Video Abstracts -- brief videos summarizing key findings of new articlesCerebral Palsy in Extremely Preterm Infants
Abstract
BACKGROUND AND OBJECTIVES: The risk of cerebral palsy (CP) is high in preterm infants and is often accompanied by additional neurodevelopmental comorbidities. The present study describes lifetime prevalence of CP in a population-based prospective cohort of children born extremely preterm, including the type and severity of CP and other comorbidities (ie, developmental delay and/or cognitive impairment, neurobehavioral morbidity, epilepsy, vision and hearing impairments), and overall severity of disability. In this study, we also evaluate whether age at assessment, overall severity of disability, and available sources of information influence outcome results.
METHODS: All Swedish children born before 27 weeks’ gestation from 2004 to 2007 were included (the Extremely Preterm Infants in Sweden Study). The combination of neonatal information, information from clinical examinations and neuropsychological assessments at 2.5 and 6.5 years of age, original medical chart reviews, and extended chart reviews was used.
RESULTS: The outcome was identified in 467 (94.5%) of eligible children alive at 1 year of age. Forty-nine (10.5%) children had a lifetime diagnosis of CP, and 37 (76%) were ambulatory. Fourteen (29%) had CP diagnosed after 2.5 years of age, 37 (76%) had at least 1 additional comorbidity, and 27 (55%) had severe disability. The probability for an incomplete evaluation was higher in children with CP compared with children without CP.
CONCLUSIONS: Children born extremely preterm with CP have various comorbidities and often overall severe disability. The importance of long-term follow-up and of obtaining comprehensive outcome information from several sources in children with disabilities is shown.
- ADHD —
- attention-deficit/hyperactivity disorder
- ASD —
- autism spectrum disorder
- Bayley-III —
- Bayley Scales of Infant and Toddler Development III
- BFMF —
- Bimanual Fine Motor Function
- CI —
- confidence interval
- CP —
- cerebral palsy
- EP —
- extremely preterm
- EXPRESS —
- Extremely Preterm Infants in Sweden Study
- GA —
- gestational age
- GMFCS —
- Gross Motor Function Classification System
- WISC-IV —
- Wechsler Intelligence Scale for Children IV
What’s Known on This Subject:
The risk of cerebral palsy (CP) is high in children born extremely preterm; prevalence rates vary between 7% and 20%. CP is often accompanied by other neurodevelopmental comorbidities.
What This Study Adds:
Lifetime prevalence of CP in children born extremely preterm is 10.5%. Despite the majority being ambulatory, various comorbidities and often severe disability is shown. The importance of long-term follow-up and obtaining comprehensive outcome information in children with disabilities is demonstrated.
The risk of cerebral palsy (CP) is higher in infants born preterm than in infants born at term; this risk increases with decreasing gestational age (GA).1–5 The authors of outcome studies of infants born extremely preterm (EP) have reported prevalence rates ranging from 7% to 20%.4,6–10
CP describes a group of disorders of movement and posture, causing activity limitations attributed to nonprogressive disturbances in the developing fetal or infant brain. The motor disorders of CP are often accompanied by disturbances of sensation, cognition, communication, perception, and behavior and/or by seizure disorders.11–14
Few studies of children with CP born EP have included the combination of population-based prevalence with categorization of severity, type of CP, and comorbidities.15–18 Comorbidities are unequally distributed in children with CP and might be associated with specific neurologic subtypes and motor function status.16,19–21 A correlation between decreasing GA and more severe motor impairment and a higher percentage of comorbidities has been described in children with CP.15 In a systematic review, including 30 population registry studies on all GA groups of CP, 49% of children had intellectual disability (in 28%, it was severe), 31% could not walk (Gross Motor Function Classification System19 [GMFCS] level IV or V), 26% had behavioral disorders, 35% had epilepsy at some point (in 24%, it was active), 11% were functionally blind, and 4% had severe hearing impairment or were deaf.22
The national Extremely Preterm Infants in Sweden Study (EXPRESS), comprising all infants born before 27 weeks in Sweden during a 3-year period, revealed neurodevelopmental outcomes at 2.58 and 6.5 years of age.23 The point prevalence of CP was 7.0% and 9.5%, respectively. Comorbidities associated with CP were not described. The primary aim of this study is to describe the lifetime prevalence of CP up to the age of 6.5 years in the EXPRESS cohort of children, including the type and severity of CP and additional comorbidities, such as global developmental delay and/or cognitive impairment, neurobehavioral morbidity, epilepsy, and vision and hearing impairments, as well as the overall severity of disability. A secondary aim is to evaluate if the age at assessment, severity of disability, and sources of information influence outcome results in children with CP.
Methods
In the EXPRESS study, all infants born in Sweden before 27 gestational weeks and between April 1, 2004, and March 31, 2007, were included. There were 1011 infants born EP, of whom 707 were live born and 494 survived to 1 year of age (Fig 1). Perinatal and neonatal data were collected prospectively, and the results have been published.24,25 Survivors were clinically assessed at a corrected age of 2.5 years (n = 415) and at 6.5 years (n = 382). At 2.5 years, the Bayley Scales of Infant and Toddler Development III (Bayley-III)26 was used to assess cognitive, language, and motor development (n = 398). At 6.5 years, intellectual ability was measured with the Wechsler Intelligence Scale for Children IV (WISC-IV)27 (n = 359). These results were compared with the mean and SD of a group of matched control children born at term (Supplemental Table 6). At 6.5 years, motor function was assessed with a clinical examination, the Movement Assessment Battery for Children II28 (n = 344), and a modified Touwen examination (n = 369).29 In children with CP, the GMFCS and the Bimanual Fine Motor Function (BFMF) classification30 were used to assess functional level at 6.5 years. In 40 children at 2.5 years and in 60 children at 6.5 years, who were not clinically examined, medical charts were reviewed by the pediatrician in charge (original chart review). Details and results from these investigations have previously been published.8,23
Flowchart of the EXPRESS population, available information, and CP diagnosis at different assessment points. a Includes both those with clinical examination and original medical chart review; b Includes 3 children identified at the extended chart review; c Includes 3 children with significant spasticity, previously not diagnosed with CP.
In the current study, all available information was initially scrutinized for (1) an established CP diagnosis at the 2.5- or 6.5-year examinations; (2) signs of possible CP, such as spasticity (at least 2 of hyperreflexia, increased passive tone, and/or altered body position5,31) and/or affected motor function (motor developmental scores at the 2.5-year Bayley-III examination <−2 SD of controls, Movement Assessment Battery for Children II centiles <5, noted at any of the examinations or in the original chart review); and (3) potential risk factors for CP defined as intraventricular hemorrhage grade ≥II and/or periventricular leukomalacia.12,32 In children who fulfilled any of these 3 criteria, except for those with a well-documented absent combination of spasticity and motor problems, additional information from medical and rehabilitation charts was collected up to the age of 6.5 years (extended chart review). The combined information from the EXPRESS study data and the extended chart reviews was thereafter independently evaluated by 2 experienced pediatric neurologists (M.H., B.S.). A lifetime prevalence of CP was defined as children with an established CP diagnosis and/or children with a combination of spasticity and motor problems diagnosed at any time up to the age of 6.5 years. A search for additional information on neurodevelopmental comorbidities such as global developmental delay and/or cognitive impairment, neurobehavioral morbidity, epilepsy, and vision33 and hearing impairments was also included in this evaluation.
The GMFCS and BFMF classifications were applied at the 6.5-year examination. If this was missing, the classification was accomplished retrospectively by using available information.
To determine developmental level and/or cognitive function, the WISC-IV, full-scale IQ score at 6.5 years of age23 was used. If WISC-IV results were not available, the lowest score from the cognitive and language scale of the Bayley-III examination at 2.5 years8 was used. The child’s test results at 6.5 and 2.5 years, respectively, were thereafter categorized relative to the mean and SD interval of the term controls into 4 cognitive level categories ranging from normal to severe deficit (Supplemental Table 6). If this information was missing, the development level and/or cognitive function estimated from the extended chart review was used.
In children fulfilling the CP definition, the classification of overall severity of disability was assessed in accordance with the criteria shown in Table 1.
Classification of the Overall Disability Severity in Children With CP
Possible associations between GA (linear continuous variable) and CP (yes/no) and between GMFCS subtypes (linear continuous variable) and developmental and/or cognitive deficit (<−2 SD) (yes/no) were assessed by using linear logistic regression analysis. Fisher’s exact test was used to evaluate binary outcomes for those born at 22 to 24 vs 25 to 26 weeks, GMFCS I to II versus III to V, and normal or mild development impairment versus moderate-severe developmental impairment, respectively. Spearman ρ was used to investigate the correlation between Bayley-III cognitive scale categories 1 to 4 at 2.5 years with WISC-IV cognitive categories 1 to 4 at 6.5 years; the corresponding 95% confidence interval (CI) was obtained by using bootstrapping (2000 bootstrap samples). Statistical analyses were performed by using SPSS version 23 (IBM SPSS Statistics, IBM Corporation, Armonk, NY) and Gauss version 10 (Aptech Systems Inc, Maple Valley, WA). P < .05 was regarded as statistically significant.
The Regional Ethics Review Board, Lund, Sweden, approved the study, and parents provided written consent.
Results
In the EXPRESS study, 494 of 707 live-born (70%) children born EP were alive at 1 year of age. Neonatal information from 490 infants, including cranial ultrasound results in 486, was found. Neurodevelopmental outcome data at 2.5 and 6.5 years of age for 455 and 442 children, respectively, have previously been published.8,23 By combining these data, outcome information was collected for 680 of 707 (96.1%) live-born infants and for 467 of 494 children (94.5%) alive at 1 year of age (Fig 1). Seventy-five of the 467 children met the criteria for an extended chart review, and for 64 this information was included; 2 had died, 5 had declined participation at 6.5 years, and 4 had no extended chart information available.
CP Prevalence, Type, and Functional Outcome
Forty-nine children were identified as having a lifetime diagnosis of CP up to the age of 6.5 years, which corresponds to 10.5% (49/467) of those alive at 1 year of age with a known outcome. The figures were similar across the GA groups, and no linear association between GA and CP was indicated (P = .63). None of the 5 children born at 22 weeks and alive at 1 year had CP. Seventy-six percent (37/49) were ambulant without aid (GMFCS levels I–II). The BFMF classification showed levels corresponding to those of GMFCS. Spastic CP was found in 45 (91%); 35 of them had a bilateral spastic type (Table 2). All children with unilateral CP were ambulant without aid, compared with 71% of the children with bilateral spastic CP. Two children had a probable postnatal explanation for their CP; 1 had paraparesis caused by a spinal injury after neonatal surgery and the other had sequelae after inflicted violence.
The Description of EXPRESS Cohort of Children Born EP and Outcome in Children With Lifetime CP According to GA at Birth
Neurodevelopmental Comorbidities and Overall Severity of Disability
A complete WISC-IV assessment was available for 26 children with CP (and for these children, the Bayley-III results from 2.5-year examination were also available). For 17 children, only the results from the Bayley-III cognitive and languages scores at 2.5 years were available, and for still another 6 children, only data from extended chart reviews were used for approximate developmental- and/or cognitive-level categorization. Among the 26 children with lifetime CP who attended both assessments, there was a positive correlation between the Bayley-III development categories at 2.5 years and the corresponding cognitive categories based on the WISC-IV full-scale results (Spearman ρ: 0.67, 95% CI: 0.37–0.86; P < .001). Fifty-four percent had scores in a lower category on the WISC-IV assessment compared with the scores on Bayley-III.
Thirty of 49 (61%) children had a developmental and/or cognitive level <−2 SD, including all children with GMFCS levels IV and V (Table 3). A linear association between the GMFCS level and the occurrence of a developmental and/or cognitive level <−2 SD was indicated, but this was not statistically significant (odds ratio for 1 GMFCS unit increment: 1.65, 95% CI: 0.92–2.97; P = .10).
Comorbidities According to Motor Function (GMFCS Level) in 49 Children With CP
Of children born at 22 to 24 weeks, 87% had a developmental and/or cognitive level <−2 SD, compared with 50% among those born at 25 to 26 weeks (P = .03) (Table 2).
Eight children with CP were diagnosed with autism spectrum disorder (ASD) and 2 had attention-deficit/hyperactivity disorder (ADHD). An additional 6 children had highly suspicious neurobehavioral symptoms (ie, inattention, hyperactivity, or autistic features). Six children were presently on treatment for epilepsy and 3 had been previously treated. Four children with CP were blind, and 7 had a moderate visual impairment. One child had a severe hearing impairment, and 4 children had moderate hearing problems. Seventy-six percent (37/49) of the children had at least 1 additional neurodevelopmental comorbidity. Comorbidities in relation to motor function and developmental level are shown in Tables 3 and 4. Nine of 11 children with visual impairment had a developmental and/or cognitive level <−2 SD (P = .043). No additional associations were found, in relation neither to motor function nor to developmental level and/or cognitive function.
Comorbidities According to the Developmental Level and/or Cognitive Function in 49 Children With CP
Ten of 49 children (20%) with CP had an overall mild disability, and 27 of 49 (55%) had a severe disability (Table 5). All children with CP born at 22 to 24 weeks (n = 15) had moderate-severe disability, compared with 71% (24/34) of those born at 25 to 26 weeks (P = .02) (Table 2).
Sources of Information at the 6.5-Year Assessment of the EXPRESS Cohort According to the Occurrence of Lifetime CP (the Information From the Extended Chart Review is Included) and Disability
Influence of Sources of Information, Age at Assessment, and Severity of Disability on Outcome Results
The combined information from all sources in this cohort identified a CP diagnosis in 49 children up to the age of 6.5 years. At the 2.5-year assessment, 32 were identified.8 The extended chart review at 6.5 years revealed another 3 children who already had a CP diagnosis at 2.5 years, which was not registered then. At the 6.5-year assessment, 42 were identified,23 8 of them by the original chart review. In addition, 1 child with a CP diagnosis at 2.5 years who died before 6.5 years, 3 who had a CP diagnosis at the 2.5-year assessment and declined participation in the 6.5-year assessment, and 3 children with significant clinical findings of spasticity and motor developmental problems not previously considered to have CP (1 solely by the extended chart review information) were identified and included in the lifetime prevalence rate. The extended chart reviews also identified additional details concerning other comorbidities. Twenty-nine percent (14/49) were diagnosed with CP after 2.5 years, and all were ambulant without aid (GMFCS levels I or II).
At the 6.5-year assessment, the probability for not having a complete clinical and formal cognitive assessment was higher in children with CP (22/48; 46%) compared with children without CP (80/413; 19%) (P = .0001). This was true for 62% (16/26) of children with CP and severe disability, compared with 27% (6/22) of those with CP and mild or moderate disability (P = .02) (Table 5). Among 10 children with CP and ASD, blindness, and/or deafness, only 1 had a complete evaluation at 6.5 years.
Only 1 of the 11 children with GMFCS III-V was cognitively assessed by WISC-IV at 6.5 years. In the 16 surviving children with CP and a severe developmental deficit on the Bayley-III test (<−3 SD) at 2.5 years, only 5 children were assessed with WISC-IV at 6.5 years.
Discussion
In this population-based cohort of children born before 27 weeks’ gestation and surviving beyond 1 year of age, the lifetime prevalence of CP up to the age of 6.5 years was 10.5%. The majority (76%) had ambulatory gross motor function. These results are favorable compared with those of other studies.4,10,22,34 No difference in CP prevalence was seen with regard to GA in this population.31 Seventy-six percent had at least 1 neurodevelopmental comorbidity in addition to CP; global developmental delay and/or cognitive impairment was the most frequent (61%). Severe disability was seen in 55% and was more common in children born at 22 to 24 weeks than in children born at 25 to 26 weeks; none of the 15 children with CP born before 25 weeks were ambulant and without additional comorbidities. These results correspond to those of studies on CP in children in general populations15,16 and in outcome studies on children born EP.9,35 This underlines the importance of describing outcomes as comprehensively as possible.22,36
The importance of not ruling out CP too early in life is revealed by the difference in point prevalence in the EXPRESS cohort. At 2.5 years,8 the reported CP prevalence was 32 in 456 (7.0%), and at 6.5 years,23 the prevalence was 42 in 441 (9.5%). All children diagnosed with CP after the 2.5-year assessment were ambulant. This is in accordance with the authors of studies on CP registers, who recommend inclusion only of children older than 4 years of age to minimize false diagnoses and to include those with milder symptomatology.11,14
CP is a clinical descriptive term and its exact diagnostic criteria are under debate.11 The distinction from other motor impairments, such as developmental coordination disorder and global developmental delay, can be challenging.14,37 Because available extended information up to the age of 6.5 years was scrutinized by 2 experienced pediatric neurologists in accordance with predefined criteria for CP, a more standardized evaluation was achieved, and 3 previously undiagnosed children were identified. In addition, 3 children with CP who declined participation at 6.5 years and 1 who died before the 6.5-year assessment were included. The total number of children with CP in this cohort of children born EP increased from a point estimate of 42 at 6.5 years23 to a lifetime (up to the age of 6.5 years) number of 49. This includes 2 children with probable postnatal causes of CP. Although their etiologies were not directly related to extreme prematurity, infants born EP might be more prone to this type of injury. If only children who were clinically assessed at 6.5 years had been included, the number of children with CP would have been 34, because 8 were assessed solely by chart reviews. This study revealed that the probability of not having a full clinical and cognitive assessment is increased in children born EP with CP (46%), compared with those without CP (19%), with the highest probability in children with the most severe disability (62%). The importance of combining data sources to gain more accurate estimates of prevalence rates of CP has been shown and discussed,4,16,18,38 and with our results, we underline the difficulties and importance in obtaining sufficient outcome information in children with disabilities.7,11,35
With our design combining several sources of information, including clinical assessments and chart reviews; a low drop-out frequency; and follow-up to 6.5 years, we created unique possibilities for assessing lifetime prevalence of CP, functional level (GMFCS), and additional comorbidities in children born EP. Because a majority of studies on CP prevalence are based on registers, the differences in completeness of data might explain some of the variation reported.16,38
Strengths of the Study
This is a population-based study including all infants born EP in Sweden before 27 weeks’ gestation during a 3-year period. Neonatal data were prospectively collected, and all results from assessments up to the age of 6.5 years were available. In children with CP and suspected CP, additional detailed medical information was collected and scrutinized. With this combination of information, we were able to provide reliable outcome data regarding CP in 96% of live-born infants and information on additional comorbidities in those with CP.
Limitations of the Study
Complete information was not available from either the clinical data at the 2 assessments or the extended chart scrutiny. The fact that only 54% of the children with CP had a cognitive assessment with WISC-IV at 6.5 years and that 54% of those with both WISC-IV and Bayley-III assessments scored at a lower developmental and/or cognitive level on the WISC-IV increases the uncertainty regarding their intellectual ability at 6.5 years.21 The proportion of children with cognitive impairment might thus be higher than reported. Twenty-four percent of the children with CP did not have a clinical GMFCS and BMFM classification at 6.5 years of age, and this might have influenced the accuracy of the functional level. The numbers of children with ASD and ADHD is probably underestimated because the evaluations at 2.5 and 6.5 years were not primarily designed to detect neurobehavioral symptoms. Moreover, milder forms of neurobehavioral disorders are difficult to diagnose early in childhood, and children with severe disability are difficult to assess.5,37
Conclusions
In this population-based follow-up study of children born EP, 10.5% had a lifetime diagnosis of CP up to the age of 6.5 years. The majority (76%) of the children with CP were ambulatory, 76% had at least 1 additional comorbidity, and 55% had an overall severe disability. With the extended study design of lifetime CP, we identified 49 children with CP, compared with the point-estimate findings at 2.5 and 6.5 years of 32 and 42 children, respectively. This underlines the importance of long-term follow-up and of obtaining outcome information that is as comprehensive as possible from several sources.
Acknowledgments
All professionals who participated in the EXPRESS study are gratefully acknowledged: Pediatricians: Dr Mats Blennow, Dr Mikael Norman, Dr Brigitte Vollmer, Dr Uwe Ewald, Dr Lena Hellström-Westas, Dr Gunnar Sjörs, Dr Vineta Fellman, Dr Kristina Rosengren-Forsblad, Dr Lennart Stigson, Dr Ulla Lindskog, Dr Elisabeth Olhager, Dr Eva Lindberg, and Dr Andreas Ohlin. Ophthalmologists: Dr Gerd Holmström, Dr Kerstin Hellgren, Dr Ann Hellström, Dr Peter Jakobsson, Dr Kristina Tornqvist, Dr Gunnar Lindgärde, Dr Agneta Wallin, Dr Kent Johansson, and Dr Pia Lundgren. Psychologists: Karin Stjernqvist, PhD, Johanna Månsson, PhD, Anette Carnemalm, MSc, Christina Helgason, MSc, Milly Marken, MSc, Marie Adamsson Johansson MSc, Ylva Fredriksson, MSc, Kari Ylimäinen, MSc, Anna Lönegren, MSc, Margreth Ericsson, MSc, Anna Nyrén, MSc, Carin Johansson Wiemerö, MSc, Birgitta Böhm, PhD, Eva Eklöf, MSc, Christina Lindqvist, MSc, Irmgard Obwexer, MSc, and Claudia Aulin-Villa, MSc. Local study coordinators: Barbro Fossmo, RN, Cecilia Ewald, RN, Christina Fuxin, RN, Lena Swartling-Schlinzig, RN, Ann-Cathrine Berg, RN, Cecilia Tobiasson, EN, and Pia Lundqvist, RN, PhD. We thank Professor Emeritus Ingemar Kjellmer for invaluable sharing of vast knowledge and support to the main author. The English language of the manuscript was revised by Mrs Nancy Eik-Nes.
Footnotes
- Accepted October 17, 2017.
- Address correspondence to Maria Hafström, MD, PhD, Department of Pediatrics, St Olav’s Hospital HF, Postboks 3250 Sluppen, 7006 Trondheim, Norway. E-mail: maria.hafstrom{at}ntnu.no
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
FUNDING: Supported by the Swedish Research Council grants 2006-3858, 2009-4250, and 523-2011-3981; the Uppsala-Örebro Regional Research Council grants RFR-10324 and RFR-23351; the Health Care Subcommittee, Region Västra Götaland grant RFR-66881; a grant from the Research Council South East Region of Sweden; grants to Researchers in the Public Health Care from the Swedish government; the Ann-Mari and Per Ahlqvist Foundation; and the Märta and Gustaf Ågren Foundations. Dr Hafström was also supported by St Olav’s Hospital–Trondheim University Hospital grants RFR 16/9564-123. Dr Källén reported support from the Evy and Gunnar Sandberg and from the Birgit and Håkan Ohlsson Foundations. Financial support was also provided through regional agreements between the University of Umeå and Västerbotten County Council, Uppsala County Council and Uppsala University Children’s Hospital, and Stockholm County Council and Karolinska Institute (grant ALF-20160227), which have a regional agreement on medical training and clinical research. Dr Ådén also got support from the Marianne and Marcus Wallenberg Foundation (grant 2011.0085), the Swedish Order of Freemasons in Stockholm, and the Swedish Brain Foundation and through a regional agreement. Finally, the study was supported by the Lilla Barnets Fond children’s fund.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
References
- Copyright © 2018 by the American Academy of Pediatrics
