Prevalence and risk factors for attention and working memory impairment in children with epilepsy attending the University of Benin Teaching Hospital, Benin City: a comparative cross-sectional study

¹University of Benin Teaching Hospital, Benin City, Edo state, Nigeria

^&Corresponding author
Paul Ehiabhi Ikhurionan, University of Benin Teaching Hospital, Benin City, Edo state, Nigeria

Introduction    

Children with epilepsy (CWE) experience a wide range of developmental and functional challenges, including cognitive and behavioral problems [1]. Cognitive impairment occurs more frequently in CWE compared to non-epileptic controls [2-7]. Cognitive impairment may result in academic challenges and other functional difficulties in these children [8,9]. Regardless of intelligence scores, specific cognitive domains problems are reported in children with epilepsy, and this may ultimately affect the prognosis and quality of life of these children [5,10]. Problems with attention and memory are among the most common problems reported by parents of CWE, as well as the result of several studies [5,11-14]. Some researchers have suggested that significant attention deficit hyperactivity disorder (ADHD) symptoms are present in more than one-third of CWE with inattentive subtype being the most commonly diagnosed [15-17]. Memory and learning difficulties have also been shown to be common in CWE and affect both the verbal and visual domains [5]. A prevalence rate of 32.1% was reported among 106 CWE in the United States and Canada, compared to (1.9%) in the general population [3]. Studies from Israel [5], Portugal [14] and Jamaica [18] also show significant difference in memory scores between CWE and controls. Although, data from Africa are still very scanty, a study by Bioh et al. [7] reported that there was no significant difference in memory function among 50 Ghanaian CWE compared to healthy controls. Several factors contribute to cognitive problems in children with epilepsy. These include sociodemographic factors, aetiology of disease; seizure characteristics and treatment-related factors. The age at seizure onset [6,11,19] and type of seizure [15,20] has been identified as likely factor contributing to attention and memory impairment in children with epilepsy. Other factors include duration of active epilepsy [11], seizure frequency [6], and treatment. Although, many studies have shown significant association of attention and memory problems with epilepsy in children, there remains a dearth of knowledge on the prevalence and associated factors in Nigeria. Unrecognised impairment may result in poor academic performance and ultimately poor quality of life. The current study sought to determine the prevalence of attention and memory impairments in children with epilepsy and compare them with controls. The study also examined the influence of epilepsy-related variables (type of seizure, age at epilepsy onset, duration of active epilepsy, frequency of seizures, and treatment) on attention and memory functions. The findings will give a picture of attention and memory impairment in CWE. This understanding is crucial to guide the management of CWE as well as providing academic support for those with impairment.

Methods     

Study design: a comparative cross-sectional study of children with epilepsy (CWE) and healthy controls was conducted by comparing cognitive scores between groups. This design was employed to suit the nature of the study.

Study setting: the study was conducted at Paediatric Neurology Clinic of the University of Benin Teaching Hospital between October 2017 and November 2018. The University of Benin Teaching Hospital is an 860-bed capacity tertiary health facility located in Benin city, Edo state, Nigeria. Edo State is strategically located between the south-western and south-eastern subregions of Nigeria and lies within latitude 50 30' and 70 25´ north of the equator and longitudes 50 00´ and 60 50´ east of the Greenwich Meridian. The 2006 national census put the population of Edo State at 3,233,366 people [21]. The University of Benin Teaching Hospital serves as the major referral centre for hospitals within the state and from neighbouring states of the southern region of Nigeria. The paediatric neurology clinic, which serves about 500 children with neurologic conditions from infancy till they turn 18 years. Neurologic conditions such as epilepsy and epileptic encephalopathies, cerebral palsies, developmental disorders, movement disorder, autistic spectrum disorders, attention deficit hyperactivity disorders, paediatric stroke, neurocutaneous syndromes and the muscle dystrophies are seen in the clinic. Services offered include diagnostic workup, pharmacologic treatment, monitoring/follow up, and specialized care. The clinic is run by two consultant paediatricians specialised in paediatric neurology, supported by four paediatric trainees, six nurses, one occupational therapist, laboratory and pharmacy staff on clinic days.

Study population: subjects included children with idiopathic epilepsy (defined as epilepsy for which no cause is identified) aged six to 16 years. The subjects comprised of 80 CWE who were already on treatment for at least three months and 80 CWE who were newly diagnosed and not yet commenced on treatment. The control group composed of children without epilepsy presenting at the General Practice Clinic for medical examinations for school entry and placement, as well as those presenting for follow-up visit after brief acute illnesses within the study period. A semi-structured pro forma was used to collect information such as the socioeconomic, demographic characteristics, seizure related data, and physical examination. Children with a history suggestive of intrauterine hypoxia, bilirubin encephalopathy, neonatal seizures, complicated meningitis and cerebral malaria as well as those with hearing and/or visual impairment that limited their ability to take the test were excluded. Children on anxiety or depression treatment at the time of testing were also excluded because of the influence of non-antiepileptic drugs (non-AED) on neuropsychological function. No exclusion was made based on intellectual capacity in order to increase generalisation.

Sampling: Epi Info version 7.0 (Centre for Disease Control and Prevention, Atlanta, Ga) was used to calculate the sample size using the formular for unmatched case-control qualitative data. The prevalence of cognitive impairment in CWE was set at 14.2% and the prevalence of impairment in health control was set at 4.7% based on previously conducted research [22,23]. To have a power of at least 80% to detect meaningful differences between groups (CWE vs controls), 160 cases and 80 controls were necessary. The nature of the study population frame was not in favor of randomization; hence, a convenience sampling technique in recruiting the study participants was adopted. We consecutively enrolled participants who fulfilled the inclusion criteria for the study groups until the sample size was achieved. Cases and controls were recruited at the same time.

Study variables: independent variables that were considered included age at seizure onset (<5 years, >5 years), duration of epilepsy (<5 years, >5 years), type of seizure (Generalised vs. Focal), seizure frequency in previous 2 years (<1 seizure per year and >1 seizure per year), AED status (CWE on treatment versus newly diagnosed CWE yet to be commenced on AED) and duration of treatment (<5 years, > 5 years). The duration of epilepsy was calculated as the difference between the current age and the age at onset of epilepsy (in months) and thereafter, categorised into the two groups. Seizure frequency and antiepileptic drug (AED) status were classified as at the time of neuropsychological testing.

Procedures: the Iron psychology computerised test battery [24] was used to assess attention and memory function in subjects and controls. The test contains a set of computerised tasks designed to assess a variety of cognitive functions: motor performance and reaction time; attention and concentration; memory and laterality. The memory function was assessed using the recognition memory test (RMT - word and figure) while attention was assessed using the fixed binary choice reaction. The test was administered in a reasonably quiet and well-lit room at a room temperature between 20°C to 25°C. The subject sat at 40 cm to 60 cm from the visual display screen of the computer. Effort was also made to ensure adequate brightness and contrast of the screen with adequate sound of the computer speakers.

Attention: the binary choice reaction test was used to assess attention and mental speed. The patient has to react to squares resented on either side of the computer screen. A red square was presented on the left side of the screen, while a green square was presented on the right side. Only one square was presented at a time on the screen. The interval and sequence of presentation was varied randomly. A total of sixty stimuli presentation constituted a test session. Reaction time was a measure of divided attention and the decision-making process. The test score is the binary choice reaction time in milliseconds. The criterion for impairment was fixed at two standard deviations (SD) worse than the mean of age-matched controls.

Memory test: verbal and visual working memory was assessed using the word and figure recognition test of the Fepsy computerised test battery. The tests involved a learning phase and a test phase. Four words and four figures are presented for one second during the learning phase and then disappear from the screen. After a delay of two seconds, the screen shows one of these words/figures between three other distracters. The target item has to be identified among the distracters. The test score is the percentage of correct responses out of 24 items. Similarly, the criterion for impairment was fixed at two standard deviations (SD) worse than the mean of age-matched controls.

Bias: to limit the challenge with recall bias, information for the CWE on treatment were extracted from their case notes. Those not on drugs had mainly recent onset epilepsy and thus their information were still quite recent. The psychometric assessments of the participants were performed by one of the authors in the same order, using the same computer to limit bias.

Data analysis: the collected data were organised, tabulated and statistically analysed using the International Business Machines Corporation (IBM) Statistical Package for the Social Sciences (SPSS) version 21.0 (SPSS for Window Inc; Chicago, LL, USA) Statistical Software. Level of cognitive function was dichotomised into normal and impaired. Attention and working memory function were impaired if the performance was more than two standard deviation (SD) worse than the mean of age-matched controls. The socio-economic class of study subjects and controls were classified into the upper, middle and lower socio-economic class based on the fathers´ profession and mothers´ level of education according to the method described by Olusanya et al. [25]. The independent variables included age at seizure onset, duration of epilepsy, type of seizure, seizure frequency in previous 2 years, AED status, duration of treatment and number of AEDs. Seizure frequency and antiepileptic drug (AED) status and number were classified as at the time of neuropsychological testing. The strength of association was assessed by calculating the odds ratio while multivariate analysis was used to determine which independent variable(s) were associated with impairment. A two-tailed p value of less than 0.05 was accepted as the level of statistical significance.

Ethical consideration: informed consent to participate in the study was obtained from parents or legal guardians after explanation of the study objectives and relevance, risks and benefits. Approval was obtained from the Ethics and Research Committee of the University of Benin Teaching Hospital, Benin City. Participation in the study was entirely voluntary and participants were advised that they could withdraw from the study at any point. Personal identifiers like name and addresses were not collected; rather each participant was given a study code which was linked with their case file number.

Results     

Socio-demographic characteristics: one hundred and sixty children with epilepsy and 80 healthy controls were enrolled for this study. The mean age of the subjects was 10.65 ± 2.90 years and the mean age of the control was 11.01± 2.61 years. The male to female ratio for subjects and control was 1.03 and 1.05 respectively. Majority of the participants were from the upper socio-economic class (Subject=42.5%; controls=77.5%). The summary of the demographic and clinical characteristics of the study population is provided in Table 1.

Prevalence of impaired attention and memory function: thirty-five (21.9%) subjects and three (3.8%) controls had impaired attention while, 17 (10.6%) subjects and one (1.3%) control had impaired verbal memory. Impairment of visuo-spatial memory was recorded in five (3.1%) subjects and four (5.0%) controls. The odds ratios (OR) were 7.19 (95% confidence interval (CI)=2.14-24.17, p<0.001) and 9.39 (95%CI=1.23-71.90, p=0.009) for impaired attention and verbal memory respectively (Figure 1).

Factors associated with attention and memory impairment in children with epilepsy: the bivariate analysis and odds ration are presented in Table 2 (for attention) and Table 3 (for verbal memory). Subjects whose seizures started before their fifth birthday were more likely than those whose seizures started later to have impaired attention (OR=6.4; 95%CI=2.7-15.4; p<0.001). Seizure duration less than five years (OR=0.3; 95%CI=0.16-0.74; p=0.005) and focal seizures (OR=0.27; 95%CI=0.08-0.95; p=0.031) were associated with less likelihood of impaired attention than those with seizure duration greater than five years and generalised seizures respectively (Table 2). Similarly, seizure onset before five years of age (OR=7.25; 95%CI=1.99-26.37; p=0.001) significantly increased the likelihood of having verbal memory impairment while duration of disease less than five years (OR=0.84; 95%CI=0.77-0.91; p=0.002) reduced the likelihood (Table 3). However, in multivariate analysis, only age at seizure onset less than five years remained significantly associated with impaired attention (OR=4.794; 95%CI=2.070-11.105; p<0.001) and impaired verbal memory (OR=0.222; 95%CI=0.069-0.716; p=0.012).

Discussion     

Our findings indicate that attention and verbal memory problems are significantly higher in CWE than in controls. In contrast, the prevalence of visuo-spatial memory impairment appears to be similar in CWE and controls. Analysis of the results according to seizure characteristics revealed that children with seizure onset before their fifth birthday had a higher odd of impairment than those with later onset. Children with duration of epilepsy less than 5 years were less likely than those with longer duration of disease to have impairment of attention and verbal memory function. The type of seizure only had significant influence on attention but not on verbal memory impairment. In our sample, children with epilepsy had significantly higher prevalence of impairment relative to healthy control. The prevalence of impaired attention in CWE was more than five times the prevalence in controls. This observation confirms the findings from other studies which show significant association between epilepsy in children and impaired attention [13,26,27]. Berl and colleagues [12] in Washington using the Test for Everyday Attention for children (TEA-Ch) battery and excluding children whose FSIQ was <70 reported significantly worse performance of complex and timed attention tasks in CWE compared to controls. The prevalence of impaired attention found in this study was lower than that reported by Berl [12] (64 - 68%). The lower prevalence might be because in this study impairment was defined as performance worse than 2SD below the average for age whereas, Berl defined impairment as scores less than one standard deviation (SD) below the average for age.

Regarding working memory task, the prevalence of verbal memory impairment in CWE was more than eight times the prevalence in healthy control group. However, the prevalence of impaired visual memory in CWE was comparable to controls in this study. This result agrees with findings by other researchers. Kernan et al. [6] Henkin et al. [5] and Rantanen et al. [28] in separate studies reported that children with epilepsy had poorer performance in domains of verbal memory but not visuo-spatial memory compared to controls. Lopes and co-workers however, examined memory abilities in 90 children with common childhood epilepsy syndromes and compare them with 30 matched controls [14]. They found significant deficit in both verbal and visuospatial memory in children with frontal lobe epilepsy. The selective effect of epilepsy on verbal memory and not visuo-spatial memory in the current study is not clearly understood. The age at onset of epilepsy was significantly associated with impairment of attention and verbal memory in CWE in this study. Subjects with seizure onset in the first five years of life were more likely than those whose seizure started later to have impairment of attention and verbal memory. A strong relationship between early onset epilepsy and impairment of attention and verbal memory has been previously reported in the literature [6,11,19]. The maturation of the frontal lobe and neural network responsible for attention and verbal working memory begins in early childhood and continues throughout adolescence. The occurrence of seizures at an earlier age, therefore, interferes with developmental changes and adversely affects structural and cognitive functions associated with the frontal lobe. This study thus provides further evidence of the importance of the age at onset of seizures for cognitive functioning.

Limitations: our study has some important limitations which should be addressed. The impact of socio-economic status on cognition is established [18,28-31]. The inadvertent addition of more children from the upper socio-economic class in the control group compared to the subjects may account for some variations observed. The cross-sectional nature of this study clearly limits its ability to establish causal relationship between seizure characteristics and impairments observed.

Conclusion     

The present study demonstrates that children with epilepsy show significant deficits in attention and verbal memory but not visuo-spatial memory. In addition, earlier age at epilepsy onset and longer duration of epilepsy were associated with attention and memory problems. Our findings underscore the need for attention and memory function assessments in children with epilepsy especially those with an early age at epilepsy onset. This is in order to provide appropriate and timely intervention when needed.

Funding: the research was self-funded by the authors and no funding was received from any funding body or organisation.

Competing interests     

The authors declare no competing interests.

Authors' contributions     

Paul Ehiabhi Ikhurionan; conceptualization and design, data collection, data analysis, interpretation of data, drafting of article and final approval of the version to be published. Olusola Peter Okunola; conceptualization and study design, data collection, interpretation of data, revision of draft critically for important intellectual content; and final approval of the version to be published Blessing Imuetinyan Abhulimhen-Iyoha; conceptualization and design, data analysis, interpretation of data, drafting the article and final approval of the version to be published Gabriel Egberue Ofovwe; conceptualization and design, interpretation of data, revision of draft critically for important intellectual content and final approval of the version to be published.

Acknowledgments     

The authors acknowledge the contribution of the following persons for their assistance in data collection and manuscript preparation. Ughiemosomhi Pamela Ikhurionan, Jonathan Amake, Emmanuel Eyo-Ita, Ifueko A. Eyo-Ita, Catherine Oside. The study was self-sponsored by the authors and no funding was received from any funding body for the study.

Tables and figure     

Table 1: demographic and clinical characteristics of study population

Table 2: association between seizure characteristics and impaired attention in subjects

Table 3: association between seizure characteristics and impaired verbal memory in subjects

Figure 1: prevalence of impairment of attention and memory in subjects and controls

References