Investigating Age-Related Neural Compensation During Emotion Perception Using Electroencephalography

Abstract

Previous research suggests declines in emotion perception in older as compared to younger adults, but the underlying neural mechanisms remain unclear. Here, we address this by investigating how "face-age" and "face emotion intensity" affect both younger and older participants' behavioural and neural responses using event-related potentials (ERPs). Sixteen young and fifteen older adults viewed and judged the emotion type of facial images with old or young face-age and with high- or low- emotion intensities while EEG was recorded. The ERP results revealed that young and older participants exhibited significant ERP differences in two neural clusters: the left frontal and centromedial regions (100-200 ms stimulus onset) and frontal region (250-900 ms) when perceiving neutral faces. Older participants also exhibited significantly higher ERPs within these two neural clusters during anger and happiness emotion perceptual tasks. However, while this pattern of activity supported neutral emotion processing, it was not sufficient to support the effective processing of facial expressions of anger and happiness as older adults showed reductions in performance when perceiving these emotions. These age-related changes are consistent with theoretical models of age-related changes in neurocognitive abilities and may reflect a general age-related cognitive neural compensation in older adults, rather than a specific emotion-processing neural compensation.

Keywords: ageing; electroencephalogram; emotion perception; neural compensation.

Figure 1

Example of emotional facial stimuli of anger (a) and happiness (b) used in the experiment. Hard task conditions contain facial stimuli with lower emotional intensities (15% and 30%), and easy task conditions contain facial stimuli with higher emotional intensities (60% and 75%).

Figure 2

An outline of a trial. In each trial, a face was presented for 500 ms, followed by a blank screen with a fixation cross for 500 ms, and then followed by a prompt, asking participants to provide a response on the emotion category, happy or angry or neutral. The inter-trial interval was 1500 ms.

Figure 3

Older (patterned bars) and younger (solid bars) participants’ perceptual performance (accuracy) in anger (a) and happiness (b) experimental trials at different task difficulties (neutral (baseline), easy and hard). Bars in red represent trials with older face stimuli, bars in blue represent trials with younger face stimuli. (1) Anger perception (a): older group performed worse than younger group in both easy (old face stimuli condition and young face stimuli condition) and hard (old face stimuli condition and young face stimuli condition) levels. (2) Happiness perception (b): older adults performed poorer in hard condition using younger face stimuli, but not in other conditions. (3) In addition, within-group comparison revealed that younger participants’ performance on young and old face stimuli were significantly different for only hard level of anger and happiness perception tasks, with superior performance on younger face stimuli in happiness perceptual tasks and superior performance on older face stimuli in anger perceptual tasks. Error bars represents S.E. (*** represents p < 0.001. ** represents p < 0.01. * represents p < 0.05).

Figure 4

Significant clusters of the nonparametric cluster randomisation test comparing the two groups (young minus old) on neutral condition. First cluster time window: 100–200 ms, left-frontal and centromedial area (F1, F3, F5, F7, Fc5, Fc5, C1, C3, Cp1, Pz, Cpz, Fc4, Cz, C2, C4); second cluster time window: 250–900 ms; frontal area (AF3 F1, F3, F5, Fc1, Fz, F2, F4, Fcz).

Figure 5

Older (bars in blue) and younger (bars in red) participants’ cluster one mean ERP amplitude for anger (a) and happiness (b) experimental trials at different task difficulties (neutral (baseline), easy and hard). Solid bars represent trials with older face stimuli; patterned bars represent trials with younger face stimuli. The overall left-frontal and centromedial mean ERP amplitude (100–200 ms) of easy condition was significantly higher than the hard condition in anger task, regardless of group. Error bars represents S.E. (*** represents p < 0.001).

Figure 6

Older (bars in blue) and younger (bars in red) participants’ cluster two mean ERP amplitude for anger (a) and happiness (b) experimental trials at different task difficulties (neutral (baseline), easy and hard). Solid bars represent trials with older face stimuli; patterned bars represent trials with younger face stimuli. (1) Older participants exhibited significantly higher frontal ERP amplitudes (250–900 ms) in both easy and hard levels of anger and happiness perceptual tasks compared to younger participants. (2) Within-group comparison revealed that younger participants’ mean frontal ERP amplitudes (250–900 ms) for the easy condition was significantly higher than the hard condition in anger perception, but not significantly different in the happiness perceptual tasks; whereas older participants did not show significant different ERPs across easy and hard task conditions in both the happiness and anger perceptual tasks. Error bars represents S.E. (significances are marked by *** p < 0.001, ** p < 0.01, and * p < 0.05).