EEG Oscillation Evidences of Enhanced Susceptibility to Emotional Stimuli during Adolescence

Abstract

Background: Our recent event-related potential (ERP) study showed that adolescents are more emotionally sensitive to negative events compared to adults, regardless of the valence strength of the events. The current work aimed to confirm this age-related difference in response to emotional stimuli of diverse intensities by examining Electroencephalography (EEG) oscillatory power in time-frequency analysis.

Methods: Time-frequency analyses were performed on the EEG data recorded for highly negative (HN), moderately negative (MN) and Neutral pictures in 20 adolescents and 20 adults during a covert emotional task. The results showed a significant age by emotion interaction effect in the theta and beta oscillatory power during the 500-600 ms post stimulus.

Results: Adolescents showed significantly less pronounced theta synchronization (ERS, 5.5-7.5 Hz) for HN stimuli, and larger beta desynchronization (ERD; 18-20 Hz) for both HN and MN stimuli, in comparison with neutral stimuli. By contrast, adults exhibited no significant emotion effects in theta and beta frequency bands. In addition, the analysis of the alpha spectral power (10.5-12 Hz; 850-950 ms) showed a main effect of emotion, while the emotion by age interaction was not significant. Irrespective of adolescents or adults, HN and MN stimuli elicited enhanced alpha suppression compared to Neutral stimuli, while the alpha power was similar across HN and MN conditions.

Conclusions: These results confirmed prior findings that adolescents are more sensitive to emotionally negative stimuli compared to adults, regardless of emotion intensity, possibly due to the developing prefrontal control system during adolescence.

Keywords: adolescence; event-related desynchronization (ERD); event-related synchronization (ERS); negative stimuli; susceptibility; time-frequency analysis (TFA).

FIGURE 1

The diagram of our experimental task (Yuan et al., 2015).

FIGURE 2

The accuracy for adults and adolescents during HN, MN, and Neutral stimuli.

FIGURE 3

The reaction time (RT) for adults and adolescents during HN, MN, and Neutral stimuli.

FIGURE 4

(Left) The averaged EEG oscillatory activities during HN, MN, and Neutral conditions at Fz over time (x-axis, 0 denotes picture onset) and frequency (y-axis, 3–12 Hz). Blue color indicates power decrease relative to pre-stimulus baseline and red color indicates power increase relative to the baseline. (Right) The time-frequency map and the scalp distribution of the emotion effect across HN and MN stimuli in 500–600 ms (Right, red color denotes smaller p values, and stronger significance level).

FIGURE 5

(A) The scalp distribution of the emotion effect on theta for HN and MN stimuli in 500–600 ms (red color denotes smaller p value and stronger significance level); (B) The scalp distribution of the emotion effect on beta for HN and MN stimuli in 500–600 ms (red color denotes smaller p value and stronger significance level).

FIGURE 6

(Left) The averaged beta rhythm oscillations during HN, MN and Neutral conditions at Cz over time (x-axis, 0 denotes picture onset) and frequency (y-axis, 14–30 Hz). Blue color indicates power decrease relative to pre-stimulus baseline and red color indicates power increase relative to the baseline. (Right) The time-frequency map and the scalp distribution of the emotion effect across HN and MN stimuli in 500–600 ms (Right, red color denotes smaller p values and stronger significance level).

FIGURE 7

(Left) The averaged alpha rhythm oscillations during HN, MN, and Neutral conditions at Oz over time (x-axis, 0 denotes picture onset) and frequency (y-axis. 7–17 Hz). Blue color indicates power decrease relative to pre-stimulus baseline and red color indicates power increase relative to the baseline. (Right) The time-frequency map and the scalp distribution of the emotion effect across HN and MN stimuli in 850–950 ms window (Right, red color denotes smaller p values and stronger significance level).