Tested and reported executive problems in children and youth epilepsy

Abstract Objectives Executive problems in children and youth with epilepsy influence their ability to handle important aspects of daily life activities. The present study sought to explore factors associated with executive problems for patients with epilepsy in this age group. Methods The cohort consisted of 97 consecutive patients at the National Centre for Epilepsy in Norway, aged 10–19 years, with focal or genetic generalized epilepsy. All underwent tests of executive functions (D‐KEFS), the Behavior Rating Inventory for Executive Function (BRIEF), and screening for psychiatric symptoms, using the Strengths and Difficulties Questionnaire (SDQ). Results Parent‐reported cognitive executive dysfunction (BRIEF, Metacognitive Index) was the strongest independent predictor for tested executive dysfunction and vice versa. Furthermore, male gender correlated strongest with parent‐reported behavioral regulation problems (BRIEF, Behavioral Regulation Index) along with borderline/pathological score on the SDQ and parent‐reported cognitive executive dysfunction. Conclusions A strong association between parent‐reported cognitive executive dysfunction and tested executive dysfunction was found. Male gender correlated strongest with parent‐reported behavioral regulation problems. The latter was probably related to a higher frequency of symptoms associated with psychopathology among the boys than the girls. The frequency of executive deficits according to the different modes of measurement varied from 16% to 43%, suggesting that they capture different aspects of behavior under the executive umbrella.

One population-based study with emphasis on executive functions (EF) found that the patient group performed poorer on seven out of eight EF measures compared with the healthy comparison group (Hoie et al., 2006), and that executive problems contributed to school difficulties beyond intellectual dysfunction. EF includes behaviors necessary for successful coping with everyday challenges, such as initiation, planning, organization, purposive actions, self-monitoring, and self-regulation (Lezak, Howieson, Bigler, & Tranel, 2012). Thus, identification of factors associated with executive dysfunction in CWE is of importance for development of preventive measures and interventions to improve social and academic function.
The Behavior Rating Inventory of Executive Function (BRIEF) (Gioia, Isquith, Guy, & Kenworthy, 2000), a standardized parentrated measure assessing eight domains of EF, represents an effort to assess both behavioral and cognitive aspects of EF in everyday life. The BRIEF has been used to assess EF in several pediatric samples, including traumatic brain injury (Mangeot, Armstrong, Colvin, Yeates, & Taylor, 2002;Vriezen & Pigott, 2002), attention-deficit-hyperactivity disorder (ADHD) (Jarratt, Riccio, & Siekierski, 2005), hydrocephalus (Mahone, Zabel, Levey, Verda, & Kinsman, 2002), and autism (Gilotty, Kenworthy, Sirian, Black, & Wagner, 2002;Gioia, Isquith, Kenworthy, & Barton, 2002). In pediatric epilepsy populations, poor scores on the BRIEF have shown association with everyday executive dysfunction (Campiglia et al., 2014) and poor health-related quality of life . Slick, Lautzenhiser, Sherman, and Eyrl (2006) found that a substantial proportion of children with intractable epilepsy display significant EF deficits measured by BRIEF and called for research into the relationship of BRIEF scores to other measures of EF in children with epilepsy to further clarify its clinical utility. Only a few studies have adressed the relationship between tested and reported information about EF in children with epilepsy. Parrish et al. (2007) looked at the BRIEF, completed by parents, and test results from aspects of the Delis-Kaplan Executive Function System (D-KEFS) and found in newly diagnosed children with epilepsy characterized by good seizure control that the two modes of measurement were significantly correlated on Metacognition Index, but not on the Behavioral Regulation Index.
In a study by MacAllister et al. (2012), similar correlation between tested and reported EF deficits was not found when EF was tested with the Tower of London test (Anderson, Anderson, & Lajoie, 1996). They also found that the Tower of London performance but not ratings on the BRIEF could be predicted by epilepsy-related variables, and discussed possible differences in the validity of these two modes of measuring EF.
Inconsistent findings between tested and reported executive deficits in children with epilepsy call for further investigation. Thus, the present study sought to analyze epilepsy-related correlates to tested as well as reported behavioral and cognitive executive problems in children and youth referred to a tertiary epilepsy center. Based on referenced findings, we expected to find (Hoie et al., 2005) that tested executive problems would be associated with early seizure debut and reported everyday cognitive executive problems (Hoie et al., 2006), that behavioral aspects of reported everyday executive problems would be best correlated to psychiatric symptoms, and (Berg et al., 2008)

| Patient inclusion and clinical data
Patients between 10 years and 19 years hospitalized at The National Centre for Epilepsy, the only tertiary epilepsy center in Norway, were included consecutively from January 2012 to June 2014 ( Figure 1). Informed written consent was obtained from parents or participants of legal age. The study was approved by the Regional Ethics Committee (2011/1636/REK).
A detailed description of the inclusion and exclusion criteria in this study has been described elsewhere (Alfstad et al., 2016

| Behavioral evaluation
The Strengths and Difficulties Questionnaire (SDQ) (Goodman, 2001) is a 25-item brief behavioral screening questionnaire that was completed independently both as self-report by the patients and by the parents. The questionnaire includes five subscales, covering emotional, conduct, hyperactivity, and peer relationships, as well as prosocial behavior. The items are rated "not true" (0), "somewhat true" (1), or certainly true" (2), resulting in a total difficulties score from 0 to 40. According to available norms (Goodman, 2001), the total scores were classified as normal 0-15, borderline 16-19, or abnormal >19 in the self-report and correspondingly in the parent report 0-13, 14-16 and >16. Other research groups have previously used this methodological approach to classify scores on the SDQ (Alfstad et al., 2016;Heiervang et al., 2007).
The Behavior Rating Inventory of Executive Function (BRIEF) (Gioia et al., 2000), a standardized 86-item parent-rated measure, was employed to assess both behavioral and cognitive aspects of EF in everyday life. The subscales of inhibit, shift, and emotional control make up the summary Behavioral Regulation Index (BRI), while the Metacognition Index (MCI) is comprised of initiate, working memory, plan/organize, organizations, and monitor subscales. Scores are standardized based on a normal distribution with a mean of 50 and SD of 10. We examined scores on both the BRI and MCI and according to the manual (Gioia et al., 2000) T-scores ≥ 65 were considered abnormal.

| Neuropsychological assessment
General cognitive ability was measured by the Wechsler Abbreviated Scale of Intelligence (WASI) (Wechsler, 1999).

| Statistical analysis
First, descriptive statistics of the demographic, clinical, behavioral, and cognitive characteristics of the patient population was com- puted. Group comparisons were tested using Pearson's Chi-square for categorical variables and independent sample t tests for continuous variables. Odds ratios for occurrence of tested and reported executive problems were estimated using logistic regression analysis.
Univariate analyses were first performed for dichotomized variables reported in Tables 2-4, and significant factors were included in the multivariate analysis. Results of multivariate analysis are presented with odds ratios (OR) with 95% confidence intervals (CI) and p values. All tests were two-sided and performed at a 5% significance level. The statistical Package for Social Sciences (spss version 24) was used.

| RE SULTS
Demographical, epilepsy-related, and behavioral background characteristics are provided in Table 1. Table 2 shows that patients scoring below and above cutoff on the Executive Score are similar on all the demographical and epilepsy-related variables and significantly different on all the cognitive and behavioral measures with poorest results for the group with Executive Score below cutoff. Table 3 shows similar results for all the cognitive and behavioral measures for the groups obtaining abnormally high score versus normal score on the Metacognitive Index. Additionally, it was found that patients with abnormally high score on the Metacognitive Index had significantly earlier seizure debut (7.0 vs. 8.7 years, p = .03) than those with scores in the normal range. In Table 4

| D ISCUSS I ON
The main findings of this study are that parent-reported everyday cognitive executive dysfunction best correlated with tested executive dysfunction and vice versa, that tested executive dysfunction was best correlated with parent-reported everyday cognitive executive dysfunction. Furthermore, male gender showed the strongest association with parent-reported everyday behavioral regulation problems along with borderline/pathological score on the SDQ and parent-reported everyday cognitive executive dysfunction.
Borderline/pathological score on the SDQ was also significantly associated with parent-reported everyday cognitive executive dysfunction. Executive deficits according to the three different modes of measurement was found in 16% based on tested executive dysfunction, 43% based on parent report of everyday cognitive executive problems, and 28% based on parent report of everyday behavioral regulation difficulties.
The strong association between parent-reported everyday cog-  (Anderson et al., 1996), which is a measure of EF. This finding fits well with studies that evaluate the association between reported and tested EF in children with other neurologic conditions. For instance, Anderson, Anderson, Northam, Jacobs, and Mikiewicz (2002)   Early seizure onset, defined as below mean for the group (below 8 years of age) was associated with higher and more pathological scores on both the Metacognitive Index and the Behavioral Regulation Index on the BRIEF, but not on the composite Executive Score. As mentioned above, the BRIEF correlates highly with SDQ, which is associated with psychiatric symptoms. Thus, this finding is probably in line with our previous finding of high association between early seizure debut and psychiatric comorbidity (Alfstad et al., 2016). Most previous studies have not found this association (Alfstad et al., 2011). The divergent findings may be related to a host of factors, including different study populations, different measurement methods and which cutoff for early versus late seizure debut that has been employed. The finding that tested EF was not associated with early seizure debut fits with findings, suggesting that early seizure debut is primarily associated with lower IQ in pediatric populations (Hoie et al., 2005). Other seizure variables, such as seizure frequency, showed no association with any of the three modes of measuring executive problems, employed in this study. AED monotherapy can have cognitive and executive side effects (Aldenkamp et al., 1993;Hessen et al., 2006Hessen et al., , 2007a and AED polytherapy may have an additional increased risk for cognitive side effects (Aldenkamp et al., 1993). However, in this study, we found no association between AEDs and EF, in particular AED polytherapy and executive deficits.
Executive function is an umbrella term comprising very different cognitive processes and behavioral competencies including initiation, planning, verbal reasoning, problem-solving, the ability to sustain attention, resistance to interference, self-monitoring, Another limitation of the study is the recruitment bias. All patients were included consecutively while hospitalized at The National Centre for Epilepsy, that is, a tertiary epilepsy center. They may be representative for patients fulfilling our selection criteria, but not for children with epilepsy in a population-based setting, as the majority of persons with epilepsy are well controlled, mostly seizure free (Kwan & Sander, 2004).
In conclusion, the main finding of this study was a strong association between parent-reported everyday cognitive executive dysfunction and tested executive dysfunction. Furthermore, male gender showed the strongest association with parent-reported everyday behavioral regulation problems, probably related to a higher frequency of symptoms associated with psychopathology among the boys than the girls. Despite strong associations between the modes of measurement, the frequency of executive deficits according to the three different modes varied much, from 16% to 43%, suggesting that they capture different aspects of behavior under the executive umbrella. These diverse findings based on measures that all are named as executive measures raise the question of whether the term EF may be too broad to be meaningful.

ACK N OWLED G M ENTS
Financial support was obtained from The National Centre for Epilepsy and The Department of Research, Division of Surgery and Clinical Neuroscience, Oslo University Hospital, and from the Norwegian Chapter of the ILAE. We thank all participants and parents for the time and effort and for their trust and openness.

CO N FLI C T O F I NTE R E S T
None of the authors has any conflict of interest related to this article to disclose.