Elsevier

Archives of Oral Biology

Volume 45, Issue 6, 1 June 2000, Pages 461-474
Archives of Oral Biology

Quantification of human chewing-cycle kinematics

https://doi.org/10.1016/S0003-9969(00)00015-7Get rights and content

Abstract

This study introduces new methods of quantifying and evaluating the human chewing cycle. These methods were validated on a sample of 26 young adults (11 women and 15 men) between 20–35 years of age. Movements of the mandibular central incisors were recorded (100 Hz) using an optoelectric computer system while the participants chewed gum. A subsample of 10 cycles was automatically selected, based on multiple objective criteria to ensure representative cycles for each individual. Once representative cycles had been identified, multilevel statistical models were used to evaluate and describe the sample’s kinematic patterns. The multilevel procedures allow for missing observations, they do not assume equal intervals, and variation can be partitioned hierarchically. Two-level models showed significantly shorter cycle duration for males (835 msec) than females (973 msec). Inferior–superior (IS) cycle range was 2.6 mm larger and maximum IS velocity was 19.6 mm/sec faster in males than females. There were no significant differences in medial–lateral (ML) and anteroposterior (AP) excursive ranges or velocities. With the exception of cycle duration and ML ranges of motion, random variation was three to five times larger between individuals than between cycles. The three-level models showed that eighth-order polynomials were necessary to describe IS, AP, and ML chewing movements of the entire cycle. The models identified highly significant sex differences in cycle kinematics (excursions, velocities, accelerations, etc.) for each aspect of movement (AP, IS, and ML). It is concluded that this approach provides several important advantages over existing methods, including (a) its objectivity, (b) a more complete description of kinematic patterns, (c) a hierarchical description of variation, and (d) its ability to test hypotheses statistically.

Introduction

The form or shape of the human masticatory cycle has been of interest for many years. At a basic level the form and timing of the cycle has been analysed to understand better the mechanisms controlling masticatory behaviour, and how these mechanisms might differ between the sexes (Gerstner and Parekh, 1997, Neill and Howell, 1986, Youssef et al., 1997a), change with age (Gibbs et al., 1982, Karlsson and Carlsson, 1990, Huang et al., 1994, Snipes et al., 1998), and adapt to different varieties of foods (Deat et al., 1995, Huang et al., 1993, Lucas et al., 1986, Peyron et al., 1997, Takada et al., 1994, Thexton et al., 1981, Van der Bilt et al., 1991). At the clinical level, the masticatory cycle has been analysed for abnormalities in patients with malocclusions (Ahlgren, 1967), dentures (Feine et al., 1994), dentofacial deformities (Youseff et al., 1997b), xerostomia (Hamlet et al., 1997), temporomandibular disorders (Kuwahara et al., 1995a, Kuwahara et al., 1995b, Kuwahara et al., 1995c), and juvenile rheumatoid arthritis affecting the temporomandibular joint (Kjellberg et al., 1995). There is very little information about changes in the masticatory cycle after treatment (Youseff et al., 1997b).

Historically, the first step necessary to carry out these kinds of studies was the development of equipment to record the rapid and complex path and timing of the masticatory cycle. The earliest analyses used clutches that attached to the teeth and traced a path on paper, similar to a contemporary axiograph. These early systems were limited to tracing movement in only two dimensions and were unable to record timing information along with the trace. More modern clutch systems recorded time and movement information into a computer (Gibbs et al., 1971), but still were relatively cumbersome to use.

Cinematography eliminated the cumbersome clutches (Ahlgren, 1966, Ahlgren, 1967) but generally recorded movement in a single plane, and each frame of film represented an instant in time, allowing calculation of timing within each chewing cycle. Because the mandible was covered by mobile skin and the teeth were mostly hidden behind the lips, the accuracy of the cinematographic record was limited.

Cinematography combined with a fluoroscope (cineradiography) allowed visualization of the facial skeleton and teeth and, like cinematography, recorded timing information (Hamlet et al., 1997, Cleall, 1965, Ostry and Munhall, 1994). Although often used to study animal mastication, its use in human studies has been limited because of dosing with ionizing radiation.

One of the most common methods of recording jaw movements today is the magnetic tracking system (Yamada and Yamamura, 1996; Kuwahara et al., 1995,1996; Nakamura et al., 1988, Deat et al., 1995, Takada et al., 1994; Youssef et al., 1977a,b). A small magnet is attached to the lower incisors and its position in three dimensions can be tracked by a set of Hall-effect antennae fitted over the participant’s head. The X,Y, and Z coordinates of the magnet, relative to the head, can be recorded by computer at speeds of more than 500 frames/sec. Importantly, the magnet represents only a single point on the mandible, and so information about movements of other teeth or the mandibular joint is lost. In addition, it has been shown that the magnetic tracking systems are subject to distortion errors at wider gaps (Balkhi et al., 1991, Throckmorton et al., 1992), perhaps restricting their application to chewing studies.

The other common method of recording jaw movements is the optoelectric system (Karlsson et al., 1991, Karlsson and Carlsson, 1990, Gerstner and Parekh, 1997, Feine et al., 1994, Snipes et al., 1998; Peryon et al., 1997). This system records the position of multiple markers on the head and mandible, at speeds up to 200 frames/sec. Depending on the system the markers either emit infrared radiation or reflect it towards a set of at least two, and in most cases three, cameras. The advantage of this system is that markers can be grouped on rigid bodies that are attached to the head and mandible. Because the rigid bodies consist of multiple points, special computer programs can track any point on the mandible, provided the coordinates of that point are known relative to the rigid body mounted on the mandible. Therefore, these systems can record not only movement of an incisor tooth or the chin, but also any other teeth or the condyles. The three-dimensional data from all moving points are stored in the computer and selected movements, including velocities and accelerations of those movements, can be easily plotted.

The second necessary step in the analysis of the human chewing cycle was deciding how to analyse the large amount of data available from even a single experimental session. Each chewing cycle includes information about timing and excursion. The analysis is further complicated when the excursions are measured along three axes (inferior–superior, medial–lateral, anterior–posterior). Over the years, many different measurements of the timing and excursions of the chewing cycle have been developed. Almost every study has measured the average cycle duration. Many studies have broken the chewing cycle into two (open–close), three (open–occlusal–close) or four (slow open, fast open, fast close, power stroke) phases (Yamada and Yamamura, 1996, Karlsson et al., 1991, Karlsson and Carlsson, 1990, Nakamura et al., 1988, Snipes et al., 1998, Peyron et al., 1997, Takada et al., 1994, Youssef et al., 1997a, Youssef et al., 1997b).

The weakness in many of these studies is the somewhat arbitrary definitions of the boundaries of the different phases. The boundary between the open and close phases, for example, is usually defined as maximum opening. But in most studies maximum opening is determined along the inferior–superior axis, without considering the incisor position in the anteroposterior and medial–lateral directions. Many studies have used an arbitrary opening distance as the boundary between the occlusal phase and the opening and closing phases (Gibbs et al., 1971, Gibbs et al., 1973, Gibbs et al., 1981, Messerman, 1963, Kiliaridis et al., 1991, Kjellberg et al., 1995, Suit et al., 1976, Ferrario et al., 1990, Woda et al., 1979, Wood et al., 1981). Studies defining four phases have used arbitrary changes in velocity along the inferior–superior axis as boundaries between the slow and fast phases (Crompton et al., 1977, Hiiemae, 1978, Hiiemae and Crompton, 1985, Hiiemae and Kay, 1973, Oron and Crompton, 1985, Snipes et al., 1998, Youssef et al., 1997a, Youssef et al., 1997b).

In order to analyse mandibular excursions, almost all studies have recorded the maximum position of the mandible along the inferior–superior axis. Fewer studies have recorded the maximum position of the mandible along all three orthogonal axes (Karlsson and Carlsson, 1990, Snipes et al., 1998, Deat et al., 1995, Youssef et al., 1997a, Youssef et al., 1997b) or have recorded maximum positions at each phase of the chewing cycle (Youssef et al., 1997a, Youssef et al., 1997b, Snipes et al., 1998). A few studies have calculated a three-dimensional vector for maximum opening (Karlsson et al., 1991, Deat et al., 1995). Only a few studies have recorded excursions of points other than the incisor or chin (Piehslinger et al., 1994, Ostry and Munhall, 1994).

Other features of the chewing cycle that have been measured include: (1) velocities and accelerations (Kuwahara et al., 1995, 1996; Karlsson et al., 1991, Karlsson and Carlsson, 1990, Gerstner and Parekh, 1997; Takawada et al., 1994); (2) the angle of approach of the mandible into centric occlusion (Karlsson et al., 1991, Gibbs et al., 1973, Gibbs et al., 1981, Suit et al., 1976, Ogawa et al., 1997, Ogawa et al., 1998); and (3) timing of specific landmarks on the chewing cycle (Gerstner and Parekh, 1997, Peyron et al., 1997, Piehslinger et al., 1994). These measurements are quantitative and can be easily analysed with simple statistical tests. However, a great deal of information about the overall form of the chewing cycle is discarded in these types of analyses.

Another quite different approach to the analysis of the chewing cycle is to classify subjectively pathways taken by the incisor in the sagittal and frontal planes (Kuwahara et al., 1996, Karlsson and Carlsson, 1990, Takada et al., 1994, Ahlgren, 1966, Ahlgren, 1967, Proschel, 1987). This approach has the advantage of using more of the information in the pathway, but deciding into which category to place a particular trace is somewhat arbitrary. In addition, a relatively small number of cycles is used to classify each individual’s chewing pattern. Statistical analysis is reduced to comparing the number of chewing cycles with a particular pattern, rather than analysis of the overall shape of the chewing cycle.

Regardless of the approach used by previous studies, only part of the information available was usually analysed statistically. The most common approach is to calculate mean values for the duration and maximum excursion variables and use either t-tests or simple ANOVAs to test for significant differences between samples. Most studies done in this way carry out a separate statistical test on each variable measured in the study (Youssef et al., 1997a, Youssef et al., 1997b, Takada et al., 1994, Deat et al., 1995, Kuwahara et al., 1995a, Kuwahara et al., 1995b, Kuwahara et al., 1995c, Kuwahara et al., 1996, Ostry and Munhall, 1994, Hamlet et al., 1997, Yamada and Yamamura, 1996, Nakamura et al., 1988, Snipes et al., 1998, Huang et al., 1993, Huang et al., 1994, Karlsson and Carlsson, 1990, Karlsson et al., 1991). However, it has not been established that the various measures of timing and excursion are independent from each other. It may be that independently testing each variable is inappropriate. To date, only Gerstner and Parekh (1997) have used a multifactorial analysis to describe the many different features of the chewing cycle.

A robust method for describing and comparing chewing cycles should have the following features. First, it should include the total three-dimensional surface of the chewing envelope. Second, it should be able to provide information about chewing speeds, including rate and acceleration. Finally, it should include a systematic method for reducing the high between-cycle variation usually seen within each individual participant.

Our purpose here is to demonstrate the applicability of novel methods for evaluating mandibular jaw movements while chewing. The focus will be on (1) improved methods of data preparation for analysis, (2) new strategies for data analysis, and (3) the application of a multilevel statistical approach for the description and comparison.

Section snippets

Materials and methods

Twenty-six young adults (11 women and 15 men, ages 20–35 years) were selected from students and staff at Baylor College of Dentistry in Dallas, Texas. All were Caucasian, with complete dentitions, excluding third molars, and had no extreme brachycephalic or dolichocephalic headforms. They all had Angle class I dental relationships with minimal overjet (<4 mm), overbite (<4 mm), and crowding (<4 mm). None had full-coverage restorations or tooth replacements, signs or symptoms of

Absolute cycle duration, distances, and velocities

Females (973 msec) showed significantly longer total cycle duration than males (835 msec). Opening duration was 92 msec longer (p<0.01); closing duration was also longer (45 msec), but sex differences were not statistically significant. Females showed less (2.61 mm) inferior–superior jaw movement than males (p<0.05). No sex differences in the ranges of medial–lateral or anterior–posterior excursion during chewing were identified. There were also no significant differences in total distances

Discussion

Our estimates of cycle duration fell within the range previously reported (Ahlgren, 1966, Jemt et al., 1979, Kiliaridis et al., 1991, Karlsson and Carlsson, 1989). We confirmed that women have significantly longer cycle duration than men (Niell and Howell, 1986; Snipes 1994). As the possibility of our committing a type I error was small, the reported lack of sex differences (Kiliaridis et al., 1991, Karlsson et al., 1991) should be reconsidered. The results also confirm that men have a larger

References (57)

  • T. Ogawa et al.

    Correlation between inclination of occlusal plane and masticatory movement

    J. Dent.

    (1998)
  • K. Takada et al.

    The effects of food consistency on jaw movement and posterior temporalis and inferior orbicularis oris muscle activities during chewing in children

    Arch. Oral Biol.

    (1994)
  • G.S. Throckmorton et al.

    Reproducibility of mandibular motion and muscle activity levels using a commercial computer recording system

    J. Prosthet. Dent.

    (1992)
  • A. Woda et al.

    Non-functional and functional occlusal contacts. A review of the literature

    J. Prosthet. Dent.

    (1979)
  • W.W. Wood et al.

    Effect of occlusal reconstruction on the reproducibility of chewing movements

    J. Prosthet. Dent.

    (1981)
  • Y. Yamada et al.

    Possible factors which may affect phase duration in the natural chewing rhythm

    Brain Res.

    (1996)
  • R.E. Youssef et al.

    Comparison of habitual masticatory activity before and after orthognathic surgery

    J. Oral Maxillofac. Surg.

    (1997)
  • R.E. Youssef et al.

    Comparison of habitual masticatory patterns in men and women using a custom computer program

    J. Prosthet. Dent.

    (1997)
  • J. Ahlgren

    Mechanism of mastication: a quantitative cinematographic and electromyographic study of masticatory movements in children, with special reference to occlusion of the teeth

    Acta Odontol. Scand.

    (1966)
  • J. Ahlgren

    Pattern of chewing and malocclusion of the teeth. A clinical study

    Acta Odontol. Scand.

    (1967)
  • K.M. Balkhi et al.

    Error analysis of a magnetic jaw-tracking device

    J. Craniomandib. Disord.

    (1991)
  • A.W. Crompton et al.

    The activity of the jaw and hyoid musculature in the Virginia opossum, Dideplphis virginiana

  • D.G. Deat et al.

    Association between the interarch distance and food bolus size in the early phase of mastication

    J. Prosthet. Dent.

    (1995)
  • J.S. Feine et al.

    Within-subject comparisons of implant-supported mandiblar prostheses: evaluation of masticatory function

    J. Dent. Res.

    (1994)
  • V.F. Ferrario et al.

    Analysis of chewing movements using elliptic Fourier descriptors

    Int. J. Adult Orthod. Orthognath. Surg.

    (1990)
  • G.E. Gerstner et al.

    Evidence of sex-specific differences in masticatory jaw movement patterns

    J. Dent. Res.

    (1997)
  • H. Goldstein

    Multilevel mixed linear model analysis using iterative generalized least squares

    Biometrika

    (1986)
  • H. Goldstein

    Multilevel Models in Educational and Social Research

    (1987)
  • Cited by (111)

    • An MRI evaluation of the effects of qat chewing habit on the temporomandibular joint

      2018, Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology
    View all citing articles on Scopus
    View full text