Children reared in institutional environments present emotional and cognitive affectations, as well as changes in the anatomy and functioning of certain cerebral areas, such as the prefrontal cortex. However, knowledge of the long run effect of these affectations remains scarce. Therefore, this study was designed to correlate the cortical electroencephalographic (EEG) and psychopathological parameters. Resting EEG from 10 institutionalized (IA) and 10 never-institutionalized adolescents (NIA) was obtained and this was parametrically correlated with the intensity of symptoms of depression, post-traumatic stress disorder (PTSD) and anxiety. Results showed that the IA group showed higher indices of depression, anxiety and PTSD, accompanied by low relative power (RP) in the fast bands, high activity in the slow bands in frontal areas, and higher alpha2 RP in temporal areas. In addition, IA showed higher EEG-correlations among frontal areas but lower EEG-correlations among prefrontal, parietal and temporal areas in the gamma band. We also found correlations between the EEG parameters and psychopathological scores. These EEG findings can contribute to improve our understanding of how stress experienced early in life may impact the development of the cortex, and how this, in turn, may predispose individuals to develop psychiatric conditions.
Institutionalization is defined as the confinement of an infant in any large congregate care facility where round-the-clock professional supervision supplants the role of family-like caregivers [
Recently, some authors have reported that institutionalized children are more prone to developing certain psychiatric conditions in the future, including symptoms of attention-deficit/hyperactivity disorder [
It is well-known that adequate cerebral functioning depends on the coordinated work of several brain structures [
In relation to the effects of institutionalization on EEG, McLaughlin et al. [
These EEG findings reported for institutionalized children could be associated with psychopathological problems in later stages of life. Some psychiatric illnesses, like autism, are currently attributed, theoretically, to a lack of connectivity between the prefrontal cortex and different cortical and subcortical areas [
Considering that mood modulation depends on the integrity of the prefrontal cortex and its functional interaction with other, associated cortical areas, and that institutionalization has been associated with emotional alterations in children, the aim of the present study was to correlate prefrontal, parietal and temporal EEG activity with psychopathological symptoms in institutionalized adolescents compared to community-reared adolescents.
Participants were 10 male adolescents housed in private or government-sponsored orphanages (institutionalized adolescents, IA), who were compared to 10 never-institutionalized adolescents (NIA). All participants were right-handed, aged 13 - 16 years, with no current neurological or psychiatric treatment. They were attending middle school or high school, had IQ ≥ 70 as evaluated by the brief form of the Wechsler Intelligent Scales [
With regard to exclusions, 4 institutionalized teenagers were not included in the final analysis due to excessive artefacts in the EEG recordings. IA participants were recruited from 4 different institutions located in 3 states in Mexico by direct invitation from their respective Principal. All institutions were registered under the Mexican Official Norms for Assistance Services for Children (Nom-167-SSA1-1997), which ensured that the adolescents’ housing, clothing and food needs and health services would be fully-covered while they participated in the project. Once permission was obtained from parents or caregivers, an appointment for applying the psychometric tests was scheduled. Control group was recruited from a secondary school of one of the same states and was invited through an informational meeting with their parents. Each adolescent from control group live in a nuclear family and did not present any sign of maltreatment. Individuals of both groups were matched according to age and IQ scores. The study took along one year to be completed approximately.
The study was conducted in two sessions. During the first meeting, the attention, concentration, IQ, Depression, Anxiety and PTSD tests were applied. All EEG recording was performed in the second session at baseline with eyes open.
All procedures involved in this experiment were approved by the Ethics Committee of the Institution in accordance with the ethical standards established in the 1964 Helsinki Declaration. All participants gave their informed consent prior to inclusion. The demographic characteristics of the sample are summarized in
EEG recordings were taken at eight derivations of the International 10 - 20 System. Electrodes were placed in prefrontal, frontopolar (Fp1, Fp2) and dorsolateral (F3, F4) areas, and in temporal (T3, T4) and parietal (P3, P4) regions. EEG-recording lasted for 3 min with subjects at rest with eyes open. All derivations were referred to linked earlobes with the ground electrode placed on the forehead. Electrooculograms were recorded to detect eye-movement artefacts using a monopolar montage with electrodes placed at the outercanthi of both
Demographic data | IA (n = 10) | NIA (n = 10) | Comparisons | Cohen’s d | |||
---|---|---|---|---|---|---|---|
M | S.E. | M | S.E. | t | p | ||
Age | 14.8 | ±1.033 | 14.1 | ±0.738 | 1.74 | 0.098 | 0.379 |
IQ | 88.9 | ±14.07 | 85.0 | ±8.62 | 0.75 | 0.46 | 0.174 |
Months of institutionalization | 153.9 | ±17.09 | --- | --- | --- | -- |
eyes. The EEG signals were amplified digitally by a Neuroscan-Nu Amps digital amplifier (Compumedics Neuroscan), with EEG filters set at 1 and 50 Hz. Impedance for the EEG electrodes was kept below 10 kV using Neuroscan sintered silver chloride electrodes (Compumedics Neuroscan). EEG signals were simultaneously recorded in a computer using the Scan 4.3 Software program (Compumedics Neuroscan). The EEG epochs that were identified visually as being contaminated by noise were removed manually with the CHECASEN computer program [
The Children’s Depression Inventory (CDI) Spanish version was applied [
The Children’s Post-Traumatic Stress Scale-Spanish version (CPSS) [
The Spence Children’s Anxiety Scale-Spanish version (SCAS) [
As mentioned above, the final analysis included 10 institutionalized and 10 never-institutionalized adolescents. A t-test for psychopathological scores between groups was performed. In the EEG analyses, a t-test for the RP values of each electrode recorded in each frequency band was carried out. The same analysis was used in the correlation tests for each pair of interhemispheric (Fp1-Fp2; F3-F4) and intrahemispheric derivations (Fp1-F3, Fp2-F4, F3-T3, F4-T4, F3-P3, F4-P4). Also, in order to determine the relation between EEG characteristics and psychopathological symptoms in each group, Pearson correlation analyses between EEG parameters and participants’ CPSS, SCAS and CDI scores were performed for the entire sample.
Participants’ scores on the CDI for the sub-domain dysphoria and the total score and sub-domain scores for re-experimentation and arousal on the CPSS were significantly higher in IA than NIA, though no significant between-group differences appeared on the SCAS.
The IA group showed higher RP in the theta (t = 2.05, p = 0.05, Cohen’s d = 0.435) and alpha1 (t = 2.98, p = 0.008, Cohen’s d = 0.568) bands at the right frontopolar (Fp2) derivations than NIA. Also, IA presented lower RP in the faster bands: bilaterally in beta1 in the frontopolar (Fp1 t = −2.095, p = 0.05, Cohen’s d = 0.442; Fp2 t = −2.099, p = 0.05, Cohen’s d = 0.443) and dorsolateral
Psychopathological scales | IA (n = 10) | NIA (n = 10) | Comparisons | Cohen’s d | CSS IA/NIA | |||
---|---|---|---|---|---|---|---|---|
M | S.E. | M | S.E. | t | p | |||
CDI | ||||||||
Total | 48.3 | ±2.34 | 42.11 | ±2.24 | 1.826 | 0.085 | 0.394 | 0/0 |
Dysphoria | 52.6 | 1.63 | 42.44 | ±1.99 | 3.979 | 0.001* | 0.683 | 0/0 |
Low self-esteem | 45.7 | 3.01 | 44.44 | ±2.75 | 0.306 | 0.763 | 0.071 | 0/0 |
CPSS | ||||||||
Total | 18.3 | ±5.24 | 5 | ±1.8 | 2.402 | 0.035* | 0.492 | 3/0 |
Re-experimentation | 6.2 | ±1.67 | 1.5 | ±0.528 | 2.665 | 0.022* | 0.531 | 4/0 |
Arousal | 4.8 | ±1.54 | 4.87 | ±0.517 | 2.449 | 0.037* | 0.498 | 5/0 |
Avoidance | 6.8 | ±2.3 | 1.8 | ±0.8 | 2.054 | 0.064 | 0.435 | 5/0 |
SCAS | ||||||||
Total | 59.2 | ±2.021 | 51.2 | ±4.09 | 1.754 | 0.097 | 0.381 | 3/1 |
Obsessive compulsive disorder | 62.2 | ±2.107 | 56 | ±13.52 | 4.28 | 0.210 | 0.71 | 3/3 |
Social phobia | 57 | ±2.35 | 50 | ±4.28 | 1.499 | 0.151 | 0.331 | 2/2 |
Panic/agoraphobia | 55.4 | ±3.53 | 55.3 | ±4.33 | 0.018 | 0.986 | 0.004 | 2/4 |
Separation anxiety | 58.2 | ±3.35 | 59.2 | ±2.55 | −0.24 | 0.815 | 0.056 | 4/4 |
Physical injury | 50.2 | ±3.1 | 53 | ±2.81 | −0.67 | 0.512 | 0.155 | 2/1 |
Generalized anxiety | 57 | ±2.18 | 54.5 | ±3.13 | 0.656 | 0.520 | 0.152 | 1/1 |
frontal areas (F3 t = −2.19, p = 0.041, Cohen’s d = 0.46; F4 t = −2.08, p = 0.042, Cohen’s d = 0.458), as well as in alpha2 in the left dorsolateral frontal (F3) area (t = −2.51, p = 0.021, Cohen’s d = 0.5). IA also had lower RP in the alpha2 band at both temporal lobes (T3 t = −3.001, p = 0.007, Cohen’s d = 0.577; T4 t = −2.17, p = 0.043, Cohen’s d = 0.455) (
IA participants presented a higher interhemispheric correlation in the alpha1 and alpha2 bands between frontopolar regions (Fp1-Fp2) than NIA (
IA also had a significantly-lower correlation in the left hemisphere between the dorsolateral and parietal (F3-P3), and dorsolateral and temporal areas (F3-T3), specifically in the gamma band (
In terms of intrahemispheric correlation, the IA group showed a higher correlation between the frontopolar and dorsolateral prefrontal areas in the delta, theta and alpha1 bands in the left (Fp1-F3) and right hemispheres (Fp2-F4), though in this case only in the alpha 1 band. Similarly, there was a higher EEG correlation between the left frontopolar and parietal derivations (Fp1-P3) in the delta and beta1 bands in the IA group compared to NIA.
A negative correlation between the SCAS total scores and RP in T3 (r = −0.606,
p = 0.005) and T4 (r = −0.462, p = 0.040) areas and SCAS OCD scores and alpha2 RP in T3 (r = −0.658, p = 0.002) and T4 (r = −0.547, p = 0.013) areas was found. The OCD SCAS also presented a negative correlation with the right dorsolateral frontal (F4) RP in alpha2 (r = −0.500, p = 0.025). In addition, a negative correlation was found between total CPSS (r = −0.446, p = 0.049), re-experiencing CPSS (r = −0.496, p = 0.026), and dysphoria CDI scores (r = −0.494, p = 0.032) with the right frontopolar prefrontal RP (Fp2) in beta1. Moreover, CDI total scores had a negative correlation with T3 alpha2 band (r = −0.533, p = 0.019).
As
EEGr | IA > NIA | t | p | Cohen’s d | IA < NIA | t | p | Cohen’s d |
---|---|---|---|---|---|---|---|---|
Fp1?Fp2 | Alpha-1 Alpha-2 | 2.85 2.29 | 0.01 0.03 | 0.558 0.475 | --- | |||
F3-F4 | --- | --- | ||||||
Fp1-F3 | Delta Theta Alpha-1 | 3.91 3.26 4.17 | 0.001 0.004 0.0005 | 0.677 0.61 0.7 | --- | |||
Fp2-F4 | Alpha-1 | 2.24 | 0.037 | 0.468 | --- | |||
FP1-P3 | Delta Beta1 | 2.57 2.86 | 0.019 0.01 | 0.519 0.559 | --- | |||
Fp2-P4 | --- | --- | ||||||
F3-P3 | --- | Gamma | −2.3 | 0.033 | 0.478 | |||
F3-P4 | --- | --- | ||||||
F3-T3 | --- | Gamma | −2.01 | 0.05 | 0.429 | |||
F4-T4 | --- | --- |
Behaviour/EEG r | Fp1-Fp2 Alpha1 | Fp1-Fp2 Alpha2 | Fp1-F3 Delta | Fp1-F3 Theta | Fp1-F3 Alpha1 | Fp2-F4 Alpha1 | F3-T3 Gamma | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
r | p | r | p | r | p | r | p | r | p | r | p | r | p | |
CPSS Re-experiencing | 0.625 | 0.003 | 0.522 | 0.018 | ||||||||||
Avoidance | 0.599 | 0.005 | 0.447 | 0.048 | 0.455 | 0.044 | ||||||||
Arousal | 0.540 | 0.014 | ||||||||||||
SCAS | ||||||||||||||
Panic agoraphobic | 0.482 | 0.032 | ||||||||||||
Separation anxiety | 0.452 | 0.045 | ||||||||||||
CDI Dysphoria | 0.520 | 0.023 | 0.494 | 00.032 | 00.577 | 0.010 | 0.540 | 0.017 |
(Fp1-Fp2). Avoidance CPSS presented a positive correlation with the alpha2 inter-frontopolar (Fp1-Fp2) and left intra-frontal (Fp1-F3) EEG correlations, while re-experiencing CPSS showed a positive correlation with the left intra-frontal EEG correlation (Fp1-F3) in alpha1. Panic agoraphobic and separation anxiety SCAS scores presented positive correlations with the left fronto-temporal EEG correlation in gamma. Finally, dysphoria CDI scores showed a positive correlation with the EEG correlation in the pair of derivations Fp1-F3 in the delta, theta and alpha1 bands, as well as at Fp2-F4 in alpha1 (
The results of this study indicate that adolescents who had been institutionalized for over 75% of their lifetime presented an electrical organization of the brain distinct from that of the subjects who had lived their entire lives in family settings with their parents. Some of the EEG parameters assessed showed correlations with anxiety, PTSD and dysphoric symptoms. In addition, the institutionalized adolescents showed greater symptoms of PTSD and dysphoria, and more of them presented clinically-significant symptoms of anxiety and PTSD.
The institutionalized adolescents had higher scores for dysphoria (CDI) and PTSD than those in the NIA group. Also, more of them presented clinically-significant symptoms of anxiety and PTSD. Several authors have established a relation between the time of institutionalization and the onset of mental disorders [
EEG activity in different areas in adolescents might reflect abnormal development or lack of development. Regarding EEG Relative Power for example, the negative correlation between Fp2 RP in the beta 1 band and CPSS (PTSD) and CDI dysphoria could be related to cortical excitability, since similar beta frequencies have been found in insomnia patients [
These results, which are common to a variety of disorders, including ADHD and autism, have also been described as signaling deviations in brain development [
Other findings that might be related to abnormal development are from the RP of the F3 area in relation to the alpha2 and beta1 bands. In this regard, a lower proportion of alpha2 and beta1 bands in temporal and frontal areas has been documented in resistant schizophrenic patients compared to non-resistant patients and controls [
Additionally, we observed a negative association between T3 and T4 in the alpha2 band and anxiety scores (SCAS Total score and SCAS CDI). This correlation could be related to the limbic irritability described by Teicher [
Considering that we obtained a negative correlation between the alpha2 band and anxiety scores in temporal areas, we conclude that the low RP of alpha2 reflects higher activity in temporal areas and so generates higher anxiety scores. Conversely, the higher RP of alpha2 reveals lower activity in temporal areas, which led to lower anxiety scores.
Regarding our EEG results, we found a positive relation between frontopolar EEG correlations and anxiety scores. First, there is a significant decrease in frontopolar thickness between 8- and 20-year-olds [
We also found a positive correlation between functional connectivity among prefrontal regions (Fp1, Fp2, F3 and F4) in the alpha bands and CPSS subscales like re-experiencing, avoidance and arousal. Similar results emerged from the CDI dysphoria subscale, where a positive correlation between functional connectivity among prefrontal regions in the delta, theta and alpha bands was identified. It has been suggested that brain structures like the prefrontal cortex are particularly vulnerable to stress, since they have a high density of glucocorticoid receptors and experience significant postnatal development [
Regarding the positive correlation between SCAS and the F3-T3 gamma band EEG-correlation, it is important to recall that those areas are critical for the control and regulation of social behavior [
Turning to the functional connectivity between the prefrontal and posterior cortices, our study found a lower gamma correlation in the left hemisphere between the prefronto-parietal and prefronto-temporal areas in IA. Similar findings have been observed with highly-functional autistic participants using fMRI techniques, which showed a lower correlation between prefrontal and parietal areas compared to controls [
Finally, we also found a higher EEG correlation between the anterior prefrontal and parietal areas in the left hemisphere in the delta and beta bands in IA. This EEG correlation again shares some similarities with data from autistic patients reported by Duffy and Als [
Finally, it is important to mention that the sample for this project was small due to several factors: the high number of institutionalized adolescents with low IQ, the effects of long-term psychiatric treatment, the few adolescents that had spent 75% of their lives in orphanages, the few institutions for adolescents (since most house younger children), and the need to remove EEG recordings affected by artefacts. These conditions made it necessary to recruit adolescents from 4 different institutions in 3 states in Mexico. Thus, the features of the study sample itself limited our ability to recruit more participants.
In conclusion, the IA adolescents presented more psychopathological symptoms and a brain electrical organization distinct from the controls, which had lived their entire lives with their parents. Their EEG parameters are similar to those of young people with psychiatric disorders like autism, and showed correlations with anxiety, depression and PTSD scores. We believe that these EEG findings can contribute to improving our understanding of the impact that early stress may have on the development of the cortex, and how this may predispose adolescents to develop psychopathologies.
We thank to CONAYCT with the grant 290840 and CB/2012/180981. We also thank to Procuraduría de la Defensa del Menor y la Familia (PRODEMEFA) of the Sistema Integral de la Familia (DIF) of the state of Yucatán, México and to orphan institutions: Comunidad de Restauración Integral de Adolescentes (CRIAI.A.P.), Centro de Atención Integral al Menor en Desamparo (CAIMEDE) in Mérida, Yucatán, Hogar Cabañas, in Guadalajara, Jalisco and Cancún Pro Kids in Cancún Quintana Roo, for their support.
The authors declare no conflicts of interest regarding the publication of this paper.
Hevia-Orozco, J.C. and Sanz-Martin, A. (2018) EEG Characteristics of Adolescents Raised in Institutional Environments and Their Relation to Psychopathological Symptoms. Journal of Behavioral and Brain Science, 8, 519-537. https://doi.org/10.4236/jbbs.2018.810032