ORIGINAL ARTICLE

Characterization of posed smile by using visualanalog scale, smile arc, buccal corridormeasures, and modified smile indexVinod Krishnan,a Sunish T. Daniel,b Don Lazar,c and Abin Asokc

Tamilnadu, India

Introduction: Smile evaluation and designing are important; recent emphasis on the soft tissues has beenon par with that on the hard tissues in orthodontic diagnosis and treatment planning. This importance has notyet gained proper attention, and smile analysis is often ignored in clinical examinations. We attempted acomprehensive evaluation of smile characteristics with the following aims: (1) evaluation of perceptiondifferences, if any, between dental specialists and laypersons; and (2) quantification of smile characteristicswith the smile arc, buccal corridor measurements, and a modified smile index (MSI). Methods: The sixtysubjects included in the study had an age range of 18 to 25 years (mean, 21 years) with equal numbers ofmen and women. Frontal posed smile photographs were taken of all subjects. The study had 2 parts. Analysisof perception differences between dental specialists and laypersons was performed with a visual analogscale. Quantification of smile characteristics was done with smile-arc evaluation for consonance ornonconsonance, buccal corridor measurements, the MSI, and comparisons of the MSI with the facial indexand the mandibular width-facial height index. Results: There was no perception difference between thespecialists and the laypersons on overall smile evaluation. Women had more consonant smile arcs than men,and there was high correlation between the right and left buccal corridor spaces in men and women. The MSIshowed no correlation to the facial index, but there was a negative correlation of the MSI with the mandibularwidth-facial height index. Conclusions: Smile analysis should be an important aspect of orthodonticdiagnosis and treatment planning. Orthodontists should not disturb consonant smiles but create them withproper bracket positioning. The MSI, with negative correlation with the mandibular width-facial height index,warrants further evaluation with a larger sample to validate its clinical use and to develop a predictive

approach of its relationship. (Am J Orthod Dentofacial Orthop 2008;133:515-23)

Smile, defined as a facial expression characterizedby upward curving of the corners of the mouth, isoften used to indicate pleasure, amusement, or

derision.1 The smile also influences a person’s per-ceived attractiveness and is the cornerstone of socialinteraction. Research suggests that we trust smilingpeople more than nonsmiling ones.2 There are 2 formsof smiles—the enjoyment or Duchenne smile, and theposed or social smile.3,4 Humans learned to pose asmile during evolution.5 Posed smiles gained impor-tance in dentistry and orthodontics mainly because theyare repeatable over time.6,7 The smile, which is essen-tial to express friendliness, agreement, and appreciation,

From the Department of Orthodontics, Rajas Dental College, TirunelveliDistrict, Tamilnadu, India.aAssistant professor.bSenior lecturer.cIntern.Reprint requests to: Vinod Krishnan, Gourivilasam, Kudappanakunnu PO,Trivandrum, Kerala State 695043, India; e-mail, vikrishnan@yahoo.com.Submitted, January 2006; revised and accepted, April 2006.0889-5406/$34.00Copyright © 2008 by the American Association of Orthodontists.

doi:10.1016/j.ajodo.2006.04.046

and to convey compassion and understanding,8 should notbe ignored in diagnosis and treatment planning.

Orthodontic diagnosis has come a long way andnow includes patient-driven esthetic diagnosis andtreatment planning along with its problem-orientedapproach. The reemergence of the soft-tissue paradigmin clinical orthodontics9 has made smile analysis anddesigning key elements in diagnosis and treatmentplanning.10 Much attention is given in clinical exami-nation to the display zone of smile, which is determinedby lip thickness, intercommissural width, interlabialgap, smile index, and gingival architecture.10

Esthetics denotes concern about beauty or appreci-ation of beauty. The perception of esthetics varies fromperson to person and is influenced by personal experi-ences and social environments.11 For the same reasons,there can be differences of opinion regarding beautybetween laypersons and professionals.12 One studyreported that laypersons preferred more natural profiledrawings than did dental specialists.13 Another reportpointed out an opinion difference between orthodon-

tists and their patients when the same smiles were

515

0.7

American Journal of Orthodontics and Dentofacial OrthopedicsApril 2008

516 Krishnan et al

evaluated.14 Recent studies also confirmed that there isa difference in esthetic perceptions between orthodon-tists, general dentists, and laypersons.14-16

The smile arc is defined as the relationship betweenthe curvature of the incisal edges of the maxillaryanterior teeth and the curvature of upper border of thelower lip.4,17,18 Their ideal relationship on smiling isconsidered to be parallel and is known as a consonantsmile. If the 2 are not parallel (with flatter maxillaryincisal curvature to the upper border of lower lip), it iscalled a nonconsonant smile.4,16 Orthodontists’ lack ofconsideration of the smile arc in treatment planning andmechanics was reported recently, resulting in flatteningof the smile arc and less esthetic smiles. The mainreason cited for this was that more attention was placedon tooth measurements while positioning brackets,causing compromise in the curvature of the maxillaryincisal edges.19 Ackermann et al6 conducted a comput-erized multimedia study to measure smile-arc charac-teristics in treated and untreated subjects; they reportedflattening of the smile arc in 37% of the treated ones,whereas only 5% of the untreated group had noncon-sonant smiles. Other authors found similar results.8,20

These findings were contradicted by a recent study; theauthors stated that, although orthodontic treatmentresults in flattening of the incisal edges, it does not leadto a nonconsonant smile.19 Those authors found goodcorrelation between the fitted curve of treated subjectsand the lower-lip curve at the maximum angle ofelevation.

Buccal corridors (negative or black spaces) are thespaces between the facial surfaces of posterior teeth andthe corners of lips when a person is smiling.21 Thereseems to be a difference of opinion among investigatorsabout the esthetic value of buccal corridors. Someconcluded that they have no esthetic value; others

Table I. Results of VAS: median and range of ratings,at .01 level

Specialist rating for menSpecialist rating for womenLayperson rating for menLayperson rating for women

Spearman

Groups Corr

Specialist rating: male vs female smilesLayperson rating: male vs female smilesMale smiles: specialists vs laypersonsFemale smiles: specialists vs laypersons

believe that visible buccal corridors are unattractive.16

Johnson and Smith22 evaluated the effects of premolarextraction on the visibility of negative space and foundno relationship between extraction pattern and negativespace.22 Recently, a study suggested that large buccalcorridor spaces should be included in the problem list fortreatment planning, but minimal buccal corridors can beleft as they are.23 Most recent studies reported that buccalcorridors lave little impact on smile esthetics and will notinfluence the overall rating of a smile by orthodontists,general dentists, and laypersons.16,23,24

A smile index was developed by Ackermann andAckermann10 to describe the display zone of a smile.This is determined by dividing intercommissural width

Fig 1. Consonant and nonconsonant smiles: A, conso-nant smile with parallelism between the curvature of themaxillary incisal edge and the upper border of lower lip;B, nonconsonant smile, with no parallelism between thecurves.

man rho correlation coefficients, and significance

ian Range

0 4.50-7.383 3.00-6.003 4.25-8.420 3.60-7.50

rrelations

coefficient Significance

79 No significant correlation44 No significant correlation24 Significant at 0.01 level (2 tailed)81 Significant at 0.01 level (2 tailed)

Spear

Med

6.04.95.85.5

rho co

elation

0.00.00.6

by the interlabial gap during smiling. This ratio is

American Journal of Orthodontics and Dentofacial OrthopedicsVolume 133, Number 4

Krishnan et al 517

useful for comparing smiles among patients. The lowerthe smile index, the less youthful the smile appears.6,10

Even though the index is proposed, not much researchhas been performed to validate it. With this back-ground, our aims in this study were to (1) evaluate andcompare esthetic perceptions of frontal views of smilesby dental specialists and laypersons, (2) compare smilearcs for consonance and nonconsonance in men andwomen, (3) evaluate and compare buccal corridor space(right vs left; male vs female) in the whole sample, and(4) calculate the modified smile index (MSI) value andcheck its correlation with the facial index and themandibular width-facial height index.

MATERIAL AND METHODS

This was a cross-sectional study. The 60 subjects(ages, 18-25 years; mean, 21 years), students at RajasDental College, included equal numbers of men andwomen. The following inclusion criteria were strictlyfollowed: (1) no previous orthodontic treatment ormaxillofacial surgery, (2) complete permanent denti-tion except for third molars with no missing or super-

Fig 2. Method to measure buccal corridor spaces: thedistance between the inner commissural and commis-sural lines was the buccal corridor measurement.

Fig 3. Measurements for MSI values. The vertical mea-surement was made between the upper and lowervermilion borders at the midline. The transverse mea-surement was made between the right and left lipcommissures.

numerary teeth, (3) overjet and overbite of 2 to 5 mm,

(4) no active periodontal disease and no periodontaltreatment except for routine scaling and polishing,(5) normal upper lip length (in a balanced face, thelength of the upper lip [distance from subnasale tostomion] is equal to one third of lower facial height[subnasale to menton]), (6) no craniofacial anomaliesor other pathologies, (7) no severe malocclusion (es-thetic component of index of orthodontic treatmentneed was followed so that laypersons’ perspectiveswere also considered; scores including and below 3were included), and (8) no canting of the maxillaryocclusal plane.

Informed consent forms, approved by The Tamil-nadu MGR Medical University (Tamilnadu, India)ethical committee, were signed and obtained from each

Fig 4. Classification of smiles as very good, good,average, and bad. The data from specialists and layper-sons were pooled. Note the distribution of female smilesbetween good and average. Four male smiles wererated as bad, even after strict inclusion criteria werefollowed in selecting subjects.

Fig 5. Consonant and nonconsonant smiles. Note thatmost women have consonant smile arcs.

subject before taking the frontal view photograph. The

29

American Journal of Orthodontics and Dentofacial OrthopedicsApril 2008

518 Krishnan et al

subjects were photographed with posed smiles (becauseposed smiles are the most repeatable) after seating themin a cephalostat with natural head position. Pictureswere taken in the same environment with the samelighting conditions by using Yashica Dental Eye IIcamera (Kyocera Co, Japan). The camera was fixed inposition with a tripod, and all the photographs weretaken in color. The photographs were then transferredto computer software (Adobe Photoshop, version 7,Adobe Systems, San Jose, Calif); they were croppedwith vertical (nose tip and soft-tissue pogonion) andtransverse (perpendicular drawn down from the zygo-matic prominence) limits. All images were then ad-justed to a standardized image size. Measurementsbetween 2 points (subnasale to soft-tissue menton) wereconsidered representative to check magnification error.This was then compared with clinical measurementsand was found to have a statistically significant corre-lation (0.92). The ruler in the same software was usedto obtain all measurements for this study. Averages ofthe 2 measurements made by 2 investigators (V.K. andS.T.D.) were used for the study. To verify the repeat-ability of the measurements, 15 randomly selectedsubjects were photographed again with the same set-tings. The MSI parameters measured in these photo-graphs were compared with the initial photographs ofthe same patients to check for reproducibility of posedsmiles. We obtained statistically significant correlationsbetween the posed smile photographs taken at the 2

Table II. Buccal corridor measurements

Male, right sideMale, left sideFemale, right sideFemale, left side

Pearson corr

Right vs left sides Pearson corre

Male smiles 0.865Female smiles 0.835

Independent samples

Male vs female t value

Right side 0.057Left side 0.355

Paired t test to assess diffe

Female smiles 1.454Male smiles 1.794

times (correlations of 0.84 [vertical parameter, interver-

milion distance in the midline) and 0.86 [horizontalparameter, intercommisural distance]).

A panel of evaluators was selected: 20 persons, ofwhom 10 were dental specialists (2 orthodontists, 2prosthodontists, 2 endodontists, 2 periodontists, 1 oralanatomist, and 1 oral surgeon) and 10 were laypersons(family members of patients at our outpatient clinic onthe day of test). The standardized photographs wereprojected with a projector (Philips multimedia projectorNo. LC 3631/40; Philips, Amsterdam, the Nether-lands). All 60 photographs were randomly projectedbefore the panel as a slide show to familiarize themwith the photographs, before asking the panel to eval-uate each one. The panelists were given a chart with avisual analog scale (VAS) ranging from 1 to 10 (1, verygood; 10, worst) to rate the overall appearance of thesmiles. The VAS was briefly explained to the panelmembers, with illustrations. For the evaluation, eachsmile was projected for 25 seconds, and the ratingswere tabulated. The average value for each smile wascalculated and rated from good to bad with the follow-ing classification system: �3, very good; 3-4.9, good;5-6.9, average; �7, bad.

Then the smiles were segregated and tabulated(Table I) by sex. The Spearman rho correlation coeffi-cient was calculated with SPSS software (version 12,SPSS, Chicago, Ill) to find any correlation betweenspecialists’ and laypersons’ ratings, and between maleand female smiles.

Mean (mm) SD

6.97 1.617.43 1.736.95 1.307.27 1.75

coefficients

Significance

Significant at 0.01 level (2 tailed)Significant at 0.01 level (2 tailed)

or equality of means

P value Significance

.954 Not significant

.724 Not significant

etween right and left sides

.157 Not significant

.083 Not significant

elation

lation

t test f

df

5858

rence b

29

The 60 standardized photographs were developed

American Journal of Orthodontics and Dentofacial OrthopedicsVolume 133, Number 4

Krishnan et al 519

into 4 � 6-in prints and labeled according to sex. Twolines were drawn on the photograph (1 connecting theincisal edges of the maxillary incisors and the cusp tipsof the maxillary canines, and the other through theupper border of the lower lip). Both lines were checkedfor parallelism (Fig 1). The numbers of consonant(parallel) and nonconsonant (nonparallel) smiles weredetermined separately for men and women.

Each photograph was opened in the computersoftware to draw the inner and outer commissural lineson both sides as proposed by Ackerman and Acker-man.10 The distances between these lines on both sideswere obtained as the right and left buccal corridorspaces (Fig 2). The paired t test was used to determinewhether there was any significant difference betweenthe right and left sides in both sexes. Regressionanalysis and Pearson correlation coefficients were cal-culated between right and left buccal corridor spaces inboth sexes. To check for correlation in the right and leftsides between the male and female smiles (right sides inmen and women; left sides in men and women), theindependent samples t test was used.

The smile index proposed by Ackerman et al6 andAckermann and Ackermann10 is calculated by dividingthe intercommissural width by the interlabial gap(width/height). We used a modified form of this smileindex, called the MSI. As part of the modification, toinclude the lips, we measured intervermilion distance atthe midline for height (Fig 3). We reversed the calcu-lation; we divided height by width and multiplied thatby 100 to obtain a percentage. This was done mainly tomake the calculated value comparable with other com-mon indexes such as the facial index (N-Gn/Zy-Zy X100) and the mandibular width-facial height index(Go-Go/N-Gn X 100). The formula for calculating theMSI is:

Intervermilion distance at midline

Intercommissural distance� 100

The facial index and the mandibular width-facialheight indexes were calculated directly from clinicalmeasurements with the methods proposed by Farkasand Posnick25 and Proffit and Fields.26 All values(MSI, facial index, and mandibular width-facialheight index) were tabulated without any sex differ-ence. The female and male data were pooled aftersubjecting the values to correlation and regressionanalyses separately. We observed little differencebetween these values, making pooling of data feasi-ble for this study. (Correlations of 0.291 and – 0.287were obtained for separate analyses by for the male

and female samples, respectively. Regression analy-

sis with male and female data provided ‘B’ values of0.287 and 0.291, respectively.) Regression analysisand Pearson correlation coefficients were calculatedto find any correlations between these parameters.Individual vertical (intervermilion distance andN-Gn distance) and transverse (mandibular width[Go-Go] and intercommissural distance) measure-ments were also compared for correlations withPearson correlation coefficient with the SPSS soft-ware.

A retrospective power analysis with Power and Pre-cision software (version 2.0, Power and Precision, 2000,developed by Borenstein, www.powerandprecision.com)was calculated to determine whether the correlations for

Fig 6. Regression plots comparing right and left buccalcorridor spaces: A, male smiles; B, female smiles. Notethe positive correlations between the right and left sides inboth plots, indicating predictability in relationships.

comparison between the MSI and the other indexes were

American Journal of Orthodontics and Dentofacial OrthopedicsApril 2008

520 Krishnan et al

valid. For this purpose, the P value (�) was set at .01 (2tailed).

RESULTS

The data from the VAS are summarized in Table I,which also shows the Spearman rho correlation coeffi-cient values and the significance levels at P �.01. It isclear that, when rated by both specialists and layper-sons, there is no statistically significant correlationbetween male and female smiles. A highly significantcorrelation was observed between specialists’ and lay-persons’ ratings when male and female smiles wereconsidered separately.

The classification system showed no smile in the“very good” category; 14 female smiles were rated asgood, and 16 were rated as average. In the men, 4smiles were rated as bad, and most (24) was rated asaverage (Fig 4). When male and female smiles werecompared for consonance, we found a significant differ-ence between the numbers of consonant and nonconso-nant smiles, with more female consonant smiles (Fig 5).

The paired t test with buccal corridor measurementsshowed no significant difference statistically betweenthe right and left sides for either sex, and there was highcorrelation between the right and left sides. No statis-tically significant difference was seen when indepen-dent samples t tests for equality of means were calcu-lated between male and female buccal corridor spaces.Regression plots with the data showed a positiverelationship between the right and left sides in bothsexes (Table II, Fig 6).

Table III. Mean and standard deviations of all measuremindex and results of the Pearson correlation coefficient

N-Gn distanceBizygomatic widthBigonial widthIntervermilion distance at midlineIntercommissural widthFacial indexMandibular width-facial height indexMSI

Pearson corr

Grouping

Intervermilion distance at midline and N-Gn distanceIntercommissural width and bigonial widthMSI and facial indexMSI and mandibular width-facial height index

Means, standard deviations, and Pearson correla-

tion coefficient values for the MSI, the facial index, andthe mandibular width-facial height index are given inTable III. Comparisons between measurements forcorrelating the vertical and transverse parameters dif-ferently are also shown in Table III. Clearly, there is nocorrelation between the facial index and the MSIvalues. Comparison of vertical (intervermilion distanceand N-Gn distance) and transverse (mandibular widthand intercommissural distance) parameters with thePearson correlation showed a high correlation coeffi-cient that was significant statistically (Fig 7). A nega-tive correlation of 0.363 was observed between themandibular width-facial height index and the MSI thatwas significant statistically. Regression plot analysis bycomparison of the mandibular width-facial height indexand the MSI confirmed this finding (Fig 8).

With the correlation obtained (0.363), the poweranalysis showed a power of 0.628 with a sample size of60 and the P value set at .01. This means that 63% ofstudies would be expected to generate a significanteffect such as this. (With P at .05, the power test for thissample size would be 84%, considered highly signifi-cant for a scientific study.)

DISCUSSION

We evaluated characteristics of the smile with 2methods. One was the subjective evaluation of overallappearance of the smile with ratings by 10 specialistsand 10 laypersons on a VAS. The other was quantifi-cation of smile characteristics by identifying consonantand nonconsonant smiles, by measuring buccal corri-

r MSI, facial index, and mandibular width-facial height

Mean (mm) SD

107.45 6.03115.02 11.89123.95 2.28

30.95 4.0567.42 4.6992.51 6.33

115.65 7.1346.04 6.16

coefficients

lation coefficient Significance

0.381 Significant at 0.01 level (2 tailed)0.442 Significant at 0.01 level (2 tailed)0.194 Not significant

�0.363 Significant (negative correlation)at 0.01 level (2 tailed)

ents fos

elation

Corre

dors, and by calculating the MSI and comparing it with

American Journal of Orthodontics and Dentofacial OrthopedicsVolume 133, Number 4

Krishnan et al 521

the facial index and the mandibular width-facial heightindex.

The high correlation between specialists and lay-persons in rating the overall appearance of the smilesgoes against most research in this regard.12-16 Ourresults agree with those of Flores-Mir et al,11 whostated that the level of dental-related education has littleimpact on dental and esthetic perceptions. We includedmore panelists for evaluating the smiles and went 1 step

Fig 7. Regression plots comparing the vertical andtransverse anthropometric measurements: A, correla-tion between vertical parameters (facial height [N-Gn]and intervermilion distance at midline); B, correlationbetween transverse parameters (mandibular width [Go-Go] and intercommisural distance). Note the positivecorrelation in both plots, indicating predictability inrelationships.

farther than other reports, with equal numbers of dental

specialists (from 6 backgrounds) and laypersons. Theonly drawback or shortcoming of this study was notconsidering male and female judges separately to de-termine any differences in evaluating smiles by sex. Ithas been reported that laypersons consider facial anddental arrangements in evaluating overall smile appear-ance.12-16 These and our results point to the need forthorough discussion with a patient, to learn his or herpreference of the smile and the dental arrangement,before determining a treatment plan.

In this study, we also evaluated consonance andnonconsonance in the smile-arc relationship. It is wellknown that a consonant smile arc is more attractivethan a nonconsonant one.4,17 We found more women(67%) with consonant smiles than men (40%). Todetermine whether this has any relationship to attrac-tiveness, we categorized the ratings of the panelists fora comparison with our MSI. To our surprise, no smilewas rated as “very good,” and the 14 female smilesrated as “good” all had consonant smile arcs. Thus,female smiles seem to be more attractive and consonantthan male smiles. This aspect has clinical implications.The arc of the maxillary incisal edges can be altered bytherapeutic measures—either orthodontic or restorativetreatment. In orthodontics, the brackets can be carefullypositioned (not to disturb the existing consonance or tocreate a consonant smile) to create a parallel smile-arcrelationship that is also attractive. It was proposed thatincreasing the cant of the maxillary occlusal plane tothe Frankfort horizontal plane will increase maxillaryanterior tooth display and improve the consonance ofthe smile.10 There also is a relationship between a

Fig 8. Regression plot comparing MSI and the mandib-ular width-facial height index. Note the negative corre-lation between these indexes, indicating that an in-crease in 1 will result in a decrease in the other.

patient’s arch form and smile-arc curvature. We sug-

American Journal of Orthodontics and Dentofacial OrthopedicsApril 2008

522 Krishnan et al

gest not altering the patient’s original arch form or, atleast, not creating a broader arch form that will causeless curvature of the anterior segment resulting in a flatsmile arc, because this is often unesthetic. This findingagrees with a recent report that flat smile arcs havelower attractiveness ratings.27

The impact of buccal corridors on smile estheticshas been studied in recent years. All reports show thatbuccal corridors have no effect on the esthetic evalua-tions of smiles.16,23,24,27 The results of our study, whichlooked at differences between buccal corridor spaces ofthe right and left sides in posed smiles, indicated nostatistically significant difference between the right andleft negative spaces in either sex. We also observed apositive correlation between the right and left buccalcorridor spaces in both sexes. The regression plotconfirmed this finding and suggests predictability of 1side from the other. The positive regression plot indi-cates that, as the negative space on 1 side increases ordecreases, the same change will occur on the other side.This finding was same in both sexes.

When male and female smiles were compared fortheir buccal corridor measurements, we found a highcorrelation. This finding disagreed with our VAS val-ues, which showed a significant difference betweenperceived male and female smiles. Thus, we canassume that buccal corridor space plays only a minimalrole in esthetic evaluation of a smile, and the perceiveddifference could be due to other factors such as smilearc, tooth arrangement, tooth shade, gingival architec-ture, gingival display, lip thickness, and so on. Thedrawback of this research was the controlled sampleselection with strict inclusion criteria, including attrac-tive smiles with only minimal to moderate buccalcorridor spaces. These results should be viewed cau-tiously, since a recent report stated that excessivebuccal corridor spaces are less attractive to both orth-odontists and laypersons.24

In the last part of our research, we proposed andcalculated the MSI, enabling a smile to be comparedwith time-tested indexes (facial index and mandibularwidth-facial height index). This research is the first thatattempted a comparison of facial indexes to smilecharacteristics. The anthropometric measurements madein this study helped us to determine that no relationshipexists between facial types and smile characteristics.The various facial types (euryprosopic, mesoprosopic,and leptoprosopic) as proposed by Farkas and Posnick,25

had different MSI values that could not be correlated.An interesting finding was the correlation between

vertical and transverse parameters when analyzed indi-vidually. This finding agrees with a report that vertical

development of the face causes increases in all associ-

ated parameters and affects smile characteristics.7 Wealso found correlation between transverse parameters.The high correlations indicate predictive capability;this means that a change in 1 parameter will have adefinite influence on the other and will be reflected as achange in its measurement, with the ratio maintained.When we combined both correlating variables in theform of MSI and correlated with mandibular width-facial height index, it showed a negative correlation.This suggests that any change in the mandibular width-facial height index value will have a negative impact onthe MSI value (as the mandibular width-facial heightindex increases, there will be a corresponding decreasein the MSI index value). Although the correlationcoefficient was low (0.363), this finding was significantstatistically. The clinical significance of these results isstill questionable and points to the need for thoroughresearch with larger samples and a well-planned re-search strategy. We suggest that this report is prelimi-nary in this regard, because more research should beperformed to confirm the finding and to elucidate itspredictive value.

CONCLUSIONS

1. There is no difference in evaluation of overall smilecharacteristics between specialists and laypersons.

2. More women (67%) than men (40%) have conso-nant smiles.

3. There is no significant difference between right andleft buccal corridor spaces in either sex; thesespaces have a high correlation.

4. The MSI values showed no correlation with facialindex values.

5. There are positive correlations between verticalanthropometric (N-Gn) and vertical measurementsof the MSI (intervermilion distance at midline) andtransverse anthropometric (Go-Go) and transversemeasurements of the MSI (intercommissural dis-tance).

6. The MSI has low negative correlation with themandibular width-facial height index that is signif-icant statistically.

We thank the chairman and the management ofRajas Dental College, staff members (Suresh Sathia-sekar, Biju Sebastian, Babukuttan Pillai, and DhivakarSelvaraj) of the Department of Orthodontics, statisti-cians Muralidharan Nair and Nisha Kurian, and thepanel members.

REFERENCES

1. Roget’s II: The New Thesaurus. 3rd ed. Available at: http://

www.answers.com/topic/smile. Accessed January 1, 2006.

American Journal of Orthodontics and Dentofacial OrthopedicsVolume 133, Number 4

Krishnan et al 523

2. LaFrance M, Hecht MA, Paluck EL. The contingent smile: Ameta-analysis of sex differences in smiling. Psychol Bull 2003;129:305-34.

3. Rigsbee OH, Sperry TP, Begole E. The influence of facialanimation on smile characteristics. Int J Adult Orthod Orthog-nath Surg 1988;3:233-9.

4. Sarver DM, Ackerman MB. Dynamic smile visualization andquantification: part 2. Smile analysis and treatment strategies.Am J Orthod Dentofacial Orthop 2003;124:116-27.

5. Brown WM, Palameta B, Moore C. Are there nonverbal cues tocommitment? An exploratory study using the zero-acquaintancevideo presentation paradigm Evol Psychol 2003;1:42-69.

6. Ackerman JL, Ackerman MB, Brensinger CM, Landis JR. Amorphometric analysis of posed smile. Clin Orthod Res 1998;1:2-11.

7. Hosseinzadeh-Nik T, Yazdani-Damavandi L, Kharazi-Fard MJ.The correlation of smile line with the vertical cephalometricparameters of anterior facial height. J Dent 2005;2:21-8.

8. Tjan AH, Miller GD, The JG. Some esthetic factors in a smile.J Prosthet Dent 1984;51:24-8.

9. Ackerman JL. The emerging soft tissue paradigm in orthodonticdiagnosis and treatment planning. Clin Orthod Res 1992;2:49-52.

10. Ackerman MB, Ackerman JL. Smile analysis and design in thedigital era. J Clin Orthod 2002;36:221-36.

11. Flores-Mir C, Silva E, Barriga MI, Lagravere MO, Major PW.Layperson’s perception of smile aesthetics and facial views.J Orthod 2004;31:204-9.

12. Albino JE, Tedesco LA, Conny DJ. Patient perceptions ofdentofacial esthetics: shared concerns in orthodontics and prosth-odontics. J Prosthet Dent 1984;52:9-13.

13. Bell R, Kiyak HA, Joondeph DR, McNeill RW, Wallen TR.Perceptions of facial profile and their influence on the decision toundergo orthognathic surgery. Am J Orthod 1985;88:323-32.

14. Prahl-Andersen B, Boersma H, Van der Linden FP, Mooer

AW. Perceptions of dentofacial morphology by laypersons,

general dentists and orthodontists. J Am Dent Assoc 1979;98:209-12.

15. Kokich VO, Kiyak HA, Shapiro PA. Comparing the perceptionof dentists and laypeople to altered dental esthetics. J Esthet Dent1999;11:311-24.

16. Roden-Johnson D, Gallerano R, English J. The effects of buccalcorridor spaces and arch form on smile esthetics. Am J OrthodDentofacial Orthop 2005;127:343-50.

17. Sarver D. The importance of incisor positioning in the esthetic smile:the smile arc. Am J Orthod Dentofacial Orthop 2001;120:98-111.

18. Husley CM. An esthetic evaluation of lip-teeth relationshipspresent in the smile. Am J Orthod 1970;57:132-44.

19. Wong NKC, Kassim AA, Foong KWC. Analysis of estheticsmiles by using computer vision techniques. Am J OrthodDentofacial Orthop 2005;128:404-11.

20. Mejia-Maidl M, Evans CA. Soft tissue facial considerations andorthodontic treatment. Semin Orthod 2000;6:3-20.

21. Frush JP, Fisher RD. The dynesthetic interpretation of thedentogenic concept. J Prosthet Dent 1958;8:558-81.

22. Johnson DK, Smith RJ. Smile esthetics after orthodontic treat-ment with and without extraction of four first premolars. Am JOrthod Dentofacial Orthop 1995;108:162-7.

23. Moore T, Southard KA, Casko JS, Qian F, Southard TE. Buccalcorridors and smile esthetics. Am J Orthod Dentofacial Orthop2005;127:208-13.

24. Parekh SM, Fields HW, Beck M, Rosenstiel S. Attractiveness ofvariations in the smile arc and buccal corridor space as judged byorthodontists and laymen. Angle Orthod 2006;76:557-63.

25. Farkas LG, Posnick JC. Growth patterns of the face: a morpho-metric study. Cleft Palate Craniofac J 1992;29:308-15.

26. Proffit WR, Fields HW. Contemporary orthodontics. 3rd ed.2000. C. V. Mosby; St Louis: p.159.

27. Ritter DE, Gandindi LG, Pinto Ados S, Locks A. Estheticinfluence of negative space in the buccal corridor during smiling.

Angle Orthod 2006;76:198-203.