Anatomical variations of external and internal teeth morphologies associated with ethnicity have been well documented [22]. Differences in root and root canal morphology of MSMs have been reported by several studies around the world with different percentages of each anatomical feature [2-9]. The current study presents the first description of the root and canal anatomy of Yemeni MSMs.
Yemeni MSMs had mostly separated two roots (89.6%). This is close to the findings reported in Iranians (79.2%, 81.6%, and 86.7%) [8, 23, 24], Turkish (85.4%, 90%) [11, 25], Indians (79.35%, 88.8%) [9, 26], Belgians (83.93%) and Chileans (86.61%) [10]. The observed one-rooted MSMs in Yemeni population was 0.8% which is in the same line with the findings in Turkish (1.29%) [11]. Nevertheless, higher percentages were reported in Iranians (13.3%, 19.8%) [8, 23], Indians (8.7%) [26], Chileans (8.93%) [10], Turkish (10%) [25], Belgians (14.29%) [10] and Chinese (22%) [3]. We found only 0.6% of three-rooted MSMs, which was in accordance with other studies in Iranians (0.6%) [24] and Belgians (0.89%) [10] but was higher than that reported in Koreans (0.3%) [27]. However, higher percentages were reported in Thai (1.2%) [6], Turkish (3.45%) [11], Brazilians (3.5%) [2], Chileans (3.57%) [10] and Indians (7.53%) [9]. No statistical gender- or tooth location-related differences were found in the occurrence of root numbers of molars within this study similar to Nur et al. study findings in Turkish [25].
MSMs with fused roots were 9% in this study. Close results were reported in Turkish (8.97%) [11] and south Indians (13.12%) [9]. However, higher percentages (24% and 39%) were found in Chinese [3, 4].
Root cross section of separated two-rooted MSMs was different in both roots such that mesial root was ribbon-shaped and distal root was kidney-shaped. Extreme care should be exercised while preparing root canals to avoid any complications that may occur due to thin dentin sections (danger zones).
MSMs with three orifices were the most common in this study (77%), followed by two orifices (21%). This is similar to findings in Turkish (72.8% with three orifices and 22.8% with two) [11] and Chinese (46% with three orifices followed by 38% with two) [3].
Mesial root of MSMs with separated two roots showed mainly type II canal, followed by type IV. This agrees with the results in Iranians [8, 23], but disagrees with numbers reported in Sudanese [5], Chinese [3], Indians [9], Iranians [24] and Turkish [11, 25] where mesial roots mainly had type IV canal. Our results also do not agree with findings in Belgians and Chileans [10] where type III canal was the most common type followed by type V. We found distal root had mostly type I canal which was the same as in Thai [6], Sudanese [5], Iranians [8, 23, 24], Chinese [3], Indians [9], Turkish [11, 25], Belgians and Chileans [10].
C-shaped canals in this study were found in 9% of MSMs, similar to that found in Saudis (9.1%) [28]. Closer percentages were reported in Chileans (8.93%) [10], Indians (8.1%, 9.7%, 13.12%) [9, 26, 29], Sudanese (10%) [5], Belgians (10.71%) [10] and Iranians (6.7%) [23]. However, it was far lower than results in Iranians (17.6%, 21.4%) [8, 24], Lebanese (19.1%) [30], Chinese (29%, 38.6%) [3, 4], Koreans (39.8%, 44.5%) [7, 31] and Malaysians (48.7%) [32]. Nevertheless, the incidence of C-shaped canals in Yemenis was higher than Brazilians (3.5%) [2], and Turkish (4.1%) [11].
C-shaped canals were found in females more than males with no significant difference same as reported in Chinese [4], Indians [29] and Iranians [8]. Similar results but with a statistically significant difference were found in Koreans [31], Saudis [28] and Malaysians [32]. Regarding tooth location, C-shaped canals occurred almost even on both sides in this study with no significant difference similar to reports in Chinese [4], Iranians (35), Indians [29], and Koreans [31].
MSMs with C-shaped root showed mainly both lingual and buccal grooves, followed by lingual groove only and then by buccal groove only unlike study’s findings in Saudis [28], Koreans [7] and Chinese [4, 21] which showed mainly lingual groove only. Moreover, Wadhwani et al. [29] reported MSMs mainly with a buccal groove only in Indians.
C1 was found to be the prevalent shape at orifice level which agreed with the findings in Chinese [4]. C3c was the dominant shape coronally, followed by C2. C3c was also the dominant shape in the middle third. These findings differed from those of Zheng et al. [4] where C1 followed by C3d were the dominant shapes coronally and C3d was the most prevalent in the middle third. However, C3c followed by C3d were the dominant shapes apically in both Yemenis and Chinese. A study in Iranians [8] reported different results in which C1 was the most frequent in the coronal third and C3d was the major shape in both middle and apical thirds. Moreover, Kim et al. [31] reported that C2 is the most common configuration at the orifice level.
C-shaped canal configuration remains unchanged from orifice to apical level in 4.4%. Similar findings were recorded in Iranians [8] and Chinese [4] where 4.9% and 5.9% of C-shaped canals remained unchanged along the root length, respectively. This agrees with the results of Fan et al. [21] who reported that C-shaped canals vary in shape and number along the root length. Therefore, the shape of the canal orifice cannot be considered as an indicator of the C-shaped canal anatomy along the tooth root to its apex. There was no constant change in the configuration of the C-shape canal between two adjacent root levels. This was also reported by Zheng et al. [4].
The occurrence of C1 and C2 shapes decreased from the coronal to the apical levels, however, C3c shape increased toward the middle level and C3d type increased toward the apical level. This revealed a high possibility of division of C-shaped canals into two or three canals towards the apex. Similar results were reported earlier in Chinese [21]. This necessitates the emphasis of applying the available techniques for canal debridement to ensure proper cleaning of such complex anatomy at different root levels.
The bilateral and unilateral occurrences of C-shaped canals were found to be equal. Janani et al. [8] reported a slightly higher occurrence of bilateral C-shaped canals (15.6%) than their unilateral occurrence (11.76%) in Iranians with no significant difference. However, there was a much higher percentage of bilateral C-shaped canals occurrence (81.3%) than unilateral occurrence (18.7%) in Chinese [4]. Similarly, a 71% of bilateral occurrence of C-shaped canals was reported in Koreans [31]. The unilateral occurrence of C-shaped canals was higher on Saudis (53.85%) [28]. In relation to gender, bilateral occurrence of C-shaped canals in this study was more in females than males unlike Zheng et al. [4] who showed no difference of bilateral distribution with gender. Unilateral occurrence of C-shaped canals showed no significant difference regarding gender or tooth location in this study, similar to findings in Chinese [4].
Six variants (1, 3, 6, 8, 9 and 10) were observed in the root and root canal morphology of the studied MSMs. Other studies showed more variants such as seven variants (1, 3, 4, 6, 8, 9 and 10) in Brazilians [2], and eight variants (1, 3, 4, 5, 6, 8, 9 and 10) in Chinese [3]. Variant 3 was the most common morphology in Yemenis followed by variant 1, similar to previous study reports in Brazilians [2] and Chinese [3]. Yemeni MSMs showed a higher percentage of variant 3 (71%) than that found in Thai (54%) [6], Brazilians (54%) [2], and Chinese (42%) [3].
Root and canal morphology of Yemeni MSMs when compared with different populations showed the presence of morphological differences that should be taken into consideration during clinical practice. Therefore, a thorough radiographical examination during endodontic treatment is essential to identify the canal shape at each root level and facilitate planning for canal debridement and subsequent obturation.