CI: 56.3%–64.2%) (range 49%–79%) (Q = 41.478, , and I2 = 64%), South Asia 38.6% (95%CI: 0.0%–91.7%) (range 0%–70%) (Q = 108.934, , and I2 = 97%), West Asia 37.5% (95%CI: 23.4%–52.7%) (range 0%–100%) (Q = 27.498, , and I2 = 67%), America 35.1% (95%CI: 17.9%–54.4%) (range 33%–35%) (Q = 0.004, p = 0.95, and I2 = 0%), and Europe 18.7% (95%CI: 3%–41.5%) (range 0%–60%) (Q = 19.515, , and I2 = 74%).

There was not enough information to calculate pooled estimates for Southeast Asia and Africa.

3.4. Tooth-Level Prevalence of Three-Rooted PMFM according to Sides of the Mandible (Subgroup Analysis)

Thirty-five studies allowed us to calculate the tooth-level distribution of three-rooted PMFM according to sides of the mandible. On the right side, the total number of teeth was 12604, of which 2631 were three-rooted. On the left side, the total number of teeth was 12483, of which 2043 were three-rooted.

The global prevalence of three-rooted PMFM on the right side of the mandible was 16% (95%CI: 12.2%–20.2%) (Q = 1261.665, , and I2 = 97%), ranging between 1% and 34%. On the other hand, the global prevalence of three-rooted PMFM on the left side of the mandible was 12.1% (95%CI: 9.2%–15.4%) (Q = 950.242, , and I2 = 96%), ranging from 0% to 31%. This signified the right-side predominance of the three-rooted PMFM.

3.5. Tooth-Level Prevalence of Three-Rooted PMFM according to Gender (Subgroup Analysis)

Thirty-five studies permitted us to estimate the tooth-level distribution of three-rooted PMFM according to gender. The total number of male PMFMs was 12922, of which 2393 had three roots. The total number of female PMFMs was 13313, of which 2190 were three-rooted.

The global prevalence of three-rooted PMFMs in male patients was 13.3% (95%CI: 9.8%–17.3%) (Q = 1348.191, , and I2 = 97%), ranging between 1% and 32%. On the other hand, the global prevalence of three-rooted PMFMs in female patients was 13% (95%CI: 9.7%–16.7%) (Q = 1219.977, , and I2 = 97%), ranging from 0% to 30%. This pointed out that there was no feature of sexual dimorphism in this case.

3.6. Global Tooth-Level Prevalence of Radix Entomolaris and Radix Paramolaris (Sensitivity Analysis)

To undertake sensitivity analysis, we excluded the studies which did not use the terms “RE” and “RP.” We included the studies that used the terms “RE” and “RP” in the sensitivity analysis.

Forty-three studies allowed us to estimate the tooth-level prevalence of Radix Entomolaris (RE) and Radix Paramolaris (RP). The total number of teeth was 28822. Of these, RE comprised 5056 and RP 21.

Global tooth-level prevalence of RE and RP was 12.3% (95%CI: 9.3%–15.7%) (Q = 2929.107, , and I2 = 99%), ranging from 0% to 29% and 0.1% (95%CI: 0.0%–0.1%) (Q = 61.672, p = 0.03, and I2 = 30%), running between 0% and 2%, respectively.

3.7. Publication Bias Test

Publication bias test revealed that funnel plot asymmetry was seen in Ι global tooth-level prevalence of three-rooted PMFM (Figure 9) and ΙΙΙ global patient-level prevalence of bilateral symmetry of three-rooted PMFM (Figure 10). Doi plots also showed “major asymmetry” for both results. LFK indexes were −4.02 for objective Ι and −3.19 for objective ΙΙΙ.

Symmetrical funnel plot resulted from ΙΙ global patient-level prevalence of three-rooted PMFM (Figure 11). Doi plot also pointed out “no asymmetry.” LFK index was 0.05 for ΙΙ.

1944 of records were identified through PubMed, Goggle Scholar, Research Gate, ProQuest, and LILACS database search. 1711 duplicates and irrelevant records were excluded. 233 of records were screened. 108 full-text articles were assessed for eligibility.

Thirty-six full-text articles were excluded: (1) 6 studies that did not report the age of the patients and developmental condition of root apex, (2) 5 studies of which primary objectives were MMC, MR, and IM of the teeth, (3) 9 studies that used the methods, not being CBCT, (4) 3 in vitro studies that used CBCT, (5) 3 studies that did not report research methods and settings, (6) 1 study that used the combined method (PR + Spiral CT), (7) 1 study from which the data cannot be extracted, (8) 2 studies in which 5- and 6-year-old children were sampled, (9) 4 studies whose data were overlapped, (10) 1 study of contemporary cadaver sample, and (11) 1 thesis which was later published.

125 of records were excluded due to the following: (1) 25 in vitro studies, (2) 62 case reports, (3) 16 systematic reviews, (4) 10 that used X-ray methods, (5) 1 of extraction methods, (6) 1 of treatment modalities, (7) 1 book chapter, (8) 1 that combined clinical investigation and Spiral CT, (9) 1 editorial, and (10) 7 Chinese language studies of which full-texts were not available to access.

Seventy-two studies were included in both qualitative and quantitative analyses.

4. Discussion

4.1. Summary of Main Findings

Global tooth-level prevalence of three-rooted PMFM was 9% ranging from 0% in Chilean, Italian White, Brazilian, Serbian, and Russian populations [30, 75, 76, 78, 85] to 29% in the Chinese populations [47, 93]. By the subgroup analysis of objective Ι with regard to the different geographical locations, East Asia, Southeast Asia, South Asia, West Asia, Europe, America, and Africa demonstrated 24%, 13%, 5%, more than 4%, 2%, 1.8%, and nearly 1%, respectively. The proportions were apparently downgraded from the East across Asia to the West. We believe that globalization, migration, and blended ethnicity may influence the prevalence of three-rooted PMFM, especially in the Native American population.

Global patient-level prevalence of three-rooted PMFM was 10% ranging from 0% in Chilean, Italian White, Brazilian, Serbian, and Russian populations [30, 75, 76, 78, 85] to 35% in a Chinese population [47]. By the subgroup analysis of objective ΙΙ according to the different geographical regions, East Asia, South Asia, West Asia, America, and Europe comprised 29%, 6%, more than 4%, over 2%, and 1% individually. Tooth-level data was usually lower in proportion than patient-level data, significantly in East Asia. It seems to be basically originated from which the number of teeth, if being bilaterally present, is more numerous than the number of patients to whom the teeth belong. This could affect the denominator of the pooled estimate. The higher the denominator count, the lower the resulting proportion.

Global patient-level bilateral symmetry of three-rooted PMFM was 46% ranging from 0% in Iranian, Turkish, Pakistani, and Egyptian populations [35, 41, 55, 66, 83] to 100% in an Iranian population [40]. By the subgroup analysis of objective ΙΙΙ across the different geographical locations, East Asia, South Asia, West Asia, America, and Europe displayed 60%, nearly 39%, approximately 38%, 35%, and over 18%, respectively. Although a Middle East country was ranking the highest of the range of bilateral symmetry, the East Asia group pooled up to 60% at this parameter.

For the objective ΙΙΙ, a wide range of the prevalence and broad confidence interval of each subset were investigated, especially in South Asia and West Asia. This reflects the small sample size of each subgroup in these areas.

4.1.1. Subgroup Analyses

Global tooth-level prevalence of right three-rooted PMFM was 16%, ranging from 1% in Egypt population [83] and then 2% in a Brazilian population [36] up to 34% in the Han population of China [47]. Global tooth-level prevalence of left three-rooted PMFM was 12% fluctuating between 0% in both Egypt [83] and an Iranian population [35] and 31% in a Western Chinese population [89]. We identify that the prevalence of three-rooted PMFM could be seen in nearly one-third of the number of teeth in East Asia. In this meta-analysis, the prevalence of three-rooted PMFM was definitely skewed towards the right side of the mandible.

Global tooth-level prevalence of three-rooted PMFM in male patients was more than 13% ranging from 1% in both Brazilian [36] and Egyptian [83] populations to 32% in Han people of China [47]. Global tooth-level prevalence of three-rooted PMFM in female patients was 13% ranging from 0% in an Egypt population [83] and then 2% in Brazilian [36], Turkish [41], and Israeli [2] populations up to 30% in a Western Chinese population [89]. As a result, the occurrence of three-rooted PMFM was not sexually dimorphic in our meta-analysis.

4.1.2. Sensitivity Analysis

Global tooth-level prevalence of Radix Entomolaris was more than 12% limiting between 0% in the Egyptian population [83] and 29% in the Han and Chengdu populations of China [47, 93]. Global tooth-level prevalence of Radix Paramolaris was 0.1% ranging from 0% in nearly two-thirds of the meta-analyzed studies, then around 1% in Indian [44], Greece [3], and Israeli [2] populations, respectively, and nearly 2% in Spanish [73] population. We postulate that RP could be less reported and underestimated due to its lesser clinical significance than RE.

4.2. Comparison with the Findings of Previous Reviews

There were an increasing number of pieces of evidence, particularly systematic reviews and meta-analyses in the subject of endodontic anatomy.

One earliest systematic review of the scope investigated that the prevalence of the third root in PMFMs across the world demonstrated 13% [24]. Our meta-analysis investigated a 4% reduction in the pooled tooth-level prevalence compared to the previous one. On the other hand, the prevalence of RE found in our analysis was comparable to the result of this former investigation. We estimate that blended ethnicities resulting from globalization may have an impact on the prevalence of the third root in the Western world. Interestingly, 2 studies and 1 case report using the CBCT method were included in this former evidence [24]. Specifically, the prevalence of three-rooted PMFM in a dental school sample of Germany and an English Caucasian population demonstrated 0.7% and 3.3%, resulting from the narrative synthesis in this previous review [18, 94]. The pooled patient-level prevalence in our present review displayed 1% ranging from 0% to 4% in Europe. So, the two reviews were in agreement at this point. Additionally, both of the European [18, 94] studies included in this review reported no bilateral symmetry of this morphology. Our meta-analysis pointed out that the pooled bilateral symmetry of three-rooted PMFM in Europe was the lowest out of all the continents.

Then, a systematic review of Brazil investigated the prevalence of three-rooted PMFMs revealed between 1.5% and 3% in such population [25]. The range in our present meta-analysis comprised from 0% to 6% in the American continents, including Brazil. Zero-prevalence studies [30], some studies’ samples restricted to the White population [43], and blended ethnicities including Asian-American [46] used in some studies may broaden the range. Surprisingly, one of the earliest investigations detected 22% of three-rooted PMFMs found in the Canadian Eskimo sample [95]. So, we conclude that the anatomy of migrants could differ from that of the original natives in the continents.

Tomaszewska and coauthors found that the pooled prevalence of three-rooted PMFMs comprised 8.9% [27] after the meta-analysis of 35 chosen studies. This finding was nearly similar to the global tooth-level prevalence of our analysis. However, the former meta-analysis did not further investigate the patient-level prevalence and bilateral symmetry of three-rooted PMFMs, subgroup, and sensitivity analyses to explore heterogeneity. Additionally, the investigators used studies of different research methodologies to pool the estimate.

Consequently, a recent systematic review of Saudi Arabia qualitatively observed that the occurrence of RE ranged from 2.9% to 6.07% [26]. The range was slightly wider than our review’s findings. In our review, the prevalence of RE in West Asia, including Saudi Arabia, featured from 2% in Israeli [2] to 5% in the Iranian population [40]. We postulate that the European descend Israeli population may be implicated to become lower prevalence. We also suspect that the previous review was limited to the Saudi population, not all West Asia area, and showed a methodological discrepancy in comparison with our review. The primary studies in the Saudi review [26] used the X-ray method and the combined methodology of CBCT and X-ray. However, 60% of RE was limitedly investigated in periapical X-rays [11]. The straight distolingual root could be overlapped by the distobuccal root. As a result, the X-ray method may lead to a deficiency in the estimation of RE.

In addition, the tooth-level prevalence of three-rooted PMFM was more numerous in females than in males in the review of Saudi people [26]. In our findings, only the Nepalese population [38] also showed female determinants of three-rooted PMFMs clearly. Alternatively, the various East Asia studies [47, 50, 56, 77] demonstrated that three-rooted PMFMs were more often seen in males than females.

However, distolingual root or two distal roots could not be strictly defined as “RE.” Centrolingual and distobuccal positions of the extra root may also be found (Figures 1(b) and 1(e)) [24]. All RE can be three-rooted. However, every three-rooted PMFMs could not be stated as “RE.” So, we contributed and categorized the variable positions of the third root, with reference to the latest available evidence (Figure 1) [24].

Importantly, patient-level prevalence and bilateral symmetry of three-rooted PMFMs were missed to report in the previous reviews. They also did not undergo the proportional meta-analysis for Southeast Asia and South Asia by subgroup analysis. They should report RP as a separate entity. All these variables should also be considered as clinically significant factors.

4.3. Significant Findings apart from the Previous Reviews

In addition to objectives ΙΙ and ΙΙΙ (patient-level prevalence), subgroup analyses of Southeast Asia and South Asia were significant aside from the previous observation.

In Southeast Asia, the pooled tooth-level prevalence of three-rooted PMFM was 13%, ranging between 8% and 22% in this meta-analysis. The lowest 8% was found in the study, which restricted the sample solely to include Malay ethnicity [67]. The highest 22% was seen in the study, in which more than 90% of the sample was Chinese [70]. Both studies were conducted in Malaysia. So, the prevalence of three-rooted PMFM depended upon the demographic characteristic of the participants even in the same country. The prevalence of three-rooted PMFMs was 10% in Burmese [96] and 19% in Thai populations [97] regardless of considering different research methods used. As a result, we conclude that the values of the previous studies were within the range of our meta-analysis.

In South Asia, the summarized tooth-level prevalence of three-rooted PMFM was 4.7%, fluctuating between 1% in Karachi people of Pakistan [66] and 9% in Nepal [38] by the subgroup analysis of our review. However, the Nepal sample was a mixture of Aryan, speaking the Indo-European language, and Mongoloid people. Chandra and coworkers [19] found that the incidence of RE was 13% in tooth-level and more than 18% in patient-level of an Indian population, being obviously outlying when compared with our investigation. The patient-level prevalence of three-rooted PMFMs in our meta-analysis indicated a 6% pooled estimate in South Asia and almost 11% in the Nepalese population [38] at the upper boundary. At this point, we suggest that ethnical diversity, sample size, and the number of selected studies could be the impact factors on pooling the combined approximation.

As a result, the Mongoloid descents were positively associated with the prevalence of three-rooted PMFMs in both Southeast Asia and South Asia regions.

Additionally, “RP” Radix Paramolaris is less clinically significant than “RE,” although having both anatomical and anthropological significance. Unfortunately, the previous reviews did not point out the description of RP, although the pooled prevalence of RP in our meta-analysis demonstrated 0.1%. RP was mostly found in nearly 2% of Spanish [73] and Greece populations of Europe [3], Jews who could be European or Arab descendants [2], and Indians [44] in our present meta-analysis. Less than 0.5% of RP was found in some Saudi [31] and Turkish [55] populations. Interestingly, RP was not reported in East Asian and African studies. Additionally, Radix Paramolaris of the mesial root of PMFM [2] (Figure 1(f)) could not be detected prominently in periapical X-ray.

4.4. Strengths and Weaknesses of the Meta-Analysis

In contrast to the previous systematic reviews, patient-level prevalence and bilateral symmetry of three-rooted PMFMs were the most striking features of the review. Additionally, the Asia continent was subdivided into four geographical locations to transparent different levels of prevalence of three-rooted PMFMs in the continent. RP was also reported as a pooled prevalence, which could be anthropologically significant.

Literature search was thoroughly done across five databases for the review (Figure 4). To overcome language bias, 4 studies [3335, 47] were translated from native languages to English (Table 1). We emphasize that findings could be more generalizable to the global population when foreign language studies were included as much as possible.

Being in vivo study was one eligible criterion for our meta-analysis. In vitro needs extraction and can lead to underestimation of the findings due to root fracture during the surgical procedure. Validity of measurement in the included studies could be properly achieved by the CBCT method. Counting the number of roots could be objectively done by viewing CBCT images, especially in axial slices. Thereby, measurement bias could be reduced.

As a weakness, there was no adequate raw data for the patient-level prevalence of Africa and Southeast Asia. And zero-prevalence studies were included in the meta-analysis (Table 1). This may skew the pooled findings to be underestimated or overestimated. To explore heterogeneity, subgroup analysis was not done through different age groups. However, the presence of the third root in PMFM could not be influenced by the age of the population. Additionally, the prevalence of three-rooted PMFMs, with regard to the different ethnicities across the world, was not explored in this meta-analysis. Blended ethnicities, migration, and globalization may not guarantee the representativeness of each ethnicity.

In Africa, the pooled estimate of our meta-analysis revealed 0.9% ranging from 0.5% in Egypt [83] to 1.1% in South Africa [87]. This finding was slightly inferior to the previous report demonstrating the prevalence of three-rooted PMFMs more than 3% in the Senegalese population [98]. Otherwise, it reached nearly an agreement with the former one reporting 0.65% of three-rooted PMFM in Egyptian residents of Saudi Arabia [99]. However, we conclude that only two studies selected for Africa in our analysis may not be generalizable to the whole continent.

Similarly, there was no eligible study for the Australian continent. More than three-quarters of the Australians were of European descent, and the remaining were a mixture of different ethnicities [100]. As a result, we postulate that the prevalence of three-rooted PMFM in Australians may not be seen different from that of the Western world. Consequently, a recent worldwide cross-sectional study [101] with a meta-analysis found that the prevalence of two distal roots in PMFM comprised 4% of the White Australian population. This finding was slightly superior to the upper bound of the prevalence of three-rooted PMFM in Europe of our meta-analysis. As a result, this finding from 214 White patients of Melbourne and a single-center study [101] could not be generalizable to the whole population of the Australian continent. Additionally, the aboriginal population of Australia should not be neglected to consider.

Consequently, the internal morphology of the third root of PMFM was not included in the objectives of this meta-analysis. More precisely, the importance of the radix molar and its associated clinical features should also be conveyed to clinicians.

4.5. Limitations

Slightly more than half of the included studies were categorized as moderate risk of bias, whereas another half had a high risk of bias. Bias may arise from the weaknesses of research methodology in some selected studies (Figure 5).

More than 80% of the studies did not formally set a sampling frame. Strictly, 4% of the eligible studies conducted random sequence generation. Only 15% of the included studies used the calculated sample size. Up to 95% of the selected picked up the data from the single centers of the city or the province. As a result, selection bias, over coverage or under coverage on the population of interest, and deficient in generalization [102] may arise and manifest on the findings of the meta-analysis.

Nearly 50% of the selected studies undertook a reliability test on measurement. And only 33% of these reported the outcomes thoroughly. Consequently, reporting bias and performance bias [102] may partly jeopardize the results of this review.

In addition, 19 studies of the review presented the primary objective, which matched the primary outcome of the review. Of these, 12 studies thoroughly reported the variables and allowed us to calculate the pooled estimates. So, incomplete outcome reporting may be investigated even in the studies in which the main objective was to assess the prevalence of three-rooted PMFM or RE.

Consequently, publication biases were investigated in cases of global tooth-level prevalence and global patient-level bilateral symmetry of three-rooted PMFMs.

In Figure 9, the small studies crowded at the funnel base. And they represented extreme values (proportion/double arcsine prevalence) at the x-axis. This means that these small studies exhibited a large effect. Also, in Figure 10, there was sparse or lack of studies at the left-hand base of the funnel plot. This means asymmetry. The asymmetrical plot and the small studies having great effect indicated publication bias.

The studies, scattering equal over both sides of the funnel plot, were detected in the case of the global patient-level prevalence of three-rooted PMFMs (Figure 11), showing “no asymmetry.”

All of the publication bias tests were additionally confirmed by Doi plots and LFK indexes.

4.6. Heterogeneity

Q statistics mean the difference in the prevalence of three-rooted PMFMs of individual studies included in the meta-analyses. I2 statistics mean the percentage of variation in the prevalence of three-rooted PMFMs of such studies. An increase in the two values indicates the progress of heterogeneity.

From the global tooth-level prevalence of three-rooted PMFMs (objective Ι), I2 demonstrated considerable heterogeneity (Figure 6). To explore heterogeneity, subgroup analysis according to the different geographical locations was undertaken. After the analysis, I2 was downgraded from the combined estimate of 99% to 0% in Africa at the upper bound and to 92% in America at the lower.

From the global patient-level prevalence of three-rooted PMFMs (objective ΙΙ), I2 dropped from the pooled estimate of 98% (Figure 7) to 55% in East Asia and to 92% in America, investigated by the subgroup analysis.

From the global patient-level prevalence of bilateral symmetry of three-rooted PMFMs (objective ΙΙΙ), I2 rose from 90% (Figure 8) to 97% in South Asia and descended to 0% in America after the subgroup analysis.

As a result, I2 declined at least 6% and as much as 99% after the subgroup analyses.

So, the heterogeneity mainly emerged from the different geographical locations. The rest of the heterogeneity was rooted partially from clinical, statistical, and methodological factors.

Most designs of the chosen studies were descriptive and retrospective in character. 19% of the selected studies were analytical designs (Table 1). Only 1 study approached prospective [38]. These contexts figured out the clinical and methodological heterogeneities. Blending of Asians and Caucasian samples [46] and limitation to select Caucasians [81] may lead to over and under estimation of the prevalence.

The sensitivity analysis drove the heterogeneity of the overall estimate of 99% to 30% in the event of the global prevalence of Radix Paramolaris (RP). It demonstrates that the prevalence around the world displayed agreement among the chosen studies in such a point.

There was no significant variable heterogeneity in the right and left and gender distributions compared with the pooled tooth-level estimate.

4.7. Direction of Future Studies

We have a plan to intervene in a future systematic review and meta-analysis to find the clinical factors connected with three-rooted PMFMs across the global population. The present meta-analysis should be updated to additionally assess variations in the position of the third root along the circumference of PMFMs in the distance future.

4.8. Take-Home Messages for Clinicians

Dental practitioners should generally keep in mind, according to the key estimates of the present meta-analysis, the following:(i)9% of PMFMs in the world population demonstrated 3 rooted PMFMs(ii)10% of the people across the world had three-rooted PMFMs(iii)45% of the people who had three-rooted PMFMs revealed a bilateral symmetrical distribution of such morphology(iv)The global prevalence of three-rooted PMFM was not identified as a sex determinant but showed right-side predominance(v)The global tooth-level prevalence of Radix Entomolaris and Radix Paramolaris was 12% and 0.1%, respectively

5. Conclusion

To the best of our knowledge, the prevalence of three-rooted PMFMs was concerned with different geographical locations all over the world and by the widespread habitation of Mongoloid descents. We suggest that globalization, blending, and interaction among ethnicities may have a huge impact on the reduction or accentuation of the anatomical significance in some populations.

Data Availability

The data that support the findings of the meta-analysis are available from the corresponding author, Nyan Min Aung, upon reasonable request.

Conflicts of Interest

The authors declare that they have no conflicts of interest.