The effectiveness and safety of mouthguards depend on the material and thickness of the sheet [1] [2] [3]. To counter the stress and strain generated during impacts, it is necessary to provide a sufficient thickness of 3 - 4 mm on the labial and buccal sides of the mouthguard [4] [5]. For single-sheet mouthguards, a 4.0-mm-thick sheet is often used. However, because the thickness is greatly reduced by thermoforming, it is difficult to secure the thickness required for impact absorption. Laminated mouthguards reliably provide a suitable thickness, but they may not be readily available to users due to the cost and fabricating time. Therefore, various fabrication methods have been investigated to secure the thickness after formation with a single sheet [6] [7] [8].

Mouthguard sheets are manufactured by extrusion or injection molding. Most of the sheets of ethylene-vinyl acetate resin (EVA), which is a mainstream product currently on the market, are manufactured by extrusion molding. Extruded sheets shrink in the extrusion direction as the strain introduced during manufacturing is released during thermoforming. This thermal shrinkage affects the mouthguard thickness depending on the sheet and model installation direction. In contrast, sheets manufactured by injection molding are not distorted during sheet manufacturing, and thus do not undergo thermal shrinkage during thermoforming [9]. Previously, the effect of the heating conditions and the shape of the sheet on the thermal shrinkage that occurs in the extruded sheet have been investigated [7].

There have been studies on the effect of model angles on the mouthguard thickness [10] [11]. These findings indicate that the presence of labial undercuts extends the mouthguard sheet and reduces labial thickness. For athletes with maxillary anterior teeth tilted to the labial side, trimming the model so that there is no undercut on the labial side tends to increase the model height. However, as the model height increases, the mouthguard becomes thinner [4] [12]. Athletes with a large overjet have a higher risk and severity of sports injuries [13] [14], indicating that the mouthguard thickness is particularly important for this type of model form. The present study was envisioned the fabrication of a mouthguard for athletes with maxillary anterior teeth tilted labially. Therefore, it was verified how much the amount of undercut in the model affected thermal shrinkage. The null hypothesis was that the thermal shrinkage of the sheet was not affected by the model form.

The working model was fabricated using a silicone rubber (Correcsil, Yamahachi Dental Mfg., Co., Aichi, Japan) impression taken from a maxillary dental model (D16FE-500A-QF, Nissin Dental Products Inc., Kyoto, Japan), into which dental gypsum (New Plastone, GC, Co., Tokyo, Japan) was poured [6] [7]. The plaster model was trimmed into the following forms using a model trimmer (MT-6, Morita, Co., Tokyo, Japan): 1) Model A, in which the angle of the model formed between the labial surface of the central incisor and the base of the working model was 90˚ with a height of 25 mm at the incisal edge of the maxillary central incisor and a height of 20 mm at the mesiobuccal cusp of the maxillary first molar (i.e., undercut amount on the labial side 0˚); 2) Model B, in which the angle was 10˚ greater than in Model A with a height of 25 mm at the incisal edge of the maxillary central incisor and a height of 25 mm at the mesiobuccal cusp of the maxillary first molar (i.e., undercut amount on the labial side 10˚); and 3) Model C, in which the angle was 20˚ greater than in Model A with a height of 25 mm at the incisal edge of the maxillary central incisor and a height of 30 mm at the mesiobuccal cusp of the maxillary first molar (i.e., undercut amount on the labial side 20˚) (Figure 1).

Mouthguards were thermoformed using EVA sheets (Sports Mouthguard, Keystone Dental Inc., Cherry Hill, NJ; 127 × 127 × 4.0 mm, clear) and a vacuum forming machine (Pro-form, T & S Dental & Plastics Co., Inc., Myerstown, PA). The sheet was mounted on the forming machine so that the sheet extrusion direction was vertical (condition V) or parallel (condition P) to the model’s centerline. The model position was 25 mm from the front of the forming unit. The sheet was softened until it sagged 15 mm, and then suction was continued for 30 s [6] [7]. The model was left in place for at least 24 h before the mouthguard was removed. Six specimens were formed for each condition; thus, a total of 36 mouthguards were fabricated (i.e., 2 extrusion direction × 3 model forms × 6 repetitions).

Mouthguard thickness was measured using a specialized caliper accurate to 0.1 mm (21-111, YDM, Co., Tokyo, Japan) without a spring, so as to prevent distortion during measurement [6] [7]. The measurement points were the left and right central incisors (10 points on the incisal edge and 20 points on the labial surface) and the first molars (8 points on the cusp and 20 points on the buccal surface) [6] [7]. The measurements were taken once for each specimen.

The differences in the reduction rate of the thickness due to model form and extrusion direction were analyzed using statistical analysis software (IBM SPSS 24.0, SPSS Japan Inc., Tokyo, Japan). The Shapiro-Wilk test for normality of distribution and Levene’s test for homogeneity of variance were also used. Each measurement exhibited normality and equal dispersion; accordingly, analysis was performed by two-way analysis of variance (ANOVA) and Bonferroni’s multiple comparison tests. All analytical methods were performed with a significance level of 5% and a detection power of 80%, and the difference was considered significant when both were satisfied.

Table 1 shows the results of two-way ANOVA for the mouthguard thickness after formation. At all measurement points, the main effects of the model form and extraction direction were significant, and their interaction was also significant. Based on the results, simple main effect tests were performed prior to multiple comparisons among levels.

Table 2 and Figure 2 show the results of multiple comparison analysis of the mouthguard thickness reduction. At all measurement sites, the rate of decrease in thickness tended to increase as the model angle increased. Differences in thickness depending on the extrusion direction of the sheet were observed in Models B and C on the labial surface and in all models on the buccal surface, and the thicknesses obtained under condition P were significantly thinner than those obtained under condition V. The thickness of the incisal edge and the cusp was not affected by the extrusion direction.

Thermal shrinkage of the mouthguard sheet occurs when the strain accumulated during manufacturing is released from the extruded sheet [9]. Because

df: degree of freedom; SS: sum of squares; MS: mean square; **P < 0.01: denotes statistically significant difference.