Executive functions and behaviour problems in preschool children with perinatal brain lesion

Bilać, Snježana. (2017). Executive functions and behaviour problems in preschool children with perinatal brain lesion. PhD Thesis. Filozofski fakultet u Zagrebu, Department of Psychology.
(Poslijediplomski doktorski studij psihologije) [mentor Tadinac, Meri].

Preview

PDF (Croatian) Download (4MB) | Preview

Abstract

Introduction Perinatal brain lesions most often affect the white matter, especially in the periventricular (PV) area, resulting in poor connections between different parts of the central nervous system (Filley, 2010), and functional deficits. Because of that, children with normal general development may have a specific cognitive impairment or difficulties in complex psychological functions, such as executive functions or behaviour (Aylward, 2005). The relation between the severity of the lesion and the functional outcome has not been fully understood. Executive functions (EF) are high level control functions related to purposeful, goaldirected behaviour in new and complex situations. We used Gioia’s empirical EF model (Gioia, Espy & Isquith, 2003), in which EF have hierarchical and integrative character, regulating cognition as well as emotional reactions, social relationships, and overt behaviour. Gioia’s model highlights their developmental and ecological aspect through the influence of EF on child’s everyday behaviour. Previous studies have shown that children with perinatal brain damage are simultaneously at risk in both, EF and behaviour; yet how the two are related has not been fully elucidated, especially in early childhood. Of particular interest is the role of EF in the regulation of behaviour in early childhood (Loe et al., 2014), with regard to the severity of perinatal brain lesion or to child’s sex. Objective: The aim of this research was to examine the differences in EF and behavioural difficulties in preschool children who experienced perinatal brain lesion of different severity degrees, and to examine the contribution of EF in predicting behavioural difficulties. In view of the fact that previous research is ambiguous, the results of this study could contribute to scientific explanation of these complex and less known psychological outcomes following perinatal brain lesion. The following objectives have been set: 1. To determine whether there are differences in EF in pre-school children with regard to the severity of perinatal brain lesion and their sex. 2. To determine whether there are differences in the prevalence of behavioural difficulties in pre-school children with regard to the severity of perinatal brain lesion and their sex. 3. To examine the predictive value of the severity of perinatal brain lesion, child’s sex, and EF in predicting behavioural difficulties. Method Participants The total research sample consisted of 162 children, aged 3 to 4 years, without major neurological impairment. The clinical sample consisted of 107 outpatients of the Regional Centre for Children with Neurological Risk in the Special Hospital Goljak. Two clinical groups were formed - one with mild (N2 = 55; 30 boys and 25 girls, average age 3 years and 11 months), and the other with more severe brain lesions (N3 = 52; 25 boys and 27 girls, average age 4 years). The group with severe brain damage, as a rare event, comprised all available children born between 2008 and 2010. The inclusion criterion was a general developmental quotient of ≤ 80 (Developmental Test Čuturić, Čuturić, 1996) at the age of 1 or 2 (clinical samples were equalized). Term and preterm patients were included, similarly in the two clinical groups (66% preterm children in the group with mild lesion; 71% in the group with more severe lesion). The groups were marginally statistically different in gestational age (GA), p = .05, and significantly in birth weight (BW), p < .05. In the group with mild lesions the GA was 33.9 w (total range from 25 w to 42 w), and BW was 2280.5 g (total range from 750 g to 4830 g); in the group with severe lesions the GA was 32.1 w (total range from 24 to 42 w), and BW was 1873.1 g (total range from 710 g to 4750 g). The comparison group consisted of typically developing children born on term (N1 = 55; 27 boys and 28 girls, average age 3 years and 9 months), who attended kindergarten in Zagreb, and had no known perinatal risk factors. The three groups of participants were matched for children’s age and sex and mother’s education level. Measures 1. The degree of brain lesions was determined on the basis of neonatal ultrasound scans in existing medical records. In the group with mild lesions were classified as follows: intracranial haemorrhage of grades I and II (haemorrhagia intracranialis, HIC), PV echogenicity/echolucency (PVE) of grades I and II, and PV leukomalacia of grade I (PVL). The group with severe lesions encompassed as follows: HIC of grade III, PVE of grade III and PVL of grades II and III (Papille et al., 1978; Pidcock & Graziani; 1990; Weiglas-Kuperus et al., 1990). The damages were mostly bilateral (over 90% of patients). The combined hypoxic-ischaemic (HI) and HIC lesions were present in 50 pc of the children. 2. Verbal intelligence, as a control variable, was assessed using the subtest Receptive Vocabulary from the WPPSI-III (The Wechsler Preschool and Primary Scale of Intelligence – Third Edition; Wechsler, 2002). 3. Executive functioning was assessed in three ways. The performance-based tests were used from the NEPSY-II - A Developmental Neuropsychological Assessment, Second edition (Korkman, Kirk & Kemp, 2007) - the Statue (measure of inhibition), Comprehension of Instructions (working memory), Visuomotor Precision (inhibition) and Visual Attention (attention and inhibition). In order to assess behavioural aspects of EF, we used the Leiter-R - Leiter Examiner Rating Scale – Revised (Roid & Miller, 2002), Cognitive/Social Composite (C/S, with 4 subscales – Attention, Organization/Impulse control, Activity level, Sociability). The BRIEF-P - Behavior Rating Inventory of Executive Function - Preschool Version (Gioia, Espy & Isquith, 2003) was used to assess the mother’s rating of the child’s everyday behaviour with regard to EF. The Global Executive Composite (GEC) consists of 5 scales, which include the primary EF components. Inhibition is the ability to inhibit behaviour; shift is the ability to switch attention and behaviour; emotional control is the child's ability to modulate emotional responses. Working memory is the child's capacity to hold information in mind with the purpose of making a response, and its integral part is the ability to sustain attention. Planning/Organizing is the ability to anticipate future events and appropriate sequencing of responses. The combined scale scores result in three broad indices: Inhibitory Self Control (Inhibition and Emotional Control), Flexibility (Shift and Emotional Control) and Emergent Metacognition (Working memory and Plan/Organize). 4. Behaviour was assessed in two ways. For the assessment of emotional regulation during evaluation, we used the Leiter-R (Roid & Miller, 2002) Emotional/Regulation Composite (E/R, with 4 subscales – Energy and Feelings, Mood and Regulation, Anxiety, Sensory Reactivity). In assessing the mothers’ rating of difficulties in everyday behaviour, we used the CBCL1,5-5 - Child Behavior Check List (Achenbach & Rescorla, 2000). The dimension Internalizing Problems consists of four syndrome scales: Emotionally Reactive, Anxious/Depressed, Somatic Complaints and Withdrawn. The dimension Externalizing Problems consists of two syndrome scales: Attention Problems and Aggressive Behaviour. The Total Problems Score (CBCL) consists of Internalizing and Externalizing scores and scores in Sleep problems and Other problems subscales. All applied EF and behavioural measures demonstrated satisfactory Cronbach-alpha reliability, matching standard norms. The results had a normal distribution, particularly the composite scores. Procedure The research was conducted in accordance with ethical principles, with the prior consent of relevant ethical commissions. The participants’ parents were informed in advance and they signed a written consent to participate. Participants in the clinical sample were examined as part of regular psychological monitoring of children with neurological risk factors. Anamnestic, demographic and medical data of the children were already available as part of the medical hospital records. The classification of brain lesions was supervised by a paediatrician neurologist. The assessment was conducted in standard conditions, optimal for the child. The child’s mother was present during the assessment, evaluating EF and behaviour on rating scales. Participants were examined individually by the main researcher over a period of 15 months. The average duration of each assessment was 30 minutes, always starting by administering the Receptive Vocabulary. The sequence of neuropsychology tests was random to avoid an impact of training or fatigue on group results. Some participants didn’t perform the Statue, 34 in all, significantly more in the clinical groups (matched in both groups, regardless of sex, more frequent among the younger participants). The researcher noted behaviours important for qualitative behavioural analyses in the NEPSY-II and for the rating of the child’s behaviour (Leiter-R). Parental counselling was conducted and a clinical psychological report was written for each child. In the comparison group, the children were assessed in similar standard conditions as the clinical groups, but adjusted to the kindergarten environment. A short report on the assessment results was written for every child and advice was given where necessary. Results and discussion To examine the differences in executive functions and behaviour in preschool children with various degrees of perinatal brain lesions, a two-way analysis of variance (ANOVA) was conducted, with the main effect of the severity of the lesion and child’s sex, as well as the effect of their interaction. In the post-hoc analysis of the differences between the groups the Tukey test was used. 1. Executive functions A significant effect of severe brain lesion was found for the majority of used EF measures (performance-based measures, the examiner’s and mother’s ratings; low effect size, ES). These measures encompass inhibition, attention and flexibility. Participants with severe lesions achieved lower results than children with mild lesions and the comparison group. This was shown by: a) results on the Statue, Visuomotor Precision and Visual Attention; b) the researcher’s ratings on Organization/Control of impulse and Activity subscales; c) the mother’s ratings on Shift and Plan/Organize scales and General Executive Composite (GEC). A significant effect of the presence of brain lesions (mild to medium ES) was partially confirmed. Participants with perinatal brain damage, regardless of severity, achieved lower results compared to the comparison group. This was shown by: a) results on the Comprehension of Instructions; b) the researcher’s rating on Attention subscale; c) the mother’s rating on Working Memory scale and Emergent Metacognition Index. The graded effect of the degree of brain lesion was confirmed only by the researcher’s ratings of EF on Sociability subscale (high ES) and Cognitive/Social Composite (medium ES). This researcher’s rating showed a stronger effect on group differentiation than did test results and mothers’ ratings. Groups with mild and more severe brain lesion differed significantly, and both achieved lower results than the comparison group. The obtained test results are in accordance with previous results on reduced attention, working memory and inhibition in young children with perinatal brain damage, regardless of the term of birth (Kostović, 2012). Lower results in preterm children were mostly linked to abnormalities of white matter (Edgin et al., 2008; Woodward et al., 2011; Young et al., 2016). During assessment they showed weaker attention, cooperation and perseverance (Clark et al., 2008; Sajaniemi et al., 2001). The behavioural aspects of EF increased the examiner’s sensitivity in differentiating groups, particularly for the Sociability subscale. Refusal to take test or poor cooperation in these children may indicate functional difficulties (Langkamp & Brazy, 1999; Mäntyen et al., 2001; Wong et al., 2014). These views are supported by our data on lower results on other tests in children who did not complete the Statue. Previous results on the mothers’ ratings also confirm more executive difficulties in preterm children on Shift scale (Baron et al., 2011; Pozzetti et al., 2014) and GEC in BRIEF-P (Gioia et al., 2003; Loe et al., 2015). Participants with severe lesions have poor adjustment in new situations with poorly organized behaviour. They often require external assistance. In our sample, mothers reported greatest difficulties on Working Memory scale and Emergent Metacognition index in both clinical samples, regardless of the severity of the lesion. This is in accordance with results in preterm children (Gioia et al., 2003; Loe et al., 2015; Pozzetti et al., 2014), as well as our results on the Comprehension of Instructions and the researcher’s rating on Attention subscale. The severity of brain lesion is not always directly related to outcome for functions which activate distributed neural networks (Bell & Wolfe, 2004). Even a mild brain damage can have negative consequences and affect child’s daily functioning (Young et al., 2016). Sex differences are marginally significant in EF in favour of girls based on performance-based measures and the researcher’s ratings (small ES). None of the measures revealed a significant interaction between the severity of the brain lesion and sex. Some previous results have been similar for visual attention (Sajaniemi et al., 2001), inhibition (Kostović, 2012; Loe et al., 2014) or activity level (DiPietro & Voegtline, 2017). The observed differences are so small that they can probably be attributed to developmental and not actual sex differences (Klenberg et al., 2001; Taylor et al., 2006). 2. Behaviour difficulties The results showed a significant effect of brain lesion for behaviour difficulties (medium to high ES). Participants with perinatal brain damage, regardless of degree, have more difficulties compared to the comparison group. This was shown by: a) the researcher’s rating on Anxiety subscale and Emotional/Regulation Composite; b) the mother’s rating on Anxiety/Depression and Attention problems subscales, as well as Withdrawal subscale and Internalizing problems. These children were more often anxious and concerned, avoided contact and responded poorly to external stimuli. They also showed a weaker task focus and reduced persistence. During the assessment, they were tense and invested a greater mental effort than typically developing children (Baron et al., 2012, 2011). The effect of severe brain lesion was partially confirmed (small ES). Children with severe lesions had more behavioural difficulties than children in the comparison group or those with mild lesions. This was shown by: a) the researcher’s ratings on Mood Regulation and Sensory Reactivity subscale; b) the mother’s ratings on Somatic Complaints and Other problems subscales and Externalizing problems as well as Total Problems Score (CBCL). Children with severe lesions were irritable, oversensitive to external stimuli and with poor adjustment to new situations. A strong biological basis for sensory regulation difficulties is probably indicative, also shown by other studies (Clark et al., 2008; Lengua, 2003; Spittle et al., 2009). The graded effect of the degree of brain lesion was found only for the researcher’s rating on Energy and feelings subscale (medium ES). Groups with mild and severe brain damage were significantly different, while both achieved lower results than the comparison group. Therefore, the researcher’s rating showed a stronger effect in differentiating groups than do the mother’s ratings. These children have frequently shown discomfort and low levels of energy and satisfaction, which is similar to other results (Clark et al., 2008; Landry et al., 1996). Our testing results are in accordance with the majority of previous results in preterm children, which showed that brain lesion further increase the presence of behavioural difficulties (Arpi & Ferrari, 2013; Young et al., 2016), particularly for those of higher degree, such as PVL or HIC of the III to IV degree (Delobel-Ayoub et al., 2006; Spittle et al., 2009). The typical profile of behavioural difficulties included internalizing problems (mostly anxiety/depression), attention and social problems (Spittle et al., 2009; Treyvaud et al., 2012; Weiglass-Kuperus et al., 1993; Young et al., 2016). These typical problems require activation of distributed neural networks subserving behaviour regulation (Bell & Wolfe, 2004), so even a mild lesion can lead to maladjusted behaviour. Sex differences were not confirmed for any behavioural measure, nor was there any observed interaction between sex and brain damage. This is in accordance with earlier results on emotional regulation in young children (Clark et al., 2008; Sajaniemi et al., 2001), while results based on mother’s ratings are inconsistent (Arpi & Ferrari, 2013). 3. Prediction of behaviour difficulties A three-step hierarchical multiple regression was carried out on separate clinical samples for the prediction of behaviour difficulties in two criterions (Emotional/Regulation Composite, E/R and Total Problems Score, CBCL). In the first model, the predictors were child’s age and sex, and the mother’s education level. The GA was added in the second model, and EF tests were added to the third model. The fourth model was done for the total clinical sample with researcher’s and mother’s ratings of EF, and the severity of the lesion as new predictors. Performance-based measures of EF (the Statue, Visuomotor Precision and Visual Attention) were significant predictors of emotional regulation during assessment (E/R) in all groups (varying according to group). In addition, the GA was a significant predictor of regulation in the comparison group (a total of 38% of explained criterion variance) and the group with severe brain lesion (48% of explained variance). The mother’s education was a significant predictor only in the group with mild lesion (43% of explained variance). In the fourth model of prediction researcher’s rating of EF (Cognitive/Social Composite) became the only highly significant predictor of emotional regulation. A total of 51% of criterion variance was explained in the comparison group and 67% in the clinical sample. The mother’s rating of EF (GEC) was not a significant predictor of the child’s regulation, as was the case with other predictors (child’s age and sex, mother’s education, GA, the severity of lesion). In the prediction of the mother’s rating of behavioural difficulties (CBCL) performance-based measures of the EF were not significant predictors in any group and neither were other predictors significant in any model (child’s age and sex, mother’s education, GA). In the fourth model of prediction mother’s rating of EF (GEC) became the only highly significant predictor of child’s everyday behaviour. GEC explained 40% of the criterion variance in the comparison group and 67% in the clinical sample. The researcher’s rating of EF (C/S Composite) was not a significant predictor of behaviour, as was the case for the other individual predictors (child’s age and gender, mother’s education, GA, the severity of lesion). In summary, the importance of attention, working memory, inhibition and flexibility has been confirmed in the prediction of behavioural difficulties, depending on the applied methods of assessment and the context in which EF and behavioural data were collected. Measures obtained in the same context – test situation or everyday life – showed a significant interrelationship. Our results are in accordance with studies that used correlations between different kinds of measures in typically developing (e.g. Hudges & Ensor, 2008), and in preterm children (Clark et al., 2008; Sajaniemi et al., 2001; Woodward et al., 2005; Young et al., 2016). The significant contribution of EF in the prediction of behaviour was identified only in few studies. Nadeau et al. (2001) established the mediation effect of objectively measured working memory on the relationship between the GA and behavioural difficulties in preterm children of early school age. Loe et al. (2015; 2014) found in 3- to 5- years-old preterm children that the mother’s rating of EF (GEC) was the mediator for the relationship between the GA and behaviour, while performance tests were not significant predictors of difficulties in everyday behaviour. These results are similar to our results. Conclusion The degree of perinatal brain lesion is important in two EF, inhibition and shifting. The very presence of a lesion is important in attention and working memory, regardless of severity. The graded effect of severity of lesion is important in sociability whereas sex is marginally significant in EF in these children. The presence of brain lesion is important in some aspects of emotional regulation and behaviour, regardless of severity. These problems are anxiety/depression symptoms and withdrawal, i.e., internalizing problems, and attention problems. The degree of brain damage is important in mood regulation, sensory reactiveness, somatic complaints and externalizing problems. Child’s sex is not found to be significant for behavioural difficulties. The importance of EF in the prediction of behavioural difficulties has been confirmed, depending on the assessment methods and context in which data were collected. The results support the Gioia’s empirical model (Gioia et al., 2003), where EF have a regulatory role in overall behaviour which indicates interdependence between EF and behaviour in their expression. We found that typical executive difficulties in attention and working memory, as well as regulatory difficulties in energy, anxiety and sociability (internalizing problems) are linked to the very presence of a brain lesion. Therefore, the psychological monitoring should include both low and high risk children with perinatal brain lesion from infancy on.

Actions (login required)

View Item