Effects of expectation congruency on event-related potentials (ERPs) to facial expressions depend on cognitive load during the expectation phase

Previous studies have shown that event-related potentials (ERPs) to facial expressions are modulated by expectation (congruency) and that the ERP effects of expectation congruency are altered by cognitive tasks during the expectation phase. However, it is as yet unknown whether the congruency ERP effects can be modulated by the amount of cognitive load during the expectation phase. To address this question, electroencephalogram (EEG) was acquired when participants viewed fearful and neutral facial expressions. Before the presentation of facial expressions, a cue indicating the expression of a face and subsequently, an expectation interval without any cues were presented. Facial expressions were congruent with the cues in 75% of all trials. During the expectation interval, participants had to solve a cognitive task, in which several letters were presented for target letter detection. The letters were all the same under low load, but differed under high load. Event-related potential (ERP) results showed that the amount of cognitive load during the expectation phase altered the congruency effect in N2 and EPN amplitudes for fearful faces. Congruent as compared to incongruent fearful expressions elicited larger N2 and smaller EPN amplitudes under low load, but these congruency effects were not observed under high load. For neutral faces, a congruency effect in late positive potential (LPP) amplitudes was modulated by cognitive load during the expectation phase. The LPP was more positive for incongruent as compared to congruent faces under low load, but the congruency effect was not evident under high load. The findings indicate that congruency effects on ERPs are modulated by the amount of cognitive load the expectation phase and that this modulation is altered by facial expression.