Smile esthetics has recently become an important demand for patients who seek orthodontic treatment (1, 2). It also has a specific importance for the orthodontists because the results of treatment are criticized by ordinary(non-dentist) people mostly based on how the smile look like (3).
People with attractive smiles are more socially acceptable and are more successful in their relationships. In fact, smile and facial aesthetics could positively affect people`s self-confidence and self-esteem (4, 5).
Several factors such as diastema, teeth size and shape, incisors position, midline discrepancies, buccal corridor, smile arc, gingival display and shape, lip height and so on have been reported to influence smile aesthetics. Various studies have been conducted to evaluate the effect of these characteristics on smile aesthetics (6, 7, 8).
Among the aforementioned factors, midline deviation is more frequent (9) and is somehow controversial. Acceptable thresholds of less than 2mm, 3mm or even 4mm have been reported for linear midline deviations (10-15).
On the contrary, there is general agreement that excessive gingival show as well as minimal teeth display are unattractive smile features (15, 16). In addition, smiles with minimal buccal corridor space have been reported as more popular (17-19). However, in our team recent study, we showed that the effect of buccal corridor width and tooth-gingiva display on smile attractiveness is influenced by subjects` facial height (20). Therefore, midline deviation might possibly follow the same trend.
In the present study, we tried to investigate whether threshold of acceptability for midline deviation is dependent to facial height.
Materials and Methods
Completed files of two Iranian women (one long and one short face) with proper midlines and aligned anterior teeth were selected from the department of orthodontics archives. Facial types were determined regarding the Frankfort-mandibular plane angle (FMA), the Jarabak index and the ratio of the middle facial one-third to lower facial one-third based on the results of lateral cephalograms analysis (FMA= 15, Jarabak index=75 and g-sn/sn-me=55/45 for the short face patient and FMA= 35, Jarabak index=55 and g-sn/sn-me=45/55 for the long face one).
Frontal facial smiling photographs of both patients were altered using Adobe Photoshop CS3 software (Adobe Systems, San Jose, CA). Four photos with different midline deviations to right (1mm, 2mm, 3mm, 4mm) and four other photos with counter clockwise midline inclination (2˚, 6˚, 10˚, 14˚) were produced. One normal photo from each situation was taken too. Patients` eyes were covered. Buccal corridors were adjusted to be the same in both sides. (Figure 1-4)
Figure 1: Illustration of different amounts of linear midline deviation in short face patient in an increasing deviation sequence, from no deviation to 4mm deviation
Figure 2: Illustration of different amounts of angular midline deviation in short face patient in an increasing deviation sequence, from no deviation to 14˚deviation
Figure 3: Illustration of different amounts of linear midline deviation in long face patient in an increasing deviation sequence, from no deviation to 4mm deviation
Figure 4: Illustration of different amounts of angular midline deviation in long face patient in an increasing deviation sequence, from no deviation to 14˚deviation
In order to determine the threshold of acceptability for linear midline deviation, the images were set in an increasing midline deviation order and printed on 18cm x 24cm papers.
The reliability of the study was evaluated through a pilot study by participating 10 ordinary persons (5 women and 5 men).
The study population included 60 Iranian ordinary people (30 men and women) with a mean age of 25.5±3.2 years old. None of them stated any history of training in dental or facial aesthetics or orthodontic treatment. They were asked to choose the photo in which the smile was not acceptable anymore.
The same method was used for evaluation of midline angular deviation.
Inter Class Correlation (ICC) test and Weighted Kappa test were used to assess the reliability.
We used Frequency table to determine threshold of smile acceptability for midline deviation. In order to confirm reliability, rating was performed by 10 examiners twice in a two-week interval and ICC test results for linear midline and angular deviations were 0.873 and 0.943, respectively.
Chi square was applied to compare the male and female examiners. Independent T Test was used to compare the threshold of midline deviation acceptability in two facial types. P values less than 0.05 were considered statistically significant.
Linear midline deviation
Results of the present study showed that mean numbers for maximum acceptable midline deviation in long face and short face patients are 2.13±0.85mm and 2.32±0.83 mm, respectively. Comparison of data using T Test revealed that there is no statistically significant difference between the limit of acceptability for midline deviation between two groups. (P-Value = 0.24)
Chi square revealed no significant difference between male and female examiners in both short face (P-Values=0.45) and long face (P- Value=0.93) cases.
Figure 5 shows the detailed data regarding linear midline deviation in both long face and short face patients.
Angular midline deviation
The mean acceptable amounts for angular midline deviation in long face and short face patients were 8.47± 3.13˚ and 6.53±2.80 ˚, respectively. There was a statistically significant difference between long face and short face images (P=0.0001), while Chi square showed no significant difference between male and female examiners in short face (P-Value= 0.42) and long face (P-Value= 0.25) cases.
Figure 6 shows the detailed data regarding Angular midline deviation in two groups
Figure 5: Percentage of votes for acceptability limitation of linear midline deviation in long face and short face patients
Figure 6: Percentage of votes for acceptability limitation of angular midline deviation in long face and short face patients
Deviation of dental midline from facial midline can adversely affect smile acceptability and facial aesthetics (6). Literature shows controversies about the level of acceptability for midline deviation from less than 2mm up to 4mm (6, 11, 12, 21- 23).
Esthetic standards for smile are mostly acquired from the images of mouth only, or photos of subjects with normal and average facial proportions. These standards are also affected by cultural, social and ethnic figures (17, 24, 25).
Considering previous studies which have shown that the effect of buccal corridor on smile aesthetics is strongly depended on facial height as well as ethnic factors (20, 26) we assumed that other smile components such as midline could possibly be affected by facial height.
In the present study, we applied full-face photos to allow the examiners evaluate midline in long face and short face subjects. Another reason that full facial image was chosen, was based on the data from Ferreira et al. who showed that ordinary people`s perception of midline is affected by the amount of adjacent structures` show (27).
Results of the present study were similar to those reported by William et al. who did not find any statistically significant differences in threshold of acceptability for linear midline deviation between long face and short face subjects (28).
In our previous study, we showed that the effects of buccal corridor and tooth show on smile aesthetics in long and short-face subjects are based on their harmony with the geometry of face in vertical and transverse dimensions. Similarly, there is a geometric relationship between midline deviation and symmetry of face shown by Silva et al. They found that in asymmetric faces, ordinary people prefer cant of dental midline in the same direction of nose and chin deviation rather than the opposite way (29).
This study had also some limitations including searching for documents of appropriate short face and long face patients and providing accurate photoshop photos.
The present study revealed another geometric association between midline deviation and facial form: higher amounts of angular midline deviation are perceived acceptable by lay people.
The authors would like to thank Dr. Ahmadreza Shamshiri for doing statistical analysis.
Conflict of Interest
The authors declare that there is no conflict of interest.