Electroencephalogram (EEG) Frontal Asymmetry and Error-Related Signals Are Potential Markers of Internalizing and Externalizing Behaviors in Children: A Meta-Analysis

Objectives: The frontal hemispheric asymmetry and error-related components, as measured by EEG, are often considered to be biological correlates or “indicators of risk” for psychopathology. Here, we performed a meta-analysis to define specific patterns of frontal hemispheric asymmetry at rest. We tested whether the association of the EEG patterns with child psychopathology, such as internalizing symptoms and externalizing behaviors, are potential risk indicators. In addition, we conducted an analysis to test whether amplitude modulation of the error-related negativity (ERN) is also related to child-internalizing symptoms and -externalizing behaviors. Methods: We conducted two distinct meta-analyses (resting states versus event-related potential) that included a total of 31 studies (N= 2,186). For each study included, we calculated Cohen’s d between internalizing symptoms/externalizing behaviors (reported with questionnaires) and the amplitude of the ERN in the Go/no-go and Flanker tasks or the alpha asymmetry at rest as measured in frontal electrodes. Standardized effect sizes (Hedges’ g) were calculated using Cohen’s d multiplied by the coefficient J, which is a correction for small sample sizes. Hedges’ g was introduced as a generic effect size in the OpenMeta Analyst software (http://www.cebm.brown.edu/openmeta/) with the corresponding variance (SE). We used the continuous random-effect analysis with the DerSimonian-Laird (DL) method. The meta-analysis software computed the effect sizes, with 95 percent confidence intervals and the pooled effect size g*. Results: We found a significant relationship between greater right frontal asymmetry and internalizing symptoms (P = 0.012). In addition, we found a significant relationship between higher ERN amplitudes and the severity of internalizing symptoms (P = 0.006). The association between externalizing behaviors and alpha frontal asymmetry at rest was not significant, probably because of the lack of statistical power (k = 2). Conclusions: Our results show that frontal asymmetry in resting states is a probable biological indicator of risk for childhood psychopathology, specifically of the development of internalizing symptoms. Results also supported that ERN amplitude was significantly related to internalizing symptoms