Antecedents and neuroimaging patterns in cerebral palsy with epilepsy and cognitive impairment: a population-based study in children born at term

Transcription

1 AOGS IGINAL RESEARCH ARTICLE Antecedents and neuroimaging patterns in cerebral palsy with epilepsy and cognitive impairment: a population-based study in children born at term KRISTINA AHLIN 1, BO JACOBSSON 1,2, STAFFAN NILSSON 3 & KATE HIMMELMANN 4 1 Department of Perinatal Center, Department of Obstetrics and Gynecology, Institute for Clinical Sciences, Sahlgrenska Academy, Sahlgrenska University Hospital/ Ostra, Gothenburg, Sweden, 2 Department of Genes and Environment, Division of Epidemiology, Institute of Public Health, Oslo, Norway, 3 Department of Mathematical Statistics, Institute for Mathematical Sciences, Chalmers University of Technology, Gothenburg, and 4 Department of Pediatrics, Institute for Clinical Sciences, Sahlgrenska Academy at the University of Gothenburg, Gothenburg, Sweden Key words Cerebral palsy, cognitive function, epilepsy, neuroimaging, risk factors Correspondence Kristina Ahlin, Department of Obstetrics and Gynecology, Perinatal Center, Sahlgrenska University Hospital/ Ostra, SE Gothenburg, Sweden. kina_ahlin@hotmail.com Conflict of interest The authors have stated explicitly that there are no conflicts of interest in connection with this article. Please cite this article as: Ahlin K, Jacobsson B, Nilsson S, Himmelmann K. Antecedents and neuroimaging patterns in cerebral palsy with epilepsy and cognitive impairment: a population-based study in children born at term. Acta Obstet Gynecol Scand 2017; 96: Received: 4 October 2016 Accepted: 5 March 2017 DOI: /aogs Abstract Introduction. Antecedents of accompanying impairments in cerebral palsy and their relation to neuroimaging patterns need to be explored. Material and methods. A population-based study of 309 children with cerebral palsy born at term between 1983 and Prepartum, intrapartum, and postpartum variables previously studied as antecedents of cerebral palsy type and motor severity were analyzed in children with cerebral palsy and cognitive impairment and/or epilepsy, and in children with cerebral palsy without these accompanying impairments. Neuroimaging patterns and their relation to identified antecedents were analyzed. Data were retrieved from the cerebral palsy register of western Sweden, and from obstetric and neonatal records. Results. Children with cerebral palsy and accompanying impairments more often had low birthweight (kg) (odds ratio 0.5, 95% confidence interval ), brain maldevelopment known at birth (p = 0.007, odds ratio ) and neonatal infection (odds ratio 5.4, 95% confidence interval ). Moreover, neuroimaging patterns of maldevelopment (odds ratio 7.2, 95% confidence interval ), cortical/ subcortical lesions (odds ratio 5.3, 95% confidence interval ) and basal ganglia lesions (odds ratio 7.6, 95% confidence interval ) were more common, wheras white matter injury was found significantly less often (odds ratio 0.2, 95% confidence interval ). In most children with maldevelopment, the intrapartum and postpartum periods were uneventful (p < 0.05). Cerebral maldevelopment was associated with prepartum antecedents, whereas subcortical/cortical and basal ganglia lesions were associated with intrapartum and postpartum antecedents. Conclusion. No additional factor other than those related to motor impairment was associated with epilepsy and cognitive impairment in cerebral palsy. Timing of antecedents deemed important for the development of cerebral palsy with accompanying impairments were supported by neuroimaging patterns. Abbreviations: CP, cerebral palsy; NE, neonatal encephalopathy; NICU, neonatal intensive care unit;, odds ratios; WMI, white matter injury. Introduction Cerebral palsy (CP), a group of disorders affecting body movement, balance, and posture, occurs in two per 1000 live births. The motor disorders of CP are often accompanied by a seizure disorder and disturbances of cognition (1,2). CP is heterogeneous, with different subtypes, motor severity, and accompanying impairments. We have 828

2 K. Ahlin et al. Antecedents to accompanying impairments in CP previously shown that antecedents to CP vary with subtype (3,4) and degree of motor severity (5) in children born at term. To further study the complexity of CP, the antecedents to accompanying impairments in individuals with CP need to be elucidated. Previous studies have linked epilepsy and cognitive impairment in CP to the type, extent and site of the lesion and have investigated the association between a few etiological factors involved in CP and neuroimaging findings (1,6 8). However, large, population-based studies investigating the association between prepartum, intrapartum, and postpartum variables, and the occurrence of epilepsy and cognitive impairment, in children born at term, are scarce. We hypothesized that additional factors, other than those related to the motor impairment, may be important for cognitive impairment and epilepsy, and that a specific timing of injury is associated with accompanying impairments in CP. A better understanding of the variables associated with cognitive impairment and epilepsy in CP and the timing of the insult during brain development is desired to depict different prognostic or treatment groups and to shed light on the pathways to brain injury. The aim of this study was therefore to analyze the association between a large set of prepartum, intrapartum, and postpartum variables, and the occurrence of epilepsy and cognitive impairment, in children with CP born at term. An additional aim was to study the relation of neuroimaging findings to the identified antecedents. Material and methods Study design Population-based cohort study. Prepartum, intrapartum, and postpartum variables, previously analyzed for their association with CP subtype and severity of motor impairment (3,4), were now also analyzed for their association with two selected accompanying impairments, epilepsy and cognitive impairment. were from twin pregnancies. Details on these cohorts have previously been published (3,4,7,9 11). Data were retrieved from obstetrical records from antenatal care and delivery and discharge charts from the neonatal intensive care unit (NICU). Two investigators (KA and BJ) scrutinized the 309 obstetric and neonatal records, recording a total of 80 variables. The variables commented on in the article are defined below. All definitions of the analyzed variables are presented in the Supplementary material (Appendix S1). Additional information on CP type, motor severity, accompanying impairments, and neuroimaging reports were retrieved from the CP register of western Sweden. Gestational age was based on ultrasound at gestational weeks in 97% and on the last menstrual period for the remaining 3%. Small-for-gestational-age: >24% weight deviation below the ultrasound-estimated weight for a given gestational age and sex, corresponding to < 2 standard deviations from the mean; the latter is a more common definition in pediatric contexts. Brain maldevelopment was identified by ultrasound before birth, and early identification of brain maldevelopment was when this diagnosis was present on the NICU discharge reports. Instrument delivery refers to cesarean delivery or vacuum extraction. NICU care was not restricted to a specific diagnosis or condition and included all children in need of observation or treatment. Neonatal encephalopathy (NE) was diagnosed according to Badawi et al. based on seizures alone, or any of the following lasting for longer than 24 h: altered consciousness, difficulty maintaining respiration (of presumed central nervous origin), difficulty feeding (of presumed central nervous origin) or abnormal tone or reflexes (12). Congenital infections that occurred were cytomegalovirus infection, toxoplasmosis, or varicella. Neonatal infection was defined as a verified infection during an NICU stay and any of the following: pneumonia, septicemia, meningitis, encephalitis, neonatal urinary tract infection or ventriculitis. Diagnosis of CP was made according to Mutch et al. as a group of non-progressive, but often changing, motor impairment syndromes that are secondary to lesions or Study population and definitions We studied all children diagnosed with CP and born at or after 37 weeks of gestation during in western Sweden. The children were at least 4 years old at the time of diagnosis (4 6) and 356 children born at term were diagnosed with CP. Postneonatal cause of CP (n = 21), spinal malformation (n = 1) and ataxic CP (n = 25), a subtype that is difficult to distinguish from other non-cp syndromes, were excluded, leaving a total of 309 children, of which 307 were singletons and two Key Message Antecedents of accompanying impairments in cerebral palsy were explored. Epilepsy and cognitive impairment share similar antecedents with severe motor impairment, namely low birthweight, brain maldevelopment, and neonatal infection. Timing of antecedents could be confirmed by identified neuroimaging patterns. 829

3 Antecedents to accompanying impairments in CP K. Ahlin et al. abnormalities of the brain arising in the early stages of development (13). Cognitive impairment documented was defined as an IQ of <70. Epilepsy was defined as a diagnosis of active epilepsy at inclusion in the register at 4 8 years of age. Neuroimaging reports of MRI and CT were retrospectively analyzed and classified by one of the investigators (KH) into five categories: brain maldevelopment, whitematter injury (WMI), cortical and basal ganglia lesions, miscellaneous, and normal finding, according to Kr ageloh-mann (14). Analyses were performed to ascertain whether identified antecedents to accompanying impairments in CP correlated with specific brain imaging patterns. The study was approved by the Regional Ethical Review Board in Gothenburg (432-13), Sweden. Data analyses The Statistical Package for Social Sciences (SPSS) for Windows version 20 was used for data analyses (SPSS Inc., Chicago, IL, USA). The significance tests were twotailed, and a significance level of 0.05 was chosen. Fisher s Exact Test was used for comparisons between groups and p-value in univariate analyses, and the Mann Whitney U-test was used for continuous variables. Binary logistic regression was performed and odds ratios () with 95% CI were calculated. The for a child with CP of being diagnosed with epilepsy, cognitive impairment, and both, respectively, was calculated for every study variable and neuroimaging pattern. In the latter analysis children without epilepsy and cognitive impairment were used as references. Multivariate analyses, using logistic regression with forward selection, were performed to investigate the influence of confounding factors. We performed logistic regression using two different models. Model A included all antecedents yielding p < 0.1 in univariate analyses. Model B excluded the antecedents Apgar <7 at 5 min, admission to NICU, mode of delivery parameters, meconium-stained amniotic fluid, and NE as these excluded antecedents are not in themselves regarded as potential etiological factors, but instead as either markers of vitality or intermediates between exposure and outcome, and when significantly related to the exposure, they may prevent earlier antecedents from becoming significant in the final model. The relation of neuroimaging findings to the identified antecedents was studied using multivariate analysis, forward selection including all antecedents displayed in Table 3. Interrelationship analyses were performed using Spearman s rank correlation. Several of the studied variables were highly interrelated. For the significant variables these relationships have been summarized in the Supplementary material (Table S1). Results Significant results from the analyses of antecedents associated with epilepsy and cognitive impairment are presented in Table 1. All results analyzing prepartum, intrapartum, and postpartum variables and associations with CP accompanied by epilepsy and/or cognitive impairment are presented in the Supplementary material (Tables S2 and S3). More than one-third had epilepsy and two-thirds had cognitive impairment. Most of the children with epilepsy also had a cognitive impairment (Figure 1). Associations with epilepsy and cognitive impairment according to CP types demonstrate that both epilepsy and cognitive impairment were particularly common in children with bilateral spastic CP, present in 64% compared with 11% of children with spastic hemiplegia, and 25% of those with dyskinetic CP (S2 3). Children with epilepsy and cognitive impairment had a severe motor impairment (Gross Motor Function Classification System level IV V) in 73% (S2 3). We found a substantial convergence of variables significantly associated with epilepsy and the variables associated with cognitive impairment when analyzing these impairments separately (S2). Therefore, the two are presented together, termed accompanying impairments. The analysis was performed in the 88 children with accompanying impairments compared with 158 children without these accompanying impairments (S3). Prepartum antecedents Early identified brain maldevelopment ( 11.5, 95% CI ), suspected small-for-gestational-age ( 4.5, 95% CI ), birth length (cm) ( 0.9, 95% CI ), birthweight (kg) ( 0.5, 95% CI ), head circumference at birth (cm) ( 0.8, 95% CI ), small-for-gestational-age ( 3.0, 95% CI ), and congenital infection (, 95% CI 1.2 ) were associated with the presence of accompanying impairments (Table 1, and see Supplementary material, Table S2). Intrapartum and postpartum antecedents Instrumental delivery ( 2.4, 95% CI ), meconium-stained amniotic fluid ( 2.9, 95% CI ), Apgar score at 5 min <7 ( 2.6, 95% CI ) and meconium aspiration (, 95% CI 2.2 ) were significantly associated with the occurrence of accompanying impairments. NE was diagnosed in 41% ( 4.3, 95% CI ) and 64% received care in the NICU ( 2.0, 95% CI ). In comparison, children without accompanying impairments had significantly fewer of all 830

4 K. Ahlin et al. Antecedents to accompanying impairments in CP Table 1. Factors associated with epilepsy (EP) and cognitive impairment (CI) in cerebral palsy (CP). CP+EP+CI (n = 88) CP EP CI (n = 157) Univariate analyses Multivariate analyses A Multivariate analyses B Variable n or mean % or SD n or mean % or SD p-value (95% CI) p-value (95% CI) p-value (95% CI) Prepartum factors Suspected small for ( ) gestational age Decreased symphyseal ( ) fundal measurements Birth length (cm) ( ) Birthweight (kg) ( ) ( ) Head circumference at ( ) ( ) birth (cm) Small for gestational ( ) age Early identified ( ) 0.02 <0.001 maldevelopment Congenital infection (1.2 ) Intrapartum factors Cesarean section ( ) Instrumental delivery ( ) Meconium-stained ( ) amniotic fluid Apgar score at ( ) 5 min <7 Meconium aspiration (2.2 ) Neonatal < ( ) < ( ) encephalopathy Postpartum factors Neonatal infection ( ) ( ) Antibiotic therapy ( ) Admitted to NICU ( ) n (%) is presented for categorical variables. Mean (SD) is presented for continuous variables. Fisher s Exact test was used for comparison between groups and for p-value (two-sided) and the Mann Whitney U test was used for continuous variables. Cesarean section (elective, acute, emergency), instrumental delivery (cesarean section, vacuum extraction). The odds ratio () and 95% CI for a child of being diagnosed with CP and epilepsy and cognitive impairment were calculated for every study variable, using children without epilepsy and cognitive impairment as references. The estimates for birthweight are calculated to show the increased risk of CP per kg of body weight. Forward logistic regression was used for multivariate analyses. Multivariate analyses model A included all variables with p < 0.1 from univariate analyses (SIII), model B included all the variables p < 0.1 from univariate analyses (SIII) except Apgar, NICU, mode of delivery parameters, meconium stained amniotic fluid and NE. intrapartum and postpartum events described (Table 1). Ten percent of the individuals with accompanying impairments had a neonatal infection ( 5.9, 95% CI ) (Table 1 and Figure 2). Multivariate analyses Multivariate analysis model A: head circumference at birth (cm) ( 0.8, 95% CI ), early identified brain maldevelopment (p < 0.001, ) and NE ( 6.6, 95% CI ), were significantly associated with accompanying impairments in the final model (S3). Multivariate analysis model B: birthweight (kg) ( 0.5, 95% CI ) as well as neonatal infection ( 5.4, 95% CI ), not displayed in the first model, were significant, apart from early identified brain maldevelopment (p < 0.001, ), which was also significant in analysis A (Table 1). Neuroimaging Neuroimaging had been performed in 236 of 309 individuals (76%), in 75 of 88 with both accompanying impairments (53 CT, 22 MRI), and in 109 of 157 without 831

5 Antecedents to accompanying impairments in CP K. Ahlin et al. Individuals with Cerebral Palsy (CP) n = 309 CP + Epilepsy n = 107 CP + Cognitive impairment n = 132 Mild motor impairment n = 18 (21 %) Mild motor impairment n = 142 (90 %) Moderate motor impairment n = 6 (7 %) CP + Epilepsy + Cognitive impairment n = 88 CP - Epilepsy Cognitive impairment n = 158 Moderate motor impairment n = 3 (2 %) Severe motor impairment n = 64 (73 %) Severe motor impairment n = 13 (8 %) Spastic Diplegia and tetraplegia n = 56 (64 %) Spastic Hemiplegia n = 10 (11 %) Dyskinetic CP n = 22 (25 %) Spastic Diplegia and tetraplegia n = 32 (20 %) Spastic Hemiplegia n = 115 (73 %) Dyskinetic CP n = 11 (7 %) Figure 1. The recruited cohort. Distribution of accompanying impairments among children with cerebral palsy born at term between 1983 and Table 2. Neuroimaging patterns in children with cerebral palsy (CP) and epilepsy (EP) and cognitive impairment (CI). CP+EP+CI (n = 75) CP EP CI (n = 108) Univariate analyses Variable n % n % p-value (95% CI) Maldevelopment ( ) WMI < ( ) Cortical/subcortical ( ) lesions Basal ganglia lesions ( ) Miscellaneous findings ( ) Abnormal scans ( ) n (%) is presented for categorical variables. % was calculated on 75 CP+EP+CI cases and 108 CP EP CI cases as 13 and 49 reports were missing. Fisher s Exact test was used for comparison between groups or dichotomous variables and for p-value (two-sided). accompanying impairments (95 CT, 14 MRI). MRI and CT reports revealed findings of maldevelopment (28%), WMI (17%), cortical/subcortical lesions (28%), basal ganglia lesions (7%), miscellaneous lesions (9%), and normal findings (11%) among the children with accompanying impairments. Neuroimaging patterns differed by the presence of accompanying impairments. The children with accompanying impairments more often had maldevelopment ( 4.3, 95% CI ) and cortical/subcortical lesions ( 3.1, 95% CI ), and were less likely to have WMI ( 0.2, 95% CI ) or normal scans ( 0.3, 95% CI ) (Table 2). The neuroimaging patterns correlated with the antecedents associated with accompanying impairments (Table 3). Children with maldevelopment had more often undergone a congenital infection ( 8.1, 95% CI ), were shorter (cm) ( 0.8, 95% CI ), had a smaller head circumference (cm) ( 0.8, 95% CI ) and less co-habitation with the father ( 3.4, 95% CI ) compared with children with other neuroimaging patterns. Furthermore, children with maldevelopment had less often experienced intrapartum and postpartum adverse events. None of these children had an Apgar score <7 at 5 min (p < 0.001), and were less likely to have meconium-stained amniotic fluid ( 0.3, 95% CI ), to have sustained NE ( 0.2, 95% CI ) or to have been admitted to the NICU ( 0.5, 95% CI ) compared with children with other neuroimaging patterns. Also in the WMI group, few children had Apgar score <7 at 5 min ( 0.3, 95% CI ), NE ( 0.2, 95% CI ) or were admitted to NICU ( 0.4, 95% CI ). A different pattern was identified for the children with subcortical/cortical and basal ganglia lesions, where meconium-stained amniotic fluid ( 2.5, 95% CI ), Apgar score <7 at 5 min ( 3.5, 95% CI ), meconium aspiration ( 9.4, 95% CI ), NE ( 5.5, 95% CI ) and admittance to NICU ( 5.2, 95% CI ) were 832

6 K. Ahlin et al. Antecedents to accompanying impairments in CP Table 3. Associations between antecedents and neuroimaging findings among children with cerebral palsy (n = 309). Maldevelopment (n = 39) WMI (n = 74) Cortical/subcortical & basal ganglia lesions (n = 53) Univariate Multivariate Univariate Multivariate Univariate Multivariate n or Variable mean % or SD p-value n or % n or % (95% CI) p-value (95% CI) mean or SD p-value (95% CI) p-value (95% CI) mean or SD p-value (95% CI) p-value (95% CI) Prepartum factors Birthweight (kg) ( ) ( ) ( ) Birth length (cm) < ( ) < ( ) ( ) ( ) Head ( ) ( ) ( ) circumference at birth (cm) Suspected SGA ( ) ( ) ( ) Small for gestational age Noncohabitation with baby s father ( ) ( ) ( ) ( ) ( ) < ( ) Congenital infection ( ) ( ) Intrapartum factors Cesarean section ( ) ( ) ( ) Instrumental ( ) ( ) ( ) delivery Meconium-stained amniotic fluid Apgar score at 5 min < ( ) ( ) ( ) ( ) 0 0 < ( ) < ( ) Meconium aspiration ( ) Neonatal encephalopathy Postpartum factors Antibiotic therapy postpartum ( ) 8 11 < ( ) < ( ) < ( ) < ( ) ( ) ( ) ( ) ( ) Neonatal infection ( ) ( ) ( ) Admitted to NICU ( ) ( ) < ( ) n (%) is presented for categorical variables. Mean (SD) is presented for continuous variables. Fisher s Exact test was used for comparison between groups or dichotomous variables and the Mann Whitney U test was used for continuous variables., Odds Ratio. The estimates for birthweight are calculated to show the increased risk of brain lesion per kg of body weight. Cesarean section (elective, acute, emergency), instrumental delivery (cesarean section, vacuum extraction). and for p-value (two-sided) and the Mann Whitney U test was used for continuous variables. Multivariate analysis, forward selection was used and all variables displayed in the table were included. 833

7 Antecedents to accompanying impairments in CP K. Ahlin et al. much more common (Table 3). As demonstrated in Table 3, multivariate analysis highlights the temporal difference among the antecedents related to the different neuroimaging patterns. Discussion The occurrence of prepartum, intrapartum, and postpartum factors has been shown to vary by the CP subtype (10,11) and severity of motor impairment (5) in a case control study of this cohort. In the present study, analyzing 80 prepartum, intrapartum, and postpartum variables, and neuroimaging patterns we found that poor intrauterine growth, maldevelopment, and infection postpartum, which were previously shown to be associated with increasing severity of motor impairment (5), were associated with epilepsy and cognitive impairment. Epilepsy and cognitive impairment, in the context of CP, do not seem to be manifestations of additional antecedents other than those related to the motor impairment. Moreover, our data suggest that CP and its accompanying impairments probably stem from multiple causes and that the causal pathways encompass different time periods during gestation. Similar etiologic heterogeneity has previously been reported (15,16). We found that cognitive impairment shared the same antecedents as epilepsy when analyzing epilepsy and cognitive impairment separately, probably due to the great overlap between these two outcomes. Of prepartum variables, the children with accompanying impairments more often had early identified brain maldevelopment, and lower birthweight than the children without these accompanying impairments. Our results agree with those of previous studies (3,17,18). Fetal growth is dependent on genetic, placental, and maternal factors (19,20). Regarding intrapartum factors, NE was strongly associated with epilepsy and cognitive impairment in children with CP. Neonatal seizures have been shown to be a useful marker for future epilepsy in children with CP (7,21), and our results are in accordance with these findings. It is debated whether neonatal seizures, themselves may cause injury to the developing brain (22). Regarding postpartum factors, neonatal infection was associated with epilepsy and cognitive impairment in children with CP. Wallace, in 2001 (22) and Wu et al. in 2013 (23) have shown that infection is associated with epilepsy in CP (7). The link between infection and CP has been described in the present data set (4), as well as in a review by McIntyre et al. (24). Clearly, infection early in life seems to be a potent risk factor for permanent injury. In the present study more than a third of the children with accompanying impairments were not admitted to the NICU. These children apparently showed no definite signs of injury at this early point. The number of children with severe impairments not admitted to the NICU and the severity of their impairments have been underestimated in the past and warrant greater attention, as highlighted by McIntyre (25). Maldevelopment and cortical/subcortical lesions were more often found in children with accompanying impairments whereas WMI and normal scans were less likely to be found, as previously described by Himmelmann and Uvebrant (6) and Carlsson et al. (7). Epilepsy arises in the cortex, hence the grey-matter injury patterns increase the risk of epilepsy. Staudt showed in 2010 that axonal bypasses can be reorganized after periventricular lesions in the preterm brain (26). This may partly explain why children with WMI patterns may have fewer accompanying impairments (26). Different neuroimaging patterns are related to different timing of compromise, albeit not the cause. Specific regions and cells are susceptible to insults during various periods of brain development (14). Our results highlight the temporal difference among the antecedents related to the different neuroimaging patterns. Prepartum antecedents were most frequent in the groups with maldevelopment, a neuroimaging pattern of prepartum origin (14), whereas intrapartum and postpartum antecedents were most prevalent in the children with subcortical/cortical injury and basal ganglia lesions, of presumed late prepartum or intrapartum and postpartum origin (14). We found no significant association between intrauterine growth restriction and WMI, in contrast to previous findings (27). However, the occurrence of children being small-for-gestational-age was highest in this group (14%). Meconium staining in the amniotic fluid was associated with WMI. Meconium can induce vasoconstriction causing hypoxemia ischemia in the chorionic plate of the placenta (28), which might induce prepartum intrauterine stress. Maternal antibiotic usage postpartum was significantly associated with subcortical/cortical and basal ganglia lesions, giving us a clue to one possible pathogenesis behind those brain-imaging patterns. The strength of this study is the large number of cases drawn from a population-based register, based on welldefined diagnostic criteria. There were no missing records. Because of the number of variables analyzed and the case-only design, the study is limited to being mainly hypothesis generating. Given the time period studied, the neuroimaging reports were old and neuroimaging was performed with CT in 85% of children, which may have led to lower diagnostic accuracy. Furthermore, in the analyses of associations between antecedents and neuroimaging findings we did not have access to population level data for comparison. These limitations 834

8 K. Ahlin et al. Antecedents to accompanying impairments in CP notwithstanding, this study provides valuable insight as it points to mechanisms involved in causing serious brain damage, epilepsy and cognitive impairment among individuals with CP. Conclusion No additional factor other than those related to motor impairment, namely low birthweight, brain maldevelopment, and neonatal infection were associated with epilepsy and cognitive impairment in term-born children with CP. Hence, epilepsy and cognitive impairment, in the context of CP, do not seem to be manifestations of additional antecedents. Rather, it seems that epilepsy and cognitive impairment reflect an increased severity of brain injury. Nor was there a specific timing of adverse events, although the neuroimaging pattern was more likely to reflect later, rather than early derived injury to the developing brain. The timing of injury suggested by the antecedents found, was in agreement with neuroimaging patterns. Acknowledgments We gratefully acknowledge Mattias Molin at Statistiska konsultgruppen, G oteborg, for assistance with statistics, and Dr. Elisabet Hentz for expert advice on neonatal issues. Funding We acknowledge Linnea and Josef Carlsson s Foundation, the G oteborg Medical Society, FOU-Unit in S odra Alvsborg, Swedish government grants (ALFGBG ), the Swedish Medical Society ( ), and Queen Silvia s Jubilee Fund for funding. The sponsors had no involvement in the study design; in the collection, analysis or interpretation of data; nor in the writing of the report or in the decision to submit the article. References 1. Beckung E, Hagberg G. Neuroimpairments, activity limitations and participation restrictions in children with cerebral palsy. Dev Med Child Neurol. 2002;44: Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M, Damiano D, et al. A report: the definition and classification of cerebral palsy April Dev Med Child Neurol Suppl. 2007;109: Ahlin K, Himmelmann K, Hagberg G, Kacerovsky M, Cobo T, Wennerholm U, et al. Non-infectious risk factors for different types of cerebral palsy in term-born babies: a population-based, case control study. BJOG. 2013;120: Ahlin K, Himmelmann K, Hagberg G, Kacerovsky M, Cobo T, Wennerholm U, et al. Cerebral palsy and perinatal infection in children born at term. Obstet Gynecol. 2013;122: Ahlin K, Himmelmann K, Nilsson S, Sengpiel V, Jacobsson B. Antecedents of cerebral palsy according to severity of motor impairment. Acta Obstet Gynecol Scand. 2016;95: Himmelmann K, Uvebrant P. Function and neuroimaging in cerebral palsy: a population-based study. Dev Med Child Neurol. 2011;53: Carlsson M, Hagberg G, Olsson I. Clinical and aetiological aspects of epilepsy in children with cerebral palsy. Dev Med Child Neurol. 2003;45: Himmelmann K, Beckung F, Hagberg G, Uvebrant P. Gross and fine motor function and accompanying impairments in cerebral palsy. Dev Med Child Neurol. 2006;48: Hagberg B, Hagberg G, Olow I. The changing panorama of cerebral palsy in Sweden. VI. Prevalence and origin during the birth year period Acta Paediatr. 1993;82: Hagberg B, Hagberg G, Olow I, von Wendt L. The changing panorama of cerebral palsy in Sweden. VII. Prevalence and origin in the birth year period Acta Paediatr. 1996;85: Hagberg B, Hagberg G, Beckung E, Uvebrant P. Changing panorama of cerebral palsy in Sweden. VIII. Prevalence and origin in the birth year period Acta Paediatr. 2001;90: Badawi N. Antepartum risk factors for newborn encephalopathy: the Western Australian case-control study. BMJ. 1998;317: Mutch L, Alberman E, Hagberg B, Kodama K, Perat MV. Cerebral palsy epidemiology: where are we now and where are we going? Dev Med Child Neurol. 1992;34: Kr ageloh-mann I. Imaging of early brain injury and cortical plasticity. Exp Neurol. 2004;190: El-Tallawy H, Farghaly W, Shehata G, Badry R, Rageh T. Epileptic and cognitive changes in children with cerebral palsy: an Egyptian study. Neuropsychiatr Dis Treat. 2014;10: Kułak W, Sobaniec W. Risk factors and prognosis of epilepsy in children with cerebral palsy in north-eastern Poland. Brain Dev. 2003;25: Arpino C, Curatolo P, Stazi M, Pellegri A, Vlahov D. Differing risk factors for cerebral palsy in the presence of mental retardation and epilepsy. J Child Neurol. 1999;14: Bilder D, Pinborough-Zimmerman J, McMahon W. Prenatal and perinatal factors associated with intellectual disability. Am J Intellect Dev Disabil. 2013;2: Blair E, DeGroot J, Nelson K. Placental infarction identified by macroscopic examination and risk of cerebral 835

9 Antecedents to accompanying impairments in CP K. Ahlin et al. palsy in infants at 35 weeks of gestational age and over. Am J Obstet Gynecol. 2011;205: Halliday H. Neonatal management and long-term sequelae. Best Pract Res Clin Obstet Gynaecol. 2009;23: McIntyre S. Causal pathways to cerebral palsy in term and near-term singletons: Analysis of a total population case control study. PhD Thesis. Sydney: Faculty of Medicine, The University of Sydney, Bruck I, Antoniuk S, Spessatto A. Epilepsy in children with Cerebral Palsy. Arq Neuropsiquiatr. 2001;59: Wallace S. Epilepsy in cerebral palsy. Dev Med Child Neurol. 2001;43: Wu C, Pedersen L, Miller J, Sun Y, Streja E, Uldall P, et al. Risk of cerebral palsy and childhood epilepsy related to infections before or during pregnancy. PLoS ONE. 2013;8:e McIntyre S, Taitz D, Keogh J, Goldsmith S, Badawi N, Blair E. A systematic review of risk factors for cerebral palsy in children born at term in developed countries. Dev Med Child Neurol. 2013;55: Staudt M. Reorganization after pre- and perinatal brain lesions. J Anat. 2010;217: Wu Y, Croen L, Shah S, Newman T, Najjar D. Cerebral palsy in a term population: risk factors and neuroimaging findings. Pediatrics. 2006;118: Naeye RL, Lin H. Determination of the timing of fetal brain damage from hypoxemia-ischemia. Am J Gynecol. 2001;184: Supporting information Additional Supporting Information may be found in the online version of this article: Appendix S1. Definitions of study variables. Table S1. Interrelationship between the identified antecedents among children with cerebral palsy (n = 309). Table S2. Associations with epilepsy and cognitive impairment. Table S3. Associations with cerebral palsy (CP) together with both epilepsy (EP) and cognitive impairment (CI). 836