Dynamic concordance between subjective and facial EMG hedonic responses during the consumption of gel-type food

Abstract

The concordance between subjective and facial hedonic responses while eating is informative, both practically and theoretically. Recent psychophysiological studies reported that hedonic ratings during the consumption of gel-type food were negatively associated with facial electromyography (EMG) signals recorded from the corrugator supercilii and positively associated with those from chewing-related muscles. However, the relationships were tested in a static manner, and the dynamic subjective-facial concordance remains untested. Therefore, we investigated this by assessing participants' dynamic valence ratings and recording their facial EMG from the corrugator supercilii, zygomatic major, masseter, and suprahyoid muscles while they chewed and swallowed gel-type food stimuli of various flavors. Cross-correlations with dynamic valence ratings were negative for corrugator supercilii EMG signals and positive for zygomatic major, masseter, and suprahyoid EMG signals during both chewing and swallowing. These findings indicate that subjective hedonic experiences and facial EMG signals are dynamically coupled during the consumption of food.

Keywords: Facial electromyography (EMG); Gel-type food; Gustometer; Valence.

Graphical abstract

Fig. 1

Experimental setup used in the present study. Gel-type food stimuli were delivered via the mouthpiece to participants with facial electromyography electrodes. The participants chewed and swallowed the food stimuli according to instructions on a monitor and rated their momentary valence experience using a mouse. The individual depicted is an amateur model who did not participate in the experiment and provided informed consent for her image to be published in a scientific journal.

Fig. 2

Electrode placement for electromyography recording of the corrugator supercilii, zygomatic major, masseter, and suprahyoid muscles.

Fig. 3

Group mean (with standard error) waveforms of the subjective valence ratings and facial electromyography signals in response to generally positive and negative flavors and no flavor.

Fig. 4

Representative trial-based waveforms of subjective valence ratings and facial electromyography signals in response to generally positive and negative flavors and no flavor.

Fig. 5

Mean (with standard error) intra-individual cross-correlation coefficients between the dynamic valence ratings and facial electromyography signals during chewing.

Fig. 6

Mean (with standard error) intra-individual cross-correlation coefficients between the dynamic valence ratings and facial electromyography signals during swallowing.

Fig. 7

Mean (with standard error) individualized maximum cross-correlation coefficients within ±3 lags (i.e., indices of dynamic concordance) between the dynamic valence ratings and facial electromyography signals. ∗∗∗, p < 0.001.

Fig. 8

Mean (with standard error) individualized Pearson correlation coefficients (i.e., indices of static concordance) between the averaged dynamic valence ratings and averaged facial electromyography signals. ∗∗∗, p < 0.001; ∗∗, p < 0.01.