Measuring lip force by oral screens Part 2: The importance of screen design, instruction and suction

Abstract The aim of this study was to find a reliable method for measuring lip force and to find the most important factors that influence the measurements in terms of magnitude and variability. The hypothesis tested was that suction is involved and thus the instruction and the design of the oral screen are of importance when measuring lip force. This is a methodological study in a healthy population. This study was conducted in a general community. The designs of the screens were soft and hard prefabricated screens and 2 semi‐individually made with a tube allowing air to pass. The screens and the instructions squeeze or suck were tested on 29 healthy adults, one at a time and on 4 occasions. The test order of the screens was randomized. Data were collected during 4 consecutive days, and the procedure was repeated after 1 month. The participants were 29 healthy adult volunteers. The instruction was an important mean to distinguish between squeezing and sucking. The design of the screen affected the lip force so that it increases in relation to the projected area of the screen. A screen design with a tube allowing air to pass made it possible to avoid suction when squeezing. By measuring with and without allowing air to pass, it was possible to distinguish between suction related and not suction related lip force. The additional screen pressure when sucking was related to the ability to produce a negative intraoral pressure. In conclusion lip force increases in relation to the projected area of the screen, sucking generally increases the measured lip force and the additional screen pressure when sucking is related to the ability to produce a negative intraoral pressure.

using a soft oral screen, Ulmer large, by testing on two occasions (Sjögreen et al., 2011). The impact of the size of the screen and the lip force executed has been shown to be of great importance (Wertsén & Stenberg, 2017). A small screen scores the smallest value and the largest screen the highest (Table 1).
Furthermore, it was shown that by dividing the measured lip force by the screen size, measured as the projected area, the influence of screen size could be eliminated. By measuring oral screen pressure (OSP) in a pressure unit, for example, kPa (kilopascal) measurements from screens with different areas can be compared (Wertsén & Stenberg, 2017). When sucking, apart from the perioral muscles, a great number of muscles are involved. Thus, the measured force will not reflect the force performed by the perioral muscles. In the previous study, the screens were made with a small tube allowing air to pass in to the oral cavity thus preventing the test person to use suction during the measurement (Wertsén & Stenberg, 2017). This option was not available in other studies (Hägg et al., 2008;Sjögreen et al., 2011). Eklund and Eklund who used the screen OS/II in a pilot study also discussed the influence of sucking.
They found that some of the test persons had a tendency to suck the screen. This, as well as clenching the jaws, increased the measuring values markedly (Eklund & Eklund, 2007). They also stated that the measurement value was not influenced by different head or seating positions. In other studies, the subjects were instructed to keep the screen as long as possible inside the lips resisting the increasing force for as long as possible (Hägg et al., 2008;Sjögreen et al., 2007). This instruction makes it uncertain whether the subjects have been able to avoid suction or not. However, the studies are few in numbers and have been performed with different kinds of oral screens, hard, soft, and semi-individually made. Thus, the values reported are difficult to compare.
Changes in intraoral pressure are required to implement swallowing, mastication, or speech. Engelke proposed that several structures operating as biofunctional compartments and valves are participating during swallowing. These collaborate in a coordinated interaction following a dynamic pressure gradient (Engelke, Jung, & Knösel, 2011). Santander showed that intraoral pressure varied significantly depending on different bolus used and its consistency (Santander, Engelke, Olthoff, & Völter, 2013). As a consequence, it is of interest to investigate if it is possible to clearly separate the measurements with an oral screen for squeezing and sucking using different instructions. The aim of this study was to compare soft and hard prefabricated oral screens and two semi-individually made screens to get separate values for squeezing and suction.

| MATERIALS AND METHODS
The Ethics Committee of the University of Gothenburg approved the study , and it was performed in accordance with the Declaration of Helsinki.

| Test subjects
Twenty-nine healthy adults, 20 females and 9 males, aged 31-68 years with ordinary morphology of the face, normal oral motor function, and occlusion were recruited on voluntary basis for the study. They mainly comprised dental health personnel at the Public Dental Service clinic in Mölndal, Sweden. All the individuals gave their informed consent to participate and continued through the entire testing period.

| Procedure for measuring projected area of the oral screen
The area of the different screens was determined by projecting as in a previous study (Wertsén & Stenberg, 2017). To investigate if the deformation of the soft screen, when pulling, would influence on the projected area, it was fixed with transparent tape in a pellucid tube.
The screen was then exposed to the same forces that were measured with the testing subjects.

| Measurement procedure
A dental chair with arm and foot rests was used. The examiner started by demonstrating the measuring procedure and gave a verbal instruction. The measuring procedure was carried out with and without suction. Instruction without suction: "Hold the oral screen in your mouth and squeeze your lips as firmly as you can, while I pull it out." Instruction with suction: "Hold the oral screen in your mouth and suck as hard as you can while I pull it out." The test person placed the oral screen in the vestibulum. The wire was stretched in a straight angle, and an assistant started the measuring period of 10 s. The examiner pulled the wire and gradually increased the power until the oral screen was pulled loose. The maximum value was noted. The procedure was repeated three times in a sequence, and all the values were used in the statistical analyses.

| Data collection
The order, in which the screens were tested, was randomized. Data were collected on approximately the same hour during four consecutive days. Four oral screens were tested for each of the 29 subjects, one each day. Two instructions were given, and three measurements for each instruction were carried out. Each screen type was tested at two occasions 1 month apart. To squeeze was always the first instruction given at each session. The test persons rested for 3 min between the two instructions. All measured values were registered in Newton. In total, 174 measurements were carried out for each oral screen and instruction. The same investigator made all measurements.

| Statistical analysis
The dataset was first analyzed in MS Excel for calculating the OSP by dividing force values with appropriate screen areas. Ninety-five percent confidence levels of the different means were calculated from t statistics with df = 173 for the overall means, df = 28 for the distribution of individual means, and df = 5 for individual means.
The sample correlation coefficient (r 2 ) was calculated in the MS Excel graphs. The individual means data were analyzed in SPSS for normality by the Shapiro-Wilk test. Confidence ellipses for the bivariate Gaussian distribution were calculated from the chi-square distribution with elliptical axes = 2kσ with k = 1.177 for 50% confidence and k = 2.448 for 95% confidence.

| Pilot vacuum experiment
One of the screens (Screen C) was modified to permit measurement of the established vacuum in the oral cavity during lip force measurement. A silicone tube was used to connect the low-pressure side of a differential pressure transducer (Freescale Semiconductor MPX2050) to the central tube of the screen. The high-pressure side of the sensor was exposed to the room atmosphere. The sensor output voltage was amplified by an instrumentation amplifier (INA126) and recorded in a personal computer by a multifunction USB device (National Instrument USB-6008). By using a silicone tube filled with water, the sensor could be calibrated by well-defined hydrostatic water columns. The sensor sensitivity was found to be 0.109 V/kPa. Estimated uncertainty based on sensor data was a maximum error of less than ±0.2 kPa in the range 0-50 kPa. A small control group of six healthy adults were recruited on a voluntary basis for this test. The measurements were carried out as described previously with three measurements for each of the two instructions at the same (one) occasion. The intraoral negative pressure was recorded during the lip force measurement with suction, and the maximum value of the differential pressure was selected for further analysis.

| RESULTS
The labelling and projected area of the different screens used is presented in Table 2. The area of the soft screen was initially 10.9 cm 2 and reduced to 8.3 cm 2 when it was subjected to a load of the same magnitude as the average lip force (18-25 N). The pressure was subsequently calculated with the reduced area. The error in area measurement for all screens is estimated to maximum 5%.
When evaluated as an OSP, the measurements for the different screens could be compared. In Table 2, an overall picture is presented for the instruction "Squeeze the lips as firmly as you can." Here, the overall mean value for each screen is based on 174 measurements.
We can see that the mean values are grouped in two, one value around 22.9 kPa for Screens A and B and another lower value around 15.6 kPa for Screens C and D. However, with the other instruction "Suck the screen as hard as you can," there is a different result. The holes in Screens C and D are now sealed with wax allowing suction. In Table 3, we can see that the mean values are grouped in two, but now the Screens A and B are giving the lower group value.
In Figure 2, the same data set as in Tables 2 and 3 is presented  The results in Figure 2 indicate that when suction is possible as for Screens A and B, there is a possibility that some suction is executed even during the instruction "Squeeze." In order to check the influence of suction on OSP measurements, the data were further processed. In Figure 3, the additional OSP from   suction P Su+ is evaluated as the difference between measured OSP at instruction "Suck" P Su and OSP measured at instruction "Squeeze" P Sq .
The two clusters can still be identified, but there is now a significant difference between screens with and without a hole preventing suction. For Screens C and D, there is almost no correlation between P Su+ and P Sq . The trend line for Screen C is almost horizontal with a coefficient of determination r 2 = 0.0024 indicating that the parameters P Su+ and P Sq are practically uncorrelated. On the other hand, with Screens A and B, it is not possible to extract the contribution from suction. The trend line for Screen B has a negative slope with a much higher coefficient of determination of r 2 = 0.264 indicating that suction may be present with the squeeze instruction.
The results from Screens C and D motivate the definition of two independent OSP parameters P Su+ and P Sq . These parameters are summarized in Table 4 based on measurements from Screen C. Both parameters may well be normally distributed because the p values in the Shapiro-Wilk tests were greater than 0.05. In Figure 4, individual means for the test group is shown together with estimated confidence ellipses around the overall mean. These regions correspond to 50% and 95%, respectively, of a healthy control group.
It seems natural to investigate in what respect the new parameter P Su+ is related to the intraoral pressure that is present during suction. In Figure 5, the result from the pilot experiment is shown where the measured maximum negative differential pressure P Vac is plotted against the OSP parameter P Su+ . It was found that there seems to be a linear relation between P Vac and P Su+ , but in absolute  Same data as in Figure 2. The additional oral screen pressure from suction P su+ is evaluated as the difference between P su (oral screen pressure "Suck") and P sq (oral screen pressure "Squeeze"). The different screens were Screen A (□), Screen B (○), Screen C (■), and Screen D (•). Solid line is the trendline for Screen C (r 2 = 0.0024), and dashed line is trendline for Screen B (r 2 = 0.264) Properties of individual oral screen pressure means for the control group (N = 29) and Screen C divided into two independent parameters: Oral screen pressure from squeezing (P Sq ) and additional oral screen pressure from sucking (P Su+ )  In previous studies, prefabricated screens have been used, but no discussion has been made about how the possibility to mix "Suck" and "Squeeze" affects the measurement (Hägg & Tibbling, n.d.;Hägg et al., 2008;Sjögreen et al., 2011).

| Oral screen design
The projected area for the soft oral screen decreased with 25% under load with the same forces that were measured with the testing subjects. This means that the active area is substantially smaller than the projected and the measured value of screen force might thus be smaller compared to a hard screen of similar size. This aspect has not been taken into account in studies performed with soft prefabricated oral screens (Sjögreen et al., 2007;Sjögreen et al., 2011).
When measuring lip force using prefabricated oral screens without a hole to let air in, it becomes difficult for the patient to execute the instruction "Squeeze." However, when using a semiindividual screen with a tube allowing air to pass, it is likely to get clearly separated values for the abilities squeezing and sucking.
The screen design is obviously of great importance for the outcome.
No studies have been found where oral screens with a similar design have been used.

| Impact on test value
A problem when studying a normal population is that in healthy adults, the force when sucking is so high that the measuring procedure might be interrupted because of pain from the mucosa. This may result in a poorer performance in the first but mainly in the second session, the subjects having the experience of the first measurement in mind. In this study, the measured values for Screen D showed this phenomenon. Moreover, with Screen D, there was a great variation between the pressures performed by different healthy individuals compared to Screen C. This indicates some kind of impact on the measured values but no further analysis of this observation has been done in this study.
However, a person with impaired oral motor function will probably never reach these high values.

| Cut off value
To evaluate oral motor function in patients, it is desirable to have a cut off value for lip force that is independent of the equipment used. To make the value useful, it should be comparable to values from other studies. Therefore, a value of OSP expressed in kPa is preferable. In this study, we have proposed two independent parameters of oral screen pressure. One is related to the mechanical strength in the perioral muscles, P Sq , and one is also related to suction, P Su+ . It is essential to have threshold values for both these parameters. A plot of longitudinal data in a two-dimensional diagram such as Figure 5 could be a useful tool to diagnose and analyse therapeutic effects in patients. In other studies, values have been expressed in Newton, and no reflections have been made that values might differ with the size of the screen (Hägg & Tibbling, n.d.;Hägg et al., 2008;Sjögreen et al., 2007;Sjögreen et al., 2011).

| Oral cavity pressure
The additional OSP as defined in this work is not identical to the vacuum in the oral cavity resulting from suction. One possible explanation could be how much tissue area actually is in contact with the screen during suction. The mechanism of how these parameters are related is not the subject of this work and should be interesting to study in future works. Santander et al. (2013) showed that the intraoral compartment pressures mainly are negative when swallowing (Santander et al., 2013). When examining patients with swallowing impairment, it is important to be able to diagnose if there is a problem creating negative FIGURE 5 Pilot experiment showing the relation between the negative pressure inside the mouth P Vac and the additional oral screen pressure from suction P Su+ as measured from the lip force measurements. Mean values from three measuements are used for all data points intraoral pressure and/or to execute enough pressure in the perioral musculature. This study shows that when using a screen with a tube allowing air to pass, it is possible to get clearly separated values for the abilities squeezing and sucking. The significance of this opportunity has to be further investigated in future studies on patients with impaired oral motor function.

| Diagnostic possibilities
5 | CONCLUSIONS 1. The instruction: a. To suck generally increases the measured lip force.
b. The instruction is an important mean to get the correct measurements to distinguish between squeezing and sucking.
c. Measurements with the prefabricated screens indicate that the subjects mix squeezing and sucking in spite of the instructions.
2. The design of the oral screen: a. The lip force increases in relation to the projected area of the screen.
b. It is possible to avoid suction when squeezing if the screen is designed with a tube allowing air to pass.
3. The additional screen pressure when sucking is related to the ability to produce a negative intraoral pressure.