184

The mesiodistal tooth size of the maxillaryand mandibular arch must relate to each other in orderto obtain an excellent occlusion at the completion ofthe orthodontic treatment.

Before 1949, nobody expressed the differences intooth size between the upper and lower arch in a ratio.Neff1 developed the anterior coefficient, a method tocompare the widths of the anterior teeth in the upperarch with the teeth in the lower arch. The value of thiscoefficient was associated with the amount of overbiteat the incisor level; an ideal overlap of 20% of the lowerincisor crown height was associated with a coefficientof 1.20 to 1.22. Bolton2 analyzed a group of 55 excel-lent occlusions. He introduced mathematical tooth sizeratios, which were supposed to be helpful in diagnosisand treatment planning. A determined tooth size ratiomay predict the functional and esthetic outcome of thecase. His tables for anterior and overall tooth size ratiosare still used today.

As a result of secular changes, mesiodistal toothwidth increases.3 In an epidemiologic study on poten-

tial orthodontic patients in the US Army (Fort Meadestudy4), an important amount (13.4% overall and30.6% anterior ratio) of patients showed ratios beyond2 standard deviations (SD) as calculated by Bolton.

There are many therapies proposed for resolvingtooth size discrepancy problems. Interproximal strip-ping5-7 or extraction of 1 lower incisor is mentioned bymany investigators.8-10

The maximum amount of interproximal strippingthat one can obtain is 2 to 3 mm in the anterior seg-ment.6,11 According to Paskow7 and Hussels andNanda,12 interproximal stripping of teeth is not a treat-ment on its own but a last resort when alignment ofteeth and occlusion fails.

Artistic positioning of the maxillary incisors isanother option; it provides 1 additional millimeter pertooth by tipping the root 6° distally. This also dependson the shape of the incisor.10 Moreover, torquing ofmaxillary incisors may close small diastemas in thefront area.10-12 Beyond orthodontic solutions, restora-tive dentistry may be helpful in this respect.11

Many investigators claimed that mesiodistal toothsize discrepancy (TSD) is a very important elementin diagnosis, which should be measured in eachorthodontic patient before the start of orthodontictreatment.3,4,13-15 It remains questionable whetherTSD always affects the final orthodontic treatmentoutcome. According to Bolton,2 the overall tooth sizeratios varied between 87.5 and 94.8, with a meanvalue of 91.3 (SD = 1.91). Stifter16 evaluated

CONTINUING EDUCATION ARTICLE

The effect of tooth size discrepancy on occlusion: Anexperimental study

Marianne Heusdens, DDS,a Luc Dermaut, DDS, PhD,b and Ronald Verbeeck, PhDc

Gent, Belgium

The purposes of this experimental study are the following: (1) to compare the anterior and overall tooth sizeratios reported by Bolton to values reported in epidemiologic studies, (2) to assess the accuracy of tooth sizediscrepancy measurements, (3) to investigate to what extent generalized tooth size discrepancy affectsocclusion, (4) to investigate the effect of leveling the curve of Spee, and (5) to evaluate the effect of extractiontherapy of 4 premolars on occlusion. For the first part of the study, Bolton’s mean anterior and overall toothsize ratio (as well as the extraction values) were compared with calculations derived from 4 publicationsreporting mean mesiodistal tooth width by using the t test (P ≤ .05). The second part of the study was carriedout on a setup made from 1 dental cast of a male patient judged to have an ideal occlusion at the end of anonextraction treatment. Our data reveal (1) no significant difference between the overall tooth size ratios ofthe Bolton values compared to 4 studies, but the anterior ratios were significantly different, (2) highreproducibility (99%) of tooth size discrepancy measurements, (3) that severe tooth size discrepancy affectsthe occlusion only a little, (4) that an excessive (6 mm) curve of Spee creates the poorest setup result, and (5)that extraction therapy only slightly affected the final occlusion. The effect of generalized tooth size discrepancyappears to be limited. (Am J Orthod Dentofacial Orthop 2000;117:184-91)

From the Department of Orthodontics and the Department of Dental MaterialScience, School of Dentistry, University of Gent.aPostgraduate student, Department of Orthodontics.bProfessor and Chair, Department of Orthodontics.cProfessor, Department of Dental Material Science.Reprint requests to: Luc Dermaut, DDS, PhD, Professor of Orthodontics,Department of Orthodontics, School of Dentistry, University of Gent, De Pin-tenlaan 64, 9000 Gent, Belgium.Copyright © 2000 by the American Association of Orthodontists.0889-5406/2000/$12.00 + 0 8/1/98619

American Journal of Orthodontics and Dentofacial Orthopedics Heusdens, Dermaut, and Verbeeck 185Volume 117, Number 2

Bolton’s ratios on a group of patients with 34 normaland 24 ideal occlusions and arrived at similar results.Hunter and Priest17 evaluated the experimental errorsand discrepancies involved in the measurement oftooth size on plaster casts and made several sugges-tions about these measurements.

Earlier studies only used plaster material for mea-suring TSD. We feared, however, that repeated mea-surements might damage the teeth and would make themesiodistal measurements inaccurate. Therefore, as apart of this study, other materials will be tested to makethe measurements as accurate as possible.

It remains questionable which deviations of theoverall and anterior values proposed by Bolton affectthe final treatment outcome. Pegg-shaped lateralincisors are easily detected and express most often aTSD. Esthetic correction (crowns, veneers, etc) is oftenthe first choice of treatment. Hidden TSD from a gen-eralized discrepancy in tooth width between upper andlower teeth is less detectible at first sight and maycause a less favorable treatment outcome. In this study,a created generalized TSD was tested on a setup in anattempt to investigate to what extent generalized TSDdoes affect the final occlusion.

There are more parameters that affect occlusion.According to Andrews,18 the absence of any 1 or moreof the “Six keys” results in an occlusion that deviatesproportionally from normality. The closure ofdiastemas, the correction of tooth rotations, and treat-ment toward a Class I molar relationship are wellaccepted treatment goals. As has been explained here,changing the crown angulation and inclination is atreatment procedure proposed as a therapy for resolv-ing TSD problems.

Leveling the curve of Spee to a flat or a slightlycurved occlusal plane is a common treatment objectivein orthodontic therapy. According to Proffit and Acker-man,19 a clinical rule of thumb is that 1 mm of arch cir-cumference is needed for each millimeter of curve ofSpee depth present. Baldridge20 and Germane et al21

stated that the ratio between these 2 variables was notlinear and less than 1 to 1. Lower incisor proclinationoften associated with leveling is primarily the result ofimproper orthodontic mechanics.22 There is a naturaltendency for the depth of the curve of Spee to decreasewith age.23-26 Although it is generally accepted in clin-ical orthodontics that leveling of the curve of Spee isone of the treatment objectives, there is to our knowl-edge no scientific evidence that this change does guar-antee a better occlusion and in the long term a more sta-ble treatment result.

Extraction therapy is a common procedure in clini-cal orthodontics. Lundström27 proposed extraction

therapy in cases where there is a genetic discrepancybetween the size of the jaws and the teeth. To keep theBolton ratio constant, it becomes necessary to removemore tooth material in the maxillary arch than in thelower jaw. Bolton28 presented practical examples ofdentitions with an ideal overall ratio in which premolarextraction was carried out. He concluded that thereshould be a difference in tooth size between lower andupper premolars of 1 mm per tooth in order to keep theoverall ratio constant. Because most often the differ-ence in mesiodistal width between upper and lowerpremolars is less than 1 mm, Bolton proposed that theoverall ratio should not be used as a specific guide tothe predicted occlusion after the removal of 4 premo-lars. Although a difference of 1 mm between upper andlower premolars does not affect the overall ratio in thesame manner, Bolton correctly stated that extractiontherapy is affecting the suggested overall ratio value of91.3. After the extraction of 4 premolars, patients with-out TSD would have an overall ratio ranging between87 and 89.28 Saatçi and Yukay29 used a mean 88 valueas a norm to measure TSD in different combinations ofextractions. The removal of the 4 first premolars cre-ated the most severe TSD, whereas the extraction of all4 second premolars created a lesser discrepancy. Itwould be interesting to know whether this value sug-gested by Bolton is a useful ratio to establish a goodocclusion in extraction cases. Therefore in this study,the measured mesiodistal tooth widths reported in 4publications4,14,30,31 will be used to calculate TSD inthe 4 possible extraction therapies. Moreover, it seemsinteresting to know whether extraction therapy doescreate a TSD of clinical importance.

The purposes of this study are:1. To compare the anterior and overall ratio values reported

by Bolton (ideal occlusions) to the calculated values fromdata in other epidemiological studies (mesiodistal widths)

2. To assess the accuracy of TSD measurements on dentalcasts compared with epoxy models

3. To investigate to what extent generalized TSD affectsocclusion

4. To investigate the effect on occlusion of leveling the curveof Spee

5. To investigate the effect of extraction therapy on the finalocclusion

MATERIAL AND METHODS

Bolton’s mean overall (with and without premolarextraction therapy) and anterior ratios were comparedwith calculations derived from 4 publications4,14,30,31

reporting mean mesiodistal size of teeth. Weightedmean values and SDs were calculated according tothe formulas:

186 Heusdens, Dermaut, and Verbeeck American Journal of Orthodontics and Dentofacial OrthopedicsFebruary 2000

Weighted mean value =

Standard deviation = ��Bolton used only ideal occlusions to define his mean

value. Most epidemiologic studies report on themesiodistal width of teeth in a selected group of patients.The ratios calculated by using these values were com-pared to Bolton’s standards by using the t test (P ≤ .05)and were considered to be statistically significant.

The measurements for the second part of the studywere taken from 1 dental cast of a male patient with anideal occlusion at the end of a nonextraction treatment(Fig 1). To minimize errors induced by duplication ofdental casts, silicone molds were made of both themaxillary and mandibular arch and of each separatedtooth. The separation of the teeth from the dental castwas carried out by a figure saw without touching thecontact points. Finally the teeth were separated bybreaking.32 The separated teeth were duplicated with 3materials (plaster, Epoxy-Die, and Araldit) to comparethe various shrinkage percentages. To avoid damage tothe duplicated teeth during measurement proceduresand in order to create a precise amount of TSD, atougher material than plaster would be convenient.Sliding callipers (permitting readings to the nearest0.02 mm) were used to determine the largest mesiodis-tal diameter of the separated teeth from first molar tofirst molar, measured perpendicularly to the occlusalsurface. The total mesiodistal size on dental cast (plas-ter) of the maxillary and mandibular teeth (from firstmolar to first molar) was calculated by needle pointdividers (0.1 mm accuracy) with the points of thedividers parallel to the occlusal plane.

The mesiodistal width of the teeth was transferredto a piece of paper, and the total width was measuredwith sliding callipers.32 The latter procedure was car-

1�� �

S12�

� �S

x12�

�� �

S1

2�

ried out by 3 observers. All measurements wererepeated 10 times. Interobserver and intraobserver reli-ability were tested. Mean anterior and overall ratioswere calculated, and 1 way analysis of variance(ANOVA) was used.

In order to evaluate the effect of an artificially cre-ated TSD on occlusion in a setup, different amounts ofapproximal tooth material were equally ground offfrom the anterior area (AA), from the total dental arch(DA), and from the buccal area (BA). In the experi-mental setup, the inclination and angulation of thefront teeth were considered to have an effect on thefinal setup.12 Therefore, a self-made inclination indexwas proposed indicating that the inclination wasexpressed in a score that varied between 0 and 2. Ascore of 0 was considered to be visually nice, 1 was anacceptable inclination, and 2 was an unacceptableinclination. Separately, the anteroposterior angulation

Fig 1. Dental cast of an ideal occlusion.

Table I. Single arches with their specifications

Upper/lower Extractions

arch Grinding UA LA Spee

0 / / / 3 mm1 2 mm AA / / 3 mm2 4 mm AA / / 3 mm3 6 mm AA / / 3 mm4 4 mm DA / / 3 mm5 6 mm DA / / 3 mm6 8 mm DA / / 3 mm7 2 mm BA / / 3 mm8 4 mm BA / / 3 mm9 6 mm BA / / 3 mm

10 / 2 × P1 2 × P1 0 mm11 / 2 × P2 2 × P2 0 mm12 / 2 × P1 2 × P2 0 mm13 / 2 × P2 2 × P1 0 mm14 / 2 × P1 2 × P1 3 mm15 / 2 × P2 2 × P2 3 mm16 / 2 × P1 2 × P2 3 mm17 / 2 × P2 2 × P1 3 mm18 / / / 0 mm19 / / / 3 mm20 / / / 6 mm

AA, Anterior area; DA, dental arch; BA, buccal area; UA, upperarch; LA, lower arch.

American Journal of Orthodontics and Dentofacial Orthopedics Heusdens, Dermaut, and Verbeeck 187Volume 117, Number 2

was scored likewise. The PAR index (Peer AssessmentRating)33,34 was used to score the occlusion on setup.This index was chosen because of its objective featuresin describing and evaluating a malocclusion. The sec-ond molar was not evaluated according to Bolton’sratio and the PAR index. The PAR index and the self-made inclination index and angulation index werescored blind twice, with an interval of 2 weeks by 10different observers. Consistency in the measurementsby these observers was tested. An attempt has beenmade to establish ideal setup situations according to the“6 keys to normal occlusion” of Andrews.18 In caseswith TSD different setups can be made to compensatefor the discrepancy. We have chosen to establish idealocclusions in the buccal segments (Class I) and to com-promise on anterior occlusions. By doing so, the PARvalues may be affected to the highest extent because ofthe importance of overjet and overbite in the calcula-tion of the PAR index. The overjet and overbite valuesin the PAR index calculations have an impact factor of6 and 4 compared with the buccal occlusion factor of 1.To obtain comparable setup situations, the socket of themodels (including some information of the alveolarprocess but without teeth) was multiplied and used foreach setup. This procedure was followed in an attemptto obtain comparable dental arches. All discrepanciesthat were simulated in the upper jaw were set up on theideal lower dental arch (model 0, see Table I), whereas

the discrepancies induced in the lower jaw were set upon the ideal upper model (model 0, see Table I).

To evaluate the effect of an artificially created TSDon occlusion in a setup, in a first series of models, 2, 4,and 6 mm of approximal tooth material was groundfrom the anterior teeth in the upper jaw. These 3 mod-els were put together with lower setup models in whichno stripping was carried out. The same procedure wasfollowed in the lower jaw (Table I).

In a second series of models, 4, 6, and 8 mm oftooth material was removed from each tooth from firstmolar to first molar. In a third series of models, 2, 4,and 6 mm of tooth material was stripped away in thebuccal segments. Stripping was carried out by meansof a fraise in a working bench (Fig 2). To remove 2 mmof tooth material in the anterior segment, 0.17 mm oftooth material on the approximal surfaces was groundaway. A total of 4 mm grinding in the entire anteriorsegment stands for the removal of 0.33 mm of eachinterproximal surface, whereas a total of 6 mm standsfor 0.5 mm removal of each interproximal surface. Bydoing so, 18 arches were made and put together in asetup situation by 1 examiner (Table II).

Three setups (upper and lower arches 18 to 20 inTable I) were made with the original teeth with 3 vari-ations in curve of Spee (a flat curve, a curve 3 mmdeep, and one 6 mm deep).

The different premolar extraction therapies (4 firstpremolars, 4 second premolars, 2 first maxillary and2 second mandibular premolars, and 2 second maxil-lary and 2 first mandibular premolars) were simulatedin setups with a flat curve of Spee as well as with aslight curve of 3 mm depth (upper and lower arch 10to 17 in Table I).

All created experimental models are listed in TableI. Combinations between various upper and lowermodels resulted in 30 setups (Table II).

To test the reproducibility of the construction of thesetups, the construction of 4 setups selected in eachsubgroup (see later) was repeated 5 times by 1 observerand scored separately by all the observers.

RESULTS AND DISCUSSION

Table III lists the mean Bolton anterior and overallratios compared with the calculated values deductedfrom 4 other studies. Anterior ratios were comparedwith Bolton’s mean 77.2 value. The overall ratios forthe nonextraction therapies as well as for the 4 extrac-tion therapies, for males and females, are comparedwith the mean Bolton value of 91.3. Only the nonex-traction overall ratios were found not to be statisticallydifferent between Bolton and the other studies. Thisfinding indicates that the overall ratio proposed by

Fig 2. Reduction of approximal tooth material.

188 Heusdens, Dermaut, and Verbeeck American Journal of Orthodontics and Dentofacial OrthopedicsFebruary 2000

Bolton and deducted from a selected sample with idealocclusions is not different from the calculated valuesfor this ratio in this population. This does not hold goodfor the anterior ratio. It is a well-known fact that thereis a large variation in the size of the front teeth. Thesize of the upper lateral incisor particularly may differwithin the population. This phenomenon could accountfor this finding. Although the size of the front teeth isinvolved in the overall ratio, this variability has mathe-matically less effect on the overall ratio. The differ-ences in overall ratio in the extraction combinationswere significantly different from the Bolton values(Table III). Saatçi and Yukay2 came to a comparableconclusion. Moreover, they found that extraction of the4 first premolars was creating the largest TSD. In thisstudy, the differences calculated for all extraction ther-apies were not significant from each other.

Mean Bolton values and SDs of the separated teethin 3 different materials were not statistically signifi-cant. There were only slight differences in SDs for thedifferent materials and therefore, all materials werefound to be usable for creating TSDs in setups. Intraob-

server reliability was tested, and no significant differ-ences (ANOVA) were found between the measuringson casts. Intraobserver reliability was high (99%).

Various setup situations and the correspondingaverage PAR indices for the 10 different observers aregiven in Table IV. During the pilot study, the PARindices were found to be very small. Therefore differ-ent observers were introduced in this study in anattempt to minimize individual errors. The PAR rank-ing did not coincide with the registered Bolton valuesindicating that the evaluation of the occlusion (by thePAR index) was not determined by the anterior or over-all ratios according to Bolton. In other words, theseverity of the measured TSD is not reflected in thePAR index ranking.

The mean PAR index consistency was 76.2%. Dou-ble measurements indicating a difference >1 were con-sidered as inconsistent measurements. This standard isvery high if one understands that an error in the mea-surement of the overjet of 1/2 mm can lead to a PARindex difference of 6. Therefore the consistency ofscoring was very acceptable. The inclination index andangulation index turned out to be rather inconsistent(62.5%). It seemed that inclination as well as angula-tion were a matter of the taste of the clinician.

All average PAR index ratings in the main studywere again very low, indicating that the induced TSDgenerally has a minor effect on the occlusion. Threegroups are represented in Table IV: the first with a parindex between 0.05 and 1, the second from 1.1 to 2.35,

Table II. Combinations with upper and lower arches forsetups

UA LA

Setup 1 0 0Setup 2 0 1Setup 3 0 2Setup 4 0 3Setup 5 0 4Setup 6 0 5Setup 7 0 6Setup 8 0 7Setup 9 0 8Setup 10 0 9Setup 11 1 0Setup 12 2 0Setup 13 3 0Setup 14 4 0Setup 15 5 0Setup 16 6 0Setup 17 7 0Setup 18 8 0Setup 19 9 0Setup 20 10 10Setup 21 11 11Setup 22 12 12Setup 23 13 13Setup 24 14 14Setup 25 15 15Setup 26 16 16Setup 27 17 17Setup 28 18 18Setup 29 19 19Setup 30 20 20

Table III. Comparison of Bolton values with the meananterior and overall ratios calculated from mesiodistaltooth width in 4 epidemiologic studies35-38

Bolton (n=55) x– (s) 4 Studies x– (s) n t Test

Anterior ratioMale 77.2 (1.65) 78.9 (2.92) 97 SFemale 78.9 (2.53) 110 S

Overall ratioMale 91.4 (2.68) 97 NSFemale 91.6 (3.11) 108 NS

4 × P1Male 91.3 (1.91) 90.0 (3.02) 97 SFemale 90.4 (2.94) 108 S

4 × P2Male 89.3 (2.95) 97 SFemale 89.5 (2.92) 108 S

2 × P1+, 2 × P2-Male 89.9 (2.99) 97 SFemale 90.0 (2.97) 108 S

2 × P2+, 2 × P1-Male 89.4 (2.98) 97 SFemale 89.4 (2.89) 108 S

S, Significant; NS, not significant (P ≤ .05).

American Journal of Orthodontics and Dentofacial Orthopedics Heusdens, Dermaut, and Verbeeck 189Volume 117, Number 2

and the third from 4.35 to 7.2. By testing the repro-ducibility of the setup (4 setups were made 5 times),the PAR index rating was sometimes a bit different, butthe ranking of the different models remained exactlythe same (see Table V). In Fig 3, a selection of setupsis made to illustrate the various TSD cases, rankedaccording to their PAR value. Only the 4 setups ofgroup 3 were considered the worst occlusions accord-ing to the PAR index. The occlusion was affected to ahigh extent by stripping 4 and 6 mm in the lower buc-cal segments on both sides (Table IV, setup 10 and 9).

The PAR-value was the worst by stripping 6 mm in thelower anterior area (Table IV, setup 4), and by a curveof Spee of 6 mm depth (Table IV, setup 30). Onlysevere stripping and an exaggerated curve of Speeaffects the occlusion to some extent. It has to beemphasized that stripping of 4 and 6 mm in the lowerbuccal area in fact means that a total amount of 8 and12 mm of tooth material are removed. This means thata reduction of tooth material of more than 1 premolarwidth is stripped away to create a malocclusion. Themodel in which 6 mm tooth material in the lower ante-rior segment (incisor width) was stripped away alsorepresents a malocclusion in which there is a severediscrepancy. The setup consequences, however, ofthese malocclusions are illustrated in Fig 3 and werefound to be obvious (PAR = 6.9) but clinically accept-able by the various observers. The model with theexaggerated curve of Spee (PAR = 7.2) is, of course,rather a theoretical than a clinical situation. Therefore,the models with the highest PAR scores are cases withsevere generalized TSD. In these cases, some devia-tions in the occlusion had to be accepted. The choice of

Table IV. Setups listed in order of their PAR indices with overall ratios and anterior ratios

PAR Setup OR AR UA LA Spee

0.05 Setup 29 90.0 / / / 3 mm0.10 Setup 11 91.9 82.2 2 mm AA / 3 mm0.10 Setup 25 87.3 / 2 × P2 2 × P2 3 mm0.20 Setup 15 95.9 / 6 mm DA / 3 mm0.25 Setup 14 93.9 / 4 mm DA / 3 mm0.25 Setup 16 98.1 / 8 mm DA / 3 mm0.30 Setup 8 85.8 / / 2 mm BA 3 mm0.30 Setup 28 90.0 / / / 0 mm0.35 Setup 1 90.0 78.8 / / 3 mm0.40 Setup 12 93.9 86.0 4 mm AA / 3 mm0.40 Setup 27 87.9 / 2 × P2 2 × P1 3 mm0.45 Setup 23 87.9 / 2 × P2 2 × P1 0 mm 0.50 Setup 22 88.0 / 2 × P1 2 × P2 0 mm0.60 Setup 20 88.6 / 2 × P1 2 × P1 0 mm0.65 Setup 17 93.9 / 2 mm BA / 3 mm0.75 Setup 26 88.0 / 2 × P1 2 × P2 3 mm0.85 Setup 24 88.6 / 2 × P1 2 × P1 3 mm0.90 Setup 3 85.8 70.4 / 4 mm AA 3 mm1.00 Setup 5 85.8 / / 4 mm DA 3 mm1.10 Setup 6 83.8 / / 6 mm DA 3 mm1.35 Setup 13 95.9 90.1 6 mm AA / 3 mm1.40 Setup 9 102.7 / 6 mm BA / 3 mm1.50 Setup 7 81.7 / / 8 mm DA 3 mm1.70 Setup 2 87.9 74.6 / 2 mm AA 3 mm2.25 Setup 21 87.3 / 2 × P2 2 × P2 0 mm2.35 Setup 18 98.1 / 4 mm BA / 3 mm4.35 Setup 10 77.6 / / 6 mm BA 3 mm6.80 Setup 9 81.7 / / 4 mm BA 3 mm6.95 Setup 4 83.8 66.2 / 6 mm AA 3 mm7.20 Setup 30 90.0 / / / 6 mm

OR, Overall ratio; AR, anterior ratios; AA, anterior area; DA, dental arch; BA, buccal area; UA, upper arch; LA, lower arch; bold see Fig 3.

Table V. Test of the reproducibility of the setup con-struction. Four setups are reconstructed 5 times andtested by all 10 observers by means of the PAR index

×− v

Setup 22 0.6 0.4 to 0.7Setup 2 1.5 1.3 to 1.6Setup 10 4.3 3.8 to 4.6Setup 30 7.4 6.9 to 7.8

v, Variation in PAR index after reconstruction of the setups.

190 Heusdens, Dermaut, and Verbeeck American Journal of Orthodontics and Dentofacial OrthopedicsFebruary 2000

Fig 3. Setups illustrate various TSD cases ranked according to their PAR value.

American Journal of Orthodontics and Dentofacial Orthopedics Heusdens, Dermaut, and Verbeeck 191Volume 117, Number 2

the teeth to be extracted as a treatment objective wasnot critical according to the PAR index (PAR varyingbetween 0.4 and 0.85). Only the simulation of extrac-tion of 4 second premolars was scored somewhathigher (PAR = 2.25).

CONCLUSIONS

1. The overall ratio calculated by Bolton on models inpatients with an ideal occlusion is representative for thecalculated overall ratios starting from tooth width valuesreported in epidemiologic studies. However, the anteriorratio in epidemiologic studies was somewhat higher thanBolton’s ratio possibly because of a greater morphologicvariability in upper incisor width.

2. There was no statistical difference between the measure-ments on epoxy resin models or plaster models. Both mate-rials are suitable for TSD studies. The reproducibility of theTSD measurements was found to be very high (99%).

3. The PAR index in all the setup situations varied between0.05 and 7.2, indicating that only minor malocclusionswere found. Even in severe TSD cases, an acceptableClass I molar relationship with a reasonable overjet andoverbite was found.

4. An excessive curve of Spee (6 mm) creates the poorestsetup result.

5. Extraction therapy only slightly affected the final occlu-sion, whereas the calculated overall values indicated amaximum discrepancy of 2 mm.Only in severe situations of TSD will the outcome

of the final occlusion, evaluated on setup, be affectedto some extent. The conclusion of this study is thatthe effect of generalized TSD on occlusion appears tobe limited.

REFERENCES

1. Neff CW. Tailored occlusion with the anterior coefficient. Am J Orthod 1949;35:309-33.2. Bolton WA. Disharmony in tooth size and its relation to the analysis and treatment of

malocclusion. Angle Orthod 1958;28:113-30.3. Lavelle CLB. Variation in the secular changes in the teeth and dental arches. Angle

Orthod 1973;43:412-21.4. Freeman JE, Maskeroni AJ, Lorton L. Frequency of Bolton tooth-size discrepancies

among orthodontic patients. Am J Orthod Dentofacial Orthop 1996;110:24-7.5. Ballard ML. A fifth column within normal dental occlusions. Am J Orthod

1956;42:116-24.6. Hudson AL. A study of the effects of mesiodistal reduction of the mandibular anterior

teeth. Am J Orthod 1956;42:615-24.

7. Paskow H. Self-alignment following interproximaal stripping. Am J Orthod1970;58:240-9.

8. Rees DJ. A method for assessing the proportional relation of apical bases and contactdiameters of the teeth. Am J Orthod 1953;39:695-707.

9. Sperry TP, Worms FW, Isaacson RJ, Speidel TM. Tooth-size discrepancy in mandibu-lar prognathism. Am J Orthod 1977;72:183-90.

10. Tuverson RL. Anterior interocclusal relations. Am J Orthod 1980;78:361-70.11. Fields HW. Orthodontic-restorative treatment for relative mandibular anterior excess

tooth-size problems. Am J Orthod 1981;79:176-83.12. Hussels W, Nanda RS. Effect of maxillary incisor angulation and inclination on arch

length. Am J Orthod Dentofacial Othop 1987;91:233-9.13. Sanin C, Savara BS. An analysis of permanent mesiodistal crown size. Am J Orthod

1971;59:488-500.14. Garn SM, Lewis AB, Kerewsky RS. Sex differences in tooth shape. J Dent Res

1967;46:1470.15. Crosby DR, Alexander CG. The occurrence of tooth size discrepancies among differ-

ent malocclusion groups. Am J Orthod Dentofacial Orthop 1989;95:457-61.16. Stifter J. A study of Pont’s, Howes’, Rees’, Neff’s, and Bolton’s analyses on Class I

adult dentitions. Angle Orthod 1958;28:215-25.17. Hunter WS, Priest WR. Errors and discrepancies in measurement of tooth size. J Dent

Res 1960;39:405-14.18. Andrews LF. Six keys to normal occlusion. Am J Orthod 1972;20:671-90.19. Proffit WR, Ackerman J. Diagnosis and treatment planning in orthodontics. In: Graber

TM, editor. Orthodontics: current principles and techniques. St. Louis: CV Mosby;1986. p. 64.

20. Baldrigde DW. Leveling the curve of Spee: its effect on mandibular arch length. JPract Orthodont 1969;III:26-41.

21. Germane N, Staggers JA, Rubenstein L, Revere JT. Arch length considerations due tothe curve of Spee: a mathematical model. Am J Orthod Dentofacial Orthop1992;102:251-5.

22. Braun S, Hnat WP, Johnson BE. The curve of Spee revisited. Am J Orthod Dentofa-cial Orthop 1996;110:206-10.

23. Brown VP, Daugaard-Jensen I. Changes in the dentition from the early teens to theearly twenties. Acta Odontol Scand 1951;9:177-92.

24. Knott VB. Size and form of the dental arches in children with good occlusion studied lon-gitudinally from age 9 years to late adolescence. Am J Phys Antropol 1961;19:263-84.

25. Björk A. Prediction of mandibular growth rotation. Am J Orthod 1969;55:585-99.26. De Kock WH. Dental arch depth and width studied longitudinally from 12 years of

age to adulthood. Am J Orthod 1972;62:56-66.27. Lundström A. Intermaxillary tooth width ratio and tooth alignment and occlusion.

Acta Odontol Scand 1954;12:265-92.28. Bolton WA. The clinical application of a tooth size analysis. Am J Orthod

1962;48:504-29.29. Saatçi P, Yukay F. The effect of premolar extractions on tooth-size discrepancy. Am J

Orthod Dentofacial Orthop 1997;111:428-34.30. Lavelle CLB. Maxillary and mandibular tooth size in different racial groups and in

different occlusal categories. Am J Orthod 1972;61:29-37.31. Howe RP, McNamara JA, O’Connor KA. An examination of dental crowding and its

relationship to tooth size and arch dimension. Am J Orthod 1983;83:363-73.32. van der Linden FDGM, Boersma H. Diagnostiek en behandelplanning in de ortho-

dontie. Alphen a/d Rijn: Samsom Stafle; 1986. p. 111-7.33. Richmond S, Shaw WC, O’Brien KD, Buchanan IB, Jones R, Stephens CD, et al. The

development of the PAR-index: reliability and validity. Eur J Orthod 1992;14:125-39.34. Richmond S, Shaw WC, Roberts CT, Andrews M. The PAR index (Peer Assessment

Rating): methods to determine outcome of orthodontic treatment in terms of improve-ment and standards. Eur J Orthod 1992;14:180-7.

35. Lysell L, Myrberg N. Mesiodistal tooth size in the deciduous and permanent denti-tions. Eur J Orthod 1982:113-22.

36. Moorrees CFA. The aleut dentition. Cambridge MA: Harvard University Press, 1957.37. Howe RP, McNamara JA, O’Connor KA. An examination of dental crowding and its

relationship to tooth size and arch dimension. Am J Orthod 1983;83:363-73.38. Lavelle CLB. Maxillary and mandibular tooth size in different racial groups and in

different occlusal categories. Am J Orthod 1972;61:29-37.