MABC-2 indicates Movement Assessment Battery for Children–Second Edition.
eTable 1. Characteristics of eligible extremely preterm children (< 27 gestational weeks) with and without results from assessment with Movement Assessment Battery for Children-Second edition (MABC-2).
eTable 2. Characteristics of eligible term born controls with and without results from assessment with Movement Assessment Battery for Children-Second edition (MABC-2).
eTable 3. Sex differences in the prevalence of developmental coordination disorder (DCD) among extremely preterm children and term born controls. Both groups were without major disabilities
eTable 4. Extremely preterm children with developmental coordination disorder compared with the whole term control group; unadjusted comparison; parental reports on poor motor skills and behavioral problems and comparison of mean difference in cognitive scores.
eTable 5. Extremely preterm children with developmental coordination disorder compared with extremely preterm children without developmental coordination disorder. Unadjusted odds ratios (95% CI) of parental reports on motor skills and behavioral problems, and comparison of mean differences in cognitive scores.
Customize your JAMA Network experience by selecting one or more topics from the list below.
Bolk J, Farooqi A, Hafström M, Åden U, Serenius F. Developmental Coordination Disorder and Its Association With Developmental Comorbidities at 6.5 Years in Apparently Healthy Children Born Extremely Preterm. JAMA Pediatr. 2018;172(8):765–774. doi:10.1001/jamapediatrics.2018.1394
How common are motor impairments that interfere with daily life and associated comorbidities at early school age in apparently healthy children who are born extremely preterm?
This national cohort study found that more than one-third of children who were born extremely preterm in the modern neonatal era had developmental coordination disorder at 6.5 years of uncorrected age, which was associated with considerable adaptive behavioral and perceptual problems and worse cognition.
Motor impairments are common at early school age in children who are born extremely preterm and are associated with several comorbidities, which is important for clinicians planning and implementing follow-up programs.
There are concerns that apparently healthy extremely preterm children face a risk of developing motor impairments, such as developmental coordination disorder.
To evaluate the prevalence of developmental coordination disorder and associated comorbidities in a national cohort of apparently healthy children born at 22 to 26 gestational weeks, compared alongside term-born peers.
Design, Setting, and Participants
This prospective, population-based cohort study included all children who were consecutively born at 22 to 26 gestational weeks in Sweden from April 1, 2004, through March 31, 2007. At 6.5 years, 441 preterm children were evaluated alongside 371 controls. A total of 275 preterm children (62.4%) and 359 term-born children (96.8%) did not have neurodevelopmental disabilities. Motor assessments were completed for 229 of 275 preterm children (83.3%) and 344 of 359 (95.8%) term-born children, who composed the final study sample.
Main Outcomes and Measures
Developmental coordination disorder was defined as a score of the fifth percentile or lower on the Movement Assessment Battery for Children–Second Edition scale, using control group scores. Assessment tools included the Wechsler Intelligence Scale for Children-Fourth Edition, the Brown Attention-Deficit Disorder Scales, the Five to Fifteen questionnaire, and the Strengths and Difficulties questionnaire.
Of the 229 extremely preterm children and 344 term-born controls who underwent motor assessments, 115 (50.2%) and 194 (56.4%) were boys, respectively. Developmental coordination disorder was present in 85 of 229 (37.1%) preterm children and in 19 of 344 controls (5.5%) (adjusted odds ratio [OR], 7.92; 99% CI, 3.69-17.20). When preterm children with developmental coordination disorder were compared with term-born peers, the risk was increased for total behavioral problems, internalizing, externalizing, attentional problems, hyperactivity, perceptual problems, executive dysfunction, and poor social skills, with adjusted ORs varying from 2.66 (99% CI, 1.09-6.48) for time concepts to 9.06 (99% CI, 3.60-22.8) for attentional problems (all P < .01). When preterm children with and without developmental coordination disorder were compared, preterm children with developmental coordination disorder had more behavioral problems; the adjusted OR for total behavioral problems was 2.71 (99% CI, 1.15-6.37); for externalizing problems, 2.80 (99% CI, 1.10-7.12); for inattention, 3.38 (99% CI, 1.39-8.18); and for combined attention/hyperactivity problems, 3.68 (99% CI, 1.47-9.16) (all P < .01). Parents underestimated the children’s motor problems and only a few of the children had received psychological care or physiotherapy.
Conclusions and Relevance
Children who were born extremely preterm faced a high risk for developmental coordination disorder with associated comorbidities. Our findings support the importance of a structured follow-up of motor function, behavior, and cognition.
Children who are born extremely preterm face an increased risk of developing motor impairments,1-3 the most severe being cerebral palsy. However, many preterm children who do not develop cerebral palsy are still at high risk of developing motor impairments, such as developmental coordination disorder (DCD).3-5 This disorder involves impaired motor function that interferes with daily activities or academic achievement and cannot be explained by other impairments.6 Developmental coordination disorder has been associated with mental health problems,7,8 learning difficulties, problems with psychosocial adjustment9 and lower cognitive function.10 In most children, DCD persists until at least early adulthood11 and may result in secondary associations with self-esteem12 and poor physical health.13,14
In typically developing populations, the most frequently reported prevalence of DCD is 5% to 6%,15 and in preterm populations, the reported prevalence has varied between less than 10% and more than 50%5,16-18 depending on the definition of DCD, the sample size, and the sample composition.5 Prevalence rates are mostly based on cohorts born in the 1980s and 1990s4,16,19,20 and may not reflect outcomes after the advent of active perinatal care of the most immature children.21 In addition, there are very few recent studies that have reported associations between DCD and other comorbidities in children born extremely preterm.
The main aim of this study was to investigate the prevalence of DCD at 6.5 years in a large national cohort of children born at 22 to 26 weeks of gestation who were without major neurodevelopmental disabilities compared with healthy term-born controls, and to explore the prevalence of behavioral problems and cognitive dysfunction among the preterm children with DCD compared with (1) term-born peers and (2) preterm children without DCD. A secondary aim was to assess the parents’ perception of their preterm child’s motor difficulties and the use of health care services.
This study formed part of the Extremely Preterm Infants in Sweden Study (EXPRESS), which included all children born at fewer than 27 weeks of gestation from April 1, 2004, to March 31, 2007, in Sweden.22 The recruitment procedures, definitions of neonatal characteristics, and neurodevelopmental and ophthalmological outcomes at ages 2.5 and 6.5 years have previously been reported.22-25 Of the 486 survivors, 441 (90.7%) were followed up along with 371 term-born controls (37-41 weeks) at 6.5 years. The perinatal and sociodemographic characteristics and disability rates were reported to be similar among the mothers and children who were and were not followed-up.23
The inclusion criteria for this study were children born before 27 gestational weeks or at term and without cerebral palsy, cognitive impairment (defined as less than −2 SDs compared with the means and SDs of the controls), and any visual or hearing impairment. Thus, 275 of 441 children (62.4%) born extremely preterm and 359 of 371 controls (96.8%) fulfilled the inclusion criteria.
At 6.5 years of uncorrected age, the children were evaluated with a clinical examination, parental questionnaires, and a psychological assessment, as previously described.23 All parents gave their written, informed consent and the regional ethics review board of Lund approved the study.
Motor performance was assessed with the Movement Assessment Battery for Children–Second Edition (MABC-2),26 which is a standardized test that has high reliability and validity and is extensively used in both clinical settings and research. It assesses gross and fine motor function using 3 tasks to evaluate manual dexterity, 3 tasks to evaluate balance, and 2 tasks to evaluate ball skills. Higher scores indicate better motor function. Movement Assessment Battery for Children–Second Edition scores of the 15th percentile or lower are considered borderline abnormal and scores of the fifth percentile or lower as definitely abnormal.26 We chose the more restrictive criterion to identify children with DCD to ensure that we explicitly included children who had motor impairments that were likely to interfere with daily living.27
Because Swedish norms are unavailable and mean values on the MABC-2 differ between populations,28 we related the cutoff to the distribution of the control group. To enable comparisons with studies that defined DCD as motor function corresponding to a score of the 15th percentile or lower on the MABC-2 scale, the impairment rates for the sixth to 15th percentile were also calculated.
Cognitive function was assessed with the Wechsler Intelligence Scale for Children, Fourth Edition (Q-Interactive). It consists of 4 subtests—verbal comprehension, perceptual reasoning, working memory, and processing speed—that added together give a full-scale intelligence quotient (FSIQ). Cognitive impairment was defined as an FSIQ of less than −2 SD (≤76) of the mean of the term-born controls.
The Five to Fifteen Questionnaire (FTF) comprises 181 statements that explore the difficulties that participants experience with a certain task when they are compared with other children of the same age and sex.29 Higher scores indicate more problems. A cutoff of more than the 90th percentile compared with the control group was used to define clinically relevant problems.30 The FTF has been validated31 and has been used in preterm populations.32 We used the parts of the FTF that assess the parents’ perception of the child’s gross and fine motor skills, executive function, social skills, and perceptual problems.
The Strengths and Difficulties Questionnaire (Youthinmind)33 identifies children with behavioral and emotional difficulties in clinical and community populations.33 It comprises 25 items on 5 scales: emotional symptoms, conduct problems, hyperactivity/inattention, peer problems, and prosocial behavior. A total difficulties score is calculated by combining the first 4 scales, with higher scores indicating more difficulties. Scores of more than the 90th percentile compared with the control group were considered clinically relevant problems.34
Attentional problems were assessed using the Brown Attention-Deficit Disorder Scales (Psychological Corporation),35 which were administered as a parental questionnaire. They consist of 6 clusters; 5 of these can be combined to provide a total inattention score and combining all 6 forms a total combined score that includes inattention and hyperactivity. The raw scores are converted into T scores and a score of 55 or higher indicates clinically relevant problems.35
The visual assessment was carried out as previously described, as was the information about hearing impairments and sociodemographic data.23,24 Information on sociodemographic data and the use of health care services was obtained from the parental questionnaire.
For groupwise comparisons of descriptive and background data, the t test and Mann-Whitney U test were used for continuous variables and the χ2 or Fisher exact test for categorical data, as appropriate. Odds ratios were estimated using logistic regression and mean differences in scores were estimated using analysis of covariance. Analyses were adjusted for age at assessment, the mother’s country of birth (Nordic/non-Nordic), and the mother’s education in comparisons between preterm children with DCD and term controls, and for sex, gestational age, age at assessment, and the mother’s country of birth (Nordic/non-Nordic) in comparisons between preterm children with and without DCD. Covariates were chosen based on background differences that were considered clinically important. To account for any correlation between multiple births, a complex samples design36 was applied to the analyses described previously. To account for multiple comparisons, the level of statistical significance was set at a 2-sided P value of <.01. The data analyses were performed using SPSS, version 25.0 (IBM Corp).
Motor assessment was completed for 229 of 275 (83%) of extremely preterm children and for 344 of 359 (96%) of term-born controls and these comprised the final study sample (Figure). The characteristics were similar between children who were or were not assessed except that more mothers were born in non-Nordic countries among preterm children who were not assessed with MABC-2 (eTables 1 and 2 in the Supplement).
Of the extremely preterm children, 85 of 229 (37.1%) were classified as having DCD, compared with 19 of 344 (5.5%) in the control group; the rate was consistent with the applied fifth percentile cutoff (Table 1). The risk for scoring at the fifth percentile or lower compared with the control group was highest for the manual dexterity component of MABC-2 (Table 1). The prevalence of DCD among 49 of 115 preterm boys (42.6%) was not significantly different from that of 36 of 114 preterm girls (31.6%) (eTable 3 in the Supplement). Borderline motor impairment, corresponding to the sixth to 15th percentile on the MABC-2 scale, was found in 35 of 229 preterm children (15.3%) and in 26 of 344 term-born controls (7.6%) (P < .001).
The children’s characteristics and their socioeconomic backgrounds are reported in Table 2. The preterm children with DCD had marginally lower gestational ages, lower birth weights, and more neonatal morbidities but similar sociodemographic characteristics as the preterm children without DCD.
The parents of preterm children with DCD reported more motor, behavioral, and perceptual problems and poorer executive functions and social skills than the parents of term-born controls (Table 3; unadjusted results in eTable 4 in the Supplement). The mean-adjusted FSIQ in preterm children with DCD was 12.5 (99% CI, 9.4-15.3) points lower than in the control group, and the 4 cognitive index scales were also lower.
The parents of preterm children with DCD reported more motor problems than the parents of preterm children without DCD (Table 4; unadjusted results in eTable 5 in the Supplement). They were also more likely to rate their child in the abnormal range for behaviors such as inattention, hyperactivity, and combined hyperactivity and inattention (Table 4). The mean-adjusted FSIQ in preterm children with DCD was 3.9 points (99% CI, 0.4-7.32; P = .004) lower than in preterm children without DCD (Table 4). Of the cognitive index scales, the perceptual reasoning and processing speed were significantly lower in preterm children with DCD. The adjusted odds ratios (aORs) for poor fine motor skills (aOR, 2.11; 99% CI, 0.82-5.45), executive function (aOR, 2.04; 99% CI, 0.87-4.81), and relation in space (aOR, 2.06; 99% CI, 0.84-5.04) all differed at a P value of <.05, but they did not meet the significance level specified for this study (P < .001). According to the parents’ reports, during the preceding 12 months, 10 of 85 (12%) of extremely preterm children with DCD vs 14 of 144 (10%) of extremely preterm children without DCD had seen a physiotherapist and 12 of 85 (14%) vs 16 of 144 (11%) had seen a psychologist.
This study assessed the prevalence of DCD and its association with developmental comorbidities at 6.5 years in children without major disabilities who were born extremely preterm in the 2000s and were part of the EXPRESS study. To demonstrate the challenges that preterm children with DCD can experience in their daily lives in relation to their peers, we compared them with term-born controls. To highlight specific features of DCD within the preterm group, we compared preterm children with and without DCD.
This study confirmed earlier research1,2,19 that motor impairment is common in preterm children. According to the study criteria, DCD was present in 85 of 229 (37.1%) of the children born extremely preterm, and another 35 of 229 (15.3%) had borderline motor function. We also showed that the associations of DCD extended beyond the motor function domain into the behavioral, cognitive, social, and perceptual domains.
The MABC-2 is commonly used for detecting DCD4,6 and children who score at the fifth percentile or lower, when compared with a control group, are expected to display significant limitations in activities related to daily living. This cutoff has been considered as appropriate for research settings and has been used in studies of preterm children.1,19,20 Children born preterm with MABC-2 scores in the sixth to 15th percentile interval are also at risk for developing DCD and several studies have used this cutoff point.37-39 As we lacked a measure to evaluate the association of DCD with daily activities, we opted for the stricter definition. However, we have also reported the prevalence of motor impairment based on the 15th percentile cutoff.
We chose MABC-2 cutoffs according to reference data obtained from our control group. This was in line with recommendations to use contemporaneous control groups for comparisons in follow-up studies.40 Such control groups allow children to be compared with peers who have a similar sociodemographic background, and they are also sensitive to changes with time that may also affect motor skills. Moreover, the Swedish version of the MABC-2 is based on UK norms that may not be applicable to other countries.28 However, it has been recognized that local norms, rather than the test norms, tend to identify a greater proportion of children as having motor impairments.3 If the normative cutoffs had been used, 12.5% of the children born extremely preterm would have had DCD.
A systematic review3 reported a pooled estimate of DCD diagnosis of 19.0 of 100 (range, 9.5%-34%) in children born at fewer than 37 gestational weeks in the 1990s. More specifically, Roberts et al20 used local norms and the fifth percentile cutoff and reported a prevalence of 16% in extremely preterm or extremely low-birthweight children born in the 1990s. Goyen and Lui37 used a −1.5 SD cutoff and found that 42% of apparently normal extremely preterm children, who were also born in the 1990s, had DCD, compared with 8% in the term control group.
Research has shown that the sensitivity of parental questionnaires that were designed to detect motor problems is poor compared with the MABC.4,20 In this study, less than half (49%) of the parents of extremely preterm children with DCD reported clinically relevant motor problems. Moreover, pediatric examinations are not sensitive enough to detect DCD in preterm children4 and formal testing of motor function is essential.
In this study, the risk for scoring at the fifth percentile or lower, compared with the control group, on the MABC-2 was particularly high for the manual dexterity component. This is important, because fine motor skills predict later school achievement.41 There were more preterm boys than preterm girls with DCD, but the sex difference was not significant. Some previous studies have shown an increased risk for boys,18,19 whereas others have shown no difference between the sexes.16,37,42-44 In contrast, most reports on term-born populations indicate that boys face a larger risk of DCD.15
Despite the fact that children with an FSIQ of less than −2 SD were excluded, the adjusted FSIQ in preterm children with DCD was 12.5 points, about 1 SD, lower than the control group. Furthermore, within the preterm group the FSIQ was 4 points (about 0.4 SD) lower in children with DCD. The differences were mainly due to a lower processing speed and perceptual reasoning, representing the cognitive aspects of executive abilities. Although the differences may seem trivial, shifts of population means of this magnitude are clinically important.45 Executive dysfunction is a crucial determinant of academic underachievement,46 and similar findings have been reported from children with DCD who were born at term.10
Preterm children with DCD had more behavioral and perceptual problems than the term-born controls on almost all aspects. When preterm children with and without DCD were compared, the excess risk for several outcomes remained even though the magnitude decreased. In the extremely preterm group, 35 of 84 children (42%) with DCD had behavioral problems and 33 of 85 (39%) had combined inattention and hyperactivity problems. Similar findings have been reported from some,4,19 but not all,43 preterm populations.
The strengths of this study included the national prospective cohort study design with a high follow-up rate. The sample size was larger than in similar studies,5,16,20,37 and we believe that the results could be generalized to other preterm populations born under the same circumstances. We comprehensively assessed the children for motor skills and developmental comorbidities by using standardized and validated questionnaires and tests. To allow for comparisons with peers, we included a term-born control group that was assessed by the same investigators that assessed the preterm group. Furthermore, we also compared the characteristics of children with and without DCD within the preterm group.
A possible limitation of the study was the early age at assessment, and the emotional and intellectual effect of DCD may be higher in later childhood.47 Because Swedish children start school at age 6.5 years, our study precluded information on school achievements. Masking was not possible because some of the assessments were part of follow-up visits and this might have introduced expectation bias. We did not control neurocognitive outcomes for IQ, as this is controversial48 and the focus of this study was descriptive rather than analytical. Our objective was not to analyze the prevalence of comorbidities in term-born children with and without DCD. Thus we cannot know whether the pattern of comorbidities found in the preterm group could also be detected among children born at term. The choice to adjust for multiple comparisons by adopting a significance level of P < .01 might have increased the risk for type II errors.
Developmental coordination disorder persists into adolescence and beyond,2 and problems are likely to increase with greater demands at school and in everyday life. Developmental coordination disorder has been described as a hidden problem,4 as the impairment is overshadowed by more severe physical and intellectual impairments. For example, in this study, fewer than 15% of the preterm children with DCD had been seen by a physiotherapist or child psychologist during the 12 months before the study. Yet this study, and other studies, have shown that children with DCD have weaknesses that increase their risks of poor school performance19,20,44 and developing poor self-esteem.12 In addition, poor participation in physical activities has been shown to contribute to poorer cardiorespiratory fitness14 and obesity.13
It is the responsibility of health professionals to identify children with DCD, and interventions based on everyday motor skills have been found to be effective.49 In this study, more than 40% of the children with DCD already had behavioral problems at 6.5 years of age, illustrating the need for multidisciplinary follow ups and interventions early in life.
The prevalence of DCD was much higher in apparently healthy extremely preterm children than in term-born controls. Parental questionnaires alone are inadequate for identifying DCD and structured testing is needed to establish a diagnosis at an early age. Because these children are more likely to have behavioral, perceptual, attentional, hyperactivity, and intellectual comorbidities, all these factors need to be taken into account when these children are evaluated.
Accepted for Publication: April 19, 2018.
Corresponding Author: Jenny Bolk, MD, Department of Women’s and Children’s Health, Karolinska Institutet, Tomtebodavägen 18a, 171 77 Stockholm, Sweden (firstname.lastname@example.org).
Published Online: June 4, 2018. doi:10.1001/jamapediatrics.2018.1394
Author Contributions: Dr Serenius had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: All authors.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Bolk, Serenius.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Bolk, Serenius.
Obtained funding: Bolk, Farooqi, Hafstrom, Aden, Serenius.
Administrative, technical, or material support: Hafstrom, Serenius.
Supervision: Hafstrom, Aden, Serenius.
Conflict of Interest Disclosures: None reported.
Funding/Support: This study was supported by grants 2006-3858, 2009-450, 523-2011-3981, and 2017-03043 from the Swedish Research Council, Sachs’ Children and Youth Hospital, the Kronprinsessan Lovisas Foundation, Hjärnfonden (grant FO2017-0131), the regional agreement on medical training and clinical research between Stockholm Count Council and the Karolinska Institutet (grant ALF SLL 20170243), and the Lilla Barnets Fond and Frimurarorden. Dr Farooqi was supported by the Jerringfonden. Dr Hafström was supported by grants from the health care subcommittee, Region Västra Götaland (grant RFR 66881), the Ann-Mari and Per Ahlqvist’s Foundation, the Märta and Gustaf Ågren Foundations, Torbjörn Jebner’s Foundation for Neuropediatric Research, the Adlerbert Research Foundation, and St Olav’s Hospital–Trondheim University Hospital grant RFR 16/9564-123.
Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: We thank the obstetricians (Karel Marsal, MD, principal investigator, Lund University), pediatricians (Bo Strömberg, MD, Uppsala University; Mats Blennow, MD, Karolinska Institutet; Mikael Norman, MD, Karolinska Institutet; Uwe Ewald, MD, Uppsala University; Lena Hellstrom-Westas, MD, Uppsala University; Gunnar Sjors, MD, Uppsala University; Vineta Fellman, MD, Lund University; Kristina Rosengren-Forsblad, MD, Lund University; Lennart Stigson, MD, Gothenburg University; Ulla Lindskog, MD, Linköping University Hospital; Elisabeth Olhager, MD, Linköping University; Eva Lindberg, MD, Örebro University; and Andreas Ohlin, MD, Örebro University), ophthalmologists (Gerd Holmstrom, MD, Uppsala University; Kerstin Hellgren, MD, Karolinska Institutet; and Kristin Tornqvist, MD, Lund University Hospital), psychologists (Karin Stjernqvist, PhD, Lund University; Johanna Mansson, PhD, MSc, Lund University; Milly Marken, MSc, Umeå University Hospital; Ylva Fredriksson Kaul, MSc, Uppsala University Children’s Hospital; Anna Lonegren, MSc, Linköping University Hospital; Anna Nyren, MSc, Linköping University Hospital; Birgitta Bohm, PhD, Karolinska Institutet; and Eva Eklof, MSc, Karolinska Institutet), physiotherapists (Cecilia Montgomery, MPT, Uppsala University; Annika Isberg, DDS, PhD, University of Umeå; Maria Bergström, MOT, Linköping University Hospital; and Annika Lundkvist Josenby, PhD, RPT, Lund University), and local study coordinators (Barbro Fossmo, RN, University of Umeå; Cecilia Ewald, RN, Uppsala University; Christina Fuxin, RN, Linköping University Hospital; Lena Swartling-Schlinzig, RN, Karolinska Institutet; Ann-Catherine Berg, RN, Lund University; Cecilia Tobiasson, EN, Gothenburg University; and Pia Lundqvist, RN, PhD, Lund University) who participated in the EXPRESS study. The English language of the manuscript was revised by Annette Whibley, MD, Wizard Communications. Ms Whibley was compensated for her contribution from Dr Farooqi, and the psychologists and physiotherapists mentioned received compensation from the principal investigators of the EXPRESS study and/or from one of the authors.
Create a personal account or sign in to: