Biomechanical properties of masticatory balance in cases with RPDs—The influence of preferred and nonpreferred chewing side: A pilot study

Abstract Objectives Removable partial dentures (RPDs) are inserted with the aim to restore masticatory function. There is however inconsistent evidence supporting the alleged improvements, posterior occlusal contacts being one of the decisive factors. We hypothesized that the distribution of abutment teeth in RPDs influences masticatory performance and functional parameters. To evaluate the masticatory performance and functional parameters in patients with a RPD using a single mathematical parameter (tilting index [TI]) for both jaws that predicts biomechanical behavior on the basis of the distribution of abutment teeth. Materials and Methods Masticatory performance was measured in patients wearing long‐time adapted RPDs using the standardized test food optocal, yielding the mean particle size (X 50). Mastication on the preferred and nonpreferred chewing sides was analyzed. Total muscle work (TMW) was calculated using bipolar electromyographic recordings of the masseter and anterior temporalis muscle. Functional parameters were subjected to multiple linear regression analysis including X 50 as a dependent variable and functional units (FU), the number of teeth, bite forces, and sagittal and frontal components of TI (TI α and TI β) as independent variables. Results When the preferred chewing side was tested, none of the investigated parameters correlated significantly with X 50. In contrast, chewing on the nonpreferred side was correlated significantly with performance for most variables (p < .05). This means that increased dental support improved chewing performance with RPDs under these conditions. Conclusions In well‐adapted RPDs, the distribution of abutment teeth as expressed by the tilting index seems to be of subordinate importance for masticatory performance.


| INTRODUCTION
With a rapid increase in life expectancy, the number of partially dentate patients is also growing. In particular, caries and periodontal diseases lead to tooth loss and/or shortened dental arches.
Depending on the number and location of lost teeth, removable partial denture (RPD) is used to restore masticatory function, esthetics, and phonetics. Previous studies have shown that the masticatory performance of patients with RPD is less than for fully dentate subjects (Ikebe et al., 2012). Above all, gender, bite force, location of remaining teeth, and the number of functional units (FU) affect the masticatory performance of patients with RPD (Tumrasvin et al., 2006).
Studies on the biomechanical effects of incorporated RPD have furnished divergent results, depending on the measure of performance used and the number of lost teeth. It has been reported that patients with an extremely shortened dental arch have a poor masticatory performance, which can be improved by wearing a removable partial denture (Arce-Tumbay et al., 2011). In contrast, other studies found that RPD has no effect on masticatory performance if the premolar regions are intact (Ikebe et al., 2011;Peyron et al., 2004).
For patients with unilateral shortened dental arches, the correlation between bite force and masticatory performance is stronger on the dentate side (Tumrasvin et al., 2005), and the greatest correlation between masticatory performance and bite force is observed for the first molar region (Lujan-Climent et al., 2008). It can generally be stated that the decrease in masticatory performance among elderly patients is mainly caused by tooth loss, and less by reduced bite force and muscle forces (Ikebe et al., 2011;Peyron et al., 2004).
Monitoring of electric muscle activity (EMG) reveals that symmetrical activation of the masticatory muscles improves chewing performance (Garrett et al., 1995) and that the duration of EMG activity during the chewing cycle is inversely proportional to the stability of the dentition (Balkhi et al., 1993). It has, furthermore, been reported that EMG activity (area under the curve, or integral) during chewing of foods of different hardness is not significantly different for elderly people with full dentition and young subjects. This means that the elderly are still able to adapt excellently in the submaximum range of muscle activation, despite the loss of maximum bite-force capacity (Peyron et al., 2004). Reduced chewing performance of the elderly might be caused by tooth wear, which reduces the biomechanical effectiveness of the teeth (Giannakopoulos et al., 2014).
In previous work, biomechanical conditions that affect the chewing performance of partially edentulous patients have been studied by use of classic measures, for example, the "Kennedy" or "Eichner" classification (Ikebe et al., 2010;Rehmann et al., 2015), which characterize linear, triangular, and quadrangular dental support conditions as well as FU (Figure 1). Essential physiological data for the masticatory system, in particular EMG and bite force measurements, have also been used to determine the neuromuscular effect of tooth loss on the functional behavior of the system. Criteria for evaluation of the biomechanical balance of prostheses on the basis of the arrangement of the residual teeth and other support elements for both jaws, by use of a single model, are not available. Such an approach might enable more consistent grading of the biomechanical conditions for prosthetic reconstructions, and, more importantly, better characterization of the biomechanical balance. Additional basic condition in this context represents the adaptability of the neuromuscular system. The target for a realistic assessment of the neuromuscular capacity of the masticatory system is unquestionably the physiology of mastication.
Mastication is a complex process involving food breakdown, moistening and dilution in saliva, bolus formation, and swallowing.
Masticatory performance was proposed to be the result of two processes: selection of food particles and breakage (Lucas & Luke, 1983). These processes are fundamentally dependent on learned motor patterns, and both should be reflected by the method of testing experimental performance. Commonly, the masticatory performance is investigated in the short term after the incorporation of dentures. However, this experimental design cannot picture the realistic chewing performance after long-term motor adaptation.
Previous investigations substantiated the adaptability of the jaw motor system and have also shown that the motor behavior adapts to new motor tasks by training (Hellmann et al., 2011). This adaptation needs a certain time (Goiato et al., 2010). In this context, it can be supposed that in the fully adapted chewing system the nonpreferred chewing side, which is challenged by unfamiliar chewing, may respond with different performance.
The purpose of this investigation was twofold. First, it correlates the masticatory performance of partially dentate subjects wearing RPD for more than 3 years with two variables, the distribution of

| Biomechanical model
Traditional classification schemes, for example, the "Eichner" or  F I G U R E 2 Model for evaluation of the support situation. At the positions of the teeth, springs (stiffness k i/j ) are attached to a rigid plate in the occlusal plane. Δz denotes the initial deflection of the spring ends toward the occlusal plane necessary to produce a preload the plate remains in the horizontal position. The more the situations in the maxilla and mandible differ from this state, the larger becomes the tilting, that is, more unfavorable is the static situation of the system. The vertical displacement at the position of the resultant force in each jaw (z rf ) is a measure of total stiffness differences of the supports in the maxilla and mandible and will always be oriented toward the more resilient side.
Clinically, the algebraic signs of the three variables are possibly misleading; absolute values of |z rf |, |α|, and |β| should therefore be used. In statistical analysis of the results from this study, however, only the TI, |α|, and |β|, were analyzed whereas the effect of total stiffness differences (corresponding with z rf ) was not taken into account.

| Masticatory performance
After completing a test cycle, the patients performed three different masticatory performance tests. Each test entailed 15 chewing strokes for habitual and unilateral chewing on the right and left sides of the jaw. Habitual chewing was repeated two times. Standardized artificial test food (Optocal) was used (Pocztaruk Rde et al., 2008) and 17 5.6-mm cubes formed a portion. Optosil, plaster, alginate, vaseline, and toothpaste were blended in a standardized mixing ratio.
The hardness and texture of Optocal make it similar to natural test food, but it can also be chewed by patients wearing a prosthesis (Pocztaruk Rde et al., 2008;Slagter et al., 1993) and it has been used in previous studies (Eberhard et al., 2018).
In this context, habitual chewing enabled the identification of the preferred chewing side. To identify the preferred chewing side, three methods were employed. First, total muscle work (TMW) ratios were used. The side that showed greater muscle activity (TMW) was determined to be the preferred chewing side. Similar methods were applied by Yamasaki et al. (2016) and Ratnasari et al. (2011). Second, observation by an examiner was used as an indicator. Finally, the patients were interviewed and their preferred chewing side was documented. Related to this, other studies have used a questionnaire or a visual analog scale to document the subjective preferred chewing side (Diernberger et al., 2008;Rovira-Lastra et al., 2016).
It is known that EMG activity is greater on the preferred chewing side during deliberate unilateral chewing (Stohler, 1986). Thus, muscle activity for habitual chewing and deliberate unilateral chewing were compared, and the greater activity was chosen. This pattern is reflected in the measurements from all three performance tests; usually, it is most pronounced during habitual chewing. The activity of the masseter muscle was decisive for determining the chewing side, because this muscle is mainly responsible for the power stroke, whereas the temporalis muscle is more important for the coordination of movements and positioning of the mandible. This is also in agreement with Yamasaki et al. (2015), in which masseter activity was used to determine the actual chewing side.
For silicone test food, the standard approach is to determine the weight of the particles retained by sieves and to curve-fit the cumulative weight by the use of the Rosin-Rammler equation. In the present study, in contrast to classical sieving methods, optical scanning analysis methods for measuring masticatory performance were used (Eberhard et al., 2012). The approximate weight distributions were determined by the use of the Rosin-Rammler equation and by a least-squares method (Olthoff et al., 1984;Rosin & Rammler, 1933;Slagter et al., 1993), by use of a MatLab tool (MatLab tool "Rosin Rammler diagram v 1.0" by Ivan Brezani, 2010). The Rosin-Rammler function used was: In this equation, Q represents the volume percentage of the particles with a size smaller than X. X 50 is the median particle size or the size of the theoretical sieve through which 50% of the volume of particles can pass. The variable b represents the broadness of the size distribution. A high X 50 -value means that the chewing performance is poor. The particle size is overall large. A low value, on the other hand, means that the particles are quite small and the chewing performance is good (Mowlana et al., 1994;Speksnijder et al., 2009;van der Bilt et al., 1993).

| Electromyography
The electromyographic activity of the chewing muscles was recorded

| Experimental procedure
After the installation of the EMG recording device, the three masticatory performance tests were conducted in one session. First, the subjects performed three maximum bites in maximum intercuspation. After a test cycle for habitual chewing to familiarize them with the procedure, subjects were asked to chew habitually, followed by chewing on the right side and, finally, on the left side. Habitual chewing was used to identify the preferred chewing side and was repeated two times. Under each condition, 15 chewing cycles were performed. The minced food was spat out and the mouth was rinsed with water to collect all the particles in a filter bag.
The bitefork was adjusted with silicone impression material.
Initially, while holding the bitefork in the mouth and stabilizing it by hand, the patient was asked to bite on the bite blocks and to perform some bites to become familiar with the feedback device. If the subject was able to control the bite force by watching a bar on the feedback screen, he/she was asked to bite three times with 50, 100, and 150 N on the device.

| Data analysis
The collected and dried chewed artificial test food was scanned and analyzed by use of a validated procedure (Eberhard et al., 2012). X 50 values were calculated by use of the Rosin-Rammler algorithm. The EMG data were analyzed by use of AcqKnowledge 3.9.1 software and a semiautomatic Matlab program. Root mean square (RMS) normalized recordings adjusted to maximum biting EMG were used to compute the area under the curve (integral) of the EMG bursts.
TMW for all bilaterally measured muscles was summed for the 15 chewing cycles and analyzed. The ratios of side-specific TMW (SMW) for the preferred and nonpreferred chewing sides were also analyzed. Bite force recordings were evaluated separately for each side of the jaw.
The tilting indices TI α and TI β categorize the overall biomechanical balance of the jaws for the respective RPD-reconstructed dentition. The number of remaining teeth was also counted, and documented as totals for the preferred and nonpreferred chewing sides for each subject. In the same way, FU (antagonistically contacting teeth) for each person were matched with the chewing sides.
To achieve a realistic comparison of the functional data with the remaining dentition, all analyzed recordings were adjusted in relation to the preferred or nonpreferred chewing side (e.g., the EMG data for F I G U R E 4 Bitefork with specific characteristics. Bilateral sensors are placed between bite blocks individualized with silicone impression material. S1 and S2: individualized bite blocks left-side chewers were interchanged in comparison with those for the right-side chewers). Means and standard deviations (SD) were calculated for all results.

| Statistics
Spearman correlation was used to analyze correlations between the variables X 50 , TI, TMW, FU, bite forces, and the number of teeth.
Multiple linear regression analysis was also used. Preferred and nonpreferred chewing sides were analyzed separately. Chewing side differences for TMW, SMW ratios, and bite forces were tested by repeated-measures analysis of variance. The level of significance was set at p < .05.

| Ethical approval
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional review board and with the 1964 Helsinki declaration and its later amendments.

| RESULTS
A broad distribution of the total number of remaining teeth was observed for the maxillae and mandibles; the minimum was three and the maximum 20 ( Figure 5). The findings revealed that 22 of the subjects chewed preferentially on the right and seven on the left. The values obtained for TI α and TI β are listed in Table 2; high and low values represent the most unfavorable and favorable cases, respectively. FU could be detected for 16 subjects but were absent in 13.
The X 50 values presented in Table 3 ranged from 0.50 to 4.82 mm; high scores are indicative of relatively poor and low scores for relatively good performance, respectively. The X 50 values for the nonpreferred chewing side were smaller than those for the preferred side but did not differ significantly (p = .68). Values of X 50 and the tilting index β (tilting around a sagittal axis) were significantly positively correlated (r = .36, p = .05) for chewing on the nonpreferred side (Table 4). No significant correlation (r = −0,04, p = .85) with TI β was observed for the preferred chewing side, however. Correlations between X 50 and TI α values of both preferred and nonpreferred chewing sides were not significant (r = .17, p = .39 and r = .23, p = .51, respectively) ( Table 4).
On the preferred chewing side, the mean number of teeth was 6.52 (SD = 2.6); on the nonpreferred side the mean was 6.10 (SD = 2.7). A significant negative correlation was observed for the total number of teeth and the X 50 values for both the preferred and nonpreferred chewing sides (r = −.42, p = .03 and r = −.37, p = .05).
None of the bite forces for the preferred and nonpreferred chewing sides correlated significantly with the X 50 values (p > .05), and no significant differences between bite forces for either side were found for any force level (

| DISCUSSION
The objective of this study was to quantify the functional effects of different abutment teeth distributions for a sample of patients restored with RPDs that had been worn for at least 3 years. The performance of the nonpreferred chewing side was assumed to correspond to a nonadapted situation, for instance, a condition immediately after tooth loss or a short-term response after modifying static tooth distribution in the context of various prosthetic reconstruction techniques. In the context of this study, we refer to adaptation in a general sense as the reaction of the organism to an alteration of internal or environmental conditions. Several physiological variables were included in the statistical analysis. For the preferred chewing side, the main result of this investigation is that-on the basis of the rather small data set of this pilot studymasticatory performance correlates neither with a static biomechanical balance of the restored chewing system, characterized by the TI nor with the number of FU, the number of teeth, TMW, or bite force. A small, nonsignificant influence on the number of functional units can be seen on the preferred chewing side as well.
We assume that in a larger patient collective, this effect might be more pronounced. Even so, for long-term adapted RPD and the preferred chewing side static biomechanical variables seem to be of subordinate significance for performance.
In contrast, if the nonpreferred chewing side is used by the patients, the static balance of restorations, characterized by the TI, the FU, and the number of teeth, correlate significantly with X 50 values. The TI α, representing the tilting of the denture in the sagittal plane, is of greater importance than the TI β, referring to the frontal plane. Summed up, the distribution of abutment teeth has a significant effect on the nonpreferred side. We would attribute this effect to the unfamiliarity of this condition to the patient. The lower TMW indicates that patients may have exerted greater caution when chewing on the nonprefered side. As their masticatory performance tends to increase over an adaptation period (Giannakopoulos et al., 2017), we assume that in an unfamiliar situation, the influence of the distribution of abutment teeth is higher in proportion to overriding factors and thus produces a significant result. It can be argued, whether this unfamiliar situation is generalizable to nonadapted states as described above (e.g., new dentures). To validate this claim, longitudinal studies involving a change in the number of abutment teeth would have to be performed.
On the basis of our findings, the initially stated hypothesis that specific distributions of abutment teeth of RPDs in the maxilla and mandible significantly affect masticatory performance must be rejected for the adapted restored jaw.
The results do, however, support the notion that for adapted neuromuscular systems motor control strategies are adjusted in such a way as to enable them to perform with the best available biomechanical effectiveness. This may imply that, in the long term, the masticatory system does not correspond to static loading conditions in intercuspation, as is generally assumed. This is also supported by the findings that the ratios of SMW (i.e., working side vs. balancing side; 1.5 for chewing on the preferred side vs. 1.1 for chewing on the nonpreferred side) differ significantly under both chewing conditions; this was not observed for the TMW values. The ratios for the preferred side correspond well to those for natural dentition during unilateral chewing (Proschel & Morneburg, 2010). In the context of our study, it must be considered that during chewing the food bolus is placed unilaterally between the opposing jaws, resulting in bite forces substantially different from those of symmetrical static loading of the dentition or prosthetic reconstruction. Thus, our results challenge a mechanistic approach that is still commonly used for the prediction of the denture dynamics of reconstructions.
Several limitations of this study must be considered. First, the study sample was relatively small and the genders were unequally distributed (this issue should be considered in future investigations).
The broad range of abutment teeth distribution may, however, have can be expected for the planned denture or prosthesis immediately after incorporation. Based on these results, the dentist could decide to add another implant or abutment tooth to achieve a smaller tilting index and greater stability. A major advantage of the developed model is that the tilting index can be calculated individually for each patient time-saving and efficient.
On the basis of this pilot study, and despite its inherent limitations, it might also be of future interest to evaluate the previously preferred chewing side of patients, to enable comparison of this preference with the preferred chewing side in long-term use, in particular, because for the human population and all age groups chewing side preference seems to be normal behavior (Barcellos et al., 2012;Nayak et al., 2016;Nissan et al., 2004). For future planning of RPD, this information might be of benefit for obtaining recommendations for placing additional load-bearing structures (i.e., implants) in the jaws for immediate optimum chewing performance or in patients with a known history of bruxism, to better distribute the generated pathophysiological forces. However, prospective studies have to prove these conceptual hypotheses.

ACKNOWLEDGMENTS
This investigation was conducted as part of a thesis. The first and second authors contributed equally to this study. The work received no funding.

CONFLICTS OF INTEREST
The authors declare no conflicts of interest.

INFORMED CONSENT
Informed consent was obtained from all individual participants included in the study.

DATA AVAILABILITY STATEMENT
Supporting Information, such as raw data, is available from the corresponding author upon reasonable request.