Original papers
Mastication Characteristics of Cooked Rice Prepared from High-quality Cultivars
2019 Volume 25 Issue 4 Pages 507-517
Details
2019 Volume 25 Issue 4 Pages 507-517
The objective of this study was to clarify whether the natural mastication behavior of Japanese who habitually consume rice changes among high-quality rice cultivars or not. A newly developed cultivar in Yamagata Prefecture, “Yukiwakamaru”, as well as “Haenuki”, “Tsuyahime”, and “Koshihikari” were selected. Rice samples were produced in Yamagata city in 2017, milled, cooked, and served under the same conditions. The cooked rice samples had similar amylose and protein contents, moisture, and specific volume. Instrumentally measured textural parameters of a grain of cooked rice were also similar when grain size was considered. Electromyography (EMG) from both masseter and suprahyoid muscles was recorded during natural mastication of nine-g rice samples by eleven healthy subjects. More than fifty EMG variables were examined and none differed significantly among cultivars. The results suggest that the texture of cooked rice prepared from high-quality cultivars is similar and does not significantly influence the natural mastication behavior.
Japonica type rice (Oryza sativa L.) is widely consumed in Japan as a staple food. The texture of cooked rice contributes more to palatability than does the flavor (Matsumoto and Matsumoto, 1977) and is an important factor in determining the quality of paddy rice cultivars (Okadome, 2005; Suzuki et al., 2006; Champagne et al., 2010). Most of the cooked rice preferred by the Japanese has a good textural balance of hardness and stickiness. Such rice is made from premium Japonica cultivars harvested within the year, with the white rice milled at 90% (w/w) yield and without long-term storage after milling (Food Agency, 1968; Okadome, 2005; Suzuki et al., 2006; Kohyama et al., 2014). Many Japanese have their preferred method to prepare cooked rice for every meal. Typical cooked rice has an approximate moisture content around 60% (w/w) (Kohyama et al., 2014). Rice samples are served shortly after cooking to maintain a good texture with minimal staling (Otahara et al., 2018).
Various evaluation methods for cooked rice have been developed. Sensory evaluation procedures of cooked rice have been presented (Champagne et al., 2010). In Japan, a standard method of the Food Agency has been often used. As textural attributes, hardness and stickiness are involved (Food Agency, 1968). Instrumental measurements of cooked rice texture are also widely conducted. Okabe (1979) proposed the two-bite test using a General Food Texturometer. The modified two-bite test using an Instron-type, universal testing machine has been more common, as parameters obtained from the test where the downward and upward movements are at a constant speed are more meaningful (Bourne, 1976; Nishinari et al., 2013). Okadome et al. (1999) developed multiple measurements, i.e., low at 25% compression (LC) test followed by high at 90% compression (HC) test, and a continuous progressive compression (CPC) test of a single grain of cooked rice using a Tensipresser®. The first bite (LC) examined the physical properties of the surface of the rice grain and the second bite (HC) measured those for the entire grain (Okadome et al., 1999). Parameters such as hardness, stickiness or adhesiveness correlated well with sensory assessments of Japanese rice cultivars (Okadome et al., 1999; Okadome 2005; Suzuki et al., 2006; Kohyama et al., 2016a). Japanese consumers prefer rice with a high ratio of stickiness to hardness, termed the degree of balance (Okademe et al., 1999; Okadome, 2005).
It is known that the amylose content of starch, protein content, and moisture of cooked rice influence the texture of cooked rice. The texture becomes harder and less sticky, and the rate of hardening increases with the amylose content and decreases with moisture content because of starch retrogradation (Okadome et al., 1999; Kohyama et al., 1998; 2014; 2016a; Okadome, 2005; Otahara et al., 2018). The protein content becomes high with nitrogenous fertilizers and increases the surface hardness of a cooked rice grain (Okadome, 2005).
More recently, electromyography (EMG) of masticatory muscles of humans has been introduced as a physiological method to quantitatively and objectively analyze textural changes during oral processing of food (Chen, 2009; Funami et al., 2014; Kohyama 2015). Surface EMG of masseter muscles, a typical jaw-closing muscle, is most common to study mastication behavior, and that of temporal muscles and/or jaw-opening muscles is sometimes utilized (Funami et al., 2014). In previous EMG studies, cooked rice was compared with different kinds of foods (Shiono et al., 1986; Sakamoto et al., 1989; Kohyama et al., 2002; Kohyama and Mioche, 2004; Shiozawa et al., 2013). A common observation from these studies is that cooked rice was consumed with less mastication effort or muscle work than the other foods, suggesting that cooked rice is a relatively soft and easy food to consume.
To the best of our knowledge, Kohyama et al. (1998) first reported an EMG study on the texture of cooked rice of various cultivars in which the rice samples had different textures. Since then, cooked rice samples of various cultivars (Kohyama et al., 2016a; Takahashi and Kawamura, 2011) under different cooking conditions (Kohyama et al., 2005; 2014; Nakayama and Kohyama, 2004; Kohyama, 2016a), and prepared from raw rice with different degrees of milling (Kohyama et al., 2014; Kohyama, 2016a; 2016b) have been studied in relation to texture. Human subjects masticate harder rice for a longer period and with a greater number of chews. The mastication effort, as estimated by the total EMG activities of the masseter muscles for the entire time of oral processing, was greater in such harder samples (Kohyama et al., 2005; 2014). The addition of starch paste to cooked rice enhanced the adhesiveness while maintaining a similar hardness; the more adhesive sample of model rice was masticated more and accompanied by a significant increase in jaw-opening muscle activities (Shiozawa et al., 2003). Jaw-opening muscle activities or the ratio of jaw-opening to -closing muscle activities increased with increased adhesiveness of rice samples (Kohyama et al., 1998; 2005). These EMG studies have been designed for obviously different rice samples with greater differences in chemical components such as moisture or amylose content and/or physical properties such as hardness or adhesiveness.
Japanese policy overseeing rice production was changed to deregulate rice production beginning in 2018i). Many breeders in each area had developed new cultivars in the preceding yearsii). Yamagata Prefecture developed “Yukiwakamaru (Yamagata112)”, which exhibited good cooked rice texture (Chuba et al., 2016), and its characteristics are also presented in the databaseii).
The objective of this EMG study was to clarify whether mastication characteristics among similar rice cultivars with high eating-qualities are alike or not. Four cultivars known as high-quality cooked rice, with similar amylose and protein contents, and cultivation, storage and milling conditions were compared. In this study, rice samples were prepared simultaneously under common conditions and served to healthy subjects in a randomized order. As cooked rice is consumed in large amounts as a staple food in Japan, mastication behavior of cooked rice may affect the overall eating behavior of every meal.
Rice Samples The four non-glutinous rice cultivars “Yukiwakamaru”, “Haenuki”, “Tsuyahime”, and “Koshihikari” were selected as they exhibit high eating-qualities (Chuba et al., 2016). General characteristics of these cultivars are available from the Institute of Crop Science, NAROii). They were sown in April 2017 and transplanted to a paddy field at 22.2 planting positions/m2 with four plants per position in mid-May (Goto et al., 2014). The fertilization condition was adjusted to be 0.6 nitrogen kg/a as it influences the protein content of rice grains (Okadome, 2005). Grains were harvested at maturity in mid-September for Haenuki and Yukiwakamaru or in late September for Tsuyahime and Koshihikari. As the standard for sensory evaluation and EMG studies, “Haenuki” cultivar produced at the Rice Breeding and Crop Science Experiment Station of Yamagata Integrated Agricultural Research Center (Tsuruoka city, Yamagata Prefecture) in the same year was used as reported previously by Tsujii et al. (2015). The brown rice grains were sieved to remove grains smaller than 1.9 mm (Goto et al., 2014). The crude protein and apparent amylose content of the brown rice were determined by a near infrared analyzer (Infratec® 1241; Foss, Hillerød, Denmark). The data are shown in Table 1.
| Cultivar | Standard (Haenuki) | Yukiwakamaru | Haenuki | Tsuyahime | Koshihikari | |
|---|---|---|---|---|---|---|
| Apparent amylosea (% w/w) | 19.7 | 21.7 | 20.6 | 21.6 | 20.5 | |
| Crude proteina (% w/w) | 7.0 | 6.8 | 7.0 | 6.3 | 6.4 | |
| Ratio of cooked rice to milled rice grainsb (w/w) | 2.29 | 2.29 | 2.28 | 2.29 | 2.29 | |
| Sensory evaluationc | Yukiwakamaru | Haenuki | Tsuyahime | Koshihikari | pd | |
| Hardness | - | 0.22 | 0.09 | −0.21 | −0.32* | 0.085 |
| Stickiness | - | 0.22 | 0.03 | 0.04 | −0.11 | 0.390 |
| Appearance | - | 0.38* | 0.09 | 0.36** | −0.04 | 0.067 |
| Aroma | - | 0.03 | 0.09 | 0.00 | −0.11 | 0.230 |
| Taste | - | 0.16 | 0.00 | 0.11 | −0.14 | 0.279 |
| Overall characteristics | - | 0.25* | −0.03 | 0.21 | −0.18* | 0.067 |
The brown rice grains were milled to white rice with a yield of 90.0 ± 0.5% (w/w) using a rice-polishing machine (VP-30T; Yamamoto Seisakusho, Chiba, Japan) (Goto et al., 2014) and the white rice samples were used within a week. Six hundred grams of white rice was cooked with 878 g of water according to the Food Agency guidelines (1968) using an IH rice-cooker (SR-HD103; Panasonic, Osaka, Japan) (Goto et al., 2014). Five samples including the control were finished cooking at the same time. Cooked rice in the vessel was weighed and mixed after standing for 10 min. The weight ratio of the cooked rice to milled rice (600 g) was calculated (Table 1). The other measurements were conducted between 0.5 to 4 h after cooking.
An expert panel of Yamagata Integrated Agricultural Research Center evaluated the cooked rice according to the Food Agency guidelines (1968). Overall characteristics, appearance, aroma, taste of cooked rice were evaluated with a 7-point scale from −3 (very bad), −2 (bad), −1 (slightly bad), 0 (similar), +1 (slightly good), +2 (good) to +3 (very good) compared to the standard rice with 0 score. In a similar manner, hardness was scored from −3 (very soft) to +3 (very hard), and stickiness was −3 (very weak) to +3 (very strong). The scores from 32 (Yukiwakamaru and Haenuki) or 28 (Tsuyahime and Koshihikari) panelists and the differences from the standard analyzed by a t-test are shown in Table 1.
Instrumental Characterization A modified two-bite test of a single grain was conducted using a Tensipresser® (My Boy System; Taketomo Electric Inc., Tokyo, Japan) (Okadome et al., 1999) as in our previous studies (Kohyama et al., 2014; 2016a). The plunger rate was set to 2.0 mm/s before touching the rice grain to determine the initial thickness. A grain of cooked rice was first compressed to 25% of its initial thickness and pulled and then re-compressed to 90% and pulled at a constant rate of 6.0 mm/s. Twenty grains from each sample were tested.
We selected only a few parameters from the force–time curve as follows: hardness values H1 and H2 (the peak force at the first and second compressions, respectively), stickiness values S1 and S2 (the negative peak force during pulling after the first and second compressions, respectively), and the degree of balance (S1/H1 and S2/H2, respectively).
Electromyography (EMG) This study was approved by the Food Research Institute Ethics Committee (No. 29NFRI-0005). Eleven subjects (6 men and 5 women, aged from 23 to 60 years) were selected from the sensory panel considering the gender and age balance, since EMG values during mastication have been reported to differ by sex (Nagasawa et al., 1997; Tamura and Shiga, 2014) and age (Kohyama et al., 2002; Kohyama and Mioche, 2004; Woda et al., 2006). None of the subjects suffered from masticatory problems.
EMG was recorded from the left and right masseter muscles (LM and RM) as jaw-closing muscles and suprahyoid muscles (SH) as jaw-opening muscle using bipolar surface electrodes (EL503; Biopac Systems Inc., Goleta, USA) (Kohyama, 2016b). As chewing side does not significantly influence the activities of jaw-opening muscles (Kohyama et al., 2005), the SH electrodes were put on the midline between the hyoid bone and chin. A ground electrode (EL503) was put on the wrist. At first, tapping, clenching and swallowing of a small amount of water were performed to check the recording condition (Funami et al., 2014). Next, a piece of chewing gum (Xylitol, Lotte, Tokyo, Japan) was chewed freely, and then the softened chewing gum was masticated at a fixed side for about 15 cycles and then with the other side following instruction by the experimenter (Fig. 1) (Kohyama, 2019). The chewing side order was randomized among subjects. These procedures helped to verify the chewing pattern of each subject. It was confirmed that the masseter activities were greater in the chewing side. The mean of more than 10 cycles (disregarding the first and last cycles) was used as the standard of masseter EMG for each chewing side.
Example of an electromyogram during chewing gum with right side and then left side.
Electromyography (EMG) from the right masseter (RM), left masseter (LM), and suprahyoid (SH) muscles. Subject is a man aged 23 years. Arrowheads at the top are put by the experimenter during the recording session when chewing method is instructed to the subject. Mean of the peak-to-peak amplitudes and muscle activities as the time-integral of EMG output in the chewing side is calculated using more than 10 cycles in the middle part for each side. They are used as the standard of amplitude and muscle activity for each subject and each side.
After one or two rice samples (9 g) was employed as practice, 9 g of rice samples including the control were served to the subjects on a plastic spoon (Kohyama, 2016a; 2016b). The subjects and recording operator were blinded to the identity of each rice sample served in a randomized order. The subjects were allowed to masticate the rice naturally and freely change the chewing side, as imposed chewing declined mastication performance (Mioche et al., 1999). The subjects were asked to press a button of the hand switch (TSD116A; Biopac Systems Inc.) using the hand without the ground electrode at every swallow, and to press it longer at the last swallow to indicate finishing (Kohyama, 2016b). Figure 2 is an example of the EMG of a rice sample. A mastication trial with 5 rice samples was performed continuously within 6 min; however, the entire EMG recording session from instruction to removal of electrodes lasted for approximately 30 min. The 10 subjects performed two series of trial, while one female performed only one.
Example of an electromyogram during natural mastication of cooked rice.
The same subject as Fig. 1. The bottom channel is output from a button switch pushed by subject at swallowing (in this case, two swallows). The last swallow is indicated by a longer signal. Arrowheads at the top are put for indicating the interested period by the experimenter after measurement. The period between the start and end signs is analyzed.
The EMG signals were filtered (10–500 Hz) and removed of 50-Hz noise from the power supply, and then amplified 1 000 times using three EMG 100C amplifiers (Biopac Systems Inc.). EMG signals and the switch output were saved on a PC using an MP-150 system (Biopac Systems Inc.) at 2,000 Hz using the supplied software (AcqKnowledge® ver. 4.4.2; Biopac Systems Inc.). For each cycle of each muscle four variables were determined: burst duration, peak-to-peak amplitude, muscle activity, which was estimated as the time integral of the EMG voltage, and cycle time (Funami et al., 2014; Kohyama et al., 2016b; Kohyama, 2019).
We attempted to analyze as many EMG variables as possible: the number of swallows by button switch (V1, EMG variables are written as V** hereafter), number of chews counted as masseter bursts (V2), number of chews before the first swallow (V3), time for oral processing (TOP, V4) as indicated by arrowheads from start to end in Fig. 2, total muscle activities or integrated EMG during TOP for each muscle (V5–V7). The 4 variables (the duration, amplitude, muscle activity, and cycle time) were averaged for both masseter muscles because the on-off timing was closely similar. They were averaged for the total number of chews: mean cycle time (V8), mean EMG durations (V9), mean amplitudes (V10), and mean muscle activities per chew (V11). Sum of the durations and muscle activities were also calculated for TOP (V12 and V13).
As a new trial in this study, masseter amplitudes and muscle activities were standardized using EMG of chewing gum. The averaged amplitudes and muscle activities of chewing gum of the fixed chewing side (Fig. 1) were used as the standard values. The side exhibiting greater standardized values was assumed to be the chewing side of rice for each chewing cycle. The mean values of amplitudes, muscle activities and sum of muscle activities (V14, V15, and V16, respectively) were obtained for the chewing side that may change during mastication.
Variables for EMGs from the suprahyoid muscles (V17–V21) were calculated the same as the variables V9–V13 from the masseters. The cycle time based on the suprahyoid EMG was not shown because it becomes similar to cycle time based on the masseter EMGs (V8). Similar to the degree of balance in the instrumental test, the ratio of suprahyoid amplitude to the preceding masseter amplitude (V22), and that for muscle activities (V23) were calculated (Kohyama et al., 1998).
Changes in these EMG variables during oral processing are of interest (Kohyama et al., 1998; 2014; 2016a). The variables V8–V21 before the first swallow were also calculated (V8–V21 with italicized numbers), assuming that 9 g of rice was in the oral cavity (Kohyama, 2016a; 2016b; Kohyama et al., 2016b). The ratios of those variables before the first swallow to those for TOP were then calculated. The ratios for the number of chews (V3/V2), mean masseter activities (V13/V13), standardized masseter activities by chewing gum (V16/V16), and suprahyoid activities (V21/V21) are presented later as (V24–V27) in Table 3 (see Results and Discussion section).
| Electromyography variables | Standard (Haenuki)a | Relative values to the standardb | Two-way ANOVA | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | Min | Median | Max | Yukiwakamaru | Haenuki | Tsuyahime | Koshihikari | Fcultivar | p | Ftrial | p | Fcultivar×trial | p | ||
| V1 | Number of swallows | 2.52 | 1 | 2 | 6 | 0.98 | 1.01 | 1.04 | 1.04 | 0.509 | 0.678 | 0.416 | 0.931 | 0.449 | 0.988 |
| V2 | Number of chews | 37.1 | 11 | 40 | 65 | 1.01 | 1.07 | 1.04 | 1.02 | 0.257 | 0.856 | 3.252 | 0.004 | 0.667 | 0.874 |
| V3 | Number of chews before the first swallow | 25.2 | 6 | 23 | 44 | 1.02 | 1.06 | 1.06 | 1.00 | 0.180 | 0.910 | 3.008 | 0.006 | 0.944 | 0.560 |
| V4 | Time for oral processing (s) | 27.0 | 8.0 | 29.1 | 47.4 | 1.02 | 1.06 | 1.04 | 1.01 | 0.189 | 0.903 | 2.599 | 0.016 | 0.559 | 0.949 |
| V5 | Total muscle activities of right masseter muscle (mV.s) | 2.03 | 0.61 | 1.62 | 5.75 | 0.86 | 0.98 | 0.92 | 0.92 | 0.475 | 0.702 | 2.530 | 0.018 | 0.317 | 0.999 |
| V6 | Total muscle activities of left masseter muscle (mV.s) | 1.91 | 0.68 | 1.76 | 4.59 | 1.09 | 1.10 | 1.16 | 0.97 | 0.628 | 0.601 | 1.269 | 0.280 | 0.737 | 0.805 |
| V7 | Total muscle activities of suprahyoid muscles (mV.s) | 1.85 | 0.64 | 1.93 | 4.30 | 1.07 | 1.17 | 1.13 | 1.11 | 0.089 | 0.966 | 2.427 | 0.023 | 0.821 | 0.710 |
| V8 | Mean cycle time (s) | 0.715 | 0.541 | 0.702 | 0.917 | 1.01 | 1.01 | 1.01 | 1.02 | 0.143 | 0.934 | 1.018 | 0.446 | 0.479 | 0.981 |
| V9 | Mean EMG duration of masseter muscle per chew (s) | 0.333 | 0.246 | 0.340 | 0.433 | 0.98 | 1.01 | 1.01 | 0.99 | 0.700 | 0.558 | 4.322 | 0.000 | 0.296 | 1.000 |
| V10 | Mean peak-to-peak masseter amplitudes (mV) | 1.45 | 0.70 | 1.46 | 2.30 | 0.97 | 0.98 | 0.97 | 0.95 | 0.482 | 0.696 | 4.984 | 0.000 | 0.383 | 0.996 |
| V11 | Mean masseter muscle activities per chew (mV.s) | 0.0354 | 0.0186 | 0.0305 | 0.0564 | 0.96 | 0.99 | 0.96 | 0.94 | 0.852 | 0.474 | 4.489 | 0.000 | 0.439 | 0.989 |
| V12 | Sum of masseter muscle duration (s) | 12.14 | 3.52 | 12.07 | 21.14 | 0.99 | 1.08 | 1.06 | 0.91 | 1.882 | 0.148 | 2.151 | 0.042 | 1.282 | 0.229 |
| V13 | Sum of masseter muscle activities (mV.s) | 2.49 | 1.09 | 2.44 | 4.38 | 0.96 | 1.06 | 1.01 | 0.97 | 0.749 | 0.530 | 3.356 | 0.003 | 0.507 | 0.972 |
| V14 | Masseter amplitudes standardized by chewing gum | 0.747 | 0.447 | 0.742 | 1.114 | 0.97 | 0.97 | 0.98 | 0.97 | 0.040 | 0.989 | 4.414 | 0.000 | 0.216 | 1.000 |
| V15 | Masseter muscle activities standardized by chewing gum | 0.674 | 0.429 | 0.673 | 1.167 | 0.95 | 0.98 | 0.96 | 0.95 | 0.309 | 0.819 | 4.231 | 0.000 | 0.309 | 0.999 |
| V16 | Sum of masseter muscle activities standardized by chewing gum | 24.45 | 9.19 | 23.24 | 44.97 | 0.96 | 1.05 | 1.01 | 0.99 | 0.388 | 0.762 | 3.233 | 0.004 | 0.316 | 0.999 |
| V17 | Mean EMG duration of suprahyoid muscles per chew (s) | 0.434 | 0.327 | 0.442 | 0.646 | 1.03 | 0.98 | 1.00 | 1.01 | 0.477 | 0.700 | 1.792 | 0.094 | 0.546 | 0.956 |
| V18 | Mean peak-to-peak suprahyoid amplitudes (mV) | 0.31 | 0.10 | 0.26 | 1.07 | 1.03 | 1.03 | 1.04 | 1.02 | 0.232 | 0.874 | 4.656 | 0.000 | 0.845 | 0.681 |
| V19 | Mean suprahyoid muscle activities per chew (mV.s) | 0.0108 | 0.0035 | 0.0079 | 0.0431 | 1.03 | 0.99 | 1.01 | 0.98 | 0.189 | 0.903 | 1.064 | 0.412 | 0.210 | 1.000 |
| V20 | Sum of suprahyoid muscle duration (s) | 15.54 | 4.87 | 14.85 | 28.90 | 1.05 | 1.03 | 1.04 | 0.91 | 0.962 | 0.420 | 0.636 | 0.774 | 0.697 | 0.847 |
| V21 | Sum of suprahyoid muscle activities (mV.s) | 0.34 | 0.11 | 0.31 | 1.08 | 1.05 | 1.04 | 1.05 | 1.00 | 0.144 | 0.933 | 1.513 | 0.171 | 0.241 | 1.000 |
| V22 | Ratio of suprahyoid amplitude to the preceded masseter amplitudes | 0.266 | 0.053 | 0.176 | 0.866 | 1.11 | 1.12 | 1.12 | 1.14 | 0.256 | 0.857 | 8.599 | 0.000 | 1.030 | 0.459 |
| V23 | Ratio of suprahyoid activities to the preceded masseter activities | 0.337 | 0.093 | 0.275 | 1.137 | 1.13 | 1.04 | 1.11 | 1.09 | 0.464 | 0.709 | 3.264 | 0.004 | 0.328 | 0.999 |
| V24 | Ratio of number of chews before the first chew to all | 0.70 | 0.32 | 0.74 | 0.98 | 1.00 | 0.99 | 1.01 | 0.98 | 0.257 | 0.856 | 1.626 | 0.134 | 1.068 | 0.418 |
| V25 | Ratio of mean masseter activities before the first chew to all | 0.70 | 0.33 | 0.71 | 0.97 | 1.03 | 1.01 | 1.04 | 0.99 | 0.558 | 0.646 | 2.102 | 0.047 | 1.075 | 0.411 |
| V26 | Ratio of standardized masseter activities before the first chew to all | 0.71 | 0.32 | 0.74 | 0.97 | 1.02 | 1.00 | 1.03 | 0.99 | 0.390 | 0.761 | 2.092 | 0.048 | 1.080 | 0.405 |
| V27 | Ratio of suprahyoid activities before the first chew to all | 0.63 | 0.33 | 0.65 | 0.99 | 1.01 | 1.00 | 1.02 | 0.97 | 0.348 | 0.791 | 2.600 | 0.016 | 1.239 | 0.260 |
A previous study (Kohyama et al., 1998) revealed that differences among rice cultivars in some EMG variables were significant only at the early stage of mastication. In the early stage, the rice bolus is less scattered and chewed mainly using a single side. The earliest continuous three cycles, except the first chew, that were chewed by a single side without changing the chewing side were analyzed as variables (V28–V40) and are shown in Table 4.
| Electromyography variables | Standard (Haenuki)a | Relative values to the standardb | Two-way ANOVA | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | Min | Median | Max | Yukiwakamaru | Haenuki | Tsuyahime | Koshihikari | Fcultivar | p | Ftrial | p | Fcultivar×trial | p | ||
| V28 | Mean cycle time (s) | 0.687 | 0.512 | 0.648 | 1.294 | 1.00 | 1.02 | 1.04 | 1.04 | 0.51 | 0.674 | 2.39 | 0.011 | 1.03 | 0.433 |
| V29 | Mean EMG duration of masseter muscle per chew (s) | 0.311 | 0.221 | 0.287 | 0.481 | 1.02 | 1.06 | 1.06 | 1.04 | 0.31 | 0.821 | 11.02 | 0.000 | 1.06 | 0.394 |
| V30 | Mean peak-to-peak masseter amplitudes (mV) | 1.550 | 0.693 | 1.607 | 2.977 | 1.05 | 1.02 | 0.98 | 0.95 | 1.75 | 0.158 | 8.36 | 0.000 | 1.31 | 0.143 |
| V31 | Mean masseter muscle activities per chew (mV.s) | 0.039 | 0.020 | 0.037 | 0.065 | 1.01 | 1.00 | 1.02 | 0.95 | 1.62 | 0.187 | 12.00 | 0.000 | 1.66 | 0.022 |
| V32 | Masseter amplitudes standardized by chewing gum | 0.778 | 0.397 | 0.716 | 1.399 | 1.01 | 1.00 | 1.00 | 0.97 | 0.36 | 0.783 | 6.96 | 0.000 | 1.35 | 0.114 |
| V33 | Masseter muscle activities standardized by chewing gum | 0.764 | 0.503 | 0.690 | 1.401 | 0.95 | 0.92 | 0.97 | 0.92 | 1.12 | 0.342 | 10.79 | 0.000 | 1.38 | 0.103 |
| V34 | Mean EMG duration of suprahyoid muscle per chew (s) | 0.380 | 0.267 | 0.340 | 0.567 | 1.01 | 0.98 | 1.04 | 1.03 | 1.32 | 0.268 | 6.61 | 0.000 | 1.47 | 0.064 |
| V35 | Mean peak-to-peak suprahyoid amplitudes (mV) | 0.344 | 0.102 | 0.273 | 1.201 | 1.10 | 1.02 | 1.03 | 1.01 | 1.94 | 0.124 | 6.96 | 0.000 | 1.25 | 0.187 |
| V36 | Mean suprahyoid muscle activities per chew (mV.s) | 0.011 | 0.004 | 0.009 | 0.044 | 1.02 | 0.98 | 1.04 | 0.95 | 1.49 | 0.219 | 13.53 | 0.000 | 1.94 | 0.004 |
| Averaged right-left masseter amplitudes | |||||||||||||||
| V37 | Ratio of suprahyoid amplitude to the preceded masseter amplitudes | 0.266 | 0.056 | 0.156 | 0.819 | 1.15 | 1.11 | 1.14 | 1.21 | 1.51 | 0.214 | 17.92 | 0.000 | 1.57 | 0.038 |
| V38 | Ratio of suprahyoid activities to the preceded masseter activities | 0.301 | 0.106 | 0.210 | 1.000 | 1.04 | 1.02 | 1.04 | 1.66 | 0.23 | 0.875 | 10.94 | 0.000 | 1.05 | 0.408 |
| Standardized amplitudes by chewing gum | |||||||||||||||
| V39 | Ratio of suprahyoid amplitude to the preceded masseter amplitudes | 0.016 | 0.005 | 0.011 | 0.067 | 1.12 | 1.09 | 1.18 | 1.17 | 1.11 | 0.346 | 26.08 | 0.000 | 1.93 | 0.004 |
| V40 | Ratio of suprahyoid activities to the preceded masseter activities | 0.945 | 0.456 | 0.942 | 1.525 | 1.09 | 1.15 | 1.11 | 1.23 | 1.31 | 0.271 | 3.81 | 0.000 | 0.90 | 0.620 |
Statistics Statistical analyses were performed using IBM SPSS® Statistics (ver. 23; Armonk, NY, USA). A p value of < 0.05 was considered statistically significant. Tukey's honestly significant difference (HSD) test was conducted on the rice cultivars for instrumental parameters. Multiple-comparison among 4 cultivars was conducted using the Kruskal-Wallis test for independent samples for sensory scores.
The EMG variables varied greatly among subjects. EMG variables from the four cultivars were converted to relative values to the standard within a subject and a trial (total 21 trials) (Kohyama et al., 2016a). The relative values were subjected to a two-way analysis of variance (ANOVA) with 4 cultivars × 21 trials.
Physicochemical Properties of Rice Samples Table 2 shows the results of the two-bite test of cooked rice. None of these parameters differed significantly between the two Haenuki samples, and many of those for other cultivars showed similar values. Yukiwakamaru exhibited significantly greater thickness than the standard sample. Values for hardness at the first compression and stickiness at the second compression (H1 and S2) were greatest in Yukiwakamaru and the cultivar showed significantly greater values than the standard.
| Texture parameters | Standard (Haenuki) | Yukiwakamaru | Haenuki | Tsuyahime | Koshihikari |
|---|---|---|---|---|---|
| Initial thickness (mm) | 2.21a | 2.35b | 2.30ab | 2.24a | 2.29ab |
| Load at 25% compression, H1 (N) | 0.882ab | 1.057b | 0.828a | 0.823a | 0.884ab |
| Negative load after 25% compression, S1 (N) | 0.303a | 0.304a | 0.246a | 0.238a | 0.305a |
| Degree of balance for 25% compression, S1 / H1 | 0.345a | 0.288a | 0.304a | 0.289a | 0.343a |
| Load at 90% compression, H2 (N) | 20.7ab | 21.4ab | 20.6ab | 18.2a | 21.6b |
| Negative load after 90% compression, S2 (N) | 4.62a | 5.30b | 4.64a | 5.08ab | 4.79ab |
| Degree of balance for 90% compression, S2 / H2 | 0.231a | 0.256ab | 0.230a | 0.291b | 0.227a |
Mean values of 20 replicated measurements.
Values with similar alphabetical letter are not significantly different among cultivars (p > 0.05) determined by Tukey's HSD test.
Cultivar Yukiwakamaru is known to produce larger grains, as the weight of 1 000 grains of brown rice was 24.2 g while that for the standard (Haenuki) was 22.7 g (Chuba et al., 2016). High hardness value for one grain was likely due to the larger grain size of Yukiwakamaru. Because of time constraints, we could not measure the sectional area of each grain. Assuming that the grains were similar in shape but different in size, the sectional area contacting the probe of an instrument during the compression test would be proportional to the square of the initial thickness. When the load values were divided by the squared thickness, the standardized values showed less variance and Yukiwakamaru did not significantly differ from any samples (supplemental data, Table S1 in WEB).
| Texture parameters | Standard (Haenuki) | Yukiwakamaru | Haenuki | Tsuyahime | Koshihikari |
|---|---|---|---|---|---|
| Squared thickness (mm2) | 4.89a | 5.53b | 5.33ab | 5.02a | 5.25ab |
| H1 / squared thickness (N/mm2) | 0.182a | 0.196a | 0.159a | 0.165a | 0.169a |
| S1 / squared thickness (N/mm2) | 0.0632a | 0.0576a | 0.0481a | 0.0480a | 0.0584a |
| H2 / squared thickness (N/mm2) | 4.312a | 3.959a | 3.946a | 3.658a | 4.129a |
| S2 / squared thickness (N/mm2) | 0.949ab | 0.968ab | 0.885a | 1.019b | 0.916ab |
Mean values of 20 replicated measurements.
Values with similar alphabetical letter are not significantly different among cultivars (p > 0.05) determined by Tukey's HSD test.
As the degree of balance is the ratio of force, it is not influenced by individual grain size. The cultivar Tsuyahime showed a significantly higher value of the degree of balance at 90% compression. This is attributable to the lowest hardness and the second highest stickiness.
Masticatory EMG Analysis Table 3 shows the EMG results. In the first part at left, mean, minimum, median and maximum values for the standard sample are presented. All valuables showed large differences among the 21 trials. The ratio of maximum to minimum was more than 2 in most cases and exceeded 10 regarding the suprahyoid amplitudes (V18, V19, V22, and V23), whereas the cycle time showed less variation among the trials (0.541–0.917 s). The median and mean values were close for many variables, but suprahyoid muscle activities per chew (V19), and the ratio of it to the masseter (V22 and V23) were not, where the median values were smaller. Except for a few variables, variances of each trial were considered to be relatively close to normal distributions.
Most of the differences in trials were because of the individual mastication pattern of a subject, since two trials by the same subject showed values closer (data not shown) as stated in previous studies (Kohyama et al., 2014; 2016a). In some subjects, the second trial tended to show a longer TOP (V4) than the first. A similar tendency was observed in the time-related variables such as number of chews (V2) and sum of muscle activities (V5–V7, V12, V13, V16, V20, and V21).
The fullness felt by subjects showed little impact because ordinarily the Japanese eat cooked rice of around 150–200 g per meal, which it is much greater than the total rice samples (9 g × 12 times) used in the EMG recording. Monotonousness or fatigue from masticating similar rice samples continuously may be more likely to have an effect. The hardening of rice within a trial would be negligible. One trial with 5 rice cultivars was finished within 5 min because the TOP was less than 48 s. Some staling effects of cooked rice were involved in inter-subject differences because the EMG recording was conducted sequentially. It is evident that the first and last subjects of a day ate rice samples with a different texture. However, as the rice samples had similar amylose content and moisture, the staling kinetics would be similar for all samples (Otahara et al., 2018). The randomized order of sample service negates time effects in variances among rice cultivars within a trial
To reduce large individual differences, values relative to the standard rice were statistically analyzed. The center part of Table 3 presents the results. The relative values were all close to 1 (0.86–1.17).
The right part of Table 3 shows the results of two-way ANOVA. Cultivar effects were not significant with high probabilities for all the variables, where the smallest value was 0.148 for the sum of masseter muscle duration (V12). The high p-values suggest that the mastication behaviors of four cultivars were not significantly different. The effects of trials were significant for some variables (V2–V5, V7, V9–V16, V18, V22, V23, V25–V28), and the cross effects were not significant for all variables. Among the variables without significant trial effects, the number of swallows (V1), cycle time (V8), and sum of suprahyoid muscle duration (V20) exhibited high values of probability. This fact suggests that chewing rhythm and swallowing behaviors are similar among the trials performed by different subjects in comparison with chewing behaviors, which had a greater influence on masseter EMG amplitudes.
Unlike a previous study in which chewing was only on one side (Kohyama et al., 1998), in the present study the chewing side was changed freely. As the chewing side is unknown, and both sides contributed to the consumption of a mouthful amount of 9 g, the averaged values of the right and left sides were used as in other rice studies (Kohyama, 2016a; 2016b; Kohyama et al., 2015; 2014; 2016a). This relatively large amount (9 g) is suitable for the analysis of natural eating behavior of cooked rice; however, it is difficult to verify the chewing side from EMGs. As shown in Fig. 2, similar amplitudes of both masseters were observed in many chewing cycles. By standardizing using chewing gum, we could mathematically define the chewing side with greater relative amplitudes. In all cases, the chewing side changed more than once during TOP. The results of averaged masseter EMGs and standardized EMGs were similar, in that there were significant effects of trials but not cultivars or cross effects. Standardization using chewing gum was not effective for cooked rice because no new findings were observed with this standardization. As cooked rice consists of grains, it can be chewed with both sides in a cycle unlike chewing gum. Subjects could chew bilaterally, though this has not been observed for all cycles (Kohyama et al., 2016b).
Textural Changes during the Mastication Process Instrumental measurement of food texture is performed for samples before consumption (Kohyama, 2015); however, the texture changes greatly during mastication. Food textural parameters that appear during oral processing are not evaluated by common instrumental measurements (Chen, 2009; Kohyama, 2015). Shiozawa et al. (2005) showed the instrumentally measured hardness and adhesiveness values of cooked rice and the boli collected at the middle and late stages of mastication from the mouths of subjects. The values of hardness and adhesiveness of boli decreased with the mastication process, and these were much smaller than for the initial rice samples. As EMG is a typical physiological method to measure food texture, it can monitor these textural changes during oral processing (Chen, 2009; Funami et al., 2014; Kohyama, 2015).
EMG variables averaged or summed for the period before the first swallow showed a similar tendency to those for TOP (data not included in the printed article, but supplemental data Table S2 showing variables from V8 to V23 in WEB). The ratios of before the first swallow to TOP for the number of chews (V24), masseter activities (V25), standardized masseter activities (V26), and suprahyoid activities (V27) are shown in the end of Table 3. The ratio values of the four variables resembled each other and were highly similar between all cultivars.
| Electromyography variables | Standard (Haenuki)a | Relative values to the standardb | Two-way ANOVA | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | Min | Median | Max | Yukiwakamaru | Haenuki | Tsuyahime | Koshihikari | Fcultivar | p | Ftrial | p | Fcultivar×trial | p | ||
| V8' | Mean cycle time calculated by suprahyoid muscle (s) | 0.727 | 0.541 | 0.699 | 0.991 | 1.11 | 1.12 | 1.12 | 1.14 | 0.256 | 0.857 | 8.599 | 0.000 | 1.030 | 0.459 |
| EMG variables before the first swallow | |||||||||||||||
| V8 | Mean cycle time (s) | 0.678 | 0.532 | 0.649 | 0.893 | 1.02 | 1.01 | 1.00 | 1.02 | 0.218 | 0.883 | 1.042 | 0.428 | 0.445 | 0.988 |
| V9 | Mean EMG duration of masseter muscle per chew (s) | 0.323 | 0.245 | 0.312 | 0.408 | 1.00 | 1.02 | 1.00 | 1.00 | 0.259 | 0.855 | 4.280 | 0.000 | 0.223 | 1.000 |
| V10 | Mean peak-to-peak masseter amplitudes (mV) | 1.47 | 0.66 | 1.52 | 2.39 | 0.97 | 1.00 | 1.01 | 0.95 | 0.636 | 0.596 | 3.034 | 0.006 | 0.259 | 1.000 |
| V11 | Mean masseter muscle activities per chew (mV.s) | 0.0355 | 0.0173 | 0.0329 | 0.0624 | 0.98 | 1.02 | 0.99 | 0.94 | 1.130 | 0.349 | 2.886 | 0.008 | 0.282 | 1.000 |
| V12 | Sum of masseter muscle duration (s) | 8.07 | 1.74 | 7.09 | 16.60 | 1.01 | 1.08 | 1.06 | 1.00 | 0.273 | 0.844 | 2.731 | 0.012 | 0.742 | 0.801 |
| V13 | Sum of masseter muscle activities (mV.s) | 1.68 | 0.62 | 1.58 | 3.38 | 0.99 | 1.08 | 1.05 | 0.95 | 0.753 | 0.527 | 3.392 | 0.003 | 0.760 | 0.781 |
| V14 | Masseter amplitudes standardized by chewing gum | 0.750 | 0.429 | 0.762 | 1.171 | 0.96 | 0.99 | 1.01 | 0.98 | 0.219 | 0.883 | 2.919 | 0.008 | 0.150 | 1.000 |
| V15 | Masseter muscle activities standardized by chewing gum | 0.678 | 0.449 | 0.627 | 1.256 | 0.97 | 1.00 | 0.98 | 0.95 | 0.303 | 0.823 | 2.957 | 0.007 | 0.241 | 1.000 |
| V16 | Sum of masseter muscle activities standardized by chewing gum | 15.97 | 5.03 | 15.75 | 28.08 | 0.98 | 1.06 | 1.04 | 0.97 | 0.290 | 0.832 | 2.938 | 0.007 | 0.476 | 0.981 |
| V17 | Mean EMG duration of suprahyoid muscles per chew (s) | 0.387 | 0.303 | 0.383 | 0.545 | 1.05 | 1.00 | 1.02 | 1.01 | 0.436 | 0.729 | 1.417 | 0.208 | 0.573 | 0.942 |
| V18 | Mean peak-to-peak suprahyoid amplitudes (mV) | 0.30 | 0.09 | 0.27 | 0.97 | 1.05 | 1.03 | 1.05 | 1.03 | 0.391 | 0.760 | 3.321 | 0.003 | 0.882 | 0.636 |
| V19 | Mean suprahyoid muscle activities per chew (mV.s) | 0.0096 | 0.0033 | 0.0079 | 0.0371 | 1.04 | 0.99 | 1.01 | 0.97 | 0.444 | 0.723 | 1.810 | 0.090 | 0.372 | 0.997 |
| V20 | Sum of suprahyoid muscle duration (s) | 9.58 | 2.20 | 8.47 | 17.43 | 1.07 | 1.05 | 1.07 | 1.00 | 0.279 | 0.840 | 1.943 | 0.067 | 0.637 | 0.899 |
| V21 | Sum of suprahyoid muscle activities (mV.s) | 0.25 | 0.06 | 0.17 | 0.70 | 1.10 | 1.07 | 1.09 | 1.12 | 0.335 | 0.800 | 4.470 | 0.000 | 0.647 | 0.891 |
| V22 | Ratio of suprahyoid amplitude to the preceded masseter amplitudes | 0.288 | 0.099 | 0.235 | 0.758 | 1.07 | 1.01 | 1.06 | 1.06 | 0.488 | 0.693 | 4.278 | 0.000 | 0.335 | 0.999 |
| V23 | Ratio of suprahyoid activities to the preceded masseter activities | 0.218 | 0.052 | 0.182 | 0.705 | 1.05 | 1.05 | 1.07 | 0.97 | 0.452 | 0.718 | 3.474 | 0.002 | 0.486 | 0.979 |
A previous study (Kohyama et al., 1998) revealed that the differences among rice cultivars in some EMG variables were significant only the early stage of mastication. Jaw-closing and -opening muscles alternately act as shown in Figs. 1 and 2, the ratio of jaw-opening muscle activities to the preceded jaw-closing muscle activities differed significantly only in the early chewing stage, for example, the first 5 cycles.
With respect to EMG variables per chew from the different mastication stages, the early stage showed greater sample differences than the late stage (Kohyama et al., 2014; 2016a). In our previous studies using a smaller mouthful amount of rice (5 or 6 g), subjects swallowed the rice samples at one time in most of the cases (Kohyama et al., 1998; 2014; 2016). The mouthful size was set as 9 g in this study because the one bite amount of Koshihikari rice in natural eating is around 9 g (Kohyama, 2016a). More than one swallow was observed in most trials (Table 3, V1). We did not compare the early, middle and late stages of mastication, as a smaller amount of rice was masticated in later stages.
As the chewing side was not obvious from EMGs like Fig. 2, we utilized two methods in the calculation of masseter amplitudes and activities. The first method used the average values of both right and left masseter EMGs, and the second took the assumed chewing side that showed the greater standardized values by the chewing gum. The earliest continuous three cycles from the second cycle, for example, 2–4 cycles, 3–5 cycles, etc., that were chewed by a single side without changing the chewing side were analyzed.
Table 4 shows these results. All variables showed strong significant effects of trials but no cultivar effects. A previous EMG study of gels revealed that greater bite size and greater number of fragments after fracture are easily chewed by both sides (Kohyama et al., 2016). Cooked rice is composed of grains that can be easily masticated using both sides from the first chew. As shown in Figs. 1 and 2, the chewing side is not visually obvious in many cases. In addition, it was difficult to pick up information on the early stage under natural mastication conditions as the chewing-side was not fixed. In some cases, the side with greater amplitude and muscle activity changes often, so that the earliest cycles without side-changes were not in the early stage of mastication, such as 10–12 cycles (the greatest numbers were 16–18 cycles). The initial textural characteristics may have disappeared at these cycles.
The present rice samples exhibited highly similar characteristics. The texture likely changes in a similar manner during oral processing. The samples did not exhibit significant differences among cultivars in any of EMG variables compared at a given mastication process.
Discussion of Mastication Similarities for all Cultivars As for the instrumental and EMG results, a multiple comparison was performed for the sensory characteristics shown in Table 1. Significant differences among samples were not observed, though there were some differences in comparison with standard Haenuki. The subjects were blinded to the identity of the rice samples and were served rice on a spoon by an assistant. However, they were part of an expert sensory panel and may have been able to distinguish samples, especially Yukiwakamaru with its larger grains and Tsuyahime with its glossy surface appearance before mastication.
It has been reported that amylose content and rice hardness are positively correlated (Okadome et al., 1999; Okadome, 2005). Previous EMG studies using different rice cultivars were conducted for a wide range of amylose content, 1.8–29.2 or 0.4–33.3% (Kohyama et al., 1998; 2016a); in such cases, cultivars with a higher amylose content were significantly harder. As the ranges of amylose and protein contents were narrow, 19.7–21.7% and 6.3–7.0%, respectively, in this study (Table 1) and although the analytical methods were not the same, the textural differences were expected to be small.
Previous EMG studies using cultivars with widely varied amylose content showed significant differences in EMG variables, except for cycle time (Kohyama et al., 2016a). The present results, in which no significant cultivar effects were observed, were attributed to smaller differences in amylose content, which resulted in minimal differences among original rice samples.
Even though the previous study showed wider variations, EMG variables were significantly correlated with amylose content but were not significantly influenced by protein content (Kohyama et al., 2016a). As protein distributes more in the surface region, it influences surface hardness but not overall hardness (Okadome, 2005). This was consistent with the findings of Kohyama et al. (2008), in which mechanical properties under a greater deformation (> 50%) correlated with EMG variables but not under small deformations.
Each rice grain differs in size and grain size is one of characteristics of rice cultivars; therefore, the texture test using a single grain is useful for the characterization of cultivars. When cooked rice is consumed, a bite size is composed of many grains and is not a single grain taken in by the mouth and chewed between the upper and lower teeth. Mastication force is applied to a wider area than a single grain and a tooth, thus the grain size did not influence textural sensing to modify the mastication behavior (Kohyama, 2015). If the mastication EMG or sensory evaluation was performed using only a single grain of cooked rice, the effect of grain size may be detected. Measurement of the cross-sectional area of rice grains during the compression test is required in future study.
The texture of cooked rice prepared with four rice cultivars exhibiting high eating quality, Yukiwakamaru, Haenuki, Tsuyahime and Koshihikari, was compared. Rice samples were produced in Yamagata city in 2017, milled, cooked, and served under the same conditions. Moisture, amylose and protein contents, as well as hardness and stickiness as determined by sensory evaluation and instrumental tests were similar. EMG during natural mastication of 9-g rice samples revealed that the natural mastication behavior of individuals does not change among these similar cultivars.
Acknowledgments Part of the collaborative study on Quality Evaluation of Paddy-field, Non-glutinous Rice Cultivars for Staple Food based on the Physicochemical and Masticatory Properties of Cooked Rice by FRI-NARO, ICS-NARO, and Yamagata Prefecture in 2017.
