Motor Impairment in Children With Congenital Heart Defects: A Systematic Review

Original studies published in English or French and examining motor outcomes in infants, children, and/or adolescents (0–18 years) who were born with a CHD requiring open-heart surgery (with or without cardiopulmonary bypass) in the first 2 years of life were included in this review. Published randomized controlled trials, observational cohort or cross-sectional studies, and case-control studies in which standardized assessments were used to assess fine and/or gross motor skills were selected. Conference abstracts were not included because of our limited ability to assess the quality of the studies and risk of bias for this type of publication. Because of the fast-evolving surgical and medical care, only studies published since 1997 were included. Studies that included individuals born preterm (ie, <35 weeks’ gestation), individuals with genetic anomalies, or individuals requiring heart transplant were excluded unless the results of the subgroup analysis without these confounders were presented separately. We excluded these specific conditions in an attempt to enhance uniformity of the participants because these conditions have been independently associated with an increased risk of developmental delays. For similar reasons, reporting of motor assessments was limited to outcomes after open-heart surgery. In the case of longitudinal studies that presented the results of motor skills assessment at 2 time points within the same age category (eg, twice during infancy), only the data for the larger sample were analyzed.

A search of studies examining gross and fine motor skills in children aged 0 to 18 years born with a CHD was performed in Embase, Medline, and the Cumulative Index to Nursing and Allied Health Literature in September 2017 and updated in September 2019. Medical Subject Headings were used and exploded when appropriate. Boolean operators were used to combine “congenital heart defect/disease” and “psychomotor performance/motor skills/motor performance/child development” (see Supplemental Table 5 for detailed search strategy). References of the selected articles were screened to identify additional studies. A librarian assisted the authors in the development of the search strategy.

The study selection consisted of a two-step process. First, titles and abstracts were screened to determine if the studies met the inclusion criteria. Then full texts of the selected articles were reviewed. Data extraction and study appraisal were independently performed by 2 reviewers (M.-E.B. and E.D.). A data extraction form was developed on the basis of the Institute of Medicine’s Standards for Systematic Reviews,¹⁰  the University of York Centre for Reviews and Dissemination’s guidelines for reviews,¹¹  and the Strengthening the Reporting of Observational Studies in Epidemiology statement checklist.¹²  The Cochrane risk-of-bias tool¹³  was used for the quality appraisal of the selected studies. This tool is used to examine the risk of selection bias, performance bias, detection bias, attrition bias, and reporting bias, which are subsequently categorized as low risk, high risk, or unclear risk on the basis of the judgement of the reviewers. In case of disagreement, the decisions were discussed until a consensus was reached. Authors of the selected studies were contacted if clarification was needed to confirm that the article met the inclusion or exclusion criteria or if additional information was required to be included in our analyses. Some of these authors generously provided unpublished data or new calculations that excluded children who were born prematurely or had genetic syndromes.

To better appreciate the difference in motor skills across a child’s developmental trajectory, we clustered the findings in the following categories on the basis of the mean age of the children at the time of assessment: (1) infancy (<1 year of age), (2) toddler years (1–<3 years), (3) preschool age (3–<6 years), (4) school age (6–<13 years), and (5) adolescence (13–<19 years). To account for the range of severity in motor impairments and to be consistent with the definition of motor impairments reported across studies, we report the prevalence of all motor impairments (ie, mild to severe) and the presence of severe motor impairment separately. Mild to severe motor skill impairment was defined as a motor score <−1 SD from the normative mean or below the 15th percentile, whereas severe motor impairment was defined as a motor score <−2 SDs from the normative mean or below the fifth percentile. Descriptive statistics were used to summarize these results.

In addition, effect size (Hedge’s g statistic) for the difference in motor skills between children with CHD and typically developing children for each study that presented means and SDs was calculated independently for each study. The magnitude of the effect sizes were interpreted on the basis of Cohen’s convention as follows: 0.2 to 0.5, small effect size; 0.5 to 0.8, moderate effect size; and >0.8, large effect size.¹⁴  Range and interquartile range (IQR) were used to summarize prevalence and effect sizes. Finally, we synthesized the results of the studies in which subgroups of children with different cardiac physiology were compared and the studies in which fine and gross motor skills were reported separately.

A total of 1192 studies were identified through the literature search. Of the 1002 studies remaining after the removal of duplicates, 805 were eliminated in the first stage of screening, which involved screening titles and abstracts, and 156 were subsequently eliminated after reading the full text. Five additional studies were identified through a manual search of selected articles’ reference lists, leaving 46 original studies to undergo data extraction and analysis in this review. The detailed study selection process is presented in Supplemental Fig 3.

Studies included in this review originated from 5 continents. Sixteen (35%) studies were conducted in the United States,^15–30  7 (15%) in Canada,^31–37  6 (13%) in Germany,^38–43  4 (9%) in Australia,^44–47  2 (4%) in the Netherlands,^48,49  2 (4%) in Switzerland,^50,51  2 (4%) in Finland,^52,53  2 (4%) in Japan,^54,55  1 (2%) in Brazil,⁵⁶  1 (2%) in India,⁵⁷  and 1 (2%) in Norway.⁵⁸  In addition, 2 (4%) studies were conducted in centers located both in Canada and the United States.^59,60  A cohort study design was used in 28 (61%) studies,^* a randomized control trial design was used in 9 studies (20%),^{16–19,22,27,28,48,59}  a cross-sectional design was used in 6 studies (13%),^{21,26,36,39,53,57}  a case-control design was used in 2 studies (4%),^42,55  and a case series design was used in 1 study (2%).⁴⁰  Studies included in this review included samples composed of children with different types of CHD, including hypoplastic left heart syndrome or other types of single ventricle defects exclusively (8 studies; 17%),^{20,22,23,26,32,52,53,59}  transposition of the great arteries (7 studies; 15%),^{16,18,19,33,40,41,49}  ventricular septal defect (2 studies; 4%),^28,55  total anomalous pulmonary venous connection (1 study; 2%),³¹  and aortic arch obstruction (1 study; 2%).⁴⁸  Twenty-seven studies (59%) included samples composed of children with mixed CHD diagnoses.^† Study participants’ ages ranged from a few days of age to late adolescence (ie, 19 years).

Overall, 14 (30%) studies presented a low risk of bias on all assessed criteria, 4 (9%) studies presented a high risk of bias in 1 category, and only 1 (2%) study presented a high risk of bias in 2 categories. Regarding possible performance bias, the assessors were blind to medical history, diagnosis, other assessment results, or treatment assignment in 17 of the 46 (37%) studies. In terms of possible attrition bias, 2 (4%) studies lost a large number of participants, and the authors did not consider attrition in their analyses, therefore presenting a high risk of bias for this criterion, whereas in 4 (9%) studies, the authors did not report attrition numbers. Finally, baseline imbalance in factors related to outcomes such as age, sex, and maternal education was present in 4 (9%) studies. Detailed information on the quality of the studies is presented in Supplemental Table 6.

The prevalence of mild to severe motor impairments (<−1 SD) ranged from 12.3% to 68.6% (IQR: 23.4%–52.2%) in children and adolescents with CHD on the basis of the results from 13 studies that presented this outcome. Severe motor impairments (<−2 SD) were reported in 22 studies and ranged from 0.0% to 60.0% (IQR: 7.2%–29.8%) in children and adolescents with CHD (Fig 1).

Regarding the extent of impairment, the effect size for motor skill differences between children with CHD and typically developing children across ages ranged from −0.07 to 2.19 (IQR: 0.52–1.34) across 38 studies.

In 9 studies, the authors examined motor skills in infants younger than 12 months after open-heart surgery for CHD.^‡ Study results and characteristics are presented in Table 1.

In this age group, the prevalence of motor impairments (<−1 SD) was described in only 1 study. In this study, the authors reported a 20.0% prevalence of motor skill difficulties.²¹  Additionally, da Rocha et al⁵⁶  reported suspected delays in 55.0% of infants with CHD on the basis of the results of a screening test. The prevalence of severe motor impairments (<−2 SD) in infants with CHD ranged from 0.0% to 54.6% (IQR: 8.4%–50.9%).^20,21,24,57  The results are represented in Fig 2.

Regarding the extent of impairment, among the 8 studies in which the mean motor scores of infants with CHD were compared to normative data or compared with the mean motor scores of controls,^{15,20,21,24,25,28,46,57}  6 found that infants with CHD obtained significantly poorer scores than their comparison. The other 2 also found lower mean scores, but the results were not statistically significant. The effect size ranged from 0.12 to 1.61 (IQR: 0.53–1.43).

Motor development in toddlers was examined in 30 studies.^§ Study characteristics and results are presented in Table 2.

The prevalence of mild to severe motor impairments (<−1 SD) ranged from 12.3% to 59.0% in toddlers with CHD (IQR: 14.5%–58.6%),^{30,44,49,55,59}  whereas the prevalence of severe motor impairments (<−2 SD) ranged from 1.5% to 60.0% (IQR: 6.0%–26.1%).^‖ One study that included toddlers with a single ventricle physiology identified a significantly higher prevalence of severe motor impairment (60%) using the Bayley Scales of Infant Development, Second Edition when compared with the other studies in this category.⁵¹  Another study defined motor impairment as scores <−1.5 SDs below the mean and therefore could not be synthesized with the other studies.³⁴  They reported 42% of motor impairment in their sample. The results are represented in Fig 2.

Regarding the extent of impairment, in 24 of the 28 studies (85.7%) that presented this outcome, the authors reported that at least 1 subgroup of children with CHD performed significantly worse on motor assessments than typically developing children. Overall, the effect size for motor skill differences between toddlers with CHD requiring open-heart surgery and typically developing children ranged from −0.07 to 2.19 (IQR: 0.47–1.49).^¶

Ten studies^# reported motor outcomes in preschool-aged children. The characteristics and results of these studies are presented in Table 3.

Only 2 studies reported the prevalence of children with motor scores <−1 SD in this age group. The prevalence of preschool-aged children with mild to severe motor impairments ranged from 32.4% to 68.6%,^35,40  whereas the prevalence of children with severe motor impairments (<−2 SD) ranged from 7.7% to 28.6%.^23,35,48  The results are represented in Fig 2.

Regarding the extent of impairment, all studies examining motor scores in preschool-aged children found a significant difference when compared with typically developing children. Overall, preschool-aged children with CHD were found to have a difference in motor scores that ranged between 0.18 and 1.38 SDs (IQR: 0.66–1.19) below the mean when compared to normative data or compared with controls.^{23,35,40,42,45,52,54}  In addition, the results of 2 studies could not be synthesized above because they did not provide a prevalence nor mean score and SD. One of these studies reported that preschool-aged children performed worse on norm-referenced motor assessments, with an average score at the ninth percentile for gross motor skills and the fourth percentile in fine motor skills.¹⁹  The second study reported significantly lower balance and aiming and catching skills in preschool-aged children with CHD when compared to the assessment’s normative data but did not find a significant difference on the assessment’s total score or on the manual dexterity index.⁵⁰ 

Motor performance in school-aged children was presented in 5 studies.^{18,39,41,43,58}  Their characteristics and results are summarized in Table 4.

The prevalence of motor skills impairments (<−1 SD) in school-aged children ranged from 26.7% to 46.1% (IQR: 30.6%–45.2%),^18,39,41,58  and the prevalence ranged from 7.5% to 25.8% for severe motor impairments (<−2 SD).^39,41,58 

Regarding the extent of impairment, mean scores in school-aged children with CHD were lower than those of typically developing children in all reviewed studies. The effect size ranged from 0.16 to 1.21 across studies (IQR: 0.24–1.14).^39,41,43,58  The results are represented in Fig 2.

In only 1 study, authors examined the extent and prevalence of motor impairment in adolescents with CHD (Table 4).

Easson et al³⁶  found that 42.4% of adolescents with CHD presented with mild to severe motor impairments (scores <−1 SD) and that 18.2% had severe motor impairments (<−2 SD). The results are presented in Fig 2.

Regarding the extent of impairment, the average total score of adolescents in that study corresponds to the fifth to 15th percentiles when compared to norms.

In 7 studies, children with a single ventricle physiology were compared with those with 2-ventricle physiologies.^{15,24,25,30,37,54,60}  Five of these reported significant differences in at least 1 age group, with the single ventricle group performing worse.^{15,24,25,54,60}  In the other 2 studies, lower scores were similarly found in infants and toddlers with single ventricle physiology, but the differences were not statistically significant.^30,37  In addition, 1 study compared children with hypoplastic left heart syndrome to other single ventricle physiology and found that they performed significantly worse.⁵³  Finally, 2 studies reported better motor outcome in toddlers with transposition of the great arteries compared with children with other 2-ventricle defects.

We compared the prevalence of fine and gross motor impairments (<−1 SD) in children with CHD. Three studies reported gross and fine motor scores separately.^35,36,40  The results of these 3 studies suggested no significant differences between the 2 motor skill domains. Overall, gross motor difficulties were present in 18.2% to 68.8% of participants with CHD, whereas fine motor impairments ranged from 23.2% to 55.8%.

In this review, we highlight the high prevalence of motor impairments throughout childhood and adolescence in children with CHD requiring open-heart surgery. Indeed, the results of our systematic review have revealed that approximately one-third of children with CHD have delayed motor skills and that these impairments are present in comparable proportions throughout development. Conversely, we found that the prevalence of severe motor impairments (<−2 SDs) was higher in younger children, which is in line with preliminary longitudinal observational studies, suggesting that motor skills may improve over time in children with CHD.^52,61  This may be explained, in part, by the prolonged hospitalization that some infants may experience perioperatively and the recommended movement restrictions that follow surgery, which together could delay the acquisition of age-appropriate motor skills. Moreover, the severity of the motor impairments may be decreased with development and after receiving rehabilitation intervention.

The extent of the differences in motor skills detected in the current review indicates that motor scores in children and adolescents with CHD requiring open-heart surgery were ∼1 SD below those of typically developing children. In daily life, these children may seem clumsy; may experience difficulty scribbling, cutting, doing buttons or zippers, or jumping; or may avoid participation in motor-based activities.^62,63  Although these motor impairments may seem subtle or mild, they can have long-lasting impacts on the psychological and physical well-being of individuals. For instance, children and adolescents with developmental coordination disorder, a disorder affecting motor coordination, have lower self-esteem and higher anxiety compared with age-matched controls.⁶⁴  In typically developing children, poorer motor coordination is associated with impaired emotion recognition and social behavior.⁶⁵  Furthermore, motor impairments are associated with reduced participation in physical activity, which can have an important impact on health.^66,67  In a recent article by Majnemer et al,⁶⁸  poorer motor performance in adolescents with CHD has also been associated with lower participation in active physical and social activities. It is therefore expected that the presence of motor impairments during childhood will have long-lasting consequences on daily function, self-determination, and well-being unless effective interventions to overcome these difficulties are provided. For infants and younger children, investments in the development and implementation of early intervention programs by children who are at high risk of developmental delays have been supported by various organizations and governmental agencies.^69,70  They have proven to be effective in other high-risk populations,⁷¹  and there is preliminary evidence supporting their effectiveness in children with CHD.⁷²  For children and adolescents, other approaches, such as the Cognitive Orientation to Daily Occupational Performance intervention as well as occupational and physical therapy, have also been shown to be effective in children with developmental coordination disorder,⁷³  who present a motor profile similar to that of children with CHD.³⁶ 

Early identification of motor impairments and timely referral to rehabilitation services is essential to optimize daily function and quality of life. However, the subtleness of the impairments identified in this review may pose a challenge to early identification. In 2012, the American Heart Association published guidelines for the developmental follow-up of children with CHD.³  In these guidelines, it was recommended that all children who underwent open-heart surgery be referred for formal developmental evaluation and early intervention by the primary care provider or subspecialist. Currently in North America, many tertiary care centers have implemented developmental follow-up clinics for children with CHD. However, other centers rely on the assessment of the cardiologist, neonatologist, or primary care providers during routine follow-up. The implementation of the guidelines has not been investigated outside the United States or in American tertiary care settings; however, one study suggests that awareness of the recommendations is limited among American primary care providers.⁷⁴  Current clinical practices often rely on general observations of motor milestones and parental reports and may not be sensitive enough to promptly detect more subtle difficulties, especially in older children who have learned to cope with and compensate for these difficulties. The results of this review demonstrate that standardized screening and evaluation of developmental delays, including motor delays, throughout childhood and adolescence are essential. Timely referral to rehabilitation services is required when a delay is identified to provide these children with the best opportunities to optimize their motor skills, particularly because this review suggests that these difficulties will not be outgrown.

This review also suggests that although all children with CHD who underwent open-heart surgery have a higher prevalence of motor impairments, their severity may differ on the basis of the complexity of the defect. Infants and toddlers with single ventricle physiology were found to have significantly poorer motor abilities than children with 2 ventricles. These findings are in line with a number of studies that have suggested that specific diagnoses may be associated with severity of motor impairments.^17,39,52,53  The poorer developmental outcomes in children with single ventricle physiology may be attributed to various risk factors, including reduced substrate delivery starting in the fetal period,^75–77  the need for multiple surgeries,^38,44,59,78  and the higher prevalence of brain abnormalities.^48,52,79  Although acquired brain injury and brain developmental malformation are frequent in neonates with CHD,⁸⁰  they, along with the other clinical factors found to be associated with motor skills, have not been consistently reported or evaluated in the reviewed studies. More studies examining risk factors associated with motor delays are needed to identify possible avenues to optimize outcomes. However, considering that all children with CHD who underwent open-heart surgery are at high risk of motor impairments across childhood and adolescence, each would benefit from close follow-up with standardized screening or evaluation of motor skills, notwithstanding the presence of additional risk factors.

This review must be interpreted in the context of its limitations. Variability in sampling and methodology between reviewed studies is the most important limitation in this review. The reviewed studies included children with different cardiac physiology, who may have different motor outcomes. In terms of methodology, the different motor evaluation tools used may have disparate results across the different studies because assessments may vary in sensitivity or types of motor skills assessed. Furthermore, different versions of the same test could yield different results. In a recent study, the authors concluded that the third edition of the Bayley Scales of Infant and Toddler Development may underestimate motor impairments.⁸¹  In this review, 17 studies used the second edition and 4 studies used the third edition of that test. Finally, several studies identified a limited ability of the Alberta Infant Motor Scales to discriminate between infants >15 months of age.^46,82,83  This ceiling effect could have confounded the results of 3 of the reviewed studies examining motor skills in toddlers and may have contributed to the variability in that age group. Nonetheless, these different tools are valid measures of motor performance and identify children with impairment at different ages.

In addition, although we included only studies that excluded children with genetic findings or syndromes, not all children were systematically tested; therefore, it is possible that some studies included a larger proportion of children with genetic variants, a subgroup that may have poorer motor outcome.^84,85  In addition, our review excluded studies that included children who had undergone heart transplant and who may present with more severe developmental difficulties. Therefore, our results cannot be generalized to that subgroup. Nevertheless, our results may include a small subset of children who required a heart transplant at a later time point in their life; however, this remains purely speculative. Although we could identify a risk of bias in some of the reviewed studies, the evaluation of the risk of bias across reviewed studies remained limited because of the few studies that were assessed as having a high risk of bias in each domain. Nevertheless, the lack of explicit blinding found in several studies constitutes an important limitation.

The results of this review highlight that infants with CHD have an increased risk of motor skill impairments across infancy, childhood, and adolescence. These findings stress the importance of the latest American Heart Association guidelines,³  which recommend the implementation of systematic standardized screening or evaluation of motor skills across childhood and adolescence in children with CHD requiring open-heart surgery to allow for timely detection of motor impairments throughout childhood and adolescence.

CHD

congenital heart defect

IQR

interquartile range