SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Aim  Motor skill impairment is a common negative outcome in children born preterm who do not develop cerebral palsy (CP). This study aimed to conduct a systematic review of current data to provide an accurate estimate of the prevalence of non-CP motor impairment in preterm children at school age.

Method  We searched the Medline, PubMed, and PsycInfo databases and relevant journals to identify all studies published post-1990 that reported the prevalence of motor impairment in school-aged children born preterm (<37wks’ gestation) using standardised motor assessment batteries. We applied a range of exclusionary criteria, with 11 studies included in the final analyses. We identified two levels of motor impairment commonly reported – mild–moderate and moderate – and conducted a random effects meta-analysis to produce a prevalence estimate for each.

Results  The pooled estimate for mild–moderate impairment in preterm children was 40.5/100. and for moderate motor impairment the estimate was 19.0/100. There was also a trend for lower motor impairment levels in samples born before 1990 compared with those born after 1990.

Interpretation  Children born preterm are at increased risk of motor impairment, with prevalence three to four times greater than in the general population. This highlights the need for improved surveillance and intervention strategies in this group of children.

List of Abbreviations
BOTMP

Bruininks–Oseretsky Test of Motor Proficiency

DCD

Developmental coordination disorder

PDMS

Peabody Developmental Motor Scales

TOMI

Test of Motor Impairment

Motor skill impairment is a common negative outcome of preterm birth, with cerebral palsy (CP) the most severe form.1 However, many preterm children who do not develop CP still present with impaired motor skills, similar to that observed in developmental coordination disorder (DCD). Such impairment is known to have a significant negative impact reaching beyond the motor domain, with deficits commonly observed in educational, behavioural, and social domains among children with DCD.2–5 However, this type of impairment is often overshadowed in the preterm population by more severe physical and intellectual impairments. To inform surveillance and intervention services for preterm children, we need accurate estimates of the prevalence of motor impairment, an understanding of the nature of this impairment, and an appreciation of associated risk factors.

The prevalence of DCD in the general paediatric population is estimated to range between 5 and 1/10.6 In contrast, the prevalence of motor impairment in preterm cohorts without CP appears higher when reports are considered from individual studies. However, given the absence of a systematic review, we are without a reliable, overall estimate of prevalence. It is possible that motor impairment is the most common form of impairment in this population. If this is the case, appropriate resources should be assigned to detect and treat these difficulties. Therefore, the aim of this systematic review was to calculate an overall estimate of the prevalence of motor impairment (excluding CP) at school age in preterm children by pooling all of the available published data.

Method

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Searching

We searched the Medline, PubMed, and PsycInfo databases for English-language journal articles published between 1990 and 2008. We used the following search keywords: ‘premature’, ‘prematurity’, ‘preterm’, and ‘low birthweight’. We combined each of these with the following terms: ‘developmental coordination disorder’, ‘motor impairment’, ‘motor skills’, ‘Movement ABC’, ‘Test of Motor Impairment’, and ‘Bruininks–Oseretsky’. We also searched the reference lists of relevant articles and manually searched the table of contents of relevant journals (e.g. Archives of Disease in Childhood, Developmental Medicine and Child Neurology, Early Human Development, and Pediatrics) during this period.

Selection

We originally included all studies that used a standardised motor assessment battery (e.g. the Movement Assessment Battery for children [Movement ABC]7) to assess the motor skills of school-aged children who had been born preterm (<37wks’ gestation; n=78). We then excluded studies that (1) reported on a selective sample of preterm children (i.e. children with a specific disorder, lesion, or condition); (2) provided only mean scores for the movement assessments, rather than providing the proportion of children scoring outside of a designated cut-off score; (3) did not provide information about the cut-off score used to define motor impairment; (4) did not provide an overall motor score, instead only providing fine and/or gross motor scores; (5) included children born before 1980, to ensure the estimate was based on reasonably contemporary cohorts; and (6) did not indicate that children with CP were excluded from analysis.

There were several studies identified where the same data were presented in two or more separate papers. In this instance, we selected the article with the greatest amount of relevant information to use. When longitudinal data were available for the same cohort,8 data from the most recent follow-up were used.

Table I provides a summary of the 15 studies that met the inclusion criteria. Note that one of these studies reported on three groups of preterm children, two of which had received intervention.9 For this study, only data from the third group (no intervention) were included. Also, although one study included children with CP, the authors also reported impairment rates for children without CP.10

Table I.   Summary of identified studies
StudyGestational age (wks)Birthweight (g)Birth yearsAge at follow-up (y)TestCut-offs for impairment
  1. aIncluded in mild–moderate estimate; bincluded in moderate estimate; TOMI, Test of Motor Impairment; BOTMP, Bruininks–Oseretsky Test of Motor Proficiency; M-ABC, Movement Assessment Battery for Children; PDMS, Peabody Developmental Motor Scales; Z-NMA, Zurich NeuroMotor Assessment; RMPI, Riley Motor Problems Inventory.

Davis et al.27a,b<28<10001991–19928M-ABC5th and 15th centiles
de Kleine et al.28a,b<32<15001992–19945M-ABC5th and 15th centiles
Foulder-Hughes & Cooke29b<321991–19927–8M-ABC5th centile
Hack et al.20a,b<10001992–19958BOTMP1SD and 2SD
Hall et al.23<150019848M-ABC10th centile of controls
Holsti et al.30a<8001982–19879BOTMP1SD
Johnson et al.9<331990–19935M-ABC25th centile of controls
Jongmans et al.19a,b<351984–19866M-ABC15th centile of controls
Powls et al.8a,b<12151980–198112–13M-ABC5th and 15th centiles
Rademaker et al.31a,b≤32≤15001991–19937–8M-ABC5th and 15th centiles
Schmidhauser et al.10<12501992–19946Z-NMA10th centile
Sommerfelt et al.32b<20001986–19885PDMS5th centile of controls
Torrioli et al.33a<15001991–19934–6M-ABC5th and 15th centiles of controls
Whitaker et al.34<20001984–198716RMPI10th centile
Wocadlo & Rieger14a,b<301987–19978BOTMP5th and 15th centiles

Data abstraction and synthesis

The studies included in the analysis were reviewed by a single reviewer (JW), who extracted all relevant data from each article. This included the assessment and cut-off used, sample size, and the number of children impaired. Any uncertainties were discussed and resolved with the other authors.

Table I demonstrates the range of measures and cut-offs used to indicate motor impairment. To combine the results from various studies regardless of measure used, we explored prevalence of motor skill impairment at two different levels of impairment –‘mild–moderate’ and ‘moderate’– based on commonly used test cut-off scores that indicate the presence of motor impairment (see below). In doing so, we included studies assessing children with the following test batteries: the Movement ABC and its predecessor, the Test of Motor Impairment (TOMI),11 the Bruininks–Oseretsky Test of Motor Proficiency (BOTMP),12 and the Peabody Developmental Motor Scales (PDMS).13

We defined moderate impairment as an impairment cut-off (or raw score equivalent) of the 5th centile on the Movement ABC, 2SD below the mean on the BOTMP, and the 5th centile on the PDMS. For the TOMI, a score of 6 or more was used. We defined mild–moderate impairment as an impairment cut-off of the 15th centile on the Movement ABC, 1SD below the mean on the BOTMP, and a score of 4.0 to 5.5 on the TOMI. Studies where only the prevalence of moderate impairment or mild–moderate impairment was reported were used directly in the corresponding analysis. In addition, the mild–moderate analyses included studies that reported prevalence of moderate and mild impairment separately by adding the proportion with mild and moderate impairment together. We then estimated the prevalence of these two levels of impairment. Note that although information is provided in Table I about studies that used the 10th centile as a cut-off score, these were not included for further analysis owing to the small number reporting this result. Inclusion of these studies in the mild–moderate or severe analyses was not deemed appropriate as, by definition, we would expect lower/higher prevalence estimates in studies using the 10th centile rather than the 15th/5th centile and hence their inclusion would skew the results in both analyses.

We used the extracted data to calculate the percentage of children impaired in each sample and the 95% confidence interval (CI) of this estimate. Data analysis was conducted using Stata 10 (Statacorp, Texas, USA) by KJL and JW. Results from the various studies were then combined using random effects meta-analysis to obtain a pooled estimate of the prevalence of mild–moderate and moderate motor impairment, allowing for between-study variability (heterogeneity). Analyses were also considered separately for each motor measure, the type of normative data used to define impairment, and studies with children born before and after 1990 to determine whether these differences led to different estimates of prevalence.

In addition to the meta-analyses, we conducted a pairwise correlation between the age of the study sample at follow-up and the prevalence of impairment using both the mild–moderate and moderate cut-offs. Where studies included children across ages, we used the mean of these ages to provide a single figure; for example, if a study included children aged 8 to 9 years, we used 8.5 years for the correlation.

Results

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References

Data extraction from the 15 studies resulted in nine estimates of prevalence at each impairment level. Table I provides details of each study, including the measure used, age at follow-up, and year of birth. Individual study sample sizes, prevalence levels, and 95% CIs for the proportion with mild–moderate and moderate impairment can be seen in Figures 1 and 2 respectively.

image

Figure 1.  Prevalence of mild–moderate motor impairment by study with the pooled estimate from a random effects meta-analysis. Tinted line, pooled estimate of prevalence. Weight (%), the weight given to each study when calculating the pooled estimate.

Download figure to PowerPoint

image

Figure 2.  Prevalence of moderate motor impairment by study with the pooled estimate from a random effects meta-analysis. Tinted line, pooled estimate of prevalence. Weight (%), the weight given to each study when calculating the pooled estimate.

Download figure to PowerPoint

Reported prevalence of mild–moderate motor impairment ranged from 22.2 to 72.2/100 (Fig. 1), with a pooled estimate of 40.5/100 (95% CI 32.1–48.9/100). [Correction added after publication 10 December 2009: in the preceding sentence the text ‘to 72.2’ was added.] Reported prevalence of moderate motor impairment ranged from 9.5 to 34.0/100 (Fig. 2), with a pooled estimate of 19.0/100 (95% CI 14.2–23.8/100). In both cases there was strong evidence of between-study heterogeneity (p<0.001), with estimates of the between-study SD of 12.0/100 for mild–moderate impairment estimate and 6.6/100 for the moderate impairment estimate (proportion of variability attributable to heterogeneity I2=91.3% and 85.9% respectively).

The two most commonly used motor assessment batteries in these studies were the Movement ABC and the BOTMP (Table I). Separate meta-analyses for the studies using each of these tests showed similar estimates of prevalence for each test (Table II).

Table II.   Pooled estimates (% impaired) by variables of interest
 Moderate impairmentMild–moderate impairment
n%CIn%CI
  1. n, Number of studies included in category; BOTMP, Bruininks–Oseretsky Test of Motor Proficiency; CI, confidence interval of pooled estimate.

Movement ABC620.212.9–27.4641.529.8–53.3
BOTMP215.011.8–18.3339.225.1–53.2
Test norms used for impairment cut-off718.512.7–24.3736.327.7–45.0
Impairment cut-off based on reference sample220.916.1–25.7257.430.0–84.8
Born before 1990323.116.7–29.6342.634.1–51.0
Born after 1990517.510.2–24.9541.727.9–55.5

Performing separate meta-analyses for the different types of impairment cut-off used suggested that studies using a local reference sample, rather than test norms, tend to identify a greater proportion of their sample as having motor skill impairment (Table II). This is particularly evident when the mild–moderate impairment cut-off was used. However, these findings should be viewed with caution owing to the small number of studies that included local reference samples.

Finally, Table II also shows the results of meta-analyses conducted separately on studies with children born before and after 1990. Note one study recruited their sample both before and after 199014 and was excluded from these calculations. These results suggest a small trend for higher prevalence levels of motor impairment in preterm children born before 1990 than those born after 1990; however, again, the small number of studies in each subgroup resulted in large and overlapping CIs.

Results of the pairwise correlation did not indicate a significant relation between age at follow-up and prevalence of motor impairment (r=−0.20 [n=9] for the mild–moderate cut-off and r=0.32 [n=9] for the moderate cut-off).

Discussion

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References

This study aimed to provide an estimate of the prevalence of motor impairment in the preterm population excluding children with CP. Given that it is thought that the prevalence of DCD, a form of motor impairment very similar to that seen in children born preterm, is approximately 5–1/10 of the general population,6 our findings highlight the inflated level of motor impairment in preterm populations. This suggests that we need to do far more to develop an understanding of such impairment in these children so that appropriate intervention programmes can be developed. The cause of this apparent inflated rate of motor impairment remains unknown, though it is likely to be related to high rates of white-matter pathology15 and altered brain development16 observed in this population.

Interestingly, despite the elevated prevalence of motor impairment in preterm samples, there is little evidence on the functional impact of this impairment or its developmental course. Because of this, it is possible for this impairment to be viewed as transient in nature, reducing its clinical recognition. Such a view was often taken with DCD before the publication of longitudinal studies in the 1990s which demonstrated that though a portion of children with DCD appeared to overcome their motor skill problems with age, for many, problems continued into adolescence and beyond.3,4 More recent studies, such as those by Cantell et al.,17 have supported those initial findings. Since that time, the wide-ranging psychosocial implications of DCD have also been well recognized,2–5 indicating that it is important not to underestimate the impact of motor impairment on preterm children. We have demonstrated here the inflated level of motor impairment in preterm children compared with the general population, which we believe is the first step in gaining greater clinical recognition for children with these difficulties. Evidence for the functional impact on daily living skills would be the next step in improving clinical services, followed by longitudinal studies, which continue into adolescence rather than stopping in late childhood, and which map the developmental course of these difficulties and provide evidence about their lasting nature.

To obtain the prevalence estimates of motor impairment, studies that were heterogeneous in terms of study design, outcomes, birthweight, and years of birth were included in this meta-analysis. We will discuss the effects of some of these factors, though our findings need to be interpreted cautiously owing to the small number of studies available.

Motor assessment battery

Most studies presented here used the Movement ABC or BOTMP. The results of separate meta-analyses conducted for these two assessment batteries found similar prevalence levels. Other motor assessments may not yield similar findings. For example, some studies excluded from this review reported on motor impairment using only the Test of Visual Motor Integration.18 This test does not assess motor development across a range of skill types and yields greatly different results to those of an assessment battery. For example, Jongmans et al.19 reported an impairment level of 30.7/100 using the Movement ABC, but according to Test of Visual Motor Integration only 3.8/100 of the same sample was impaired. This highlights the importance of using standardised motor assessment batteries to define motor impairment.

A further issue about assessment batteries is the relevance of test norms. We reported on studies here that used the BOTMP at a time when the test norms were quite dated; in one instance, the test norms were more than 20 years old.20 Although we did not find a large difference in reported prevalence between the Movement ABC and the BOTMP, outdated norms could over- or underestimate prevalence levels, and assessment batteries should be chosen with this in mind. Importantly, both the Movement ABC and the BOTMP have recently been revised and both provide revised normative data. It is recommended that the revised versions of both tools be used in future studies to ensure prevalence estimates remain accurate.

Impairment cut-off used

Our review highlights the importance of the cut-off used to classify children as motor impaired. First, Figures 1 and 2 demonstrate the large difference, which is to be expected, between prevalence when a moderate impairment cut-off is used compared with a mild–moderate cut-off. We would suggest that, for ease of comparison, future studies report impairment rates using both levels of impairment.

Further, if a reference sample is to be used to create local norms, we would recommend reporting impairment rates according to both the local and test norms. Our results, though limited by the small number of studies, indicate that reference samples might result in higher levels of impairment than test norms. Although using a reference sample to create local norms allows researchers to control, to a limited extent, environmental factors that might play a role in children’s development, they do not allow results to be replicated easily or compared with other studies. A brief inclusion of results according to test norms would assist in this respect.

Year of birth

The era in which the infant was born and treated might also influence the prevalence of motor impairment. Our meta-analysis, although again limited by the small number of studies, indicated slightly lower prevalence in children born after 1990 than in those born before. Survival rates increased significantly after 1990 owing to the introduction of surfactant and antenatal steroids,21 but reductions in mortality have not been matched by reductions in morbidity.22 Given that there are now more ‘high-risk’ infants (e.g. infants born <26wks) surviving in the post-1990 era than previously, the slight decrease in motor impairment in more contemporary cohorts is a positive finding that we hope is replicated as more cohorts are followed up.

Birthweight and gestational age

The risk of negative long-term outcomes of preterm birth is likely to increase with decreases in birthweight and gestational age. As such, it would be expected that motor impairment would be more prevalent as birthweight and gestational age decrease. Unfortunately, a few studies that directly compared across birthweights were excluded because they either reported only on estimates using the 10th centile23 or included children with mild CP in their sample.24,25 Unsurprisingly, though, these studies did provide support for the higher prevalence of motor impairment in lower birthweight groups.23–25 None of the studies included in this meta-analysis examined prevalence across gestational age categories. If future samples have sufficient numbers of participants to allow it, estimating rates of impairment in gestational age and birthweight subgroups is recommended.

Age at follow-up

We found no correlation between age of the sample at follow-up and prevalence levels. However, Marlow and et al.26 have previously demonstrated that there was improvement in the TOMI scores of a sample of children who were born with low birthweight between 6 and 8 years of age. These improvements were maintained at age 12 years,8 indicating there might be a period of ‘catch-up’ for some preterm or low birthweight children during childhood. This indicates that age at follow-up could impact upon reported prevalence: the lack of a significant relation here could be the result of the lack of a representative spread of ages in the studies included in this review.

Limitations

The variability among published studies presented the biggest limitation to our systematic review, with many of these factors discussed above. The variability in approaches also resulted in several studies either not meeting our inclusion criteria or being excluded from analyses. This included, for example, studies that reported mean scores on assessment batteries but did not provide prevalence levels.

The variability among the modest number of studies available for inclusion in our analysis also led to limitations in the depth of our analyses and our ability to determine how the factors of variability might interact. For example, did the effect of the impairment cut-off used (test norms vs control sample) differ depending on the test battery used?

Finally, we were unable to explore the impact of birthweight and gestational age on prevalence of motor impairment. The likely impacts of these factors were discussed above and it is hoped that future research will report levels of impairment in a way that allows such analysis.

Conclusion

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References

This review highlights the increased likelihood of childhood motor skill impairment in preterm samples. Our meta-analysis found average prevalences of 19/100 and 40.5/100 for moderate and mild–moderate impairment respectively in the preterm population, which is in contrast to 5 to 1/10 in the general population.6 It is well established that the effects of motor skill impairment in DCD can extend well beyond the motor domain and into educational, social, and behavioural domains.2–5 It is also important that similar research is conducted in preterm samples to ensure greater clinical recognition for these impairments. Further, it is critical that research is undertaken to determine the underlying mechanisms associated with general motor impairment in this high-risk population.

References

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. References
  • 1
    Bracewell M, Marlow N. Patterns of motor disability in very preterm children. Ment Retard Dev Disabil Res Rev 2002; 8: 2418.
  • 2
    Dewey D, Kaplan BJ, Crawford SG, Wilson BN. Developmental coordination disorder: associated problems in attention, learning, and psychosocial adjustment. Hum Mov Sci 2002; 21: 90518.
  • 3
    Geuze R, Börger H. Children who are clumsy: five years later. Adapt Phys Activ Q 1993; 10: 1021.
  • 4
    Losse A, Henderson SE, Elliman D, Hall D, Knight E, Jongmans M. Clumsiness in children – do they grow out of it? A 10-year follow-up study. Dev Med Child Neurol 1991; 33: 5568.
  • 5
    Piek JP, Dworcan M, Barrett NC, Coleman R. Determinants of self-worth in children with and without developmental coordination disorder. Int J Disabil Dev Educ 2000; 47: 25972.
  • 6
    Barnhart RC, Davenport MJ, Epps SB, Nordquist VM. Developmental coordination disorder. Phys Ther 2003; 83: 72231.
  • 7
    Henderson SE, Sugden DA. Movement Assessment Battery for Children. London: The Psychological Corporation, 1992.
  • 8
    Powls A, Botting N, Cooke RWI, Marlow N. Motor impairment in children 12 to 13 years old with a birthweight of less than 1250g. Arch Dis Child Fetal Neonatal Ed 1995; 72: F626.
  • 9
    Johnson S, Ring W, Anderson P, Marlow N. Randomised trial of parental support for families with very preterm children: outcome at 5 years. Arch Dis Child 2005; 90: 90915.
  • 10
    Schmidhauser J, Caflisch J, Rousson V, Bucher HU, Largo RH, Latal B. Impaired motor performance and movement quality in very-low-birthweight children at 6 years of age. Dev Med Child Neurol 2006; 48: 71822.
  • 11
    Stott DH, Moyes FA, Henderson SE. Test of Motor Impairment. London: Psychological Corporation, 1984.
  • 12
    Bruininks R. Bruininks–Oseretsky Test of Motor Proficiency. Circle Pines, Minnesota: American Guidance Service, 1977.
  • 13
    Folio MR, Fewell RR. Peabody Developmental Motor Scales and Activity Cards. Manual. Allen: DLM Teaching Resources, 1983.
  • 14
    Wocadlo C, Rieger I. Motor impairment and low achievement in very preterm children at eight years of age. Early Hum Dev 2008; 84: 76976.
  • 15
    Woodward LJ, Anderson PJ, Austin NC, Howard K, Inder TE. Neonatal MRI to predict neurodevelopmental outcomes in preterm infants. N Engl J Med 2006; 355: 68594.
  • 16
    Beauchamp MH, Thompson DK, Howard K, et al. Preterm infant hippocampal volumes correlate with later working memory deficits. Brain 2008; 131: 298694.
  • 17
    Cantell M, Smyth MM, Ahonen TP. Two distinct pathways for developmental coordination disorder: persistence and resolution. Hum Mov Sci 2003; 22: 41331.
  • 18
    Beery KE. The Beery-Buktenica Developmental Test of Visual-Motor Integration. Administration and Scoring Manual, 4th edn. Parsippany, NJ: Modern Curriculum Press, 1997.
  • 19
    Jongmans MJ, Mercuri E, Dubowitz LMS, Henderson SE. Perceptual-motor difficulties and their concomitants in six-year-old children born prematurely. Hum Mov Sci 1998; 17: 62953.
  • 20
    Hack M, Taylor HG, Drotar D, et al. Chronic conditions, functional limitations, and special health care needs of school-aged children born with extremely low-birth-weight in the 1990s. JAMA 2005; 294: 31825.
  • 21
    Doyle LW, Group VICS. Evaluation of neonatal intensive care for extremely low birth weight infants in Victoria over two decades. I. Effectiveness. Pediatrics 2004; 113: 5059.
  • 22
    Anderson PJ, Doyle LW, Group VICS. Neurobehavioral outcomes of school-aged children born extremely low birth weight or very preterm in the 1990s. JAMA 2003; 289: 326472.
  • 23
    Hall A, McLeod A, Counsell C, Thomson L, Mutch L. School attainment, cognitive ability and motor function in a total scottish very-low-birthweight population at eight years: a controlled study. Dev Med Child Neurol 1995; 37: 103750.
  • 24
    Hack M, Taylor HG, Klein N, Eiben R, Schatschneider C, Mercuri-Minich N. School-age outcomes in children with birth weights under 750 g. N Engl J Med 1994; 331: 7539.
  • 25
    Pharoah POD, Stevenson CJ, Cooke RWI, Stevenson RC. Clinical and subclinical deficits at 8 years in a geographically defined cohort of low birthweight infants. Arch Dis Child 1994; 70: 26470.
  • 26
    Marlow N, Roberts L, Cooke RWI. Outcome at 8 years for children with birth weights of 1250g or less. Arch Dis Child 1993; 68: 28690.
  • 27
    Davis NM, Ford GW, Anderson PJ, Doyle LW. Developmental coordination disorder at 8 years of age in a regional cohort of extremely-low-birthweight or very preterm infants. Dev Med Child Neurol 2007; 49: 32530.
  • 28
    De Kleine MJK, Nijhuis-van der Sanden MWG, Den Ouden AL. Is paediatric assessment of motor development of very preterm and low-birthweight children appropriate? Acta Paediatr 2006; 95: 12028.
  • 29
    Foulder-Hughes L, Cooke R. Do mainstream schoolchildren who were born preterm have motor problems? Br J Occup Ther 2003; 66: 916.
  • 30
    Holsti L, Grunau RVE, Whitfield MF. Developmental coordination disorder in extremely low birth weight children at nine years. J Dev Behav Pediatr 2002; 23: 916.
  • 31
    Rademaker KJ, Lam JNGP, Van Haastert IC, et al. Larger corpus callosum size with better motor performance in prematurely born children. Semin Perinatol 2004; 28: 27987.
  • 32
    Sommerfelt K, Ellertsen B, Markestad T. Low birthweight and neuromotor development: a population based, controlled study. Acta Paediatr 1996; 85: 60410.
  • 33
    Torrioli MG, Frisone MF, Bonvini L, et al. Perceptual-motor, visual and cognitive ability in very low birthweight preschool children without neonatal ultrasound abnormalities. Brain Dev 2000; 22: 1638.
  • 34
    Whitaker AH, Feldman JF, Lorenz JM, et al. Motor and cognitive outcomes in nondisabled low-birth-weight adolescents: early determinants. Arch Pediatr Adolesc Med 2006; 160: 10406.