https://orcid.org/0000-0002-6138-7080
Abstract
The World Health Organization's International Classification of Functioning, Disability, and Health framework, Children and Youth Version (ICF-CY), provides a valuable method of conceptualizing the multidomain difficulties experienced by children born very preterm (VP). Reviews investigating motor outcomes at preschool age across ICF-CY domains are lacking.
The purpose of this review is to identify and compare motor outcomes of 3- to 6-year-old children born VP and children born full-term (FT) within the ICF-CY framework.
Four electronic databases and reference lists of included and key articles were searched.
Studies comparing motor outcomes of 3- to 6-year-old children born VP (<32 weeks’ gestation or birth weight <1500 g) with peers born FT were included.
Two independent authors extracted data and completed quality assessments.
Thirty-six studies were included. Activity motor performance of children born VP was consistently poorer compared with peers born FT: standardized mean difference (SMD) was –0.71 (95% CI = –0.80 to –0.61; 14 studies, 2056 participants). Furthermore, children born VP had higher relative risk (RR) of motor impairment (RR = 3.39; 95% CI = 2.68 to 4.27; 9 studies, 3466 participants). Body structure and function outcomes were largely unable to be pooled because assessment tools varied too widely. However, children born VP had higher RR of any neurological dysfunction (Touwen Neurological Examination) (RR = 4.55; 95% CI = 1.20 to 17.17; 3 studies, 1363 participants). There were no participation outcome data.
Limitations include the lack of consistent assessment tools used in VP follow-up at preschool age and the quality of the evidence.
Children born VP experience significant motor impairment across ICF-CY activity and body structure and function domains at preschool age compared with peers born FT. Evidence investigating participation in VP preschool-age populations relative to children born at term is sparse, requiring further research.
Advances in neonatal and obstetric care have improved the survival rate of infants born preterm,1 especially those born at younger gestational ages.2 Despite these advances, preterm birth remains an important international concern due to short- and long-term morbidity.3 Children born very preterm (<32 weeks’ gestation) experience motor impairment from infancy to adolescence,4 with a prevalence of cerebral palsy (CP) alone of 43 per 1000 live births in those born very preterm (VP).5 There is increasing awareness of mild to moderate subsequent motor impairment in this population,6 and appreciation that even minor motor impairment can have implications for other areas of functioning, including cognition, academic ability, and behavior.7,8
Recent guidelines for the follow-up of children and young people born preterm highlight the importance of the multidisciplinary team in developmental surveillance, and specify the involvement of physical therapists.9 Physical therapists have an important role in identifying and treating the motor impairments of children born VP, and in developmental surveillance from infancy to preschool age.9 However, the difficulties faced by these children can be complexly intertwined; consistent and detailed classification of outcomes is needed in this population.
The World Health Organization's International Classification of Functioning, Disability, and Health, Children and Youth Version (ICF-CY) framework, provides an established and valuable method of conceptualizing the multidimensional difficulties faced by children born VP across the domains of body structure and function, activity, and participation.10 The ICF-CY uses these domains, alongside environmental and personal factors, to holistically describe the impact of disability on individual functioning and on life experiences.11 Body structures are defined as anatomical body parts, and body functions include the body's physiological processes, with deficits labelled as impairments.10 Activity and Participation domains are grouped together in the ICF-CY, with activity defined as the execution of a specific task or action, whereas participation is broadly described as involvement in a life situation.10 Difficulties in the activity and participation domains are expressed as activity limitations and participation restrictions.10
Prior to our review, the most recent systematic review and meta-analysis of motor outcomes for infants born VP or with very low birth weight (<1500 g) was published in 2009.4 The review by de Kieviet et al focused on motor outcomes using meta-analysis of only 3 standardized assessment tools,4 but did not include other commonly used measures such as the Peabody Developmental Motor Scales and the Touwen Neurological Examination. Moreover, 12 additional studies have been reported since 2009. Using the ICF-CY framework, our review provides a comprehensive synthesis of updated evidence and presents a uniquely holistic overview of the motor functioning at preschool age of children born VP.
The ICF has been increasingly used to categorize outcomes for various pediatric conditions, including developmental coordination disorder (DCD),12 CP13,14 and congenital hemiplegia,15 as well as traumatic brain injury.16 However, this framework has yet to be used to describe motor outcomes for children born VP compared with term-born peers at preschool age. Our study aims to identify motor outcomes of children born VP between 3 to 6 years of age, and compare these outcomes with children born FT. Our second aim is to describe motor outcomes within the ICF-CY domains of body structure and function and of activity and participation.
Methods
Our study was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)17 guidelines, and the study protocol is published in PROSPERO International Prospective Register of Systematic Reviews (CRD 42,016,037,753).
Data Sources and Searches
We identified relevant studies by searching 4 electronic databases (Medline, Cinahl, Embase, and PsycINFO) between March 20 and April 14, 2016, with results limited to year of publication (from 1990) and restricted to the English language. Database searching was updated on January 10, 2017. We conducted 3 discrete searches for each database using search terms specific to motor body structure or function, activity, and participation. The full search strategy is provided in eAppendix 1. Where appropriate, we used subject thesauri to map keywords to subject headings and exploded relevant terms.
Two independent authors (T.F. and A.K.) removed duplicates, screened the titles and abstracts of retrieved articles, and obtained full text articles when necessary. The same 2 authors manually screened the reference lists of 3 key articles,4,18,19 and the first author (T.F.) screened the reference lists of included studies. An independent third author (A.S.) resolved disagreements during the screening process. When 2 articles reporting results for the same study met inclusion, the most recently published article or the article with the most complete sample was included. Multiple articles for the same study were included if different motor outcomes were presented in each publication. The flow chart of searched, identified, and included studies is outlined in Figure 1.
Flow chart of the searched, identified, and included studies. * = some studies satisfied more than 1 criterion. FT = full term; VP = very preterm.
Study Selection
Studies met inclusion criteria if they compared motor outcomes of 3- to 6-year-old children born VP (<32 weeks’ gestational age) or with very low birth weight (<1500 g) with their peers born FT (≥37 weeks’ gestation and ≥2500 g). Studies defined by very low birth weight were included as a proxy for <32 weeks’ gestation since some would have involved cohorts where gestational age was more uncertain, but birth weight was known. Term participants also met inclusion criteria if they were only described as “full-term” (FT) or “term-born.” To fulfill inclusion, motor outcomes needed to be consistent with preselected ICF-CY10 body structure, body function, activity, or participation code-sets. Children born before 1980 were excluded due to changes in neonatal care and outcomes after the 1970s. The mean or median value, and 1 standard deviation (SD) above and below this value, were used when available to determine eligibility based on participant gestational age, birth weight, and age at assessment (either chronological or corrected age for VP participants). Studies reporting outcomes for children born ≥32 weeks’ gestation, in addition to VP participants, were included if outcomes were reported separately for those born <32 weeks’ gestation. Authors were contacted to determine inclusion criteria when necessary, and responses were considered until January 30, 2017. Systematic or other reviews of the literature, non-peer-reviewed literature, and studies published in languages other than English were excluded.
Data Extraction and Quality Assessment
Two independent authors (T.F. and A.K.) extracted data using a template created for our review. These 2 authors independently assessed the methodological quality of included studies using the Newcastle-Ottawa Quality Assessment Scale (NOS) for cohort studies.20 This method of quality assessment is recommended for nonrandomized studies by the Cochrane Collaboration21 and contains 3 categories of items: selection, comparability, and outcome.20 A star rating provides an assessment of study quality, with a maximum rating of 9 stars indicating the highest methodological quality. For quality assessment, we considered that subjects lost to follow-up were unlikely to introduce bias when follow-up rates were ³ 85%, or between 75% and 85% with an accompanying statement indicating no substantial differences between nonparticipants and participants within the study. A third author (A.S.) resolved disagreements during data extraction and quality assessment.
Data Analysis
When outcome data were unable to be provided by individual authors or authors were unable to be contacted, values were obtained from published figures using an online data extraction tool where possible.22 Median values were used as the best estimate of the mean23 when sample sizes were larger than 25 participants, as recommended by Hozo et al.24 When outcomes matched ICF-CY10 code-sets equally in multiple domains, the data were linked to all relevant domains. During this process, existing guidelines for ICF code-set linking25,26 and participation frameworks27–29 were considered. Overall test scores were used in meta-analysis; however, where only subscale data were presented, ICF-CY10 code-sets were used to decide which subscales best represented the domain being analyzed. When studies assessed children using more than 1 instrument within the same ICF-CY domain, studies presenting overall test scores or those with the most complete sample sizes if overall score was not available were included in the meta-analysis. We excluded outcomes from quantitative analysis when only individual item data from a standardized assessment were presented. Corrected age is calculated by subtracting the number of weeks a child was born preterm from his or her chronological age, and is commonly used in research settings.30 If both uncorrected and corrected scores were presented, we preferentially used data that were corrected for prematurity in analysis to avoid introduction of bias caused by underestimating performance of children born preterm.30 When studies presented cohort data at multiple time points within the preschool period, outcomes for the older age only were included.
Data were analyzed using Stata version 14.2 (StataCorp, Texas, United States of America), and meta-analysis was conducted with studies grouped by outcome. Statistical heterogeneity was assessed using the I2 statistic and analyzed by a fixed effects model (Mantel and Haenszel)31 if I2 was <50%. A random effects model using the method of DerSimonian and Laird31 was used when the I2 statistic exceeded 50%. All meta-analysis data are presented alongside study weight (%), 95% CI, P values, and I2 statistics, and expressed as standardized mean difference (SMD) or risk ratio. When necessary, mean values were multiplied by –1 to adjust scale direction; this applied to scores for the Movement Assessment Battery for Children, first edition (MABC-1) and measures of speed, where a high score indicated worse function. A sensitivity analysis was completed, excluding data obtained from published figures, raw data, and studies with uncertain sample sizes. Data that were unable to be pooled for meta-analysis were quantitatively analyzed to produce SMDs (95% CI) and graphically presented in forest plots.
Role of Funding Source
This work is supported by grants from the National Health and Medical Research Council of Australia (Centre of Research Excellence, ref. no. 1,060,733; Early Career Fellowship, ref. no. 1,053,787 to J.L.Y. Cheong; Career Development Fellowship, ref. no. 1,108,714 to A.J. Spittle) and from the Victorian Government's Operational Infrastructure Support Program. The funding sources had no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.
Results
Identified Studies and Study Characteristics
We identified 2812 articles from electronic database and reference list searching after removing duplicates (Fig. 1). After excluding articles based on title and abstract, 89 full text articles were screened for eligibility. Thirty-eight articles representing 36 studies met inclusion criteria, and 22 of these studies23,32–52 were pooled in 2 meta-analyses.
The characteristics of included studies and the perinatal characteristics of VP participants are detailed in the Table. Included studies are of cohort design, except 2 which are follow-up studies of randomized control trials, with results compared with term reference groups at follow-up.50,53 Included studies were conducted in Europe (n = 21),23,34–36,38,39,44,47,49,50,54–64 North and South America (n = 10),32,37,41,42,45,46,48,51,52,65–67 Australasia (n = 3),33,40,53 and the Middle East (n = 2),43,68 and involved 8409 participants (excluding unclear participant numbers).63,64 Studies investigated children throughout the preschool period: 3 to 4 years (n = 19)23,32,33,37,39,41,43,45–48,50,53,56–59,61,64,67 and 5 to 6 years (n = 15).34–36,38,40,42,44,49,52,54,55,60,62,65,66,68 Two studies reported data for children at multiple time points within the preschool period;51,63 however, only outcome data for the oldest age were included in our analysis. Nine studies presented outcome data which were corrected for prematurity.32–34,38,46,47,49,50,53
Quality Assessment
The highest NOS rating was 8 stars, achieved by 4 studies23,32,41,57 (eAppendix 2), and the median score was 6 stars (ranging from 8 down to 3). Based on participant inclusion and exclusion criteria, the VP participants were generally representative of the average preterm child in the community (n = 32, 84%), and term controls were mainly selected from the same community as VP participants (n = 32, 84%). Birth status was determined from neonatal records for all but 1 study, which was inadequately described.62 All 36 studies demonstrated that outcomes of interest were not known at study commencement, while the comparability of cohorts on basis of design or analysis was inconsistently reported across articles; 12 (32%) controlled for age and any other variable, 19 (50%) for one variable only, and 7 (18%) did not control for any variable. Outcome assessment was completed poorly, with 25 (66%) articles providing no description of independent or blind assessment, and a further 3 studies reporting unblinded assessment.38,49,52 Follow-up length was adequate for outcomes to occur in all studies; however, follow-up rates were poor overall. The majority of articles (n = 25, 66%) received zero stars for this criterion; <85% follow-up without accompanying statement indicating if substantial differences between non-participants and participants existed within the study (n = 22), absent statement of follow-up rate (n = 3).56,63,64 In addition to formal quality assessment, several studies excluded data from analysis based on unfavorable outcome (eg, CP).32,36–38,40,42,49,50,55,59,62
Body Structure and Function Domain
Fourteen different outcome measures were used to investigate body structure and function, with several of these outcomes also linked to the activity domain, including the Beery-Buktenica Developmental Test of Visual Motor Integration (VMI) and the McCarthy Scales of Children's Ability (MSCA) motor subscale. The only outcome solely measuring body structure and function, and able to be pooled in meta-analysis, was the Touwen Neurological Examination.69
Children born VP had higher relative risk (RR) of any neurological dysfunction, as assessed by the Touwen Neurological Examination (eAppendix 3) (RR = 4.55; 95% CI = 1.20 to 17.17; P = .026; I2 = 67.3%; 3 studies; 1363 participants); and performance of children born VP on the VMI was poorer than their peers born FT (SMD = –0.66; 95% CI = –0.98 to –0.34; P = .007; I2 = 71.5%; 5 studies; 739 participants). Motor coordination was assessed with 2 tools, the Motor Coordination subtest of the VMI50 and Koerperkoordinationstest fuer Kinder (whole body coordination test for children).65 Children born VP had poorer coordination compared with their peers born FT (SMD = –0.47; 95% CI = –0.76 to –0.17; P = .002; I2 = 21.6%; 2 studies; 249 participants). The MSCA motor subscale also measures upper and lower limb coordination, and performance of children born VP was again poorer than children born FT (SMD = –0.98; 95% CI = –1.38 to –0.58; P <.001; I2 = 53.8%; 2 studies; 336 participants).
We were unable to complete meta-analysis of other outcomes within the body structure and function domain, as assessment tools varied too widely. We contacted several authors for data to include in the body structure and function meta-analysis; however, authors were unable to provide data68 or were unable to be contacted.59,60 Individual SMD values for outcomes that were unable to be pooled in meta-analysis are presented in eAppendix 4; the SMD for individual items ranged from 0 to –2.14.
Activity Domain
Seven standardized assessment tools were pooled in the activity domain for meta-analysis of continuous outcome data, while 6 were pooled in meta-analysis of binary outcome data (Fig. 2). Figure 3 shows the individual SMD values for outcomes within the activity domain, and indicates which outcomes were pooled to obtain an overall SMD (95% CI). Motor performance of children born VP within the activity domain was consistently poorer when compared with peers born FT (SMD = –0.71; 95% CI = –0.80 to –0.62; P <.001; I2 = 39.4%; 16 studies; estimated 2432 participants). A sensitivity analysis excluding raw data33 and unclear participant numbers64 yielded very similar results (SMD = –0.71; 95% CI = –0.80 to –0.61; P <.001; 14 studies; 2056 participants). The overall test of heterogeneity produced an I2 statistic of 32% for this sensitivity analysis of continuous activity outcome data.
Activity domain and relative risk (RR) of activity limitation for children born very preterm compared with full-term (FT) controls. * = term data obtained from online data extraction tool.22 AGTE: FM skills = Ankara Gelisim Tarama Envanteri (Ankara Developmental Screening Inventory) fine motor subscale; ASQ: GM & FM = Ages and Stages Questionnaire gross and fine motor domains; BSID-II: motor = Bayley Scales of Infant Development second edition motor subscale; FTFQ: motor = Five to Fifteen Questionnaire motor domain; HSCS-PS mobility = Health Status Classification System- Preschool Version mobility subscale; MABC-1 = The Movement Assessment Battery for Children first edition; N = number of participants; n/N = number of events/total participant number; VP = very preterm.
Activity domain and motor skills of children born very preterm compared with full-term (FT) controls. * = outcome included in meta-analysis, † = locomotor subscale, ‡ = self-care subscale (functional skills). Note: higher MABC-1 scores indicate poorer performance. BSID-II: motor = Bayley Scales of Infant Development second edition motor subscale; GDS-II: motor = Griffiths Developmental Scales second edition motor subscale; MABC-1 = The Movement Assessment Battery for Children first edition; MABC-2 = The Movement Assessment Battery for Children second edition; MSCA: motor = McCarthy Scales of Children's Abilities motor subscale; N = number of participants; PDMS = Peabody Developmental Motor Scales; PEDI = Paediatric Evaluation of Disability Inventory; SMD = standardized mean difference; VMI = Beery–Buktenica Developmental Test of Visual Motor Integration; VP = very preterm; Wee-FIM = Functional Independence Measure for Children.
Children born VP had a higher RR of activity limitation (RR = 3.39; 95% CI = 2.68 to 4.27; P <.001; I2 = 40.9%; 9 studies; 3466 participants) (Fig. 3). A sensitivity analysis excluding term data obtained from a published figure34 demonstrated little change to RR of activity limitation in children born VP (RR = 3.71; 95% CI = 2.84 to 4.85; P <.001; I2 = 40.4%; 8 studies; 3177 participants).
Outcome measures that were unable to be pooled in the activity domain meta-analysis (n = 8 studies) are presented in eAppendix 5, with SMD ranging from –0.38 to –1.64. A subgroup analysis comparing studies which corrected test scores for prematurity with studies using chronological age at assessment did not alter the activity domain results (eAppendix 6).
Participation Domain
None of the identified outcome measures were able to be linked to participation code-sets using the definitions of participation and “life situations” proposed by Coster and Khetani.27,29 One study used a fitness questionnaire to describe participant physical activity; however, this questionnaire was not reported in the results and authors of the study were uncontactable.66
Two outcome measures include individual items that were linked to participation: the Five to Fifteen (FTF) Questionnaire70 and the Infant Toddler Quality of Life (ITQoL) Questionnaire.71 The FTF Questionnaire includes 1 question (out of 17 items) regarding sports participation: “Has difficulties or does not like to participate in game sports such as soccer/football, land hockey,” whereas the ITQoL physical abilities domain includes play participation among questions of activity limitation. However, authors did not have access to participation data45 or were unable to provide individual item data by completion of analysis for our study.44 As both the FTF Questionnaire and ITQoL Questionnaire items were predominantly linked to code-sets in the activity domain, we did not consider these questionnaires as measures of participation.
Discussion
Our study found that children born VP have poorer motor outcomes at preschool age compared with children born FT. This is evident within the body structure and function and activity domains. Within the body structure and function domain, children born VP have an almost 5-fold increased risk of neurological dysfunction, poorer performance on measures of visual-motor integration, and poorer measures of coordination compared with their peers born FT. Children born VP also have increased risk of activity limitation and poorer performance on measures of motor activity than children born at term; however, there was a lack of data in the participation domain. Awareness of motor difficulties within multiple ICF-CY domains is particularly pertinent for physical therapists and other pediatric clinicians to ensure comprehensive motor assessment, treatment, and appropriate referral for preschool-aged children born VP. Physical therapists need to not only consider body structure and function impairments, and activity limitations, but also participation. Participation is considered the ultimate outcome of rehabilitation, as well as an entry point for changes at the body structure and function and activity levels.72 We need to better understand participation in children born VP to guide intervention. Clinicians should also note that we found clear evidence of poorer motor performance for children born VP within the activity domain despite inclusion of studies correcting test scores for prematurity (6 out of 14 studies and 3 out of 9 studies in the meta-analysis of continuous and binary outcome data, respectively; see Table). Our results are consistent with existing literature concerning motor skill performance in other stages of childhood;4,5,73 however, we present several important points of difference. Our study comprehensively synthesizes updated evidence of motor outcomes within the important preschool period, presents the current evidence within in each ICF-CY domain, and highlights the clear knowledge gap of participation for this population.
Characteristics of Included Studies.a
| Study | Region | Study Design | Very Preterm Birth Characteristics | Sex (% Male) | Follow-Up | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pop | L | Gestational Age (wk) | Birth Weight (g) | VP | Term | Age (y)b | No. of Participants | Outcome Measure | ICF | |||||||||
| Mean | SD | Range | Mean | SD | Range | Mean | SD | Range | VP | Term | ||||||||
| Andersson et al, 201654 | Sweden | ✓ | 29.8 | 2.5 | 1461 | 448 | 66 | 57 | 5.8c | 0.3c | 35c | 224c | PSQd | A | ||||
| Arnaud et al, 200755 | France | ✓ | ✓ | < 32 | NR | 50 | 51 | 5.1 | 0.2 | 881 | 287 | Touwen (s)e | B | |||||
| Baron et al, 201132 | USA | 26 | 1.7 | 783 | 149 | 45 | 58 | 3.7f | 0.3f | 60 | 90 | VMI, MD | A, B | |||||
| Bucci et al, 201556 | France | ✓g | 26.4 | 864 | 650–1130 | 73 | 68 | 3.9 | 22 | 22 | Postural control | B | ||||||
| Bylund et al, 199857 | Sweden | ✓ | ✓ | 31 | 25–37 | 1150 | 545–1500 | NR | NR | 4 | 0.1 | 82 | 83 | GDSd | A | |||
| Chen et al, 200433 | Taiwan | ✓ | 29.4 | 2.6 | 1162 | 228 | 45 | 53 | 3f | 0.05f | 238 | 91 | BSID-IId | A | ||||
| Davis et al, 199358 | UK | 28h | 26–32 | 1040h | 850–2090 | 54 | 40 | 3–4.5 | 13 | 20 | Gait and ROM | B | ||||||
| De Rose et al, 201323 | Italy | ✓g | 29.2h | 26–31 | 1177h | 480–2200 | 40 | NR | 3.4h | 3–3.9 | 105 | 105 | MABC-2 | A | ||||
| Erikson et al, 200334 | Sweden | ✓ | ✓g | 28h | 23–36 | 1009h | 519–1500 | 52 | NR | 5.5f | 165 | 124 | MABC-1 | A | ||||
| Esbjørn et al, 200638 | Denmark | ✓ | ✓g | 27.5 | 1.8 | 922 | 167 | 49 | 46 | 5.1f | 0.2f | 207 | 76 | MABC-1 | A | |||
| Evensen et al, 200935 | Norway | ✓ | 28.7 | 2.5 | 1187 | 210 | NRi | NRi | 5.4 | 0.3 | 25 | 73 | PDMSd | A | ||||
| Falk et al, 199768 | Israel | 27 | 3.8 | 750 | 154 | NR | NR | 6.3 | 0.5 | 6 | 15 | WAnT, vertical jump | B | |||||
| Fallang et al, 200536 | Norway | ✓ | 29 | 2.9 | 1158 | 337 | 54 | NR | 6.1 | 0.6 | 52 | 12 | MABC-1, Touwen | A, B | ||||
| Gäddlin et al, 200859 | Sweden | ✓ | ✓ | 31.2 | 2.3 | 1212 | 202 | 55 | 52 | 4 | 77 | 81 | Dev and Neuro exam | A, B | ||||
| Halsey et al, 199337 | USA | ✓g | 27.4 | 2.4 | 815 | 121 | 34 | 34 | 4 | 51 | 30 | VMI, MSCAd | A, B | |||||
| Herrgard et al, 199360 | Finland | ✓g | 29 | 2.3 | 1392 | 421 | 48 | 48 | 5.1 | 0.1 | 60 | 60 | Neuro exam | B | ||||
| Keller et al, 199865 and 200065,66 | Canada | ✓g | 27.9b | 0.39b | 1063 b | 35b | 53 | 62 | 6.5 | 0.1 | 34 | 24 | KTK, WAnT | B | ||||
| Kerstjens et al, 201139 | The Netherlands | ✓ | 29.3 | 24–31.9 | 1299 | 505–2360 | 51 | 50 | 3.6–4.1 | 503c | 535 | ASQd | A | |||||
| Kilbride et al, 200452 | USA | 26 | 1.6 | 702 | 76 | 32 | 40 | 5.1 c | 0.2c | 22c | 22c | PDMS | A | |||||
| Lee et al, 200440 | China | ✓g | 29.1 | 2.6 | 1170 | 250 | 36 | 38 | 6 | 0.6 | 42 | 69 | PDMS | A | ||||
| Liebhardt et al, 200061 | Germany | ✓g | 30.7 | 2.4 | 1367 | 324 | 45 | 49 | 3.8 | 0.1 | 40 | 83 | MD tasks | A | ||||
| Lorefice et al, 201553 | Australia | ✓g | 27.3 | 1.5 | 1022 | 268 | 50 | 53 | 4.1 f | 0.2f | 90 | 36 | Postural control | B | ||||
| Maggi et al, 201441 | Brazil | 29.5h (IQR = 4) | 1237h (IQR = 401) | 50 | 50 | 4–4 y 11 m | 62 | 62 | MABC-2, PEDI | A | ||||||||
| Marlow et al, 200762 | UK and Ireland | ✓ | ✓g | ≤25.9 | NR | 45 | 45 | 6.3h | 180 | 158 | MABC-1d | A | ||||||
| Oliveira et al, 201142 | Brazil | 30 | 2.2 | 1201 | 178 | 39 | NR | 5.9 | 0.6 | 23 | 23 | MABC-1, DCDQ | A | |||||
| Ozbek et al, 200543 | Turkey | 30.5 | 2.7 | 1276 | 183 | 53 | NR | 4.8 | 0.4 | 15 | 36 | AGTEd | A, B | |||||
| Rautava et al, 201044 | Finland | ✓ | 29.6 | 2.4 | 1249 | 382 | 57 | 59 | 5 | 588 | 176 | FTFQd | A | |||||
| Sagnol et al, 200763,j | France | ✓g | 30.3 | 27.6–31.9 | 1391 | 234 | 100 | NR | 3 y 4 m | 9 | NR | Visuo-manual aiming | A, B | |||||
| Sagnol et al, 200763,k | France | ✓g | 30 | 26.4–32.9 | 1295 | 325 | 64 | NR | 5 | 11 | 11 | Visuo-manual aiming | A, B | |||||
| Schiariti et al, 200745 | Canada | ✓ | 30h (IQR = 3) | 1370h (IQR = 746) | 45 | 50 | 3.5 | 251 | 393 | HSCS-PS,d ITQoLd | A | |||||||
| Singer and Yamashita, 199746 | USA | ✓ | ✓ | 28.2b | 2.0b | 1077b | 222b | 48 | 50 | 3f | 168 | 95 | BSID-IId | A | ||||
| Stjernqvist and Svenningsen, 199547 | Sweden | ✓ | 26.2 | 1.8 | 755 | 109 | 40 | 40 | 4f | 0.2f | 20 | 20 | GDSd | A, B | ||||
| Sullivan and Hawes, 200748 | USA | ✓ | 27.8b | 2.28b | 1036b | 319b | 40 | 52 | 4 | 0.1 | 173 | 82 | VMI, MSCA | A, B | ||||
| Sullivan and Msall, 200767 | USA | ✓ | 27.5b | 2.3b | 1040b | 287b | 55 | 58 | 4.1 | 0.1 | 112 | 43 | Wee-FIM | A | ||||
| Torrioli et al, 200064 | Italy | ✓g | 31 | 26–34 | 1120 | 560–1500 | 42 | NR | 4.9 | 0.7 | 36 | NR | MABC-1 | A | ||||
| van Hus et al, 201449 | The Netherlands | ✓g | 28.7 | 1.5 | 1079 | 264 | 49 | 40 | 5.2 f | 0.2f | 81 | 84 | MABC-2, Touwen | A, B | ||||
| Verkerk et al, 201250 | The Netherlands | ✓g | 29.8b | 2.2b | 1269b | 334b | 50 | 51 | 3.7f | 0.04f | 151 | 42 | MC, VMI | A, B | ||||
| Vohr et al, 199251 | USA | ✓ | 28.9b | 2.0b | 1137b | 315b | 33 | 32 | 5 | 0.3 | 63 | 22 | VMI | A, B | ||||
| Study | Region | Study Design | Very Preterm Birth Characteristics | Sex (% Male) | Follow-Up | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pop | L | Gestational Age (wk) | Birth Weight (g) | VP | Term | Age (y)b | No. of Participants | Outcome Measure | ICF | |||||||||
| Mean | SD | Range | Mean | SD | Range | Mean | SD | Range | VP | Term | ||||||||
| Andersson et al, 201654 | Sweden | ✓ | 29.8 | 2.5 | 1461 | 448 | 66 | 57 | 5.8c | 0.3c | 35c | 224c | PSQd | A | ||||
| Arnaud et al, 200755 | France | ✓ | ✓ | < 32 | NR | 50 | 51 | 5.1 | 0.2 | 881 | 287 | Touwen (s)e | B | |||||
| Baron et al, 201132 | USA | 26 | 1.7 | 783 | 149 | 45 | 58 | 3.7f | 0.3f | 60 | 90 | VMI, MD | A, B | |||||
| Bucci et al, 201556 | France | ✓g | 26.4 | 864 | 650–1130 | 73 | 68 | 3.9 | 22 | 22 | Postural control | B | ||||||
| Bylund et al, 199857 | Sweden | ✓ | ✓ | 31 | 25–37 | 1150 | 545–1500 | NR | NR | 4 | 0.1 | 82 | 83 | GDSd | A | |||
| Chen et al, 200433 | Taiwan | ✓ | 29.4 | 2.6 | 1162 | 228 | 45 | 53 | 3f | 0.05f | 238 | 91 | BSID-IId | A | ||||
| Davis et al, 199358 | UK | 28h | 26–32 | 1040h | 850–2090 | 54 | 40 | 3–4.5 | 13 | 20 | Gait and ROM | B | ||||||
| De Rose et al, 201323 | Italy | ✓g | 29.2h | 26–31 | 1177h | 480–2200 | 40 | NR | 3.4h | 3–3.9 | 105 | 105 | MABC-2 | A | ||||
| Erikson et al, 200334 | Sweden | ✓ | ✓g | 28h | 23–36 | 1009h | 519–1500 | 52 | NR | 5.5f | 165 | 124 | MABC-1 | A | ||||
| Esbjørn et al, 200638 | Denmark | ✓ | ✓g | 27.5 | 1.8 | 922 | 167 | 49 | 46 | 5.1f | 0.2f | 207 | 76 | MABC-1 | A | |||
| Evensen et al, 200935 | Norway | ✓ | 28.7 | 2.5 | 1187 | 210 | NRi | NRi | 5.4 | 0.3 | 25 | 73 | PDMSd | A | ||||
| Falk et al, 199768 | Israel | 27 | 3.8 | 750 | 154 | NR | NR | 6.3 | 0.5 | 6 | 15 | WAnT, vertical jump | B | |||||
| Fallang et al, 200536 | Norway | ✓ | 29 | 2.9 | 1158 | 337 | 54 | NR | 6.1 | 0.6 | 52 | 12 | MABC-1, Touwen | A, B | ||||
| Gäddlin et al, 200859 | Sweden | ✓ | ✓ | 31.2 | 2.3 | 1212 | 202 | 55 | 52 | 4 | 77 | 81 | Dev and Neuro exam | A, B | ||||
| Halsey et al, 199337 | USA | ✓g | 27.4 | 2.4 | 815 | 121 | 34 | 34 | 4 | 51 | 30 | VMI, MSCAd | A, B | |||||
| Herrgard et al, 199360 | Finland | ✓g | 29 | 2.3 | 1392 | 421 | 48 | 48 | 5.1 | 0.1 | 60 | 60 | Neuro exam | B | ||||
| Keller et al, 199865 and 200065,66 | Canada | ✓g | 27.9b | 0.39b | 1063 b | 35b | 53 | 62 | 6.5 | 0.1 | 34 | 24 | KTK, WAnT | B | ||||
| Kerstjens et al, 201139 | The Netherlands | ✓ | 29.3 | 24–31.9 | 1299 | 505–2360 | 51 | 50 | 3.6–4.1 | 503c | 535 | ASQd | A | |||||
| Kilbride et al, 200452 | USA | 26 | 1.6 | 702 | 76 | 32 | 40 | 5.1 c | 0.2c | 22c | 22c | PDMS | A | |||||
| Lee et al, 200440 | China | ✓g | 29.1 | 2.6 | 1170 | 250 | 36 | 38 | 6 | 0.6 | 42 | 69 | PDMS | A | ||||
| Liebhardt et al, 200061 | Germany | ✓g | 30.7 | 2.4 | 1367 | 324 | 45 | 49 | 3.8 | 0.1 | 40 | 83 | MD tasks | A | ||||
| Lorefice et al, 201553 | Australia | ✓g | 27.3 | 1.5 | 1022 | 268 | 50 | 53 | 4.1 f | 0.2f | 90 | 36 | Postural control | B | ||||
| Maggi et al, 201441 | Brazil | 29.5h (IQR = 4) | 1237h (IQR = 401) | 50 | 50 | 4–4 y 11 m | 62 | 62 | MABC-2, PEDI | A | ||||||||
| Marlow et al, 200762 | UK and Ireland | ✓ | ✓g | ≤25.9 | NR | 45 | 45 | 6.3h | 180 | 158 | MABC-1d | A | ||||||
| Oliveira et al, 201142 | Brazil | 30 | 2.2 | 1201 | 178 | 39 | NR | 5.9 | 0.6 | 23 | 23 | MABC-1, DCDQ | A | |||||
| Ozbek et al, 200543 | Turkey | 30.5 | 2.7 | 1276 | 183 | 53 | NR | 4.8 | 0.4 | 15 | 36 | AGTEd | A, B | |||||
| Rautava et al, 201044 | Finland | ✓ | 29.6 | 2.4 | 1249 | 382 | 57 | 59 | 5 | 588 | 176 | FTFQd | A | |||||
| Sagnol et al, 200763,j | France | ✓g | 30.3 | 27.6–31.9 | 1391 | 234 | 100 | NR | 3 y 4 m | 9 | NR | Visuo-manual aiming | A, B | |||||
| Sagnol et al, 200763,k | France | ✓g | 30 | 26.4–32.9 | 1295 | 325 | 64 | NR | 5 | 11 | 11 | Visuo-manual aiming | A, B | |||||
| Schiariti et al, 200745 | Canada | ✓ | 30h (IQR = 3) | 1370h (IQR = 746) | 45 | 50 | 3.5 | 251 | 393 | HSCS-PS,d ITQoLd | A | |||||||
| Singer and Yamashita, 199746 | USA | ✓ | ✓ | 28.2b | 2.0b | 1077b | 222b | 48 | 50 | 3f | 168 | 95 | BSID-IId | A | ||||
| Stjernqvist and Svenningsen, 199547 | Sweden | ✓ | 26.2 | 1.8 | 755 | 109 | 40 | 40 | 4f | 0.2f | 20 | 20 | GDSd | A, B | ||||
| Sullivan and Hawes, 200748 | USA | ✓ | 27.8b | 2.28b | 1036b | 319b | 40 | 52 | 4 | 0.1 | 173 | 82 | VMI, MSCA | A, B | ||||
| Sullivan and Msall, 200767 | USA | ✓ | 27.5b | 2.3b | 1040b | 287b | 55 | 58 | 4.1 | 0.1 | 112 | 43 | Wee-FIM | A | ||||
| Torrioli et al, 200064 | Italy | ✓g | 31 | 26–34 | 1120 | 560–1500 | 42 | NR | 4.9 | 0.7 | 36 | NR | MABC-1 | A | ||||
| van Hus et al, 201449 | The Netherlands | ✓g | 28.7 | 1.5 | 1079 | 264 | 49 | 40 | 5.2 f | 0.2f | 81 | 84 | MABC-2, Touwen | A, B | ||||
| Verkerk et al, 201250 | The Netherlands | ✓g | 29.8b | 2.2b | 1269b | 334b | 50 | 51 | 3.7f | 0.04f | 151 | 42 | MC, VMI | A, B | ||||
| Vohr et al, 199251 | USA | ✓ | 28.9b | 2.0b | 1137b | 315b | 33 | 32 | 5 | 0.3 | 63 | 22 | VMI | A, B | ||||
aA = activity; AGTE = Ankara-Gelisim-Tarama-Envanteri (Ankara Developmental Screening Inventory); ASQ = Ages and Stages Questionnaire; B = body structure or function; BSID-II = Bayley Scale of Infant Development; Dev = developmental; DCDQ = Developmental Coordination Disorder Questionnaire; exam = examination; FTFQ = Five to Fifteen Questionnaire; GDS = Griffiths Developmental Scales, second edition; HSCS-PS = Health Status Classification System—Preschool Version; ICF = International Classification of Functioning, Disability and Health; IQR = interquartile range; ITQoL = Infant Toddler Quality of Life Questionnaire; KTK = Koerperkoordinationstest fuer Kinder (Whole-Body Coordination Test for Children); L = longitudinal; MABC-1 = Movement Assessment Battery for Children, first edition; MABC-2 = Movement Assessment Battery for Children, second edition; MC = Developmental Test of Motor Coordination; MD = manual dexterity; MSCA = McCarthy Scales of Children's Abilities; Neuro = neurological; NR = not reported; PDMS = Peabody Developmental Scales; PEDI = Paediatric Evaluation of Disability Inventory; Pop = population based; PSQ = Performance Skills Questionnaire; ROM = range of movement; Touwen = Touwen Neurological Examination; USA = United States of America; UK = United Kingdom; VMI = Beery-Buktenica Developmental Test of Visual Motor Integration; VP = very preterm; WAnT = Wingate Anaerobic Test; Wee-FIM = Functional Independence Measure for Children.
bPooled data.
cData obtained from authors.
dMotor items or subscales.
eTouwen (s) = short version of the Touwen Neurological Examination.
fCorrected for prematurity.
gLongitudinal follow-up of VP cohort only.
hMedian.
iNot provided for total cohort.
jYounger cohort.
kOlder cohort.
Characteristics of Included Studies.a
| Study | Region | Study Design | Very Preterm Birth Characteristics | Sex (% Male) | Follow-Up | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pop | L | Gestational Age (wk) | Birth Weight (g) | VP | Term | Age (y)b | No. of Participants | Outcome Measure | ICF | |||||||||
| Mean | SD | Range | Mean | SD | Range | Mean | SD | Range | VP | Term | ||||||||
| Andersson et al, 201654 | Sweden | ✓ | 29.8 | 2.5 | 1461 | 448 | 66 | 57 | 5.8c | 0.3c | 35c | 224c | PSQd | A | ||||
| Arnaud et al, 200755 | France | ✓ | ✓ | < 32 | NR | 50 | 51 | 5.1 | 0.2 | 881 | 287 | Touwen (s)e | B | |||||
| Baron et al, 201132 | USA | 26 | 1.7 | 783 | 149 | 45 | 58 | 3.7f | 0.3f | 60 | 90 | VMI, MD | A, B | |||||
| Bucci et al, 201556 | France | ✓g | 26.4 | 864 | 650–1130 | 73 | 68 | 3.9 | 22 | 22 | Postural control | B | ||||||
| Bylund et al, 199857 | Sweden | ✓ | ✓ | 31 | 25–37 | 1150 | 545–1500 | NR | NR | 4 | 0.1 | 82 | 83 | GDSd | A | |||
| Chen et al, 200433 | Taiwan | ✓ | 29.4 | 2.6 | 1162 | 228 | 45 | 53 | 3f | 0.05f | 238 | 91 | BSID-IId | A | ||||
| Davis et al, 199358 | UK | 28h | 26–32 | 1040h | 850–2090 | 54 | 40 | 3–4.5 | 13 | 20 | Gait and ROM | B | ||||||
| De Rose et al, 201323 | Italy | ✓g | 29.2h | 26–31 | 1177h | 480–2200 | 40 | NR | 3.4h | 3–3.9 | 105 | 105 | MABC-2 | A | ||||
| Erikson et al, 200334 | Sweden | ✓ | ✓g | 28h | 23–36 | 1009h | 519–1500 | 52 | NR | 5.5f | 165 | 124 | MABC-1 | A | ||||
| Esbjørn et al, 200638 | Denmark | ✓ | ✓g | 27.5 | 1.8 | 922 | 167 | 49 | 46 | 5.1f | 0.2f | 207 | 76 | MABC-1 | A | |||
| Evensen et al, 200935 | Norway | ✓ | 28.7 | 2.5 | 1187 | 210 | NRi | NRi | 5.4 | 0.3 | 25 | 73 | PDMSd | A | ||||
| Falk et al, 199768 | Israel | 27 | 3.8 | 750 | 154 | NR | NR | 6.3 | 0.5 | 6 | 15 | WAnT, vertical jump | B | |||||
| Fallang et al, 200536 | Norway | ✓ | 29 | 2.9 | 1158 | 337 | 54 | NR | 6.1 | 0.6 | 52 | 12 | MABC-1, Touwen | A, B | ||||
| Gäddlin et al, 200859 | Sweden | ✓ | ✓ | 31.2 | 2.3 | 1212 | 202 | 55 | 52 | 4 | 77 | 81 | Dev and Neuro exam | A, B | ||||
| Halsey et al, 199337 | USA | ✓g | 27.4 | 2.4 | 815 | 121 | 34 | 34 | 4 | 51 | 30 | VMI, MSCAd | A, B | |||||
| Herrgard et al, 199360 | Finland | ✓g | 29 | 2.3 | 1392 | 421 | 48 | 48 | 5.1 | 0.1 | 60 | 60 | Neuro exam | B | ||||
| Keller et al, 199865 and 200065,66 | Canada | ✓g | 27.9b | 0.39b | 1063 b | 35b | 53 | 62 | 6.5 | 0.1 | 34 | 24 | KTK, WAnT | B | ||||
| Kerstjens et al, 201139 | The Netherlands | ✓ | 29.3 | 24–31.9 | 1299 | 505–2360 | 51 | 50 | 3.6–4.1 | 503c | 535 | ASQd | A | |||||
| Kilbride et al, 200452 | USA | 26 | 1.6 | 702 | 76 | 32 | 40 | 5.1 c | 0.2c | 22c | 22c | PDMS | A | |||||
| Lee et al, 200440 | China | ✓g | 29.1 | 2.6 | 1170 | 250 | 36 | 38 | 6 | 0.6 | 42 | 69 | PDMS | A | ||||
| Liebhardt et al, 200061 | Germany | ✓g | 30.7 | 2.4 | 1367 | 324 | 45 | 49 | 3.8 | 0.1 | 40 | 83 | MD tasks | A | ||||
| Lorefice et al, 201553 | Australia | ✓g | 27.3 | 1.5 | 1022 | 268 | 50 | 53 | 4.1 f | 0.2f | 90 | 36 | Postural control | B | ||||
| Maggi et al, 201441 | Brazil | 29.5h (IQR = 4) | 1237h (IQR = 401) | 50 | 50 | 4–4 y 11 m | 62 | 62 | MABC-2, PEDI | A | ||||||||
| Marlow et al, 200762 | UK and Ireland | ✓ | ✓g | ≤25.9 | NR | 45 | 45 | 6.3h | 180 | 158 | MABC-1d | A | ||||||
| Oliveira et al, 201142 | Brazil | 30 | 2.2 | 1201 | 178 | 39 | NR | 5.9 | 0.6 | 23 | 23 | MABC-1, DCDQ | A | |||||
| Ozbek et al, 200543 | Turkey | 30.5 | 2.7 | 1276 | 183 | 53 | NR | 4.8 | 0.4 | 15 | 36 | AGTEd | A, B | |||||
| Rautava et al, 201044 | Finland | ✓ | 29.6 | 2.4 | 1249 | 382 | 57 | 59 | 5 | 588 | 176 | FTFQd | A | |||||
| Sagnol et al, 200763,j | France | ✓g | 30.3 | 27.6–31.9 | 1391 | 234 | 100 | NR | 3 y 4 m | 9 | NR | Visuo-manual aiming | A, B | |||||
| Sagnol et al, 200763,k | France | ✓g | 30 | 26.4–32.9 | 1295 | 325 | 64 | NR | 5 | 11 | 11 | Visuo-manual aiming | A, B | |||||
| Schiariti et al, 200745 | Canada | ✓ | 30h (IQR = 3) | 1370h (IQR = 746) | 45 | 50 | 3.5 | 251 | 393 | HSCS-PS,d ITQoLd | A | |||||||
| Singer and Yamashita, 199746 | USA | ✓ | ✓ | 28.2b | 2.0b | 1077b | 222b | 48 | 50 | 3f | 168 | 95 | BSID-IId | A | ||||
| Stjernqvist and Svenningsen, 199547 | Sweden | ✓ | 26.2 | 1.8 | 755 | 109 | 40 | 40 | 4f | 0.2f | 20 | 20 | GDSd | A, B | ||||
| Sullivan and Hawes, 200748 | USA | ✓ | 27.8b | 2.28b | 1036b | 319b | 40 | 52 | 4 | 0.1 | 173 | 82 | VMI, MSCA | A, B | ||||
| Sullivan and Msall, 200767 | USA | ✓ | 27.5b | 2.3b | 1040b | 287b | 55 | 58 | 4.1 | 0.1 | 112 | 43 | Wee-FIM | A | ||||
| Torrioli et al, 200064 | Italy | ✓g | 31 | 26–34 | 1120 | 560–1500 | 42 | NR | 4.9 | 0.7 | 36 | NR | MABC-1 | A | ||||
| van Hus et al, 201449 | The Netherlands | ✓g | 28.7 | 1.5 | 1079 | 264 | 49 | 40 | 5.2 f | 0.2f | 81 | 84 | MABC-2, Touwen | A, B | ||||
| Verkerk et al, 201250 | The Netherlands | ✓g | 29.8b | 2.2b | 1269b | 334b | 50 | 51 | 3.7f | 0.04f | 151 | 42 | MC, VMI | A, B | ||||
| Vohr et al, 199251 | USA | ✓ | 28.9b | 2.0b | 1137b | 315b | 33 | 32 | 5 | 0.3 | 63 | 22 | VMI | A, B | ||||
| Study | Region | Study Design | Very Preterm Birth Characteristics | Sex (% Male) | Follow-Up | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Pop | L | Gestational Age (wk) | Birth Weight (g) | VP | Term | Age (y)b | No. of Participants | Outcome Measure | ICF | |||||||||
| Mean | SD | Range | Mean | SD | Range | Mean | SD | Range | VP | Term | ||||||||
| Andersson et al, 201654 | Sweden | ✓ | 29.8 | 2.5 | 1461 | 448 | 66 | 57 | 5.8c | 0.3c | 35c | 224c | PSQd | A | ||||
| Arnaud et al, 200755 | France | ✓ | ✓ | < 32 | NR | 50 | 51 | 5.1 | 0.2 | 881 | 287 | Touwen (s)e | B | |||||
| Baron et al, 201132 | USA | 26 | 1.7 | 783 | 149 | 45 | 58 | 3.7f | 0.3f | 60 | 90 | VMI, MD | A, B | |||||
| Bucci et al, 201556 | France | ✓g | 26.4 | 864 | 650–1130 | 73 | 68 | 3.9 | 22 | 22 | Postural control | B | ||||||
| Bylund et al, 199857 | Sweden | ✓ | ✓ | 31 | 25–37 | 1150 | 545–1500 | NR | NR | 4 | 0.1 | 82 | 83 | GDSd | A | |||
| Chen et al, 200433 | Taiwan | ✓ | 29.4 | 2.6 | 1162 | 228 | 45 | 53 | 3f | 0.05f | 238 | 91 | BSID-IId | A | ||||
| Davis et al, 199358 | UK | 28h | 26–32 | 1040h | 850–2090 | 54 | 40 | 3–4.5 | 13 | 20 | Gait and ROM | B | ||||||
| De Rose et al, 201323 | Italy | ✓g | 29.2h | 26–31 | 1177h | 480–2200 | 40 | NR | 3.4h | 3–3.9 | 105 | 105 | MABC-2 | A | ||||
| Erikson et al, 200334 | Sweden | ✓ | ✓g | 28h | 23–36 | 1009h | 519–1500 | 52 | NR | 5.5f | 165 | 124 | MABC-1 | A | ||||
| Esbjørn et al, 200638 | Denmark | ✓ | ✓g | 27.5 | 1.8 | 922 | 167 | 49 | 46 | 5.1f | 0.2f | 207 | 76 | MABC-1 | A | |||
| Evensen et al, 200935 | Norway | ✓ | 28.7 | 2.5 | 1187 | 210 | NRi | NRi | 5.4 | 0.3 | 25 | 73 | PDMSd | A | ||||
| Falk et al, 199768 | Israel | 27 | 3.8 | 750 | 154 | NR | NR | 6.3 | 0.5 | 6 | 15 | WAnT, vertical jump | B | |||||
| Fallang et al, 200536 | Norway | ✓ | 29 | 2.9 | 1158 | 337 | 54 | NR | 6.1 | 0.6 | 52 | 12 | MABC-1, Touwen | A, B | ||||
| Gäddlin et al, 200859 | Sweden | ✓ | ✓ | 31.2 | 2.3 | 1212 | 202 | 55 | 52 | 4 | 77 | 81 | Dev and Neuro exam | A, B | ||||
| Halsey et al, 199337 | USA | ✓g | 27.4 | 2.4 | 815 | 121 | 34 | 34 | 4 | 51 | 30 | VMI, MSCAd | A, B | |||||
| Herrgard et al, 199360 | Finland | ✓g | 29 | 2.3 | 1392 | 421 | 48 | 48 | 5.1 | 0.1 | 60 | 60 | Neuro exam | B | ||||
| Keller et al, 199865 and 200065,66 | Canada | ✓g | 27.9b | 0.39b | 1063 b | 35b | 53 | 62 | 6.5 | 0.1 | 34 | 24 | KTK, WAnT | B | ||||
| Kerstjens et al, 201139 | The Netherlands | ✓ | 29.3 | 24–31.9 | 1299 | 505–2360 | 51 | 50 | 3.6–4.1 | 503c | 535 | ASQd | A | |||||
| Kilbride et al, 200452 | USA | 26 | 1.6 | 702 | 76 | 32 | 40 | 5.1 c | 0.2c | 22c | 22c | PDMS | A | |||||
| Lee et al, 200440 | China | ✓g | 29.1 | 2.6 | 1170 | 250 | 36 | 38 | 6 | 0.6 | 42 | 69 | PDMS | A | ||||
| Liebhardt et al, 200061 | Germany | ✓g | 30.7 | 2.4 | 1367 | 324 | 45 | 49 | 3.8 | 0.1 | 40 | 83 | MD tasks | A | ||||
| Lorefice et al, 201553 | Australia | ✓g | 27.3 | 1.5 | 1022 | 268 | 50 | 53 | 4.1 f | 0.2f | 90 | 36 | Postural control | B | ||||
| Maggi et al, 201441 | Brazil | 29.5h (IQR = 4) | 1237h (IQR = 401) | 50 | 50 | 4–4 y 11 m | 62 | 62 | MABC-2, PEDI | A | ||||||||
| Marlow et al, 200762 | UK and Ireland | ✓ | ✓g | ≤25.9 | NR | 45 | 45 | 6.3h | 180 | 158 | MABC-1d | A | ||||||
| Oliveira et al, 201142 | Brazil | 30 | 2.2 | 1201 | 178 | 39 | NR | 5.9 | 0.6 | 23 | 23 | MABC-1, DCDQ | A | |||||
| Ozbek et al, 200543 | Turkey | 30.5 | 2.7 | 1276 | 183 | 53 | NR | 4.8 | 0.4 | 15 | 36 | AGTEd | A, B | |||||
| Rautava et al, 201044 | Finland | ✓ | 29.6 | 2.4 | 1249 | 382 | 57 | 59 | 5 | 588 | 176 | FTFQd | A | |||||
| Sagnol et al, 200763,j | France | ✓g | 30.3 | 27.6–31.9 | 1391 | 234 | 100 | NR | 3 y 4 m | 9 | NR | Visuo-manual aiming | A, B | |||||
| Sagnol et al, 200763,k | France | ✓g | 30 | 26.4–32.9 | 1295 | 325 | 64 | NR | 5 | 11 | 11 | Visuo-manual aiming | A, B | |||||
| Schiariti et al, 200745 | Canada | ✓ | 30h (IQR = 3) | 1370h (IQR = 746) | 45 | 50 | 3.5 | 251 | 393 | HSCS-PS,d ITQoLd | A | |||||||
| Singer and Yamashita, 199746 | USA | ✓ | ✓ | 28.2b | 2.0b | 1077b | 222b | 48 | 50 | 3f | 168 | 95 | BSID-IId | A | ||||
| Stjernqvist and Svenningsen, 199547 | Sweden | ✓ | 26.2 | 1.8 | 755 | 109 | 40 | 40 | 4f | 0.2f | 20 | 20 | GDSd | A, B | ||||
| Sullivan and Hawes, 200748 | USA | ✓ | 27.8b | 2.28b | 1036b | 319b | 40 | 52 | 4 | 0.1 | 173 | 82 | VMI, MSCA | A, B | ||||
| Sullivan and Msall, 200767 | USA | ✓ | 27.5b | 2.3b | 1040b | 287b | 55 | 58 | 4.1 | 0.1 | 112 | 43 | Wee-FIM | A | ||||
| Torrioli et al, 200064 | Italy | ✓g | 31 | 26–34 | 1120 | 560–1500 | 42 | NR | 4.9 | 0.7 | 36 | NR | MABC-1 | A | ||||
| van Hus et al, 201449 | The Netherlands | ✓g | 28.7 | 1.5 | 1079 | 264 | 49 | 40 | 5.2 f | 0.2f | 81 | 84 | MABC-2, Touwen | A, B | ||||
| Verkerk et al, 201250 | The Netherlands | ✓g | 29.8b | 2.2b | 1269b | 334b | 50 | 51 | 3.7f | 0.04f | 151 | 42 | MC, VMI | A, B | ||||
| Vohr et al, 199251 | USA | ✓ | 28.9b | 2.0b | 1137b | 315b | 33 | 32 | 5 | 0.3 | 63 | 22 | VMI | A, B | ||||
aA = activity; AGTE = Ankara-Gelisim-Tarama-Envanteri (Ankara Developmental Screening Inventory); ASQ = Ages and Stages Questionnaire; B = body structure or function; BSID-II = Bayley Scale of Infant Development; Dev = developmental; DCDQ = Developmental Coordination Disorder Questionnaire; exam = examination; FTFQ = Five to Fifteen Questionnaire; GDS = Griffiths Developmental Scales, second edition; HSCS-PS = Health Status Classification System—Preschool Version; ICF = International Classification of Functioning, Disability and Health; IQR = interquartile range; ITQoL = Infant Toddler Quality of Life Questionnaire; KTK = Koerperkoordinationstest fuer Kinder (Whole-Body Coordination Test for Children); L = longitudinal; MABC-1 = Movement Assessment Battery for Children, first edition; MABC-2 = Movement Assessment Battery for Children, second edition; MC = Developmental Test of Motor Coordination; MD = manual dexterity; MSCA = McCarthy Scales of Children's Abilities; Neuro = neurological; NR = not reported; PDMS = Peabody Developmental Scales; PEDI = Paediatric Evaluation of Disability Inventory; Pop = population based; PSQ = Performance Skills Questionnaire; ROM = range of movement; Touwen = Touwen Neurological Examination; USA = United States of America; UK = United Kingdom; VMI = Beery-Buktenica Developmental Test of Visual Motor Integration; VP = very preterm; WAnT = Wingate Anaerobic Test; Wee-FIM = Functional Independence Measure for Children.
bPooled data.
cData obtained from authors.
dMotor items or subscales.
eTouwen (s) = short version of the Touwen Neurological Examination.
fCorrected for prematurity.
gLongitudinal follow-up of VP cohort only.
hMedian.
iNot provided for total cohort.
jYounger cohort.
kOlder cohort.
Our results add to the current body of evidence by using the ICF-CY to describe outcomes across the domains of body structure and function, activity, and participation. The ICF is a dynamic and reciprocal model, which reflects the nature of individual functioning and child development.10 By categorizing outcomes into ICF domains, we present a holistic and comprehensive overview of the motor function of children born VP at preschool age. The ICF allows us to highlight the most researched domains and the wide variety of assessment tools used within these domains. A consideration for researchers using the ICF to categorize outcomes is that many assessment tools were created prior to wider adoption of the ICF and the structured approach to holistically describing the impact of disability on individual functioning and on life experiences. Although the inevitable overlap of outcomes between domains prevents distinct classification, it illustrates the dynamic process of child development and function within environmental contexts, and highlights evidence gaps.
A strength of our study is the focus on preschool age. Understanding the motor outcomes at preschool age of children born VP is important, as it is a time of rapid sensory and musculoskeletal development characterized by the increasing complexity of motor tasks.53 Children develop major motor competencies19 and physical activity behaviors at preschool age.74 By presenting motor outcomes within the body structure and function and activity domains for children born VP, our study highlights multiple entry points for intervention, particularly within the health and education sectors.
Our review provides a comprehensive overview of motor outcomes in children who were born VP and/or with very low birth weight, adding to the previous systematic review and meta-analysis in this population, which used only 3 standardized assessments.4 In our review, we include 12 additional studies that were published from 2009 onward, highlighting the important role our work has in updating the evidence concerning motor outcomes for children born VP. Due to different inclusion criteria, we present results from a diverse group of studies compared with the 2009 review. Moreover, rather than including studies with a group of participants born FT, de Kieviet and colleagues stipulated inclusion of studies using normative data for 2 out of 3 assessment tools.4 Using test norms rather than a term comparison group may underestimate developmental delays; for example, the Bayley-III underestimates developmental delay in children who were extremely preterm and/or had extremely low birth weight.75 In addition to providing updated evidence of the preschool-age motor skill proficiency of children born VP compared with children born FT (activity domain), we also include body structure and function outcomes, such as neurological dysfunction, and highlight important gaps in our knowledge of participation in this population.
Despite evidence of body structure and function impairments, and activity limitations for children born VP, we did not identify any outcomes fully measuring participation. Activities and participation are grouped together in the ICF-CY,10 and the framework does not provide a clear method of distinguishing between the 2 constructs. This issue is compounded by ongoing debate regarding the definition of participation and “life situations.”27 A recent systematic review investigating definitions of participation, and language used within this construct, highlights the disconnect between language and outcome measures used in research.28 We chose to use the definition of participation proposed by Coster and Khetani,29 which is considered in other studies of participation.27,76–78 However, we acknowledge that participation is an evolving construct, and that using alternative approaches may result in different outcome linking within the participation domain. The unclear distinction between activities and participation in the ICF-CY is a limitation for research using this framework.
A limitation of our study is that the large variation in outcome measures impeded comprehensive meta-analysis, especially in the body structure and function domain. Excluding outcomes that were also linked to the activity domain, only 3 studies used a consistent outcome, the Touwen Neurological Examination, and had sufficient data to be pooled.36,49,55 Despite outcome measure consistency, the heterogeneity of these 3 studies was high once pooled and required random effects meta-analysis (eAppendix 3). The pooled results of other outcomes within the body structure and function domain (coordination and MSCA) should be interpreted with caution, due to the wide age range of participants, limited number of studies (n = 2 for each outcome), limited number of participants, and the statistical heterogeneity. Inaccessible data68 and differing methodologies53,56 also precluded pooled meta-analysis of exercise and postural control outcomes. More consistency can be seen within the activity domain, with multiple studies using the MABC-1,34,36,38,42 the Movement Assessment Battery for Children second edition (MABC-2),23,41,49 and the Peabody Developmental Scales (PDMS);35,40,52 however, the range of outcomes assessing activity is still large. A total of 19 different outcomes were identified within the activity domain, and several studies also used individual items of assessment tools,57,61,62 rather than administering the entire standardized assessment. When comparing between studies, the lack of consistent assessment tools used in follow-up at preschool age is a considerable limitation. Future research in VP populations should ensure consistent use of standardized motor assessment tools to facilitate comparison of outcomes between eras and geographical cohorts.
The methodological quality and study design of included studies should also be considered. Studies differed in respect to correcting test scores for prematurity, with 9 out of 36 studies using corrected age at assessment. This methodological variation is an important limitation to consider. However, a subgroup analysis comparing studies using corrected versus chronological age did not alter our overall conclusion regarding motor performance (eAppendix 6). Cohort size, follow-up, and outcome measurement are noticeably disparate for included studies. Sample size at follow-up varied greatly (range 2168 to 116855), and the comparability of VP and term cohorts was inconsistent. Independent blind assessment was poorly reported across studies, while follow-up rates were also low (eAppendix 2). A limitation of the NOS for cohort studies is that study designs including longitudinal follow-up of VP participants with later recruitment of term controls (n = 16)23,34,37,38,40,49,50,53,56,60–66 generally obtained a higher follow-up star rating than purely longitudinal designs. Another important trend to note is that multiple studies excluded participants based on unfavorable outcomes, and exclusion of these data may over-estimate the motor performance of all children born VP compared with controls born FT.
We found that preschool-aged children born VP have poorer motor outcomes compared with children born FT within the ICF-CY domains of body structure and function and activity. However, participation of children born VP compared with their term-born peers at preschool age is not well described. Evidence of poorer motor outcomes for older children born VP compared with peers born FT is well documented; however, we comprehensively highlight outcomes within the important preschool period and present the current evidence within in each ICF-CY domain.
Research and Clinical Implications
Further research is needed to investigate participation in VP populations at preschool age, and to decrease the gap between theoretical frameworks and participation assessment tools. Understanding participation in this population, and whether known body structure and function impairments and activity limitations ultimately influence participation, is needed to guide targeted intervention studies in the future. Researchers should continue to use the ICF-CY framework to examine outcomes in VP populations, due to the myriad comorbidities associated with prematurity; however, more work is needed to distinguish between activity and participation domains. Greater outcome measure consistency in research of VP populations at preschool age is required across ICF-CY domains. Clinically, physical therapists should consider all ICF domains when assessing and treating preschool-aged children born VP, and advocate for developmental surveillance into preschool age.
Author Contributions and Acknowledgments
Concept/idea/research design: T.L. FitzGerald, J.L.Y Cheong, J.L. McGinley, L.W. Doyle, A.J. Spittle
Writing: T.L. FitzGerald, J.L.Y Cheong, L.W. Doyle, A.J. Spittle, J.L. McGinley
Data collection: T.L. FitzGerald, A.K.L Kwong, A.J. Spittle
Data analysis: T.L. FitzGerald, J.L.Y Cheong, A.J. Spittle, L.W. Doyle
Fund procurement: J.L.Y Cheong, L.W. Doyle, A.J. Spittle
Providing facilities/equipment: A.J. Spittle
Providing institutional liaisons: A.J. Spittle
Consultation (including review of manuscript before submitting): A.K.L. Kwong, J.L.Y. Cheong, J.L. McGinley, A.J. Spittle, L.W. Doyle, T.L. FitzGerald.
T.L. FitzGerald conceptualized the research question, completed database searching, screening, data extraction, and quality assessment; drafted the manuscript; and approved the final manuscript as submitted. A.K.L. Kwong independently completed database screening and quality assessment, revised the manuscript, and approved the final manuscript as submitted. J.L.Y. Cheong, J.L. McGinley, and L.W. Doyle conceptualized the research question, revised the manuscript, and approved the final manuscript as submitted. A.J. Spittle conceptualized the research question; resolved conflicts during database screening, data extraction, and quality assessment; revised the manuscript; and approved the final manuscript as submitted.
Funding
This work is supported by grants from the National Health and Medical Research Council of Australia (Centre of Research Excellence, ref. no. 1,060,733; Early Career Fellowship, ref. no. 1,053,787 to J.LY. Cheong; Career Development Fellowship, ref. no. 1,108,714 to A.J. Spittle) and from the Victorian Government's Operational Infrastructure Support Program. T.L. FitzGerald's PhD candidature is supported by The Australian Government Research Training Program Scholarship and the Centre of Research Excellence in Newborn Medicine. A.K.L. Kwong's PhD candidature is supported by The Australian Government Research Training Program Scholarship, the Centre of Research Excellence in Newborn Medicine, and the National Health and Medical Research Council of Australia.
Disclosures and Presentations
The authors completed the ICJME Form for Disclosure of Potential Conflicts of Interest. They reported no conflicts of interest.
This manuscript was adapted in part from the following: an oral presentation given at the 12th International Conference on Developmental Coordination Disorder, Fremantle, Western Australia, July 7, 2017; a scientific poster presentation at the 71st Annual Meeting of the American Academy for Cerebral Palsy and Developmental Medicine (AACPDM), Montreal, Quebec, Canada, September 2017; and an oral presentation given at the 9th Biennial AusACPDM conference, Auckland, New Zealand, March 22, 2018.
References
Comments