Differentiation of Feeding Behaviors Based on Masseter and Supra-Hyoid Muscle Activity

Abstract

Older adults with disorders of mastication and swallowing are often fed soft foods such as jelly or puree. The texture of such semi-solid foods allows them to be squeezed between the tongue and palate rather than being chewed. However, it is difficult to visually identify such strategies for the oral processing of food. This study aimed to test the hypothesis that there is a difference in the sequential coordination between the masseter and supra-hyoid muscles, and to identify feeding behaviors such as chewing and squeezing using electromyography. Seventeen male subjects (mean age: 30.8 years) were recruited. Four kinds of gels were prepared (two kinds of fracture force and fracture strain) as test samples. Subjects were instructed to consume the gels in three ways: squeezing with the tongue, chewing with the teeth and eating freely until swallowing. The amount of squeezing/chewing and the consumption time was unlimited. The masseter and supra-hyoid muscle activity were recorded during the entire consumption time and videofluorography was simultaneously recorded during each ingestion. Lissajous figures were made from the electromyographic activity of the two groups of muscles during the first stroke, and a regression line was made to determine the gradient of each figure to compare squeezing and chewing using the Mann–Whitney U-test. The masseter and supra-hyoid muscles were active simultaneously during squeezing with the tongue. However, the masseter was active after the supra-hyoid during chewing. The gradient of the regression line from the Lissajous figures between the masseter and supra-hyoid muscle activity was positive during squeezing, but negative during chewing. Analysis of the ROC curve showed that the cutoff value of the gradient for differentiating feeding behaviors was 0.097, with a sensitivity of 95.3% and specificity of 98.4%. When we allocated 68 free intakes into squeezing and chewing according to this cutoff value, we could distinguish with good precision, and the accuracy, sensitivity, and specificity were 86.8, 91.1, and 66.7% respectively. These results suggest that certain aspects of muscle activity differed among oral processing methods. Lissajous analysis of muscle activity was useful for identifying ingestion behaviors.