ORIGINAL ARTICLE

Long-term effectiveness of the continuous and the sectional archwire techniques in leveling the curve of Spee
C. Brian Preston,a M. Brent Maggard,b Judith Lampasso,c and Orfan Chalabid Buffalo, NY Introduction: The correction of a deep overbite with the subsequent achievement of long-term stability is a difcult problem for orthodontists. The role of leveling the curve of Spee (COS) in bite opening and the success of orthodontic treatment has been well documented in the literature. The aim of this study was to investigate whether leveling the COS, by using 2 orthodontic treatment techniques, produces stable results on a long-term basis. Methods: We compared the long-term stability of leveling the COS with the straight-wire Alexander technique and the bioprogressive sectional-arch technique. The randomly selected subjects for this retrospective study were obtained from the private practices of Drs R. G. Alexander and Ruel Bench. Study models taken 2 months before treatment (T1), 2 months after treatment (T2), and after retention (T3) were evaluated. Measurements of the COS were made on the mandibular casts with a commercially available palatometer. Mandibular intercanine width, overbite, overjet, mandibular incisor irregularity, and mandibular arch length were also recorded. Results and Conclusions: Both techniques produced highly signicant reductions in the COS (T1 to T2). Statistically signicant, but clinically insignicant, postretention relapse of the COS occurred (T2 to T3). For both techniques, a statistically signicant difference was seen in the incidence of the relapse of the COS between patients who were completely leveled posttreatment and those who were not. We did not nd a correlation between pretreatment COS and relapse in any of the other occlusal traits studied. (Am J Orthod Dentofacial Orthop 2008;133:550-5)

arious techniques have been used to measure the depth of the curve of Spee (COS), which extends from the distal marginal ridges of the most posterior teeth in the arch to the incisal edges of the mandibular central incisors.1-6 Although several theories have been proposed to explain the COS in natural dentitions, its role in mandibular function has been questioned.7-9 There is a belief that an imbalance between occlusal forces can result in an unusually deep COS by allowing the mandibular incisors to overerupt, the premolars to infraerupt, and the mandibular molars to incline mesially.10-13 In the mandible, a deep COS can be leveled by extrusion of infraerupted premolars, intrusion of overerupted incisors, or a combination of these moveFrom the Department of Orthodontics, School of Dental Medicine, State University of New York at Buffalo. a Professor and chairman. b Orthodontist. c Associate professor. d Clinical professor. Reprint requests to: C. Brian Preston, Department of Orthodontics, School of Dental Medicine, State University of New York at Buffalo, Squire Hall Room 140, 3435 Main St, Bldg 32, Buffalo, NY 14214-3008; e-mail, cbp@buffalo.edu. Submitted, August 2005; revised and accepted, February 2006. 0889-5406/$34.00 Copyright 2008 by the American Association of Orthodontists. doi:10.1016/j.ajodo.2006.02.039

ments. Most straight-wire orthodontic techniques advocate a reverse COS incorporated into a continuous archwire to level the mandibular occlusal plane by extrusion of the premolars.14 A sectional arch approach to orthodontic treatment is usually aimed at intrusion of the incisors to level a deep mandibular COS.12,15 The proponents of sectional arch techniques believe that leveling a deep COS with a continuous archwire produces unstable results.16-22 On the other hand, supporters of the Tweed treatment philosophy11 believe that incisor intrusion is unstable and prone to relapse.23,24 Irrespective of the orthodontic technique used, a primary objective in the diagnosis, treatment, retention, and stability of orthodontic therapy is the attainment of a level occlusal plane.5,19,25-27 Cephalometric studies undertaken to compare continuous with sectional arch-leveling techniques found that both were successful in correcting overbites.28,29 These studies concentrated on overbite correction only; they did not analyze study models to evaluate how effectively the curves of Spee were leveled or examine the long-term stability of the results. In general, studies that measured the amount and type of dental relapse after orthodontic treatment found postretention increases in overjet, overbite, and mandibular incisor crowding along with decreases in arch

550

American Journal of Orthodontics and Dentofacial Orthopedics Volume 133, Number 4

Preston et al 551

Table I.

Characteristics of the samples

Time

Characteristic Alexander (n 31) Mean age at T1 Mean age at T2 Mean age at T3 Mean treatment time (T2-T1) Mean xed retention time Mean postretention time (T3-T2) Bench (n 13) Mean age at T1 Mean age at T2 Mean age at T3 Mean treatment time (T2-T1) Mean xed retention time Mean postretention time (T3-T2)

12 y 6 mo 14 y 11 mo 26 y 4 mo 2 y 5 mo 3 y 4 mo 11 y 5 mo 13 y 6 mo 16 y 2 mo 22 y 5 mo 2 y 6 mo 1 y 5 mo 4 y 1 mo

length and width.27-39 Although attempts to correlate posttreatment changes with pretreatment occlusal variables have been unsuccessful,6,26,34 there is a need to correlate the depth of the pretreatment COS with postretention changes in other aspects of the occlusion. Despite the perceived importance of the COS, there is little published research that compares the amounts of arch leveling with various orthodontic treatment techniques. The purpose of this investigation was to compare the effectiveness of the Alexander continuous archwire technique and the bioprogressive sectionalarch technique in leveling the COS in Class II Division 1 deepbite patients. A second purpose of the study was to determine and compare the long-term stability of the leveling of the COS with both techniques. Our third objective was to determine whether in either group of patients a relationship existed between a deep COS before orthodontic treatment and the relapse that occurs in some occlusal traits.
MATERIAL AND METHODS

The sample (Table I) consisted of the records of patients selected from the general patient populations of the private practices of Drs R. G. Alexander, Arlington, Texas, and Ruel Bench, Lancaster, California. The Alexander patients were the rst randomly selected 31 patients who fullled the selection criteria, and the Bench patients were the only ones who fullled these criteria. All patients were treated nonextraction, and all had Class II craniofacial (ANB angle 4) skeletal patterns, at least a half step Class II molar relationship, an overbite of 50% or greater as measured on the initial (TI) study models, a mandibular plane angle (Go-Gn to S-N) less than 32, and a COS 2 mm. Only patients with complete orthodontic records were selected for this study. The records used consisted of dental casts

taken at T1, after orthodontic therapy (T2), and at postretention (T3). All subjects in both the groups were treated with fully preadjusted xed orthodontic appliances with .018-in slots. Both techniques prescribe 5 of labial root torque in the mandibular incisor region. The 132 sets of study models used in this study were assigned random numbers; this made it possible for 1 investigator (M.B.M.) to examine the records blindly. The curves of Spee were measured on the left and right sides of each set of mandibular models according to the technique described by Koyama.5 One operator used a standard palatometer (GPM Anthropological Instruments, East Rutherford, NJ) to measure the curves of Spee. A paired t test indicated no signicant statistical differences (P .005) between the pairs of measurements (left and right curves of Spee) for each of the 132 models. The average of the right and left curves of Spee for each model was therefore used for statistical analysis. One operator used a standard dial caliper (Fowler Helios, Las Vegas, Nev) to measure, randomly and blindly, directly on the 132 study casts mandibular intercanine width,33 incisor overbite,33 overjet,33 mandibular incisor irregularity index,30 and mandibular arch length as the sum of the 4 arch segments.31 Thirty sets of models were randomly selected from the original samples and remeasured by the same operator 5 days later, without reference to the previous measurements. The casual error was calculated according to Dahlbergs formula40 (S2 d2/2n), and the systematic error was calculated by using the Pearson correlation coefcient for reliability. A signicance level of 0.01 was used for this part of the study. A paired t test was used to test the hypothesis that the COS remained unchanged from TI to T2 and from T2 to T3. To compare the incidence of relapse (T2 to T3) of the COS in patients who were completely level and those who were not at T2, a 2-sample z test was calculated. Paired t tests were used to analyze treatment effects, and relapse of the 5 occlusal traits was studied. A Pearson correlation coefcient and regression analysis were performed to determine the predictive power of the COS at T1 on the relapse of these 5 variables. A signicance level of 0.05 was used for these statistical analyses.
RESULTS

The casual errors did not exceed 0.5 mm, and no variable had statistically signicant (P .01) systematic errors. Indications of nonnormality of the COS data at T1, T2, and T3 (Dn 0.2336, Lilliefors P .001) and at T2 (Dn 0.4182, Lilliefors P .001) reect the high number of repeats in the measurements (ie, 12 values of 2 at T1 and 22 values of 0 at T2). In the

552 Preston et al

American Journal of Orthodontics and Dentofacial Orthopedics April 2008

Table II.

Treatment, posttreatment, and total changes in

Treatment changes (T2-T1) Posttreatment changes (T3-T2) Mean SD

the COS
Total change (T3-T1) Mean 1.93 80.62 2.11 90.56 SD

Table III. Pretreatment (T1), posttreatment (T2), and postretention (T3) model measurements T1 Measurement (mm) Alexander Mandibular intercanine width Overbite Overjet Mandibular incisor irregularity Arch length Bench Mandibular intercanine width Overbite Overjet Mandibular incisor irregularity Arch length Mean SD T2 Mean SD T3 Mean SD

Measurement Alexander COS (mm) COS (%) Bench COS (mm) COS (%)

Mean 2.30 95.43 2.21 94.85

SD

0.47 7.92 0.46 20.00

0.37 14.81 0.1 4.29

0.40 16.06 0.16 17.16

0.56 19.41 0.58 24.78

24.75 4.76 6.27 3.97 62.22

2.1 0.95 2.97 3.35 4.64

26.11 2.09 2.18 0.31 64.01

1.4 0.65 0.56 0.46 3.17

25.5 2.84 3.27 1.28 61.65

2.36 0.85 0.93 1.35 3.41

Alexander sample (n 30), the results of the t tests can be considered acceptably accurate. In the Bench sample (n 30), the results of the t tests are less reliable. The Pearson correlation coefcient used to determine systematic errors for the dental characteristics studied showed intraoperator reliability, with values ranging from r 0.92 to 0.96 and r 0.90 to 0.97 for the Alexander and the Bench groups, respectively. The mean COS at T1 for the Alexander patients was 2.41 mm, and the mean COS at T2 was 0.11 mm. The mean absolute and percentage changes for the curves of Spee of these subjects are shown in Table I. The difference between the TI and T2 curves of Spee (Table II) was highly signicant (P .0001). Twenty-two of the Alexander patients were completely level at T2. The mean COS at T3 for the Alexander sample was 0.48 mm. The difference between the T2 and T3 curves of Spee (Table II) was statistically signicant (P .0001). Of the 22 patients who were level at T2, 11 showed some relapse of the COS, whereas 8 of the 9 remaining patients had some relapse. Statistically signicant differences were found in the occurrence (P .05) and the magnitude (P .0001) of relapse in the COS of these subgroups. Eleven of 22 subjects who were completely level at T2 relapsed an average of 0.28 mm (11.68%), and 8 of the 9 who were not completely level at T2 relapsed an average of 0.39 mm. The mean T1 and T2 COS values for the 13 subjects treated with the bioprogressive sectional-arch technique were 2.33 and 0.12 mm, respectively. The difference (Table II) between the T1 and T2 curves of Spee was statistically signicant (P .0001). The mean T3 COS for this sample was 0.22 mm, and the difference between the T2 and T3 curves of Spee was also statistically signicant (P .0001). At T2, 9 of the Bench patients were completely level, and 4 had residual curves of Spee. Of the 9 who were level at T2, 1 showed some relapse, and 3 of the remaining patients ended up with deeper curves of Spee at T3.

25.77 5.44 4.26 4.13 61.47

2.27 1.11 1.29 2.3 2.98

26.67 2.38 2.27 0.63 63.83

1.51 0.91 0.9 0.7 2.43

26.17 2.9 2.51 1.36 62.56

1.57 0.77 0.85 1.04 3.07

The descriptive data for the 5 occlusal characteristics measured on the study casts at TI, T2, and T3 are shown (Table III). For each occlusal trait, statistically signicant differences (P .05) were found between their TI and T2 values, and between their T2 and T3 values. Twenty-four of the Alexander patients increased their mandibular intercanine widths during treatment (mean, 1.36 mm; P .0002), but they showed some reduction in this distance from T2 to T3 (mean, 0.61 mm; P .05). Likewise, 9 Bench patients increased their mandibular intercanine widths during treatment (mean, 0.90 mm; P .0002), and they showed signicant increases (mean, 0.66 mm; P .05) in this dimension from T2 to T3. In all Alexander subjects, overbite was reduced during treatment and increased signicantly in 21 of them at T3. In all Bench patients, overbite was reduced during treatment and increased signicantly in 4 at T3. Although overjet was reduced in all Alexander patients during treatment (P .0001), it also increased signicantly in 27 of them from T2 to T3 (P .0001). Overjet was reduced in all Bench patients during treatment (P .0001) but relapsed signicantly (P .0001) in 6 after the retention phase. Before treatment, minimal crowding (3.5 mm) was observed in 17 Alexander subjects, 11 had moderate (3.5-6.5 mm) incisor irregularity, and 3 had severe incisor (6.5 mm) crowding. In these patients, treatment produced a signicant decrease (P .0001) in their mandibular incisor irregularity, but it increased signicantly posttreatment (P .0001), with 28 subjects having minimal and 3 having moderate incisor irregularity at T3. In the Bench patients, 5 had minimal, 7 had moderate, and 1 had severe incisor irregularity at

American Journal of Orthodontics and Dentofacial Orthopedics Volume 133, Number 4

Preston et al 553

T1. Treatment produced a signicant decrease in their incisor irregularity (P .0001), but it increased significantly posttreatment (P .0001). All patients in this group had only minimal incisor irregularity at T3. Of the Alexander patients, 20 had an increase in arch length due to treatment (P .04), and 27 had a signicant decrease in arch length postretention (P .0001). Of the Bench patients, 8 had an increase in arch length from T1 to T2 (P .04), and 9 had a decrease in arch length from T2 to T3 (P .0001). The Pearson correlation coefcient calculated to compare the T1 COS with the posttreatment relapse (T2-T3) showed no signicant statistical correlations (P .05). Follow-up regression analysis based on the TI values showed no ability to predict relapse in any of the 5 dental traits measured (P .05).
DISCUSSION

Correcting a deep overbite and achieving long-term occlusal stability are 2 difcult problems for orthodontists.24 In the orthodontic literature, both deepbites and long-term occlusal stability have been linked to exaggerated curves of Spee.3-6,13,19,25,41 Two basic methods used to level a deep COS are provided by the continuous and the sectional-arch techniques.11,12 During this investigation, it became clear that there is a shortage of clinical records that reect the longterm effects of orthodontic treatment in general and, specically, on the COS. The patients studied here were treated by 2 experienced clinicians who set similar, clearly dened goals for their orthodontic treatments.25 The data from this study indicate that, with some minor exceptions, the results (T2) produced by these techniques were comparable. We concluded that the Alexander continuous archwire technique effectively levels a COS in Class II Division 1 nonextraction deepbite patients. At the end of treatment, 71% of these subjects were leveled completely, and 29% had a slight residual COS. In all of these patients, orthodontic treatment resulted in Class I molar and canine relationships, properly nished buccal occlusions, and normal overjets and overbites.25 The bioprogressive sectional-arch technique also successfully leveled the curves of Spee in Class II Division 1 deepbite patients: 69% of those studied were leveled completely, and the rest had slight residual curves at the end of treatment. These patients all had Class I molar and canine relationships, properly nished buccal occlusions, and normal overjets and overbites.25 Over a mean period of 7 years 5 months postretention, the Alexander patients had a slight relapse of their curves of Spee. It was thus concluded that it is possible with the Alexander technique to treat Class II Division

1 deepbite malocclusions so that, after treatment, the COS will either remain level or relapse only slightly. Twelve of the 31 Alexander patients were completely level from 5 to 25 years after orthodontic treatment. The 2-sample z test of the Alexander data showed a signicant difference in the incidence of relapse in the patients who were completely leveled and those who were not at the end of treatment. In addition, the results of the 2 independent samples t test showed a signicant difference in the magnitude of relapse in these 2 subgroups. It was concluded that the Alexander patients whose curves of Spee were not 100% level posttreatment had greater incidence and magnitude of COS relapse than did those who were completely leveled. Over a mean period of 4 years 1 month postretention, 5 patients treated with the Bench technique experienced a slight but statistically signicant relapse in their curves of Spee, whereas 8 of the 13 remained completely level. This study shows that it is possible with the bioprogressive sectional-arch technique to treat the exaggerated curves of Spee commonly encountered in Class II Division 1 deepbite malocclusions efciently and to nal results that are relatively stable. Within the limits of the relatively small size of the sample of Bioprogressive patients, the results of the 2-sample z test showed a signicant (P .05) difference in the incidence of relapse in the 9 subjects who were completely level at T2 and the 4 who were not. In these patients, the 2 independent samples t test did not show statistically signicant differences in the magnitude of relapse in those who were completely level at T2 and those who were not. A larger sample would make the ndings of this part of the study more reliable. Although the primary focus of this study was on the effects of orthodontic treatment on the COS, the changes in 5 other discrete mandibular occlusal traits were also recorded. The ndings recorded in this part of the study were similar for the subjects treated with the 2 techniques; therefore, they will be discussed together. Both techniques produced statistically signicant changes in each of the 5 variables measured on the study casts. As a result of the orthodontic treatment, there were mean decreases in overbite, overjet, and incisor irregularity, and mean increases in mandibular intercanine width and arch length. In our study, the changes in overbite were less than those previously reported,5 but the posttreatment changes in the overjet were similar to those previously observed in treated Class II Division 1 malocclusions.28,29,34 Although a statistically signicant posttreatment increase in incisor irregularity was observed, at T3, all patients had only minimal incisor irregularity (3.5 mm).39 Associated with treatment, all patients showed a net improvement in mandibular incisor crowding. These results echo those of

554 Preston et al

American Journal of Orthodontics and Dentofacial Orthopedics April 2008

Elms et al,33 who were, however, unable to demonstrate a statistically signicant mean increase in the arch length from orthodontic treatment. At posttreatment in the combined sample of patients, 4 of the 5 variables measured had statistically signicant relapse, whereas there was a nonsignicant (P .5050) relapse of mandibular intercanine width. Most posttreatment occlusal changes were small and within the range of the physiologic adaptations associated with aging.31,37,38 Since it was demonstrated that some relapse had occurred in the 5 occlusal traits studied, a Pearson correlation coefcient was calculated to compare the pretreatment COS with the posttreatment changes observed for each trait. The results of this test did not show a statistical correlation between the original COS and any of the 5 traits. Follow-up regression analyses did not disclose an ability to predict relapse in the occlusal characteristics studied based on depth of the pretreatment COS. This part of our study supports the ndings of De Praeter et al.6 Only well-dened Class II Division 1 malocclusions were included in this study. Thus, our ndings cannot be arbitrarily extrapolated to the results achievable with more extreme Class II Division 1 malocclusions or other categories of malocclusions. Since only study models were evaluated, we could not ascertain the exact mechanism by which the COS is leveled with either technique. For the same reason, it was impossible to establish the exact mechanism by which the relapse of the COS occurred.
CONCLUSIONS

1. The Alexander and the bioprogressive techniques can both effectively level a moderate COS in Class II Division 1 deepbite patients treated nonextraction. 2. Both techniques can be used to treat a moderate COS so that, if relapse occurs, it returns to a lesser extent than before the orthodontic treatment. 3. With both techniques, a pretreatment COS that is not completely level posttreatment has a higher incidence of relapse than one that is completely level posttreatment. 4. In this investigation, there was no ability to predict relapse in the 5 occlusal traits studied based on the depth of the pretreatment COS. 5. This study indicates that, in well-treated patients, the observed relapse in the COS is minimal, and it occurs over an extended period of time.
REFERENCES 1. Spee FG. The gliding path of the mandible along the skull. J Am Dent Assoc 1980;100:670-5. 2. Okeson JP. Management of temporomandibular disorders and occlusion. 5th ed. St Louis: Mosby; 2003. p. 135.

3. Germane N, Staggers JA, Rubenstein L, Revere JT. Arch length consideration due to the curve of Spee: a mathematical model. Am J Orthod Dentofacial Orthop 1992;102:251-5. 4. Braun S, Hnat WP, Johnson BE. The curve of Spee revisited. Am J Orthod Dentofacial Orthop 1996;110:206-10. 5. Koyama T. A comparative analysis of the curve of Spee (lateral aspect) before and after orthodontic treatmentwith particular reference to overbite patients. J Nihon Univ Sch Dent 1979;21: 25-34. 6. De Praeter J, Dermaut L, Martens G, Kuijpers-Jagtman A. Long-term stability of the leveling of the curve of Spee. Am J Orthod Dentofacial Orthop 2002;121:266-72. 7. Hitchcock HP. The curve of Spee in Stone Age man. Am J Orthod 1983;84:248-53. 8. Wheeler RC. A textbook of dental anatomy and physiology. 2nd ed. Philadelphia: W. B. Saunders; 1950. p. 329-409. 9. Osborn JW. Orientation of the masseter muscle and the curve of Spee in relation to crushing forces on the molar teeth of primates. Am J Phys Anthropol 1993;92:99-106. 10. Strang RH. A textbook of orthodontia. 3rd ed. Philadelphia: Lea and Febiger; 1950. p. 268-307. 11. Tweed CH. Clinical orthodontics. Volume II. St Louis: C. V. Mosby; 1966. p. 461-547. 12. Ricketts RM. Bioprogressive therapy as an answer to orthodontic needs. Part 1. Am J Orthod 1969;70:241-68. 13. Braun ML, Schmidt WG. A cephalometric appraisal of the curve of Spee in Class I and Class II Division occlusion for males and females. Am J Orthod 1956;42:255-78. 14. Tweed CH. Clinical orthodontics. Volume I. St Louis: C. V. Mosby; 1966. p. 135-47. 15. Ricketts RM. Facial and denture changes during orthodontic treatment as analyzed from the temporomandibular joint. Am J Orthod 1955;41:163-7. 16. Wylie WL. Overbite and vertical facial dimensions in terms of muscle balance. Angle Orthod 1944;14:13-7. 17. Bench RW, Gugino CF, Hilgers JJ. Bioprogressive therapy. Part 2. J Clin Orthod 1977;11:661-82. 18. Schudy FF. The association of anatomical entities as applied to clinical orthodontics. Angle Orthod 1966;36:190-203. 19. Baldridge DW. Leveling the curve of Spee: its effect on the mandibular arch length. J Pract Orthod 1969;3:26-41. 20. Otto RL, Anholm JM, Engel GA. A comparative analysis of intrusion of incisor teeth achieved in adults and children according to facial types. Am J Orthod 1980;77:437-46. 21. Berman MS. Straight wire mythsBJO interview. Br J Orthod 1988;151:57-61. 22. Woods NI. A reassessment of space requirements for lower arch leveling. J Clin Orthod 1986;20:770-8. 23. Schudy FF. Cant of the occlusal plane and axial inclination of teeth. Angle Orthod 1963;23:69-82. 24. Schudy FF. Vertical growth versus antero-posterior growth as related to function. Angle Orthod 1964;34:756-93. 25. Andrews LF. The six keys to normal occlusion. Am J Orthod 1972;62:296-309. 26. Fidler BC, rtun J, Joondeph DR, Little RM. Long-term stability of Angle Class II, Division I malocclusions with successful occlusal results at the end of active treatment. Am J Orthod Dentofacial Orthop 1995;107:276-85. 27. Williams R. Eliminating lower retention. J Clin Orthod 1985;22: 342-9. 28. Dake ML, Sinclair PM. A comparison of the Ricketts and Tweed-type arch leveling techniques. Am J Orthod Dentofacial Orthop 1989;95:72-8.

American Journal of Orthodontics and Dentofacial Orthopedics Volume 133, Number 4

Preston et al 555

29. Weiland FJ, Bartleon LIP, Drosehl H. Evaluation of continuous arch and segmented arch leveling techniques in adult patients: a clinical study. Am J Orthod Dentofacial Orthop 1996;110: 647-52. 30. Little RM. The irregularity index: a quantitative score of mandibular anterior teeth. Am J Orthod 1975;68:554-63. 31. Bishara SE, Jakobsen JR, Treder JE, Stasi M. Changes in the maxillary and mandibular tooth-size-archlength relationship from early adolescence to early adulthood. Am J Orthod Dentofacial Orthop 1989;95:46-59. 32. Burstone CJ. The mechanics of the segmental arch technique. Angle Orthod 1966;36:99-120. 33. Elms TN, Buschang PH, Alexander RG. Long-term stability of Class II Division 1 nonextraction cervical face-bow therapy: a model analysis. Am J Orthod Dentofacial Orthop 1996;100:271-6. 34. Glenn G, Sinclair PM, Alexander RG. Nonextraction orthodontic therapy: posttreatment dental and skeletal stability. Am J Orthod Dentofacial Orthop 1987;92:321-8.

35. Little RM, Reidel RA, rtun J. An evaluation of changes in mandibular anterior alignment from 10 to 20 years postretention. Am J Orthod Dentofacial Orthop 1988;93:423-8. 36. Puneky PJ, Sadowsky C, BeGole EA. Tooth morphology and lower incisor alignment many years after orthodontic therapy. Am J Orthod 1984;86:299-305. 37. Sinclair PM, Little RM. Maturation of untreated normal occlusion. Am J Orthod 1983;83:114-23. 38. DeKock WH. Dental arch depth and width studied longitudinally from 12 years of age to adulthood. Am J Orthod 1972;62:56-66. 39. Little RM, Wallen TR, Reidel RA. Stability and relapse of mandibular anterior alignmentrst premolar cases treated by traditional edgewise orthodontics. Am J Orthod 1981;80:349-64. 40. Dahlberg G. Statistical methods for medical and biological students. London: George Allen and Unwin Ltd; 1940. p. 122-32. 41. Hellsing E. Increased overbite and craniomandibular disordersa clinical approach. Am J Orthod Dentofacial Orthop 1990;98:516-22.