Content uploaded by Mikko Peltola
Author content
All content in this area was uploaded by Mikko Peltola on Nov 28, 2017
Content may be subject to copyright.
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
1
To appear in Developmental Psychobiology
Resting frontal EEG asymmetry in children: Meta-analyses of the effects of
psychosocial risk factors and associations with internalizing and externalizing
behavior
Mikko J. Peltolaa,b,c, Marian J. Bakermans-Kranenburgb,c, Lenneke R. A. Alinkb,c,
Renske Huffmeijerb,c, Szilvia Birob,c, & Marinus H. van IJzendoornb,c
a School of Social Sciences and Humanities, University of Tampere, Finland
b Centre for Child and Family Studies, Leiden University, The Netherlands
c Leiden Institute for Brain and Cognition, Leiden University, The Netherlands
Running head: META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
Correspondence to: Mikko Peltola, School of Social Sciences and Humanities, FIN-
33014 University of Tampere, Finland. Tel. +358503186120, e-mail:
mikko.peltola@uta.fi
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
2
Abstract
Asymmetry of frontal cortical electroencephalogram (EEG) activity in children is
influenced by the social environment and considered a marker of vulnerability to
emotional and behavioral problems. To determine the reliability of these associations,
we used meta-analysis to test whether variation in resting frontal EEG asymmetry is
consistently associated with a) having experienced psychosocial risk (e.g., parental
depression or maltreatment) and b) internalizing and externalizing behavior outcomes
in children ranging from newborns to adolescents. Three meta-analyses including 38
studies (N = 2,523) and 50 pertinent effect sizes were carried out. The studies
included in the analyses reported associations between frontal EEG asymmetry and
psychosocial risk (k = 20; predominantly studies with maternal depression as the risk
factor) as well as internalizing (k = 20) and externalizing (k = 10) behavior outcomes.
Psychosocial risk was significantly associated with greater relative right frontal
asymmetry, with an effect size of d = 0.36 (p < .01), the effects being stronger in girls.
A non-significant relation was observed between right frontal asymmetry and
internalizing symptoms (d = 0.19, p = .08), whereas no association between left
frontal asymmetry and externalizing symptoms was observed (d = 0.04, p = .79).
Greater relative right frontal asymmetry appears to be a fairly consistent marker of the
presence of familial stressors in children but the power of frontal asymmetry to
directly predict emotional and behavioral problems is modest.
Keywords: children; electroencephalogram; frontal asymmetry; psychosocial risk;
depression; maltreatment; internalizing; externalizing
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
3
Introduction
A common goal in the field of developmental psychopathology is to determine
biomarkers that show reliable associations with children’s vulnerability to emotional
or behavioral problems. Among the candidate markers, a considerable amount of
attention has been devoted to patterns of hemispheric asymmetry in frontal
electroencephalogram (EEG) alpha-band activity. The interest in frontal EEG
asymmetry in developmental research is largely due to the fairly consistent pattern of
greater relative right-sided frontal EEG asymmetry observed in currently and
previously depressed adults (e.g., Henriques & Davidson, 1990; Schaffer, Davidson,
& Saron, 1983). In the present study, we used meta-analysis on 38 studies (N = 2,523)
to test whether children’s frontal EEG asymmetry is consistently associated with a)
the presence of psychosocial risk factors such as parental depression or child
maltreatment, and b) child internalizing and externalizing behavior.
Frontal EEG alpha asymmetry refers to the difference in the amount of cortical
activity in one hemisphere relative to the other. Asymmetry scores are computed from
the EEG signal as the difference in ln-transformed EEG power within the alpha
frequency band (8-13 Hz in adults, 6-9 Hz in infants and young children; Marshall,
Bar-Haim, & Fox, 2002) between left and right frontal electrode sites (i.e., ln-right
minus ln-left). The typical experimental setup for EEG asymmetry consists of 1 to 8
minutes of resting/baseline recording during which external stimulation is minimal or
kept constant and neutral. As power in the alpha frequency band is inversely related to
neural activity in the underlying cortex (i.e., stronger power indicating less activity;
Lindsley & Wicke, 1974), positive alpha asymmetry scores are considered to reflect
greater relative left frontal cortical activity, whereas negative scores reflect greater
relative right frontal cortical activity.
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
4
The functional significance of asymmetrical frontal cortical activity is often
interpreted from the perspective of the approach/withdrawal model (Davidson,
Ekman, Saron, Senulis, & Friesen, 1990; Davidson, 1992; S. K. Sutton & Davidson,
1997) which relates asymmetries in frontal brain activity to basic motivational
tendencies, with the left frontal areas subserving approach motivation and the right
frontal areas subserving withdrawal motivation. A considerable body of research has
shown that in adults, depression is associated with greater relative right frontal
cortical activity (see Thibodeau, Jorgensen, & Kim, 2006, for meta-analytic
evidence), even in individuals in remission from depression (Gotlib, Ranganath, &
Rosenfeld, 1998; Henriques & Davidson, 1990). Right-sided frontal asymmetry may
thus be an endophenotype of a trait-like withdrawal motivation associated with
internalizing psychopathology such as depression and anxiety (Allen & Cohen, 2010;
Davidson, Marshall, Tomarken, & Henriques, 2000).
The motivational model makes a crucial distinction between motivational
direction and affective valence by arguing that asymmetrical frontal cortical activity
promotes motivational tendencies to approach and withdraw independently of the
affective valence underlying such tendencies (Harmon-Jones, Gable, & Peterson,
2010). Indeed, although left-sided frontal EEG asymmetry is associated with higher
positive emotionality (Tomarken, Davidson, Wheeler, & Doss, 1992), negatively
valenced externalizing behaviors that are related to approach (rather than withdrawal)
tendencies, such as trait and state anger, have been shown to be associated with
greater relative left frontal EEG activity as well (Harmon-Jones & Allen, 1998;
Harmon-Jones & Sigelman, 2001; Verona, Sadeh, & Curtin, 2009).
Frontal cortical asymmetry is also influenced by the early social environment,
with a particularly rich literature on the associations between maternal depression and
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
5
EEG asymmetry in infants and young children (see Field & Diego, 2008, for a
review). Several studies have observed greater right frontal EEG asymmetry in infants
of depressed vs. non-depressed mothers (e.g., Dawson, Frey, Panagiotides, Osterling,
& Hessl, 1997; Diego, Field, Jones, & Hernandez-Reif, 2006; Field, Fox, Pickens, &
Nawrocki, 1995), and similar findings have been observed in adolescents of depressed
mothers regardless of the adolescents’ own depression levels (Tomarken, Dichter,
Garber, & Simien, 2004). Research investigating the influence of other types of
psychosocial risk factors on children’s patterns of frontal EEG asymmetry is scarce. A
small number of studies have investigated whether the risks posed by insensitive
maternal caregiving (Hane & Fox, 2006), parental alcohol dependence (Ehlers, Wall,
Garcia-Andrade, & Phillips, 2001), and more severe conditions such as early
institutionalization (McLaughlin, Fox, Zeanah, & Nelson, 2011) and child
maltreatment (Curtis & Cicchetti, 2007; Miskovic, Schmidt, Georgiades, Boyle, &
MacMillan, 2009) are also associated with right-sided EEG asymmetry. While some
studies have documented strong effects of psychosocial risk factors in determining the
extent and direction of children’s frontal EEG asymmetry (Diego et al., 2006; Jones et
al., 1998; Miskovic et al., 2009), others have found no significant differences between
children experiencing high vs. low risk (Dawson, Klinger, Panagiotides, Hill, &
Spieker, 1992; Lusby, Goodman, Bell, & Newport, 2014). One highly unexplored
question concerns the ontogenetic mechanisms linking psychosocial risk to variations
in frontal asymmetry, i.e., whether genetic or experience-based effects are more
influential in shaping children’s frontal cortical activity. While the large effects
observed already in newborn infants of depressed mothers (Field, Diego, Hernandez-
Reif et al., 2004) may be taken to indicate a genetic disposition to greater right-sided
asymmetry in some individuals, longitudinal investigations aimed at producing long-
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
6
lasting changes in children’s exposure to psychosocial risks are required to address
this issue comprehensively.
Greater right-sided EEG asymmetry is commonly interpreted as a marker of
heightened risk of psychopathology and emotional dysregulation in children. To
directly test whether patterns of frontal EEG asymmetry can be considered as markers
of vulnerability to the development of emotional and behavioral problems, studies
have investigated the relations between EEG asymmetry and internalizing and
externalizing behavior outcomes in preschool and school age children, with both
concurrent and prospective study designs. In keeping with the model based on cortical
asymmetries and direction of motivational tendencies in adults (Harmon-Jones et al.,
2010), it can be hypothesized that internalizing symptoms (e.g., depression, anxiety,
and social withdrawal) would show associations with greater relative right
asymmetry, whereas externalizing symptoms (e.g., aggressive and impulsive
behavior) are expected to be related to greater left asymmetry. The directional
hypothesis has been supported by some studies (e.g., Gatzke-Kopp, Jetha, &
Segalowitz, 2014; Jones, Field, Davalos, & Pickens, 1997b; Pössel, Lo, Fritz, &
Seemann, 2008; Smith & Bell, 2010), whereas other studies have observed an
opposite pattern of EEG activation (Baving, Laucht, & Schmidt, 2002; Santesso,
Reker, Schmidt, & Segalowitz, 2006) or no direct associations between EEG and
these outcomes (Fox, Schmidt, Calkins, Rubin, & Coplan, 1996; Theall-Honey &
Schmidt, 2006).
In the present study we used meta-analysis to test a) the consistency of the
association between having experienced psychosocial risk and frontal cortical EEG
asymmetry and b) the effect sizes of the associations between frontal EEG asymmetry
and internalizing and externalizing behavior outcomes. The first hypothesis was that
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
7
higher psychosocial risk including parental depression would be related to greater
relative right frontal EEG asymmetry as compared to lower psychosocial risk. We
incorporated data on all potential psychosocial risk factors and compared the effects
of parental depression to those of other risk factors to obtain a more comprehensive
picture of the range of psychosocial influences on children’s EEG asymmetry.
Second, we tested the hypotheses stemming from the motivational direction model
(Harmon-Jones et al., 2010) that internalizing behavior outcomes are associated with
greater relative right, and externalizing behavior outcomes with greater relative left
EEG asymmetry. In addition to computing the combined effect sizes for the three sets
of meta-analyses (i.e., psychosocial risk, internalizing, and externalizing), we ran
moderator analyses to investigate whether effect size variations across studies are
associated with sample characteristics or procedural differences between studies. In
all meta-analyses, we tested the effects of participant age, gender, socioeconomic
status (SES), and resting EEG recording duration (to test whether shorter recording
durations are associated with larger effects, as in the meta-analysis of adult data by
Thibodeau et al., 2006). Additional moderator analyses tested the effects of different
types of psychosocial risk, different ways to assess psychosocial risks and child
outcomes (i.e., diagnosed vs. self-reported parental depression measures and observed
vs. reported child behavior), and the time lag between EEG recording and assessment
of internalizing or externalizing.
Methods
Literature search
Figure 1 outlines the study selection process. To obtain data for the meta-
analyses, we started with using PsycINFO and Google Scholar to search all empirical
journal articles in the English language available by August 15th, 2013, with the key
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
8
words EEG or electroencephalogra*, and asymmetr* in the title or abstract (the
asterisk indicating that the search contained the word or word fragment). We limited
the search results to studies including participants younger than 18 years old. This
search produced a total of 208 articles, the abstracts of which were screened. Studies
were excluded if they did not report data on resting frontal alpha band EEG activity,
for example if a) EEG asymmetry was reported only in response to a discrete stimulus
or event, b) asymmetry scores or differences in alpha power in the left and right
hemisphere were absent, or c) only non-frontal (e.g., parietal) asymmetry data was
reported. Articles were also excluded if they did not provide data on associations
between EEG asymmetry and psychosocial risk factors or on outcomes that could be
defined in terms of internalizing or externalizing behavior.
The majority of studies including data on psychosocial risk factors
investigated the association between maternal depression and children’s EEG
asymmetry. In Bruder, Tenke, Warner, and Weisman (2007), it was not specified
which parent was affected and, therefore, children in the high risk group of this study
had at least one parent (mother, father, or both) and at least one grandparent with a
diagnosis of major depressive disorder. For the present meta-analysis, studies
investigating more severe forms of psychosocial adversity such as childhood
maltreatment (Curtis & Cicchetti, 2007; Miskovic et al., 2009) or institutionalization
(McLaughlin et al., 2011) were also included, as well as two studies on the
associations between child frontal EEG asymmetry and maternal caregiving
insensitivity (Hane & Fox, 2006), and parental alcohol dependence (Ehlers et al.,
2001). Studies contributing to the analyses of internalizing behaviors consisted of
outcomes related to anxiety, fearfulness, depressiveness, social withdrawal, and
shyness reported by the parent or the child, or observations of facial signs of fear or
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
9
inhibited behavior in novel or ambiguous contexts. The outcomes in the set of studies
on externalizing behaviors included symptoms of aggression and oppositional
defiance, which were in all cases reported by the caregiver (or teacher; Gatzke-Kopp
et al., 2014). Whenever an article reported effects separately for female and male
participants, these were considered as separate outcomes. Studies that included data
on negative affect expressions that could not be clearly defined in terms of approach
or withdrawal motivation were excluded from the internalizing and externalizing
analyses. This resulted in the exclusion of studies investigating relations between
EEG asymmetry and crying in response to maternal separation (Davidson & Fox,
1989), sad facial expressions (Jones, Field, Fox, Lundy, & Davalos, 1997a), or more
global indices of negative emotionality (Dawson et al., 1999; Jones, McFall, & Diego,
2004; Shankman et al., 2011).
In the next step, articles were checked for partly overlapping samples and in
such cases, the article with the largest sample size was selected. This ensured that no
participants were included twice in the same meta-analysis. From the longitudinal
temperament study conducted by Fox and colleagues, we selected Henderson, Fox,
and Rubin (2001) to represent the internalizing data and Hane, Henderson, Reeb-
Sutherland, and Fox (2010) to represent the externalizing data from this project.
Although these publications do not report on the largest sample sizes in the context of
this project (Degnan et al., 2011), they were considered most representative/adequate
in terms of sample size, time between EEG measurement and outcome, and the
availability of sufficient statistical information for effect size calculations. From the
Mannheim Study of Risk Children, we selected the 8-year assessments (Baving,
Laucht, & Schmidt, 2000; Baving et al., 2002) as these represent the midpoint of a
longitudinal study from 4.5 to 11 years. After these steps, we identified 38 empirical
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
10
papers with 50 pertinent effect sizes, providing data for three sets of meta-analyses on
psychosocial risk factors (k = 20, N = 1,291), internalizing (k = 20, N = 1,299), and
externalizing (k = 10, N = 810) child behavior.
--------------------
Figure 1 here
--------------------
Moderators
Socioeconomic status (SES; low vs. middle/high) was coded as a categorical
moderator for all analyses. For the set of articles contributing to the psychosocial risk
analyses, we also coded risk type (parental depression vs. other adversity) and for the
studies measuring parental depression, the type of depression assessment (diagnosis
vs. self-report). For the associations with internalizing and externalizing behavior,
additional categorical moderators were the temporal relationship between EEG
recording and the outcome assessment (concurrent vs. predictive), and outcome
assessment type (observed vs. reported behavior). Moderator subgroups with k < 4
were excluded from the categorical contrast analyses. Continuous moderators
included age at the time of EEG measurement, age at the time of outcome assessment,
gender (% of male participants), time lag (in years) between EEG recording and
outcome assessment, and resting EEG recording duration. In cases where EEG was
recorded twice (e.g., Smith & Bell, 2010), the data were averaged across the two
assessments, as was age at the two assessment points. To assess intercoder reliability,
11 of the studies were coded by an independent coder. The agreement between the
coders across the categorical moderator variables was 98% (κs > .86) and correlations
between the continuous moderators were > .97.
Meta-analytic procedures
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
11
The meta-analyses were performed using the Comprehensive Meta-Analysis
(CMA) program (Borenstein, Rothstein, & Cohen, 2005). For each study, an effect
size (Cohen’s d) was calculated as the standardized difference between high and low
psychosocial risk conditions or between high and low manifestations of child
internalizing or externalizing behavior in resting frontal EEG asymmetry values. For
studies reporting correlational data, these were recomputed into Cohen’s d. Except for
Curtis and Cicchetti (2007) who reported the pertinent data from electrodes F7/8, and
Gatzke-Kopp et al. (2014) who reported data only from electrodes AF3/4, the effect
size calculations were based on data reported from mid-frontal electrodes F3 and F4.
For the analyses of psychosocial risk and internalizing problems, effects of greater
relative right-sided asymmetry were given a positive sign as they were in accordance
with our hypotheses. As externalizing behavior was hypothesized to be associated
with greater relative left-sided asymmetry, studies reporting effects of a greater
relative right-sided asymmetry associated with externalizing behavior were given a
negative sign (recall that due to the inverse relationship between alpha power and
neural activity, right-sided asymmetry indicates lower alpha power/greater neural
activity on the right frontal electrode sites).
CMA was used to compute combined effect sizes (weighted by the sample
sizes within individual studies) and 95% confidence intervals (CIs) around the point
estimates for the three separate sets of effects. Significance tests and moderator
analyses were performed with the Q-statistic on the basis of random-effects models
(Borenstein et al., 2005). Random-effects were favored over fixed-effects models as
they allow for the possibility that there are random differences between studies that
are associated with variations in procedures, measures, settings, that go beyond
subject-level sampling error and thus point to different study populations (Lipsey &
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
12
Wilson, 2001). In case of statistically significant combined effect sizes, the robustness
of these effects was estimated with the fail-safe number provided by the CMA
program, which estimates the number of studies with null results that would be
needed to reduce the overall significant result to non-significance. The Q-statistic was
also used to assess the heterogeneity of the effect sizes across studies. Meta-regression
was used to test the influence of continuous moderators. For each set of effect sizes,
Fisher’s Z scores were computed as equivalents for the effect size d, and the Z scores
were then standardized to screen for potential outliers. No outliers (standardized Z
scores ±3.29; Tabachnick & Fidell, 2001) were observed in the total set of studies or
the three separate sets.
To calculate the effect of potential data censoring or publication bias on the
significant outcomes of the meta-analyses, we used the trim-and-fill method. A funnel
plot was constructed of each study’s effect size on the x-axis against the inverse of the
standard error on the y-axis. The plot is expected to have the shape of a funnel
because studies with smaller sample sizes and larger standard errors have increasingly
large variation in estimates of their effect sizes as random variation becomes
increasingly influential, whereas studies with larger sample sizes have smaller
variation in effect sizes, making the top portion of the plot narrower (Duval &
Tweedie, 2000; A. J. Sutton, Duval, Tweedie, Abrams, & Jones, 2000). The plots
would be expected to be shaped like a funnel if no data censoring is present. However,
since smaller non-significant studies are less likely to be published, studies in the
bottom left hand corner of the plot are often omitted. With the trim-and-fill procedure,
the k right most studies considered to be symmetrically unmatched are trimmed and
their missing counterparts are imputed or ‘filled’ as mirror images of the trimmed
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
13
outcomes. This leads to a new estimate of the combined effect size taking into account
potential publication bias.
Finally, in case of statistically significant combined effect sizes, we performed
a power analysis with the G*Power 3.1 program (Faul, Erdfelder, Lang, & Buchner,
2007). First, we calculated the sample size required for an individual study to reach
the combined effect size (i.e., the assumed population effect size) with a power of
0.80 and a one-sided significance level of 0.05. Second, the actual power values of the
individual studies were calculated to estimate the range of power of the included
studies to detect the combined effect size.
Results
The combined effect sizes for the three sets of analyses and the primary categorical
moderator contrasts are displayed in Table 1. Tables 2 – 4 list the studies contributing
to the meta-analyses with descriptive data and forest plots representing the individual
effect sizes.
--------------------
Table 1 here
--------------------
Psychosocial risk
Within the set of studies on the associations between psychosocial risk factors
and EEG asymmetry, a significant combined effect size was observed (d = 0.36, CI
0.15 – 0.58, p < .01), indicating that the presence of psychosocial risk factors is
associated with greater relative right-sided frontal EEG asymmetry. The set of
outcomes was heterogeneous. The trim-and-fill method showed that one study had to
be trimmed and filled, while the resulting combined effect size remained basically
similar (d = 0.32, CI 0.09 – 0.54). The fail-safe number was 168, indicating a robust
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
14
effect. As can be observed from Table 1, the effect was of comparable magnitude
between different types of risk (i.e., parental depression vs. maltreatment or
institutionalization) and between different levels of SES. Within the set of depression
studies (k = 14), no difference in the magnitude of effects was observed between the
two types of depression assessment (diagnosis: k = 9, d = 0.35, CI 0.03 – 0.66; self-
report: k = 5, d = 0.54, CI 0.14 – 0.95; Q [1] = 0.56, p = .45).
The power analysis indicated that a sample size of N = 194 would be required
for an individual study to detect the combined effect size of d = 0.36 with a power of
0.80. The power values of the included studies to detect the combined effect size
ranged from 0.21 for the study with the smallest sample size (Field, Pickens, Fox,
Gonzalez, & Nawrocki, 1998) to 0.67 for the study with the largest sample size
(Ehlers et al., 2001), with the median power of the included studies being 0.39.
Meta-regression analyses with the continuous moderators revealed that the
effects were significantly moderated by gender but not age or resting EEG recording
duration (both ps > .10). Gender (the percentage of males in each sample) yielded a
significant negative regression weight (slope = -0.01, p = .04), indicating that studies
with a larger percentage of females in the sample were associated with larger effects.
When the continuous moderators were tested separately within the set of depression
studies (k = 14), age emerged as a significant moderator (slope = -0.05, p = .02), with
larger effect sizes in younger samples, while the effects of gender (p = .44) and
recording duration (p = .68) were not significant.
--------------------
Table 2 here
--------------------
Internalizing
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
15
The combined effect size (d = 0.19, CI -0.03 – 0.41, p = .08) for the
internalizing set was not significant. The set of studies was heterogeneous. Effect
sizes were not associated with SES, EEG-outcome time lag (i.e., concurrent vs.
predictive), or assessment type (observed vs. reported). Again, larger effects were
observed in samples with higher percentages of females (slope = -0.01, p = .02). The
associations between EEG asymmetry and internalizing behaviors were unrelated to
age at the time of EEG recording or outcome assessment, time lag between EEG
recording and outcome assessment, or resting EEG recording duration, all ps > .52.
--------------------
Table 3 here
--------------------
Externalizing
The combined effect size (d = 0.04, CI -0.27 – 0.35, p = .79) in a
heterogeneous set of outcomes provided no support for the hypothesis that frontal
EEG asymmetry would be related to externalizing behaviors in children. While no
associations between effect sizes were observed with all other categorical or
continuous moderators, gender was significantly associated with the magnitude of
effects (slope = 0.02, p < .001), but in the opposite direction as was the case in the
analyses of psychosocial risk and internalizing behavior. That is, stronger associations
between left-sided EEG asymmetry and externalizing behaviors were observed in
samples including relatively greater numbers of males.
--------------------
Table 4 here
--------------------
Discussion
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
16
The present meta-analytic study was designed to test whether the extent and direction
of frontal EEG asymmetry is consistently associated with a) having experienced
psychosocial risk and b) internalizing and externalizing behavior outcomes in studies
of children ranging from newborns to adolescents. The results showed that the
presence of psychosocial risk factors is significantly associated with greater relative
right frontal EEG asymmetry, with a combined effect size of d = 0.36. While this
association was of comparable magnitude between studies investigating parental
depression and child maltreatment, the effects appeared to be larger in samples with a
larger percentage of girls. Frontal EEG asymmetry showed a considerably weaker and
non-significant relation to internalizing symptoms (d = 0.19) and no significant
association with externalizing symptoms (d = 0.04).
The meta-analysis on studies of infants and children exposed to different kinds
of psychosocial risk supports the view of greater relative right frontal asymmetry as a
relatively consistent indicator of the exposure to familial stressors in children. While
the association appeared quite robust and no signs of a systematic publication bias
were observed, the studies included in the analyses of psychosocial risk were largely
underpowered, which may have increased the risk of false positive findings. The
effects of psychosocial risk were moderated by gender in that samples with a larger
percentage of girls were associated with larger effects, indicating that girls were more
susceptible (i.e., showed more right frontal asymmetry) to the presence of
psychosocial risk. The effect was however not significant when tested only within the
depression studies, and it appears to have been driven in the full set of studies by the
large effect observed in the female-only study by Miskovic et al. (2009): after leaving
out this study in an additional meta-regression, the effect was no longer significant (p
> .23). Inspection of the two child maltreatment studies (Curtis & Cicchetti, 2007;
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
17
Miskovic et al., 2009) does not provide an obvious answer as to why the two effect
sizes from maltreated female participants were highly different in magnitude. Among
the potential factors could be some of the sample-related variation between these
studies, such as different ways of documenting maltreatment history (official records
vs. self-report), age, SES, and ethnicity (predominantly African American in Curtis &
Cicchetti, 2007; no information on ethnicity was provided by Miskovic et al., 2009).
Nevertheless, the finding of potentially greater susceptibility of girls merits further
investigation as it appears to argue against findings from other developmental
domains indicating boys’ greater vulnerability to adverse experiences (Ramchandani,
Stein, Evans, & O'Connor, 2005; Sharp et al., 1995).
In addition, although observed only within studies having parental depression
as the risk factor, the moderation of the effect sizes by age is interesting as it seems to
indicate that the association between parental depression and right frontal asymmetry
may attenuate with extended exposure to parental depression (i.e., older age). In
infants, on the other hand, large effects were observed even in neonates with
obviously minimal experience of interaction with a depressed caregiver. As the
number of studies including older children is very limited, more research
documenting the effects across various age groups is needed before conclusions about
potential age differences in frontal asymmetry in response to parental depression and
other psychosocial risk factors can be made.
Apart from the common conceptualization of children’s frontal asymmetry as
a marker of vulnerability to later psychopathology, Saby and Marshall (2012) pointed
out that our understanding of the ontogenetic origins of frontal asymmetry variations
remains limited and a developmental model of EEG asymmetry has not been
constructed. The large effects observed in newborn infants and the concordance in
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
18
patterns of EEG asymmetry between newborns and their mothers (Field et al., 2004)
could be taken to indicate a genetic transmission. Field and colleagues have
suggested, however, that the neonatal effects may also emerge through intrauterine
exposure to a depressed mother’s biochemical imbalance affecting the levels of
cortisol and serotonin, which may have consequences on fetal brain development
(Field & Diego, 2008; Field, Diego, Dieter et al., 2004). Furthermore, the influence of
natural variations in maternal caregiving quality on children’s frontal EEG asymmetry
in low-risk samples (Hane & Fox, 2006; Hane et al., 2010) provide support for the
role of early interpersonal experiences with caregivers in shaping the pattern of frontal
cortical asymmetries. There is clearly a need for more research investigating the
malleability of children’s frontal asymmetry in response to changes in the social
environment, e.g., with intervention designs targeting parental caregiving behaviors.
The functional significance of variations in children’s frontal EEG asymmetry
is best understood by investigating its associations with emotional and behavioral
outcomes. In the present meta-analyses, however, the hypotheses derived from the
motivational approach/withdrawal models (Davidson, 1992; Harmon-Jones et al.,
2010) linking greater right frontal asymmetry to a greater risk of internalizing
symptoms and greater left frontal asymmetry to externalizing symptoms were not
supported. In the set of externalizing studies, in particular, the effects were rather
evenly distributed into positive, negative, and null effects, yielding a combined effect
size close to zero. The association between right frontal asymmetry and internalizing
behavior was stronger but not statistically significant. The effect sizes were also not
significantly dependent on the time lag between the EEG recording and internalizing
assessment, or assessment type (observed vs. reported). The behavioral outcome
effects were moderated by gender in directions that correspond to the higher rates of
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
19
internalizing symptoms in girls (Sterba, Prinstein, & Cox, 2007) and externalizing
symptoms in boys (Alink et al., 2006). Stronger associations between right frontal
asymmetry and internalizing symptoms were thus observed in samples with larger
percentage of girls and, conversely, left frontal asymmetry was more strongly related
to externalizing symptoms in samples with larger percentage of boys.
One reason for the lack of direct associations between frontal asymmetry and
the outcomes may be the relatively imprecise nature of the outcome measures
employed in many of the studies. For example, the Child Behavior Checklist (CBCL;
Achenbach & Rescorla, 2000) internalizing scale that was used in seven studies
included in the set of internalizing studies consists of separate subscales assessing
emotional reactivity, anxiety/depression, somatic complaints, and withdrawal. It is
possible that these subdomains of internalization are differentially associated with
withdrawal motivation and right frontal asymmetry, and the link between asymmetry
and internalizing may be attenuated when using the global internalizing scale. Indeed,
Shankman et al. (2013) recently showed that in adults, reductions in left frontal
asymmetry were uniquely associated with depression but not panic disorder, likely
reflecting a more tonic withdrawal motivation or reduced reward sensitivity
associated with depression vs. anxiety. In a similar vein, future studies with children
should be more detailed as to which specific facets of internalizing symptomatology
the patterns of frontal asymmetry are associated with (e.g., depression, anxiety, or
observed withdrawal behavior). Regarding externalizing symptoms, one yet
unexplored avenue for testing the contribution of left frontal asymmetry to
externalizing could be to investigate differences in frontal asymmetry in children who
manifest antisocial behavior but differ in the presence or absence of callous-
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
20
unemotional traits, as the aggression of children high in callous-unemotional traits
tends to be more proactive in nature (Frick & Viding, 2009).
In all three meta-analyses, SES was not a significant moderator of the effects.
The lack of moderation in the psychosocial risk studies therefore does not support the
hypothesis of Lusby et al. (2014) who argued that the fairly large effects observed in
many previous studies of infants of depressed mothers (e.g., Diego et al., 2006; Field
et al., 1995; Jones et al., 1998) may at least partly reflect the influence of other
stressors associated with the low socioeconomic status of the families in these studies.
It is also potentially important to point out that the analytical procedures of
EEG data in many studies included in these meta-analyses were not always optimal.
For example, whole-head EEG was measured in many studies with rather small
number of electrodes but nevertheless referenced offline to an average reference
configuration, which may be associated with biased estimation of the underlying
sources of electrical activity due to inadequate spatial sampling of electrodes (cf. Keil
et al., 2014). Given the often poorer signal-to-noise ratio in the EEG of infants and
small children, computation of the average reference from a low number of electrodes
may be a greater issue of concern in children than adults. To date, no studies have
reported EEG asymmetry from young children with high-density electrode montages
which provide a more complete coverage of the scalp and thereby also diminish the
risk of biases in the average reference computation apparent with a low number of
electrodes (e.g., variation in impedance and signal quality, or differences in scalp
location between homologous electrodes).
Finally, resonating the currently active discussion on power issues and
replicability in psychological and neuroimaging research (Bakker, van Dijk, &
Wicherts, 2012; Button et al., 2013), it may be a cause of concern that many of the
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
21
studies contributing to the combined effect size of psychosocial risk were highly
underpowered, with the median power of the included studies being 0.39. While the
median power exceeds the estimated typical power of 0.35 in psychology (Bakker et
al., 2012) and 0.21 in neuroscience studies (Button et al., 2013), it is nevertheless
considerably lower than the ideal threshold of 0.80. Not only do small sample sizes
decrease the possibility of detecting true effects, but, more worryingly, they may
inflate the estimated effect size of the observed group differences, leading to an
increasing likelihood of false positive findings. The problems associated with low
power become even more pressing when additional factors (e.g., gender) are included
in the statistical tests. Therefore, to be able to estimate the true effect sizes for the
influence of psychosocial risk on frontal asymmetry and the associations between
frontal asymmetry, internalizing, and externalizing, studies with larger sample sizes
(possibly through consortia integrating data from multiple sites) are needed.
Taken together, the present meta-analyses showed that while the pattern of
greater relative right frontal asymmetry is a fairly consistent marker of the presence of
familial stressors in children, the power of frontal asymmetry to directly predict
internalizing and externalizing behaviors is modest. The functional role of frontal
asymmetry in internalizing and externalizing may be more subtle and better
understood as a moderator of the influence of the environment or child dispositions on
behavioral outcomes. Indeed, studies taking such approach have indicated, for
example, that greater relative left frontal asymmetry may mitigate the influence of
maternal depression on children’s internalizing symptoms (Lopez-Duran, Nusslock,
George, & Kovacs, 2012) and greater relative right frontal asymmetry may exacerbate
the influence of inhibited temperamental disposition on later internalizing problems
(Fox et al., 1996). Frontal asymmetry thus appears to foster children’s tendencies to
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
22
approach or withdraw, but the relation of these tendencies to emotional and behavioral
outcomes may be critically dependent on the affective features of the environment or
the children themselves. Important challenges for future studies include investigating
the malleability of children’s frontal asymmetry in response to changes in parental
caregiving behaviors and associating patterns of frontal asymmetry to behavioral
outcomes more closely associated with motivational tendencies to approach and
withdraw.
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
23
Notes
MJP was supported by a postdoctoral research fellowship from the Institute for
Advanced Social Research, University of Tampere. MJB-K and MHvIJ were
supported by awards from the Netherlands Organization for Scientific Research
(MJB-K: VICI grant no. 453-09-003; MHvIJ: SPINOZA prize).
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
24
References
(Studies included in the meta-analyses are marked with an asterisk)
Achenbach, T. M., & Rescorla, L. A. (2000). Manual for the ASEBA preschool forms
& profiles. Burlington, VT: University of Vermont, Research Center for
Children, Youth, & Families.
Alink, L. R. A., Mesman, J., Van Zeijl, J., Stolk, M. N., Juffer, F., Koot, H. M., . . .
van IJzendoorn, M. H. (2006). The early childhood aggression curve:
Development of physical aggression in 10- to 50-month-old children. Child
Development, 77(4), 954-966. doi:10.1111/j.1467-8624.2006.00912.x
Allen, J. J., & Cohen, M. X. (2010). Deconstructing the "resting" state: Exploring the
temporal dynamics of frontal alpha asymmetry as an endophenotype for
depression. Frontiers in Human Neuroscience, 4, 232.
doi:10.3389/fnhum.2010.00232
Bakker, M., van Dijk, A., & Wicherts, J. M. (2012). The rules of the game called
psychological science. Perspectives on Psychological Science, 7(6), 543-554.
doi:10.1177/1745691612459060
*Baving, L., Laucht, M., & Schmidt, M. H. (2000). Oppositional children differ from
healthy children in frontal brain activation. Journal of Abnormal Child
Psychology, 28(3), 267-275. doi:10.1023/A:1005196320909
*Baving, L., Laucht, M., & Schmidt, M. H. (2002). Frontal brain activation in anxious
school children. Journal of Child Psychology and Psychiatry, 43(2), 265-274.
doi:10.1111/1469-7610.00019
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
25
Borenstein, M., Rothstein, D., & Cohen, J. (2005). Comprehensive meta-analysis: A
computer program for research synthesis. Englewood, NJ: Biostat.
*Bruder, G. E., Tenke, C. E., Warner, V., & Weissman, M. M. (2007). Grandchildren
at high and low risk for depression differ in EEG measures of regional brain
asymmetry. Biological Psychiatry, 62(11), 1317-1323.
doi:10.1016/j.biopsych.2006.12.006
*Buss, K. A., Schumacher, J. R., Dolski, I., Kalin, N. H., Goldsmith, H. H., &
Davidson, R. J. (2003). Right frontal brain activity, cortisol, and withdrawal
behavior in 6-month-old infants. Behavioral Neuroscience, 117(1), 11-20.
doi:10.1037/0735-7044.117.1.11
Button, K. S., Ioannidis, J. P., Mokrysz, C., Nosek, B. A., Flint, J., Robinson, E. S., &
Munafo, M. R. (2013). Power failure: Why small sample size undermines the
reliability of neuroscience. Nature Reviews.Neuroscience, 14(5), 365-376.
doi:10.1038/nrn3475
*Curtis, W. J., & Cicchetti, D. (2007). Emotion and resilience: A multilevel
investigation of hemispheric electroencephalogram asymmetry and emotion
regulation in maltreated and nonmaltreated children. Development and
Psychopathology, 19(3), 811-840. doi:10.1017/S0954579407000405
Davidson, R. J., Ekman, P., Saron, C. D., Senulis, J. A., & Friesen, W. V. (1990).
Approach-withdrawal and cerebral asymmetry: Emotional expression and brain
physiology. I. Journal of Personality and Social Psychology, 58(2), 330-341.
doi:10.1037/0022-3514.58.2.330
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
26
Davidson, R. J., & Fox, N. A. (1989). Frontal brain asymmetry predicts infants'
response to maternal separation. Journal of Abnormal Psychology, 98(2), 127-
131. doi:10.1037/0021-843X.98.2.127
Davidson, R. J. (1992). Emotion and affective style: Hemispheric substrates.
Psychological Science, 3(1), 39-43. doi:10.1111/j.1467-9280.1992.tb00254.x
Davidson, R. J., Marshall, J. R., Tomarken, A. J., & Henriques, J. B. (2000). While a
phobic waits: Regional brain electrical and autonomic activity in social phobics
during anticipation of public speaking. Biological Psychiatry, 47(2), 85-95.
doi:10.1016/S0006-3223(99)00222-X
Dawson, G., Frey, K., Self, J., Panagiotides, H., Hessl, D., Yamada, E., & Rinaldi, J.
(1999). Frontal brain electrical activity in infants of depressed and nondepressed
mothers: Relation to variations in infant behavior. Development and
Psychopathology, 11(3), 589-605.
*Dawson, G., Frey, K., Panagiotides, H., Osterling, J., & Hessl, D. (1997). Infants of
depressed mothers exhibit atypical frontal brain activity: A replication and
extension of previous findings. Journal of Child Psychology and Psychiatry,
38(2), 179-186. doi:10.1111/j.1469-7610.1997.tb01852.x
*Dawson, G., Klinger, L. G., Panagiotides, H., Hill, D., & Spieker, S. (1992). Frontal
lobe activity and affective behavior of infants of mothers with depressive
symptoms. Child Development, 63(3), 725-737. doi:10.1111/j.1467-
8624.1992.tb01657.x
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
27
Degnan, K. A., Hane, A. A., Henderson, H. A., Moas, O. L., Reeb-Sutherland, B. C.,
& Fox, N. A. (2011). Longitudinal stability of temperamental exuberance and
social-emotional outcomes in early childhood. Developmental Psychology, 47(3),
765-780. doi:10.1037/a0021316; 10.1037/a0021316
*Diego, M. A., Field, T., Jones, N. A., & Hernandez-Reif, M. (2006). Withdrawn and
intrusive maternal interaction style and infant frontal EEG asymmetry shifts in
infants of depressed and non-depressed mothers. Infant Behavior &
Development, 29(2), 220-229. doi:10.1016/j.infbeh.2005.12.002
Duval, S., & Tweedie, R. (2000). Trim and fill: A simple funnel-plot-based method of
testing and adjusting for publication bias in meta-analysis. Biometrics, 56(2),
455-463. doi:10.1111/j.0006-341X.2000.00455.x
*Ehlers, C. L., Wall, T. L., Garcia-Andrade, C., & Phillips, E. (2001). EEG
asymmetry: Relationship to mood and risk for alcoholism in Mission Indian
youth. Biological Psychiatry, 50(2), 129-136. doi:10.1016/S0006-
3223(01)01132-5
Faul, F., Erdfelder, E., Lang, A., & Buchner, A. (2007). G*Power 3: A flexible
statistical power analysis program for the social, behavioral, and biomedical
sciences. Behavior Research Methods, 39(2), 175-191. doi:10.3758/BF03193146
Field, T., & Diego, M. (2008). Maternal depression effects on infant frontal EEG
asymmetry. The International Journal of Neuroscience, 118(8), 1081-1108.
doi:10.1080/00207450701769067
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
28
Field, T., Diego, M., Dieter, J., Hernandez-Reif, M., Schanberg, S., Kuhn, C., . . .
Bendell, D. (2004). Prenatal depression effects on the fetus and the newborn.
Infant Behavior and Development, 27(2), 216-229.
doi:10.1016/j.infbeh.2003.09.010
*Field, T., Diego, M., Hernandez-Reif, M., Vera, Y., Gil, K., Schanberg, S., . . .
Gonzalez-Garcia, A. (2004). Prenatal predictors of maternal and newborn EEG.
Infant Behavior and Development, 27(4), 533-536.
doi:10.1016/j.infbeh.2004.03.005
*Field, T., Fox, N. A., Pickens, J., & Nawrocki, T. (1995). Relative right frontal EEG
activation in 3- to 6-month-old infants of 'depressed' mothers. Developmental
Psychology, 31(3), 358-363. doi:10.1037/0012-1649.31.3.358
*Field, T., Pickens, J., Fox, N. A., Gonzalez, J., & Nawrocki, T. (1998). Facial
expression and EEG responses to happy and sad faces/voices by 3-month-old
infants of depressed mothers. British Journal of Developmental Psychology,
16(4), 485-494. doi:10.1111/j.2044-835X.1998.tb00766.x
*Forbes, E. E., Shaw, D. S., Fox, N. A., Cohn, J. F., Silk, J. S., & Kovacs, M. (2006).
Maternal depression, child frontal asymmetry, and child affective behavior as
factors in child behavior problems. Journal of Child Psychology and Psychiatry,
47(1), 79-87. doi:10.1111/j.1469-7610.2005.01442.x
*Fox, N. A., Schmidt, L. A., Calkins, S. D., Rubin, K. H., & Coplan, R. J. (1996). The
role of frontal activation in the regulation and dysregulation of social behavior
during the preschool years. Development and Psychopathology, 8(1), 89-102.
doi:10.1017/S0954579400006982
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
29
Frick, P. J., & Viding, E. (2009). Antisocial behavior from a developmental
psychopathology perspective. Development and Psychopathology, 21(4), 1111-
1131. doi:10.1017/S0954579409990071
*Gatzke-Kopp, L. M., Jetha, M. K., & Segalowitz, S. J. (2014). The role of resting
frontal EEG asymmetry in psychopathology: Afferent or efferent filter?
Developmental Psychobiology, 56(1), 73-85. doi:10.1002/dev.21092
Gotlib, I. H., Ranganath, C., & Rosenfeld, J. P. (1998). EEG alpha asymmetry,
depression, and cognitive functioning. Cognition & Emotion, 12(3), 449-478.
doi:10.1080/026999398379673
*Hane, A. A., & Fox, N. A. (2006). Ordinary variations in maternal caregiving
influence human infants' stress reactivity. Psychological Science, 17(6), 550-556.
doi:10.1111/j.1467-9280.2006.01742.x
*Hane, A. A., Henderson, H. A., Reeb-Sutherland, B., & Fox, N. A. (2010). Ordinary
variations in human maternal caregiving in infancy and biobehavioral
development in early childhood: A follow-up study. Developmental
Psychobiology, 52(6), 558-567. doi:10.1002/dev.20461
*Hannesdóttir, D. K., Doxie, J., Bell, M. A., Ollendick, T. H., & Wolfe, C. D. (2010).
A longitudinal study of emotion regulation and anxiety in middle childhood:
Associations with frontal EEG asymmetry in early childhood. Developmental
Psychobiology, 52(2), 197-204. doi:10.1002/dev.20425
Harmon-Jones, E., & Allen, J. J. (1998). Anger and frontal brain activity: EEG
asymmetry consistent with approach motivation despite negative affective
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
30
valence. Journal of Personality and Social Psychology, 74(5), 1310-1316.
doi:10.1037/0022-3514.74.5.1310
Harmon-Jones, E., Gable, P. A., & Peterson, C. K. (2010). The role of asymmetric
frontal cortical activity in emotion-related phenomena: A review and update.
Biological Psychology, 84(3), 451-462. doi:10.1016/j.biopsycho.2009.08.010
Harmon-Jones, E., & Sigelman, J. (2001). State anger and prefrontal brain activity:
Evidence that insult-related relative left-prefrontal activation is associated with
experienced anger and aggression. Journal of Personality and Social Psychology,
80(5), 797-803. doi:10.1037/0022-3514.80.5.797
*Hayden, E. P., Shankman, S. A., Olino, T. M., Durbin, C. E., Tenke, C. E., Bruder,
G. E., & Klein, D. N. (2008). Cognitive and temperamental vulnerability to
depression: Longitudinal associations with regional cortical activity. Cognition &
Emotion, 22(7), 1415-1428. doi:10.1080/02699930701801367
*Henderson, H. A., Fox, N. A., & Rubin, K. H. (2001). Temperamental contributions
to social behavior: The moderating roles of frontal EEG asymmetry and gender.
Journal of the American Academy of Child & Adolescent Psychiatry, 40(1), 68-
74. doi:10.1097/00004583-200101000-00018
Henriques, J. B., & Davidson, R. J. (1990). Regional brain electrical asymmetries
discriminate between previously depressed and healthy control subjects. Journal
of Abnormal Psychology, 99(1), 22-31. doi:10.1037/0021-843X.99.1.22
*Jones, N. A., Field, T., Fox, N. A., Davalos, M., & Gomez, C. (2001). EEG during
different emotions in 10-month-old infants of depressed mothers. Journal of
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
31
Reproductive and Infant Psychology, 19(4), 295-312.
doi:10.1080/02646830120103374
*Jones, N. A., Field, T., Davalos, M., & Pickens, J. (1997b). EEG stability in
infants/children of depressed mothers. Child Psychiatry and Human
Development, 28(2), 59-70. doi:10.1023/A:1025197101496
*Jones, N. A., Field, T., Fox, N. A., Davalos, M., Lundy, B., & Hart, S. (1998).
Newborns of mothers with depressive symptoms are physiologically less
developed. Infant Behavior & Development, 21(3), 537-541. doi:10.1016/S0163-
6383(98)90027-3
*Jones, N. A., Field, T., Fox, N. A., Lundy, B., & Davalos, M. (1997a). EEG
activation in 1-month-old infants of depressed mothers. Development and
Psychopathology, 9(3), 491-505. doi:10.1017/S0954579497001260
*Jones, N. A., McFall, B. A., & Diego, M. A. (2004). Patterns of brain electrical
activity in infants of depressed mothers who breastfeed and bottle feed: The
mediating role of infant temperament. Biological Psychology, 67(1-2), 103-124.
doi:10.1016/j.biopsycho.2004.03.010
Keil, A., Debener, S., Gratton, G., Junghöfer, M., Kappenman, E. S., Luck, S. J., . . .
Yee, C. M. (2014). Committee report: Publication guidelines and
recommendations for studies using electroencephalography and
magnetoencephalography. Psychophysiology, 51(1), 1-21.
doi:10.1111/psyp.12147
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
32
*Kentgen, L. M., Tenke, C. E., Pine, D. S., Fong, R., Klein, R. G., & Bruder, G. E.
(2000). Electroencephalographic asymmetries in adolescents with major
depression: Influence of comorbidity with anxiety disorders. Journal of
Abnormal Psychology, 109(4), 797-802. doi:10.1037/0021-843X.109.4.797
Lindsley, D. B., & Wicke, J. D. (1974). The EEG: Autonomous electrical activity in
man and animals. In R. Thompson, & M. N. Patterson (Eds.), Bioelectrical
recording techniques (pp. 3-83). New York, NY: Academic Press.
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. Applied social
research methods series, Vol. 49. Thousand Oaks, CA: Sage.
*Lopez-Duran, N., Nusslock, R., George, C., & Kovacs, M. (2012). Frontal EEG
asymmetry moderates the effects of stressful life events on internalizing
symptoms in children at familial risk for depression. Psychophysiology, 49(4),
510-521. doi:10.1111/j.1469-8986.2011.01332.x
*Lusby, C. M., Goodman, S. H., Bell, M. A., & Newport, D. J. (2014).
Electroencephalogram patterns in infants of depressed mothers. Developmental
Psychobiology, 56(3), 459-473. doi:10.1002/dev.21112
Marshall, P. J., Bar-Haim, Y., & Fox, N. A. (2002). Development of the EEG from 5
months to 4 years of age. Clinical Neurophysiology, 113(8), 1199-1208.
doi:10.1016/S1388-2457(02)00163-3
*McLaughlin, K. A., Fox, N. A., Zeanah, C. H., & Nelson, C. A. (2011). Adverse
rearing environments and neural development in children: The development of
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
33
frontal electroencephalogram asymmetry. Biological Psychiatry, 70(11), 1008-
1015. doi:10.1016/j.biopsych.2011.08.006
*McManis, M. H., Kagan, J., Snidman, N. C., & Woodward, S. A. (2002). EEG
asymmetry, power, and temperament in children. Developmental Psychobiology,
41(2), 169-177. doi:10.1002/dev.10053
*Miskovic, V., Schmidt, L. A., Georgiades, K., Boyle, M., & MacMillan, H. L.
(2009). Stability of resting frontal electroencephalogram (EEG) asymmetry and
cardiac vagal tone in adolescent females exposed to child maltreatment.
Developmental Psychobiology, 51(6), 474-487. doi:10.1002/dev.20387
*Pössel, P., Lo, H., Fritz, A., & Seemann, S. (2008). A longitudinal study of cortical
EEG activity in adolescents. Biological Psychology, 78(2), 173-178.
doi:10.1016/j.biopsycho.2008.02.004
Ramchandani, P., Stein, A., Evans, J., & O'Connor, T. G. (2005). Paternal depression
in the postnatal period and child development: A prospective population study.
Lancet, 365(9478), 2201-2205. doi:10.1016/S0140-6736(05)66778-5
Saby, J. N., & Marshall, P. J. (2012). The utility of EEG band power analysis in the
study of infancy and early childhood. Developmental Neuropsychology, 37(3),
253-273. doi:10.1080/87565641.2011.614663; 10.1080/87565641.2011.614663
*Santesso, D. L., Reker, D. L., Schmidt, L. A., & Segalowitz, S. J. (2006). Frontal
electroencephalogram activation asymmetry, emotional intelligence, and
externalizing behaviors in 10-year-old children. Child Psychiatry and Human
Development, 36(3), 311-328. doi:10.1007/s10578-005-0005-2
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
34
Schaffer, C. E., Davidson, R. J., & Saron, C. (1983). Frontal and parietal
electroencephalogram asymmetry in depressed and nondepressed subjects.
Biological Psychiatry, 18(7), 753-762.
*Schmidt, L. A. (2008). Patterns of second-by-second resting frontal brain (EEG)
asymmetry and their relation to heart rate and temperament in 9-month-old
human infants. Personality and Individual Differences, 44(1), 216-225.
doi:10.1016/j.paid.2007.08.001
*Schmidt, L. A., Fox, N. A., Schulkin, J., & Gold, P. W. (1999). Behavioral and
psychophysiological correlates of self-presentation in temperamentally shy
children. Developmental Psychobiology, 35(2), 119-135.
doi:10.1002/(SICI)1098-2302(199909)35:2<119::AID-DEV5>3.0.CO;2-G
Shankman, S. A., Klein, D. N., Torpey, D. C., Olino, T. M., Dyson, M. W., Kim, J., . .
. Tenke, C. E. (2011). Do positive and negative temperament traits interact in
predicting risk for depression? A resting EEG study of 329 preschoolers.
Development and Psychopathology, 23(2), 551-562.
doi:10.1017/S0954579411000022
Shankman, S. A., Nelson, B. D., Sarapas, C., Robison-Andrew, E. J., Campbell, M.
L., Altman, S. E., . . . Gorka, S. M. (2013). A psychophysiological investigation
of threat and reward sensitivity in individuals with panic disorder and/or major
depressive disorder. Journal of Abnormal Psychology, 122(2), 322-338.
doi:10.1037/a0030747
Sharp, D., Hay, D. F., Pawlby, S., Schmücker, G., Allen, H., & Kumar, R. (1995).
The impact of postnatal depression on boys' intellectual development. Journal of
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
35
Child Psychology and Psychiatry, 36(8), 1315-1336. doi:10.1111/j.1469-
7610.1995.tb01666.x
*Smith, C. L., & Bell, M. A. (2010). Stability in infant frontal asymmetry as a
predictor of toddlerhood internalizing and externalizing behaviors.
Developmental Psychobiology, 52(2), 158-167. doi:10.1002/dev.20427
Sterba, S. K., Prinstein, M. J., & Cox, M. J. (2007). Trajectories of internalizing
problems across childhood: Heterogeneity, external validity, and gender
differences. Development and Psychopathology, 19(02), 345-366.
doi:10.1017/S0954579407070174
Sutton, A. J., Duval, S. J., Tweedie, R. L., Abrams, K. R., & Jones, D. R. (2000).
Empirical assessment of effect of publication bias on meta-analyses. British
Medical Journal, 320(7249), 1574-1577. doi:10.1136/bmj.320.7249.1574
Sutton, S. K., & Davidson, R. J. (1997). Prefrontal brain asymmetry: A biological
substrate of the behavioral approach and inhibition systems. Psychological
Science, 8(3), 204-210. doi:10.1111/j.1467-9280.1997.tb00413.x
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.).
Boston, MA: Allyn & Bacon.
*Theall-Honey, L. A., & Schmidt, L. A. (2006). Do temperamentally shy children
process emotion differently than nonshy children? Behavioral,
psychophysiological, and gender differences in reticent preschoolers.
Developmental Psychobiology, 48(3), 187-196. doi:10.1002/dev.20133
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
36
Thibodeau, R., Jorgensen, R. S., & Kim, S. (2006). Depression, anxiety, and resting
frontal EEG asymmetry: A meta-analytic review. Journal of Abnormal
Psychology, 115(4), 715-729. doi:10.1037/0021-843X.115.4.715
Tomarken, A. J., Davidson, R. J., Wheeler, R. E., & Doss, R. C. (1992). Individual
differences in anterior brain asymmetry and fundamental dimensions of emotion.
Journal of Personality and Social Psychology, 62(4), 676-687.
doi:10.1037/0022-3514.62.4.676
*Tomarken, A. J., Dichter, G. S., Garber, J., & Simien, C. (2004). Resting frontal
brain activity: Linkages to maternal depression and socio-economic status among
adolescents. Biological Psychology, 67(1-2), 77-102.
doi:10.1016/j.biopsycho.2004.03.011
Verona, E., Sadeh, N., & Curtin, J. J. (2009). Stress-induced asymmetric frontal brain
activity and aggression risk. Journal of Abnormal Psychology, 118(1), 131-145.
doi:10.1037/a0014376; 10.1037/a0014376
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
37
Table 1. Combined effect sizes and categorical moderators
k
N
d
95% CI
Q-W
Q-Ba
Psychosocial risk
Total
20
1291
0.36**
0.15 – 0.58
62.73**
SES
0.16
Low
12
704
0.32*
0.03 – 0.62
38.45**
Middle/high
8
587
0.42*
0.07 – 0.76
23.83**
Risk type
0.44
Depression
14
872
0.42**
0.14 – 0.69
36.10**
Maltreatment/ institution
4
224
0.22
-0.30 – 0.73
20.22**
Insensitive parenting
1
59
0.61*
0.08 – 1.13
Alcohol dependence
1
136
0.00
-0.38 – 0.38
Internalizing
Total
20
1299
0.19
-0.03 – 0.41
56.93**
SES
0.07
Low
6
518
0.24
-0.16 – 0.65
21.24**
Middle/high
14
781
0.17
-0.10 – 0.45
35.12**
EEG/outcome time lag
1.59
Concurrent
13
869
0.08
-0.19 – 0.35
37.12**
Predictive
6
314
0.39
-0.01 – 0.79
9.73
Outcome first
1
116
0.46*
0.04 – 0.88
Outcome assessment type
0.44
Observed
4
259
0.35
-0.16 – 0.86
6.28
Reported
16
1040
0.16
-0.09 – 0.41
49.18**
Externalizing
Total
10
810
0.04
-0.27 – 0.35
35.83**
SES
0.10
Low
5
479
-0.01
-0.48 – 0.46
17.88**
Middle/high
5
331
0.09
-0.39 – 0.57
17.88**
EEG/outcome time lag
Concurrent
8
711
0.02
-0.35 – 0.38
33.60**
Predictive
2
99
0.16
-0.62 – 0.94
2.08
*p < .05, **p < .01
k = number of study outcomes, N = total sample size, d = effect size (Cohen’s d), 95% CI = 95%
confidence interval around the point estimate of the effect size, Q-W = a statistic testing for the
homogeneity within a set of studies, Q-B = a moderation statistic testing for the significance of the
contrast between different sets of studies.
aSubgroups with k < 4 excluded from contrast. Note: Outcome assessment type contrast not shown for
the externalizing set because none of the studies in this set provided data on observational measures of
externalizing.
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
38
Table 2. Effect sizes and descriptive statistics for the set of psychosocial risk studies. The forest plot represents the individual effect sizes
(Cohen’s d with 95% confidence interval).
Study
d
p
N
Age
Bruder 2007
-0.81
.02
35
13.20
Curtis 2007 (females)
0.00
1.00
43
10.30
Curtis 2007 (males)
-0.63
.04
44
10.30
Dawson 1992
0.00
1.00
27
1.18
Dawson 1997
0.54
.00
117
1.15
Diego 2006
1.12
.00
66
0.18
Ehlers 2001
0.00
1.00
136
10.50
Field 1995
0.74
.04
32
0.40
Field 1998
-0.75
.08
24
0.30
Field 2004
0.62
.00
119
0.01
Hane 2006
0.61
.02
59
0.75
Jones 1997a
0.63
.05
41
0.08
Jones 1998
0.86
.00
58
0.02
Jones 2001
0.78
.02
38
0.84
Lopez-Duran 2012
0.19
.30
135
7.65
McLaughlin 2011
0.24
.41
76
3.50
Miskovic 2009
1.22
.00
61
14.24
Tomarken 2004
0.59
.09
38
13.00
Jones 2004
0.63
.01
78
0.16
Lusby 2014
0.22
.39
64
0.38
Total
0.36
.00
1291
4.41
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
39
Table 3. Effect sizes and descriptive statistics for the set of internalizing studies. The forest plot represents the individual effect sizes (Cohen’s d
with 95% confidence interval).
Study
d
p
N
Age
Ehlers 2001
0.00
1.00
136
10.50
Lopez-Duran 2012
-0.18
.30
135
7.65
McLaughlin 2011
0.48
.05
76
4.50
Baving 2002 (females)
0.81
.01
47
8.00
Baving 2002 (males)
-1.05
.01
35
8.00
Buss 2003
-0.20
.58
31
0.50
Forbes 2006
-0.50
.04
74
5.08
Fox 1996
0.00
1.00
96
4.56
Gatzke-Kopp 2014
0.24
.09
209
6.03
Hannesdóttir 2006
-0.90
.14
16
4.50
Hayden 2008
0.20
.66
22
6.16
Henderson 2001
0.24
.25
97
0.75
Jones 1997b
1.81
.02
15
3.00
Kentgen 2000
0.16
.74
18
15.50
Pössel 2008
0.90
.00
80
13.92
Schmidt 2008
1.46
.00
20
0.75
Schmidt 1999
0.28
.57
17
7.00
Smith 2010
0.70
.14
23
1.42
Theall-Honey 2006
0.00
1.00
36
4.50
McManis 2002
0.46
.03
116
11.00
Total
0.19
.08
1299
6.17
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
40
Table 4. Effect sizes and descriptive statistics for the set of externalizing studies. The forest plot represents the individual effect sizes (Cohen’s d
with 95% confidence interval).
Study
d
p
N
Age
Ehlers 2001
0.00
1.00
136
10.50
Hane 2010
0.03
.89
98
3.00
McLaughlin 2011
-0.13
.59
76
4.50
Baving 2000 (females)
-1.21
.00
33
8
Baving 2000 (males)
0.92
.03
25
8
Forbes 2006
0.77
.00
74
5.08
Fox 1996
0.00
1.00
96
4.56
Gatzke-Kopp 2014
0.28
.04
209
6.03
Santesso 2006
-0.98
.01
40
10.10
Smith 2010
0.63
.18
23
1.42
Total
0.04
.79
810
6.12
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
41
Figure Captions
Figure 1. A flow chart of the study selection process.
META-ANALYSES OF CHILDREN’S EEG ASYMMETRY
42
Figure 1.
Database and online key word search
Articles screened on basis of title and
abstract for inclusion criteria:
Resting frontal alpha asymmetry
Data on psychosocial risk or
internalizing/externalizing
outcomes
N = 140 articles excluded
Articles screened for duplicate samples
N = 208 articles found
N = 30 articles excluded
N = 38 articles included in the meta-
analyses:
Psychosocial risk: N = 19, k = 20
Internalizing: N = 19, k = 20
Externalizing: N = 9, k = 10
