| Sign In to gain access to subscriptions and/or personal tools. |
Edentulation Alters Material Properties of Cortical Bone in the Human Mandible
1 Division of Oral and Maxillofacial Surgery, University of Texas Southwestern, Dallas, TX; and Correspondence: * corresponding author, pdechow{at}tambcd.edu
Ridge resorption following edentulation has been documented clinically, but the effects of tooth loss on the material properties of mandibular cortical bone have received little study. Material properties and their structural basis are essential for our understanding of bone quality in the edentulous mandible and are of interest as a tissue-level model for functional adaptation. This study’s aim was to determine material property variability in the edentulous mandible, and to compare it with data from a previous study of dentate mandibles. Forty-four cortical samples were removed from each of 10 adult fresh edentulous mandibles. Cortical thickness and density were measured. Material properties were calculated from ultrasonic velocities. Mandibular cortical bone in the edentulous mandibles differed from that of dentate mandibles in cortical thickness, elastic and shear moduli, anisotropy, and orientation of the axis of maximum stiffness. These results suggest that cortical microstructural changes accompany ridge resorption following edentulation.
Key Words: ultrasound • cortical bone • biomechanics • function
Despite advances in caries prevention and periodontal disease treatment, tooth loss and the edentulous state remain significant health care issues. Many studies report the consequences of tooth loss and associated residual ridge resorption (Jaul et al., 1980; Hirai et al., 1993; Klemetti, 1996). However, little is known about the impact of edentulation and residual ridge resorption on changes in mandibular cortical structure, material properties, and biomechanics. In general, material properties and their variation, which are important indicators of bone quality, have not received much systematic analysis in individual bone organs. A recent study (Schwartz-Dabney and Dechow, in press) demonstrates significant regional material property variation in human dentate mandibles, some of which is associated with function. Although geometry changes are reported following edentulation, nothing is known about changes in other material properties, such as elastic and shear moduli. The two previous studies that measured material properties in both dentate and edentulous mandibles had sample sizes insufficient for statistical evaluation (Carter, 1989; Arendts and Sigolotto, 1990). Some information is available for bone density, but these data are largely derived from radiographs and other scans that do not allow density to be calculated as a true volumetric measurement. Our aim was to examine regional material property differences of edentulous mandibular cortical bone and to compare these with known properties of dentate mandibles. We hypothesized that edentulous mandibles, like dentate mandibles, differ regionally. Regarding the potential differences between dentate and edentulous mandibles in material properties, our null hypothesis was that there were no differences. An alternate hypothesis was difficult to formulate, given the lack of information in the literature on differences in material properties resulting from bone adaptation. In experimental studies of post-cranial bone, it is widely recognized that loss of or reduction in function may lead to loss of bone mass (for review, see Martin et al., 1998). However, little is known about changes in elastic properties, especially if their three-dimensional characteristics are considered. Likewise, most studies of density are radiographic and fail to separate differences due to apparent bone density from those due to thickness. Nevertheless, the important issue of bone quality in the edentulous mandible, especially given the increasingly widespread use of implants for dental restoration, led us to test whether edentulation results not only in alveolar bone loss, but also in changes in the material properties of the remaining basal bone.
Specimen Selection and Preparation Ten fully edentulous human mandibles were selected for dissection from unembalmed (fresh or frozen) cadavers donated to The University of Texas Southwestern Medical School willed body program. This study was exempt from IRB review because used tissues were from deceased individuals who had donated these tissues prior to death in accordance with the laws of the State of Texas. All mandibles had intact alveolar ridges with no evidence of open or healing sockets. Residual ridge height was significantly reduced in all mandibles, averaging between 14 and 15 mm along the corpus. Age at death ranged from 58 to 88 yrs and was not significantly different between three males (69.7 yrs, SD = 10.4) and seven females (72.1 yrs, SD = 12.6). Mandibles (1) were from Caucasian donors, (2) did not have a documented history of bone disease, and (3) possessed an indication of denture use by the presence of dentures with most cadaver remains. Craniometric dimensions of the mandibles are given in the Appendix, Table 1 (www.dentalresearch.org).
All bone preparation was carried out under a hood, with a sterile technique and continuous water drip for coolant. Specimens from 22 sites were harvested on both facial and lingual cortices (Fig. 1
Measurement Technique We measured all bone cubes with a digital caliper to the nearest 0.01 mm to verify diameter (4 mm) and determine cortical thickness, which is defined as the distance from the periosteum to the cortical-trabecular interface. Apparent density calculations, based on Archimedes’ principle of buoyancy, were measured to the nearest 0.001 g with a Mettler-PM460 analytical balance (Mettler, Toledo, OH, USA) and densitometry analysis apparatus. The principal axes of each specimen were identified by an ultrasonic technique (Schwartz-Dabney and Dechow, in press) with a Hewlett-Packard (Hewlett-Packard, Palo Alto, CA, USA) pulse generator, 2 mounted piezoelectric transducers (2.25 MHz longitudinal, Panametrics V323-SU, and 5.0 MHz shear, Panametrics V156-RM, Panametrics, Inc., Waltham, MA, USA), and an oscilloscope (Tektronix TDS-420, Tektronix, Inc., Beaverton, OR, USA). The arrowhead served as the origin for 9 measurements at rotations of 22.5°. The direction with the highest longitudinal velocity corresponded to the direction of maximum stiffness (D3), since ultrasonic velocity increases linearly with stiffness (Ashman et al., 1984; Kohles et al., 1997). The direction of minimum stiffness in the plane of the cortical plate (D2) corresponded to the slowest velocity and was 90° to the direction of maximum stiffness. Material property testing used the pulse transmission technique (Ashman et al., 1984; Ashman and van Buskirk, 1987; Schwartz-Dabey and Dechow, in press). Both longitudinal and transverse ultrasonic waves were passed through the principal axes (D2 and D3) and the cortical thickness (D1) of each specimen. We measured time delays to make a phase comparison of the signals before and after transmission. We calculated ultrasonic velocities by dividing the thickness or diameter by the time delay. Elastic properties were calculated from ultrasonic velocities according to standard techniques (Ashman et al., 1984; Dechow et al., 1993; Kohles et al., 1997). Young’s elastic modulus (E) measures a structure’s ability to resist deformation along an axis, indicated by subscripts as in E1, E2, or E3. Shear modulus (G) measures a structure’s ability to resist angular shear stresses in a plane between 2 axes indicated by subscripts as in G12, G31, or G32. Principal axes could be calculated for all specimens (N = 440). However, ultrasonic velocities in the D1 direction could not be obtained in 25 specimens, which were too thin (< 0.1 mm) for repeatable measurements to be obtained, and thus elastic properties could not be calculated. Half of these specimens came from 3 sites on the lingual ramus (#14, 15, 26). Differences between sides could not be tested at some sites (#27, 30) on the coronoid process, due to confluence of the cortical plates.
Analysis Minitab Software (Minitab Software, Pittsburgh, PA, USA) (release 13.3) was used for most statistical calculations. Data were checked for normality and hypotheses were tested with a balanced, unrestricted ANOVA with a repeated-measures design and mandible as the repeating factor. The repeated-measures design accounts for the lack of independence between multiple samples taken from a single mandible (Zar, 1996; Minitab User’s Guide 2, Release 13, pps. 3-31–3-33, Minitab, Inc., 2000). Statistical p-values were presented where differences existed between (1) sites, (2) facial and lingual cortices, or (3) dentate and edentulous mandibles. Directions of maximum stiffness, because of their unique angular distribution, required circular statistics (Zar, 1996), which we calculated with Oriana for Windows Version 1 (Kovach Computing Services, Anglesey, Wales, UK). We used Raleigh’s uniformity test (Zar, 1996) to determine whether the means themselves were significant, meaning the distribution of orientations was different from a random collection of angles. If angular means were significant, we tested differences between them with a generalization of the Watson-Williams test adapted for circular distributions (Zar, 1996).
Principal Axis Orientation The direction of the axes of maximum stiffness varied significantly between sites (p < 0.05), although several sites had no significant orientation (Fig. 2  and Appendix, Table 2 [www.dentalresearch.org]). Sites with the least variation were at the inferior corpus and posterior ramus on both cortices. Sites with the most variation were near the coronoid and condylar processes and sigmoid notch area. There were 5 paired edentulous sites with significant differences (p < 0.04) between facial and lingual sides (#3, 12, 25, 27, 31).
Between paired dentate and edentulous sites, two facial sites (#13, 28) differed significantly at p < 0.01. One site (facial #6) differed in that it was oriented in the edentulous but not in the dentate mandibles, while two sites (facial #24 and lingual #13) were oriented in the dentate but not edentulous mandible. Four sites (facial #12, 21, 26, and lingual #12) approached significance (p < 0.1) and are interesting because they were adjacent to sites with significant differences, suggesting a difference in a larger region of the mandible than that indicated by a single site.
Cortical Thickness
Throughout most of the edentulous mandibles, cortical bone was significantly thinner (p < 0.05) than in dentate mandibles. Facially, only 2 sites were thicker on average in the edentulous mandible (#2, 3). Lingually, the edentulous corpus tended to be thicker than that of the dentate, while much of the ramus was thinner, except for the angle and the condylar neck.
Density
Elastic Moduli
E3 values in edentulous mandibles were, on average, significantly stiffer (p < 0.04) than in dentate mandibles (22.8 GPa, SD = 5.4). Only 6 sites, all in the ramus (facial #13, 14, 21 and lingual #16, 24, 27) (Fig. 4
Shear Moduli Edentulous shear moduli were not significantly different between facial and lingual sides, but were significantly different among sites within each cortex (p < 0.01) and between directions (G12, G31, and G23) (p < 0.001) (Appendix, Table 4 [www.dentalresearch.org]). The grand means for edentulous shear moduli were 7.1 GPa (SD = 1.1) for G23, 5.3 GPa (SD = 1.0) for G31, and 4.5 GPa (SD = 0.9) for G12. G12 values in edentulous mandibles were, on average, modestly but significantly less stiff (10%, p < 0.01) than in dentate mandibles (5.0 GPa, SD = 0.6). Only 2 sites (#1, 2) on the facial cortex were stiffer. Average dentate G23 values (7.4 GPa, SD = 0.8) and G31 values (5.5 GPa, SD = 0.7) were not significantly different from those of the edentulous mandible.
Other Elastic Properties
Principal Axis Orientation Our results show orientation differences between dentate and edentulous mandibles in the retromolar and superior ramus regions. For edentulous mandibles, the direction of maximum stiffness of retromolar sites is less oblique, being about 15-30° closer to the inferior border orientation. The structural significance of these differences is difficult to determine, since little is known about the internal arrangements of osteonal bone in these regions. Likewise, functional correlates cannot be inferred, because we have inadequate knowledge of in vivo deformations of the relevant mandibular regions during oral activities. Ogata and Satoh (1995), in a clinical study, demonstrate one functional difference between dentate and edentulous individuals that may be important. A small sample of edentulous subjects exhibited increased bilateral occlusion compared with the more prevalent unilateral chewing and biting of the dentate. Bilateral occlusion reduces the transfer of masticatory and bite forces across the symphysis and mandibular twisting moments, which would increase the relative amount of bending during chewing and biting. Since an effective way for bone to resist deformation is for the direction of maximum stiffness to be aligned with the direction of load (principal stress) (Dechow and Hylander, 2000), then the less oblique direction of maximum stiffness may be a structural adaptation to resist relatively greater amounts of bending and decreased shear in the retromolar region. Other than our work on dentate mandibles (Schwartz-Dabney and Dechow, in press), no measurements of the direction of maximum stiffness are available for any bones. It is interesting that the direction of maximum stiffness deviates from the anatomical axes of the mandible at many sites. This may differ from diaphyses of the post-cranial skeleton, where the stiffest directions are thought to parallel the long anatomical axes. Also of great interest is the variation in the direction of maximum stiffness between sites. Some sites have a more constant orientation, while other sites have no statistically significant mean orientation between individuals. If material orientation is functionally related, this suggests that variations in loading may correspond to individual differences in mandibular form and muscular function.
Cortical Thickness A possible functional corollary in dentate mandibles is the relationship between mandibular cortical thickness and facial form. Cortical thickness beneath the molars varies from thicker in short-faced subjects to thinner in long-faced subjects (Masumoto et al., 2001). This difference may be directly related to masticatory loads, which are greater in short-faced individuals (Throckmorton et al., 1980).
Density Our results are apparent volumetric representations of density at an intermediate tissue level, and they may differ from the results of density studies on a microscopic scale, since they are influenced by the amounts of both microporosities, such as resorption spaces, and mineralization in bone. Our results agree with those of Henrikson and Wallenius (1974), who use a technique similar to ours and find no differences in densities between dentate and edentulous individuals, or between genders. Our study extends these results to other regions of the mandible, where likewise no significant differences are found. The lack of differences suggests that cortical bone density following edentulation is maintained despite changes in structure, stiffness, and anisotropy.
Elastic Properties The regional differences in maximum stiffness are also of great interest. In much of the mandible, cortical bone may partially compensate for a thinner cortex with increased stiffness. However, this increase in stiffness may come at a cost. While the mandible may deform less for a given load, the cortex may be more brittle and thus fail more quickly under the largest and most rapidly applied loads. On average, the edentulous mandible compared with the dentate mandible has (1) greater maximum stiffness coupled with thicker bone in the lingual corpus, (2) greater maximum stiffness anteriorly in the facial corpus and at the lingual condylar neck, (3) lesser maximum stiffness and thickness in portions of the ramus, and (4) different directions of the axes of maximum stiffness at some sites, especially in the buccal retromolar region. These differences imply adaptations to altered patterns of regional loading and deformation. In the ramus, these differences correspond to decreased EMG activity, muscular atrophy, and (presumably) muscle load in edentulous individuals (Raustia et al., 1996). An understanding of these differences and their impact on clinical restorations of the mandible and its dentition requires greater knowledge of oral function in the edentulous, as well as the links among material properties, the intermediate or microstructure of cortical bone, and the processes of bone adaptation. These findings also suggest the importance of considering the three-dimensional aspects of cortical bone structure and material properties in studies of patterns and mechanisms of bone adaptation.
We thank Drs. Peter Buschang, Richard Harper, and Gaylord Throckmorton and the three anonymous reviewers for their comments regarding this study. This project was supported in part by a VA Dental Research Fellowship and by NIH grants DE05691 and DE07256.
A supplemental appendix to this article is published electronically only at http://www.dentalresearch.org. Received for publication July 25, 2001. Revision received June 14, 2002. Accepted for publication June 17, 2002.
Journal of Dental Research, Vol. 81, No. 9,
613-617 (2002)
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
) indicates a significant difference between facial and lingual cortices.
