Journal of Oral Rehabilitation 2006 33; 423–429

Deviation from the Broadrick occlusal curve following posterior tooth loss

*Lecturer in Restorative Dentistry, Leeds

Dental Institute, Leeds, UK, Professor in Restorative Dentistry, Liverpool Dental School, Liverpool, UK and ‡Senior Lecturer, Leeds Dental Institute, Leeds, UK

SUMMARY Production of an appropriate occlusal curve for dentitions which have become deranged because of tooth loss, overeruption, tipping and drifting can present challenges for the dental technician. An earlier paper (J Oral Rehabil, 2005 32: 895–900.) demonstrates that the use of the Broadrick flag method for producing the occlusal curve is relatively accurate for most intact arches. This study demonstrates that when a posterior tooth has remained unopposed for 5 years or more positional changes, which cause deviation from the Broadrick curve, occur. The extent of the deviation may be

extreme, potentially leading to difficulties in restoring a harmonious occlusal scheme. This study also demonstrates that the Broadrick curve may provide an accurate reproduction of the occlusal curve, even when the tooth forming the posterior determinant of the curve is tipped. A moderate degree of tipping of this tooth has little effect on the radius of the Broadrick curve. KEYWORDS: Broadrick occlusal curve, posterior tooth loss Accepted for publication 24 September 2005

Patterns of dental disease have changed dramatically over the last quarter of a century, with most patients keeping at least a partial dentition for their entire lives. The increasing incidence of tooth wear, together with changes following tooth loss, may mean that the patient’s natural occlusal curve is not evident. Indeed following a deranged occlusal curve, while carrying out occlusal reconstruction, may result in occlusal interferences that could lead to destruction of restorations and tooth tissue. Derangement of the occlusal curve because of tooth positional changes after loss of a posterior tooth have not been quantified in the literature and this investigation aims to determine the extent of deviation from an ‘ideal’ curve for a sample of partially dentate patients.

Review of the Literature Over the last one and a half centuries, authors have attempted to explain the geometry of the occlusal curve
ª 2006 Blackwell Publishing Ltd

by reference to various anatomical landmarks. Ferdinand Graf Spee (1), originally described the form of the occlusal curve, and postulated its anatomical determinants in 1890. He postulated that the occlusal curve forms part of a circle or ‘cylinder’, the axis of which is formed by the crista lachrymalis posterior and the horizontal midorbital plane (Fig. 1). Since then, the relevance and development of the curve of Spee has been studied by a number of authors. Page (2) presented evidence that the occlusal curve originates from a combination of hinge axes controls and the mandibular angle, and that it is distinct from the curve of Spee with its theoretical distal extension to the head of the condyle. The occlusal curve demonstrated by Page ends at the most posterior cusp of the most posterior tooth, whereas the curve of Spee is postulated to extend to the anterior surface of the condylar head. A further complication in this discussion is the occlusal plane, whose definition is ‘the plane established by the occluding surfaces of premolars and molars (3)’. This is a trapezoidal surface extending from the distal cusps of the most distal lower teeth to the tips of the canines,
doi: 10.1111/j.1365-2842.2005.01587.x


H . L . C R A D D O C K et al.

Fig. 2. Progressive differential angle, mean angle between ideal tangential direction and long axis of different posterior mandibular teeth.

Fig. 1. Determination of the curve of Spee.

and is obviously curved in many cases. This is in conflict with the true definition of a plane, which must be flat, and brings into question the use of terminology currently in use. Ferrario et al. (4) compared what they described as the ‘curve of Spee’ (which was in fact the occlusal curve) on male and female groups of human subjects. They determined that there was a difference in the curves between the sexes and that there was a difference in the curves on each side of the arch. The male curves were larger than female and the left side was larger than the right in both groups. These researchers used computer graphics plotted onto images of patient study models to determine and measure the curves. Therefore as there was no distal extension of the curve to the head of the condyle, they were in actual fact plotting the occlusal curve. Orthlieb (5) used cephalometric images to study the relationship of the curve of Spee to the axial angulations of the posterior teeth. He concluded that the mandibular incisors followed the tangent law. He noted, however, that the posterior mandibular teeth exhibited a progressive differential angle with the direction of the tangent. This investigation was able to determine the true curve of Spee, extending to the condylar head, using radiographic tracings. A large sample group of (470) subjects was studied. As well as mathematically defining the differential angulation of the posterior teeth (Fig. 2), the study also found significant differences in the curve, relative to the degree of overbite, and the skeletal classification. This study shows that the development of the curve of Spee is likely to be multi-factorial, and that the curve

is likely to vary significantly between individuals. Closure of proximal contacts and mesial drift was felt by Orthlieb to be attributable to the forces generated by this progressive angulation of the curve of Spee. Both of the above studies demonstrate that the original ‘Tangent Law’ for the determination of the angulation of the mandibular posterior teeth to be invalid, and that significant variation can occur both between individuals, and between each side in any given subject. It follows that if we are to measure deviations of an individual tooth from the occlusal curve, we must be able to accurately determine the curve for that particular dental quadrant in our specified subject. Looking at the occlusal curve and its role in occlusal rehabilitation and reconstruction, Weinberg (6) advocated maintaining the patients natural occlusal plane, arguing that optimum transmission of occlusal forces are most likely if this is followed. He argued that if an inappropriate curve were created, the crown root ratio would be altered, together with the torquing forces transmitted. Following extensive tooth loss, the patients’ occlusal plane may become distorted because of tipping, drifting and overeruption and prosthodontic restoration may be complicated by these changes. Lynch and McConnell (3) considered this problem, and suggested the use of a ‘Broadrick flag’, in order to reproduce the ideal curve of Spee for an individual, utilising Graf Spee’s original principles. They utilized dental landmarks to create the curve, and introduced modifications for different skeletal patterns. Although this method did not allow for the incremental nature of the occlusal curve described by Orthleib, it may

ª 2006 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 33; 423–429

provide useful guidance in clinical restorative procedures.


Sources of subjects One hundred consecutive patients with one or more unopposed posterior teeth were invited to take part in the study. One hundred age and sex matched subjects, with matching interdental bone height at the sites examined were recruited as controls. The age for each subject was matched Æ2 years with a control of the same sex. Bone height was determined by measuring from the amelocemental junction on bitewing or periapical radiographs taken using paralleling techniques. Matching was to within 1 mm of that of the subjects under investigation. The sample consisted of 50 female and 50 male patients.

Aim of the study To compare the deviation of the clinical occlusal curve with a theoretical ideal, for patients with and without the presence of an unopposed posterior tooth.

Samples tested Following consultation with a statistician, a sample size calculation showed an adequate sample size for this study would be 90 subjects and 90 controls. This was calculated to detect a difference of 0Æ5 mm, at 90% power. Two hundred adult patients were examined and had study models made of their upper and lower dentitions. These models were later scanned and recorded as digital images.

Measurement of deviation from the Broadrick occlusal curve Alginate impressions were taken of the upper and lower dentition. After casting in dental stone, the study models were trimmed in a seven-sided configuration, so that the buccal plane would lie parallel to the scanner * surface when the models were placed on it. The models were then scanned on a Black Widow 3696 USB scanner at a resolution of 350 DPI. For the subject group only the side with tooth loss was scanned. The same side was scanned for the control. Using the Broadrick flag method described by Lynch and McConnell (3), the ideal occlusal plane was created using the scanned models. The anterior determinant was the mid point on the distal slope of the lower canine tooth, and the posterior determinant was the mid point on the distal slope of the disto-buccal cusp of the lower second molar. These points were used as the centre of circles from which arcs were drawn (Fig. 3). In class 1 incisal relationships, an arc of 4-inch radius was drawn from the centre of the distal slope of the incisal edge of the lower canine. The same size arc was also drawn from the tip of the distal cusp of the lower second molar tooth. The intercept of these arcs was used to determine the centre of a circle, which was drawn to lie against the points on the lower canine and second molar already described. For class 2 incisal relationships the initial arc diameter was 3Æ75 inches, and class 3 relationships a 5-inch arc was used. The degree of overjet measured on the study models was used to determine the incisal classification. The distance of the furthest cusp tip from the Broadrick curve was measured along the long axis of the tooth for each individual. Where the deviation was
*Devcom Ltd, Sterling, UK.

Selection criteria Local Ethical Committee approval was gained. Patients from Leeds Dental Institute, with one to three adjacent unopposed posterior teeth were invited to take part in the study. For the purposes of this study ‘posterior’ is defined as an upper or lower first or second premolar, or first, second or third molar. The control group had intact dentitions, without unopposed teeth. Adult patients of 18 years or above were recruited. As documentary evidence of when the missing teeth were lost was unavailable, subjects in the test group had documented evidence that the missing teeth had been absent for 5 years or more. There was no specified upper age limit. Written informed consent was obtained from all subjects following a written and verbal explanation of the purpose of the study and methods to be used in the investigations.

Exclusion criteria Unless documentary evidence was available showing that the missing tooth/teeth had been lost for a minimum of 5 years, patients were excluded from the study. Teeth that were prevented from independent movement, by involvement in support or retention for fixed or movable prostheses were also excluded.

ª 2006 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 33; 423–429


H . L . C R A D D O C K et al.

Fig. 3. Creation of Broadrick curve where tipping has occurred. Fig. 5. Effect of tipping on radius of Broadrick curve.

Fig. 4. Direction of deflection from the Broadrick curve.

outside the existing curve a positive notation was given, if the deviation was inside the curve a negative notation was given. Where no deviation was detected a value of zero was awarded. This is demonstrated in Fig. 4. This procedure was carried out for both unopposed subjects and control patients.

Statistical analysis Paired t-tests were used to compare the findings between subjects and controls. Significance was assessed at the 0Æ05 level. Bland Altman plots were used to determine examiner reliability.

Examination of the effect of tipping of the posterior determinant tooth on the radius of the Broadrick Curve Following the determination of the extent of tipping of molar teeth distal to the site of tooth loss, it was decided

that it would be appropriate to investigate the effect that the mesial movement of the tipped tooth had on the radius of the Broadrick curve. Using a model of an intact lower arch, with a class 1 occlusion, the outline of the lower buccal teeth were traced digitally (Fig. 5). The first permanent molar was erased from the drawing and a Broadrick curve was created using the mid-point on the distal slope of the canine as the anterior determinant and the mid point on the distal slope of the disto-buccal cusp of the second molar as the posterior determinant (Tooth shown in red). Using these canine and molar reference points, a compass with a circle radius (red line) determined by the incisal relationship was placed on each point, and intersecting arcs drawn (Red arcs). Using the intersection of these arcs as the circle centre, an arc was drawn from the canine tip, along the occlusal plane to determine the Broadrick occlusal curve (again shown in red). The second molar tooth (green tooth) was then tipped, about a fulcrum 1/3 of the root length from the apex, mesially on its long axis by 18°. The mean tip of the tooth was determined by measuring the tip of 76 molar teeth distal to extraction sites). The fulcrum of tip

ª 2006 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 33; 423–429

was placed 1/3 of the root length from the apex. The Broadrick curve was then re-drawn (in red) as described above, and the maximum difference between the two radii measured. This is illustrated in Fig. 5.
Table 2. Deviation from the Broadrick curve when upper teeth are unopposed Group Subjects Controls Mean (mm) 2Æ20 0Æ12 Standard deviation 1Æ69 0Æ87 Range (mm) 0–8Æ26 )1Æ65–2Æ5


Sample The age range of the matched groups was 20–79 years. The mean age was 50Æ26 years, with a standard deviation of 13Æ05.

Table 3. Deviation from the Broadrick curve when lower teeth are unopposed Group Subjects Controls Mean (mm) 0Æ47 )0Æ06 Standard deviation 1Æ79 0Æ70 Range (mm) )1Æ99–4Æ55 )1Æ69–1Æ30

Deviation from the Broadrick curve Deviation from the Broadrick curve was found to be marked in subjects who had unopposed posterior teeth, while fairly minimal in the control group (Table 1). Mean of difference between subject and control groups ¼ 1Á75 (standard deviation ¼ 1Á93Þ 95% confidence interval of the difference is À 2Á22 to À 1Á28 The results show that there is a statistically significant difference in the deviation from the Broadrick curve between patients who have lost posterior teeth and the control group.
Table 1. Deviation from Broadrick occlusal curve Group Subjects Controls Mean (mm) 1Æ98 0Æ23 Standard deviation 1Æ87 0Æ76 Range (mm) )1Æ99–8Æ26 )1Æ69–2Æ5

Figure 6 demonstrates that for the group with missing posterior teeth 77% had a tooth or teeth deviated more than 1 mm from the Broadrick curve, while only 14% of the control group had a deflection of this magnitude. 26% of the subject group had deviated in excess of 2 mm, with none of the controls with this extent of deflection. When comparing the deviation, in the upper and lower arches, it was found that arches with unopposed upper teeth are associated with a significant deflection from the Broadrick occlusal curve (Tables 2 and 3). Mean of difference between upper arch subject and control groups 2Á08 (standard deviation ¼ 1Á96Þ 95% confidence interval of the difference is 1Á55--2Á60 Mean of difference between lower arch subject and control groups 0Á53 (standard deviation ¼ 1Á69Þ 95% confidence interval of the difference is À 0Á27 to 1Á32 Unopposed lower teeth are not associated with a significant deflection from the Broadrick occlusal curve.

Deviation from the Broadrick curve 60 50 Number of subjects 40 30 20
27 50


subjects controls

Change in the Broadrick curve following tipping The difference in radii of the occlusal curves drawn before and after the tooth distal to the extraction site had hypothetically tipped was 0Æ57 mm.

10 0




10 4







'-0.2 to '-1.01







Examiner reliability When a single examiner is used, it is important to determine intra-examiner reliability. 10% of the subject and control models were re-examined in order

Extent of deviation in mm

Fig. 6. Extent of deviation from the Broadrick occlusal curve.

ª 2006 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 33; 423–429


H . L . C R A D D O C K et al.
Bland altman plot of deviation from ideal curve
0.18 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 –0.02 –0.04 –0.06 –0.08 –1.5








Mean of measurement

Fig. 7. Bland Altman plot of examiner agreement.

to draw up a Bland Altman plot. The difference between the initial and repeated measurements were plotted against the mean of the two. Reliability is demonstrated by at least 95% of the plots lying within two standard deviations of the mean of the measurement differences. Figure 7 shows the Bland Altman plot obtained, demonstrating good inter-examiner reliability.

The subject and control groups were age and sex matched and were found to be of a comparable age range to patients having lost posterior teeth in previous studies on adult tooth loss (7). Obviously, a population of this age group are likely to have experienced some restoration in the arches under investigation, and it is recognized that this could have had an effect on the results obtained. It would have been virtually impossible to find totally unrestored arches, with one or more missing teeth, to use in this study. This method of model measurement has previously been described by Craddock and Youngson (8), and has been found to be reliable. The use of widely available equipment and ‘user friendly’ software for model measurement mean this method could have a wide range of applications in dental research. The Broadrick Flag technique (3) was designed as an instrument to provide a guide to the location of the centre of the Curve of Spee, from which a curve could be created to facilitate the restoration of a posterior quadrant. The Broadrick curve is used in this investigation as a hypothetical ideal curve for each individual patient, from which the deviation of individual teeth can be measured. Craddock et al. (9) found that the

occlusal curve produced by the Broadrick flag method closely approximated the natural occlusal curve. When assessing the deviation from the Broadrick curve, it is important to appreciate exactly what is being measured. The investigation centred around determining whether patients with missing posterior teeth were more likely to have one or more teeth, which deviated significantly from the ideal Broadrick curve, relative to the control group with intact arches. The only measurement recorded for each individual was the maximum deviation from the Broadrick Curve for an individual tooth, and that therefore it is likely that the majority of teeth along the curve are likely to be a closer fit. Deviation for an individual tooth may be caused by several factors, including tooth fracture, tooth wear, poorly contoured restorations, tooth tipping and drifting and failure of complete eruption, and may not be representative of the fit of the curve for the remainder of the quadrant. Deviation from the curve could be because of overeruption (8, 10), or tipping of teeth adjacent to the site of tooth loss (11, 12). Undesirable effects of tooth movement or malposition may include non-axial tooth loading, complicated paths of insertion for restorations, the presence of functional occlusal interferences and non-ideal tooth preparation for restorations. All of these factors may have adverse effects on the occlusion, the periodontium and the vital pulp. It was beyond the remit of this study to assess the relationship between the deviation in the Broadrick curve and the presence of occlusal interferences. The occlusal factors affecting the presence of an interference are complex and multi-factorial, and lend themselves to a further ongoing investigation by the same authors using statistical modelling to explain this complex area of study. The aims of restorative treatment for those patients with unopposed posterior teeth include the restoration of occlusal and dental form, the restoration of function and the maintenance and restoration of aesthetics. The existence of a scientifically evaluated tool, which restores the occlusal form for an individual, allows the restorative team to make evidence-based decisions when designing and restoring occlusal schemes. The extent by which individual teeth deviate from the Broadrick curve following posterior tooth loss is statistically significant. What is more important is whether these findings are of clinical significance. Although the mean difference between the subject

Difference in measurements

ª 2006 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 33; 423–429

and control groups was only in the order of 1Æ8 mm, when the range of deviations was examined, some teeth in the tooth loss group had large deviations (maximum 8Æ26 mm), which would be of even greater clinical significance. The results demonstrated that 26% of subjects with posterior tooth loss had a deviation from the Broadrick plane of 2 mm or more, and in terms of restoring normal occlusal form and contact, this is not without difficulty. To truly restore normal function, in terms of tooth loading, orthodontic uprighting of many tilted posterior teeth may be indicated (13). The degree of deviation from the Broadrick plane is most significant both clinically and statistically when upper teeth remain unopposed, in other words following lower tooth loss. This is likely to be because of overeruption of the unopposed teeth, and tilting of the teeth adjacent to the extraction site. Although there is no significant difference in the mean deviations for subjects with missing upper posterior teeth, the results showing extreme deflections of up to 4Æ55 mm are likely to be of clinical significance when restoring posterior occlusal form and function for these patients. Another part of a more extensive study by the same authors into tipping of teeth adjacent to the extraction site determined that teeth distal to the extraction site are likely to tip mesially following loss of their mesial interdental contact. The mean tip was 18°. The forward component of tipping would inevitably move the posterior determinant mesially, thereby decreasing the radius of the Broadrick curve. The models were not assessed for open contacts distal to the tipped teeth as the majority of these teeth were second molars and were therefore the last standing tooth in the arch. Mesial drift may compensate for loss of contact because of both tipping and drifting, and therefore the presence of space may not be a reliable indicator of tooth movement in relation to the tooth distal to the extraction site. In order to determine how much effect this movement had on the curve produced, a simple geometric exercise was carried out, the results of which can be seen graphically in Fig. 5. Accurate measurement of the difference in the curve, at a point approximately half way between the teeth bounding the extraction site, reveals a change of <0Æ6 mm. This finding somewhat validates the usefulness of this method for clinical use when restoring some discrepancies in the occlusal curve.


1. Following posterior tooth loss, there are both statistically and clinically significant deviation of individual teeth from the Broadrick occlusal curve. 2. Loss of a lower posterior tooth is likely to result in a greater degree of deviation from the curve than loss of an upper tooth, however, the extremes found in both upper and lower arches may be of clinical significance. 3. Tipping of the posterior determinant tooth of the Broadrick curve has little effect on the size of the radius of the curve.

1. Biedenbach MA, Holz M, Hitchcock HP. The gliding path of the mandible along the skull. J Am Dent Assoc. 1980;100:670–675. 2. Page HL. The occlusal curve. Dental Digest. 1952; Jan: 19–21. 3. Lynch CD, McConnell RJ. Prosthodontic management of the curve of Spee: use of the Broadrick flag. J Prosthet Dent. 2002;87:593–597. 4. Ferrario VF, Sforza C, Miani A Jr, Colombo A. Tartaglia GMathematical definition of the curve of Spee in permanent healthy dentitions in man. Arch Oral Biol. 1992;37:691–694. 5. Orthlieb J-D. The curve of Spee: understanding the sagittal organisation of mandibular teeth. J Cranio. 1997;15: 333–340. 6. Weinberg LA. The occlusal plane and cuspal inclination in relation to incisal-condylar guidance for the protrusive excursions. J Prosthet Dent. 1958;July/Aug:607–618. 7. Meskin LH, Brown LJ. Prevalence and patterns of tooth loss in US employed adult and senior populations, 1985–8. J Dental Education 1988;52:686–691. 8. Craddock HL, Youngson CC. A study of the incidence of overeruption and occlusal interferences in unopposed posterior teeth. Br Dent J 2004;196:341–348. 9. Craddock HL, Lynch C, Franklin P, Youngson CC, Manogue M. A study of the proximity of the Broadrick ideal occlusal curve to the existing occlusal curve in dentate patients. J Oral Rehabil 2005;32:895–900. 10. Kiliaridis S, Lyka I, Friede H, Carlsson GE, Ahlqwist M. Vertical position, rotation, and tipping of molars without antagonists. Int J Prosthodont. 2000;13:480–486. 11. Kaplan P. Drifting, tipping supraeruption and segmental alveolar bone growth. J Prosthet Dent. 1985;54:280–283. 12. Gragg KL, Shugars DA, Bader JD, Elter JR, White BA. Movement of teeth adjacent to posterior bounded edentulous spaces. J Dent Res. 2001;80:2021–2024. 13. Norton LA, Proffitt WR. Molar uprighting as an adjunct to fixed prostheses. J Am Dent Assoc. 1968;76:312–315.
Correspondence: H. L. Craddock, Room 6129, Leeds Dental Institute, Clarendon Way, Leeds LS2 9L, UK. E-mail:

ª 2006 Blackwell Publishing Ltd, Journal of Oral Rehabilitation 33; 423–429

Sign up to vote on this title
UsefulNot useful