Early developmental trajectories of preterm infants

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Research in Developmental Disabilities journal homepage: www.elsevier.com/locate/redevdis

Early developmental trajectories of preterm infants Maya Yaaria, David Mankutab, Ayelet Harel- Gadassia, Edwa Friedlandera, Benjamin Bar-Ozc, Smadar Eventov-Friedmanc, Nimrod Manivd, David Zuckerd, ⁎ Nurit Yirmiyaa, a

Department of Psychology, The Hebrew University of Jerusalem, Israel Department of Obstetrics and Gynecology, Hadassah Hebrew University Hospital, Israel Department of Neonatology, Hadassah Hebrew University Hospital, Israel d Department of Statistics, The Hebrew University of Jerusalem, Israel b c

AR TI CLE I NF O

AB S T R A CT

Number of reviews are 2

Background and objectives: Preterm infants are at risk for neuro-developmental impairments and atypical developmental trajectories. The aims of this study were to delineate early developmental trajectories of preterm and full-term infants. Methods: The cognitive, language, and motor development of 149 infants – 19 extremely preterm (EPT), 34 very preterm (VPT), 57 moderately preterm (MPT), and 39 full-term (FT) – was evaluated using Mullen Scales at 1, 4, 8, 12, and 18 months. Mixed models were applied to examine group differences. Gender, maternal education, and neurobehavior were included as predictors of developmental trajectories. Results: The EPT and VPT infants achieved significantly lower scores than the FT infants in all domains, with a significantly increasing gap over time. The MPT infants' trajectories were more favorable than those of the EPT and VPT infants yet lower than the FT infants on the Visual Reception, Gross, and Fine Motor subscales. Male gender and lower maternal education were associated with lower scores that declined over time. Abnormal neonatal neurobehavior was associated lower Mullen scores and with less stability in scores over time. Conclusions: The EPT and VPT infants were found to have disadvantages across all domains. The MPT infants revealed more favorable developmental trajectories yet displayed vulnerability compared to the FT infants. Gender, maternal education, and neonatal neurobehavior are important in predicting the developmental outcomes of preterm infants.

Keywords: Preterm infants Developmental trajectories Cognitive development Language development Motor development Neonatal neurobehavior

1. Introduction Preterm birth, which is defined as a birth that takes places before the completion of 37 weeks of pregnancy, occurs in 12–13% of live births in the USA and in 5–9% of live births in other developed countries. It is the leading cause of perinatal morbidity and mortality in developed countries (Beck et al., 2010; Goldenberg, Culhane, Iams, & Romero, 2008). The biological vulnerability conferred by preterm birth, which is potentially amplified by environmental disadvantage, profoundly affects development and has consequences that extend across the lifespan (Aylward, 2005; Johnson & Marlow, 2014; Moster, Lie, & Markestad, 2008). As the longlasting effects of preterm birth on health, well-being, school performance, and participation are significant and costly, it is important to study the precursors of these difficulties and thus facilitate their early identification and intervention.

⁎

Corresponding author at: Department of Psychology, The Hebrew University of Jerusalem, Mount Scopus, Jerusalem, Israel. E-mail address: [email protected] (N. Yirmiya).

http://dx.doi.org/10.1016/j.ridd.2017.10.018 Received 17 April 2017; Received in revised form 17 October 2017; Accepted 17 October 2017 0891-4222/ © 2017 Elsevier Ltd. All rights reserved.

Please cite this article as: Yaari, M., Research in Developmental Disabilities (2017), http://dx.doi.org/10.1016/j.ridd.2017.10.018

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

In their comprehensive review of the developmental pathways that follow preterm birth, Sansavini, Guarini, and Caselli, 2011 describe the etiology underlying the adverse developmental outcomes of preterm infants as having a “complex interaction between biological and socio-environmental constraints associated to preterm birth which occurs in a critical period of rapid development of the neural system and thus leads to an atypical development” (p 102). Typical fetal development occurs under specific sensory conditions that facilitate neurological development and maturation. Premature birth interrupts these physiological processes, with the neonatal intensive care unit (NICU) environment imposing artificial conditions on the developing premature infant. In the NICU, a premature newborn is both under- and over-stimulated and may also experience elevated stress due to having to undergo an average of 10 painful procedures per day. These sub-optimal conditions during a critical period of brain development may result in atypical developmental trajectories (Als et al., 2004; Carbajal et al., 2008; Sansavini et al., 2011). In severe cases, preterm birth is associated with neurological injury (e.g., periventricular leukomalacia, intraventricular hemorrhage, and hydrocephalus) that results in neuro-sensory, motor, and severe cognitive disabilities (Volpe, 2009). However, even in the absence of identifiable severe neurological damage, evidence of atypical brain development and alterations to brain structure and organization exists (Aylward, 2005). As a result, even though preterm children do not frequently exhibit severe damages, many children with a wide range of gestational ages can present with low-severity impairments (Aylward, 2005; Sansavini et al., 2011). In their review, Sansavini, et al. (2011) further describe changes in the literature exploring prematurity outcomes; these changes include a shift from studying high-severity, low-prevalence neuro-motor and sensory disabilities to exploring low-severity, highprevalence cognitive deficiencies. Cognitive abilities, including measures such as developmental and IQ tests, are now common broad measures of prematurity outcomes (Aarnoudse-Moens, Weisglas-Kuperus, Van Goudoever, & Oosterlaan, 2009; Arpino et al., 2010; Saigal & Doyle, 2008). A meta-analysis that examines IQ differences between preterm and full-term children reported a considerable mean difference of 10 IQ points between preterm and full-term children (Bhutta, Cleves, Casey, Cradock, & Anand, 2002). Disadvantages were most pronounced in children born extremely preterm (EPT), but they were also evident in those born very preterm (VPT), moderately preterm (MPT), and late preterm (LPT) (Chyi, Lee, Hintz, Gould, & Sutcliffe, 2008; Johnson et al., 2015; Larroque et al., 2008). Researchers have also shifted their attention from broad outcomes such as general intelligence to more specific domains to further characterize the vulnerabilities and difficulties of preterm children (Sansavini et al., 2011). The application of various neonatal measures has led to the documentation of alterations in neurobehavior among preterm infants, including abnormal muscle tone and movements, asymmetry, and difficulties with attention and arousal regulation (Brown, Doyle, Bear, & Inder, 2006). The motor domain has also been identified as an area of vulnerability among preterm infants. Motor development is affected by both biological (e.g., neurological injury and brain maturation) and environmental (e.g., postural limitations) constraints in the neonatal period. The prevalence of cerebral palsy is significantly higher among preterm infants (Platt et al., 2007). Moreover, differences between preterm and full-term children have been reported in relation to gross motor skills, fine motor skills, and visual motor integration. The results of a meta-analysis that focuses on preterm infants’ motor outcomes indicated delays in the attainment of early motor milestones during the first years of life. Although these children reached these milestones by their second and third years of life, difficulties related to more advanced motor skills persisted throughout their childhood and adolescence (De Kieviet, Piek, AarnoudseMoens, & Oosterlaan, 2009). The language domain has also been identified as being vulnerable to long-term consequences of preterm birth. As in the motor domain, differences between preterm and full-term infants have been observed regarding various pre-language and early linguistic skills as early as in the first year of life (De Schuymer, De Groote, Striano, Stahl, & Roeyers, 2011). Although some evidence of catchup gains over time exists in basic language skills such as receptive language (Luu, Vohr, Allan, Schneider, & Ment, 2011), many researchers have also reported growing differences in complex language functions (Sansavini, Guarini, Savini et al., 2011; Van Noortvan der Spek, Franken, & Weisglas-Kuperus, 2012; Woodward et al., 2009). Longitudinal studies that follow preterm infants’ development at several timepoints early in life are important for understanding individual trajectories and enhancing early identification and intervention (Blauw-Hospers & Hadders-Algra, 2005; Nordhov et al., 2010). Long-term follow-up studies typically include EPT and VPT infant cohorts, whereas many studies that include MPT infants are cross-sectional. In the present study, we therefore applied the Mullen Scales of Early Learning with the aim of examining cognitive, language, and motor skills throughout the first 18 months of life in all four cohorts, namely EPT, VPT, MPT, and full-term (FT) infants. Data were collected at five timepoints throughout this period. We hypothesized that an association between gestational age will be revealed, where the EPT infants will be the most vulnerable, with the lowest MSEL scores, and the MPT infants will have more favorable outcomes compared to the EPT and VPT infants, yet lower that the FT infants. The development of preterm infants not only tends to be less favorable than that of FT infants; it is also characterized by greater variance, which suggests that the risk associated with preterm birth is not homogenous and additional conditions may enhance or attenuate initial risk. Male gender and lower socioeconomic status (SES) have been identified as risk factors associated with less favorable developmental outcomes among preterm infants (Janssen et al., 2011; Sansavini et al., 2011). Neonatal neurobehavior has also been examined as a potential predictor of developmental outcomes, but with inconsistent results (Harijan et al., 2012; Picciolini et al., 2016; Stephens et al., 2010). The second aim of this study is therefore to examine demographic characteristics and neonatal neurobehavior as potential predictors of preterm infants’ developmental trajectories beyond gestational age (GA). We hypothesized that male gender, lower maternal education and more neonatal neurobehavior abnormalities will be associated with lower MSEL scores over time.

2

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

Table 1 Demographic and medical characteristics of the sample. Characteristics

EPT (n = 19)

VPT (n = 34)

MPT (n = 57)

FT (n = 37)

Gender: Female, n (%) Male, n (%) Maternal age, years mean (SD), range Paternal age, years mean (SD), range Number of siblings, mean (SD), range

9 (47%) 10 (53%) 32 (7), 19–41 34 (8), 24–50 1 (2), 0–6

21 (62%) 13 (38%) 30 (5), 21–38 32 (6), 21–45 2 (2), 0–7

33 (58%) 24 (42%) 32 (6), 22–51 34 (5), 23–46 1.5 (2), 0–7

22 (59%) 15 (41%) 33 (4), 27–41 35 (6), 4–54 2 (2), 0–5

Maternal education: Non-academic n (%) Academic, n (%)

7 (37%) 12 (63%)

17 (50%) 17 (50%)

16 (28%) 41 (72%)

8 (22%) 29 (78%)

Income: Below median, n (%) Median range, n (%) Above median, n (%)

6 (32%) 10 (53%) 3 (15%)

9 (26%) 22 (65%) 3 (9%)

8 (14%) 39 (68%) 10 (18%)

3 (8%) 28 (72%) 8 (20%)

Medical GA at birth, mean (SD), range, weeks*** Birth-weight mean (SD), range, gr*** Apgar at 1 min, mean (SD), range*** Apgar at 5 min, mean (SD), range*** Number of days in the NICU, mean (SD), range*** Number of days of oxygen support, mean (SD), range*** SGAa n (%)***

26.5 (1.0), 24.2–28 888 (238) 490–1250 5.8 (2), 1–9 7.8 (2), 2–10 103 (44) 55–205 69 (56) 0–205 4 (21%)

30.3 (1.0) 28.1–32 1410 (338) 850–2340 7.5 (2), 2–9 8.8 (1), 7–10 46 (20) 17–111 12 (16) 0–73 4 (12%)

33.2 (0.6) 32.1–34 1865 (320) 910–2400 8.4 (1), 4–9 9.4 (1), 5–10 24 (12) 9–80 3 (7) 0–34 11 (19%)

39.8 (1.0) 37.7–41.2 3373 (346) 2600–4258 8.9 (0.3), 7–9 9.9 (.3), 9–10

BPD: None Mild Moderate Severe*** Sepsis n (%)***

4 (22%) 6 (34%) 4 (22%) 4 (22%) 10 (53%)

30 (88%) 3 (9%) 1 (3%) 0 4 (12%)

55 (96%) 2 (4%) 0 0 2 (4%)

–

ROP: No Level 1–3*** Level 4+

6 (32%) 8 (43%) 5 (27%)

27 (79%) 7 (21%) 0

57 (100%) 0 0

–

IVH: No Level I-II Level IV PVL n (%)

15 (80%) 2 (10%) 2 (10%) 0

29 (85%) 3 (9%) 2 (6%) 3 (9%)

56 (98%) 1 (2%) 0 2 (4%)

–

– –

–

–

EPT- extremely preterm, VPT – very preterm, MPT − moderately preterm; FT- full-term; NICU – neonatal intensive care unit; SGA- small for gestational age; BPD – bronchopulmonary dysplasia; ROP – retinopathy of prematurity; IVH – intra ventricular hemorrhage; PVL – periventricular leukomalacia. a SGA- defined as < 10th percentile of birthweight for gestational age. *** Significant group-differences, p < 0.001.

2. Methods 2.1. Participants Inclusion criteria were singleton infants with no genetic or congenital anomaly and with Hebrew speaking parents. Preterm infants, born at ≤34 weeks' gestation with birthweight ≤2500 g, were recruited in the NICU. Full-term infants, born at 37–41 weeks' gestation with birthweight > 2500 g following normal labor and delivery were recruited at hospital nurseries. Preterm infants were divided into three GA-based groups for the purpose of the current study: EPT infants, born at GA ≤ 28 weeks, VPT infants, born at GA 29–32 weeks and MPT infants, born at GA 33–34 weeks. A total of 149 infants who were born between the years 2009–2013 at Hadassah Medical Center were enrolled in the study, including 19 EPT, 34 VPT, 57 MPT, and 39 FT infants. The final sample included 141 infants who were examined at least four times. Three infants who were eventually diagnosed with severe sensory-motor impairments (i.e., cortical blindness and severe Cerebral Palsy which prevented valid administrations of the MSEL) were excluded (1 VPT, 2 MPT infants) and five infants (1 EPT, 1 VPT, 1 MPT and 2 FT) were lost to follow-up (3.4% attrition). Demographic and medical characteristics of the participants are presented in Table 1. Infants' age and ELC scores at each timepoint are presented in Table 2.

3

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

Table 2 Descriptive statistics for ELC scores and child's age. N

corrected age at assessment (months) mean (SD), range

ELC score mean (SD), range

1 month

EPT VPT MPT FT Total

18 32 53 37 140

1.17 1.18 1.13 1.21 1.17

(0.17), (0.24), (0.25), (0.18), (0.22),

0.95–1.61 0.82–1.93 0.75–1.80 0.98–1.57 0.75–1.93

94.22 (9.50), 81–113 99.97 (7.44), 87–125 101.36 (7.33), 85–128 103.30 (10.49), 90–134 100.64 (8.91), 81–134

4 months

EPT VPT MPT FT Total

18 31 53 37 139

4.30 4.26 4.29 4.19 4.26

(0.24), (0.23), (0.26), (0.23), (0.24),

3.91–4.76 3.91–4.90 3.81–4.99 3.87–4.66 3.81–4.99

88.33 (10.69), 63–99 94.45 (8.33), 73–107 98.15 (8.36), 80–119 100.19 (7.55), 86–114 96.60 (9.20), 63–119

8 months

EPT VPT MPT FT Total

18 32 52 37 139

8.25 8.19 8.30 8.17 8.23

(0.40), (0.31), (0.29), (0.22), (0.30),

7.61–8.82 7.61–8.82 7.74–8.82 7.84–8.62 7.61–8.82

81.61 (12.09),56–99 88.91 (11.75), 61–112 96.73 (10.54), 72–115 102.09 (8.38), 78–120 94.40 (12.43), 56–120

12 months

EPT VPT MPT FT Total

18 31 53 37 139

12.15 12.20 12.15 12.11 12.15

(0.53), (0.44), (0.39), (0.32), (0.40),

11.31–13.18 11.61–13.64 11.38–13.18 11.70–12.92 11.31–13.64

91.61 (10.41), 73–108 102.45 (11.97), 80–123 106.13 (9.55), 82–127 107.95 (9.81), 87–130 103.91 (11.42), 73–130

18 months

EPT VPT MPT FT Total

16 32 52 37 137

17.73 18.04 17.99 18.08 17.99

(0.38), (0.39), (0.39), (0.36), (0.39),

17.05–18.46 17.41–18.85 17.11–18.98 17.48–19.02 17.05–19.02

85.38 (12.09), 72–120 91.13 (9.92), 69–113 100.87 (11.57), 79–126 103.59 (10.74), 88–129 97.52 (12.76), 69–129

EPT- extremely preterm, VPT – very preterm, MPT – moderately preterm; FT- full-term.

2.2. Procedure The study was approved by the hospital's Institutional Review Board committee (249-09). Parents signed a consent form prior to enrollment. The Mullen Scales of Early Learning (MSEL, Mullen, 1995) was administered at the corrected age (CA) of 1, 4, 8, 12, and 18 months. Preterm infants’ ages were corrected for prematurity at all time-points- i.e., the number of weeks a child was born prematurely is added to her/his chronological age. The Rapid Neonatal Neurobehavioral Assessment (RNNA, Gardner, Karmel, & Freedland, 2001) was administered at the one-month CA assessment. Assessments were conducted when infants were awake and alert and re-administered within a few days if the infant was too tired or irritable to complete the evaluation. Parents completed a demographic questionnaire at the first assessment. Maternal education was categorized as high school and/or some professional non-academic education versus academic education (i.e., BA or above). Ninety-nine (67%) mothers reported on having academic education. Medical data regarding pregnancy, delivery, and postnatal hospitalization were collected via computer-based hospital records after assessments were completed. Assessments were conducted by trained graduate students who were familiar with the families at the NICU and were therefore coded from the videotapes by trained independent coders who were unaware of infants' medical history and group assignment. Families were reimbursed for travel costs and received a videotape and a report of the assessments. 2.3. Measures The MSEL is a norm-referenced, standardized instrument for assessing developmental abilities of infants from birth through 68 months (Mullen, 1995). It is in common use with infants who raise developmental concerns (Burns, King, & Spencer, 2013). It includes five subscales: Visual Reception, Receptive Language, Expressive Language, Gross Motor and Fine Motor, and an Early Learning Composite (ELC) Summary Score. Each scale yields a raw score based on the items the child has completed, which is converted to an age-adjusted scaled score. The scaled scores are normalized with M = 50 and SD = 15 for subscales and with M = 100 and SD = 15 for the ELC score. Test-retest reliability and inter-rater reliability of the MSEL were reported as high, as was congruent validity with other measures (Mullen, 1995). The Rapid Neonatal Neurobehavioral Assessment (RNNA) is an evaluation of newborn neurobehavioral functioning at postmenstrual age of 34–48 weeks (Gardner et al., 2001). Ten behavioral categories are coded as normal, mildly abnormal, or very abnormal. These categories include: attention (auditory and visual), sensory symmetry, head/neck control, trunk tone, extremity movement and tone, motor symmetry, jitteriness, feeding, quality of movements, and state regulation (alertness and peak excitement). RNNA scores discriminated infants with brain injuries of different severities and predicted developmental outcomes (Gardner et al., 2001, 2006). 4

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

2.4. Statistical analyses The mixed models approach, which is also known as “hierarchical” or “multilevel” modeling (Raudenbush & Bryk, 2002), was used to analyze the developmental trajectories of EPT, VPT, MPT, and FT infants. The analysis was conducted using SAS PROC MIXED. Scaled ELC and subscale scores of the MSEL were the dependent variables in the models. The fixed effects part of the model included time, GA group, and GA x time interaction. Time was handled as a factor variable, so that the analysis was in the style of a repeated measures ANOVA. The statistical significance of the fixed effect terms was assessed using Wald tests. To gain an understanding of specific group differences and trajectories, pairwise comparisons between GA groups were conducted to examine the overall differences between groups (i.e., across time), differences between groups with respect to MSEL score changes over time (i.e., between the first assessment at 1 month and the last assessment at 18 months), and differences between groups at specific timepoints. The aim of these multiple analyses was to investigate patterns in the data, rather than to run multiple tests and focus only on those with statistical significance. Therefore, we did not adjust for multiple comparisons in these analyses and the p-values we report are unadjusted p-values. The random effects part of the model included a random child-specific intercept and an error term for each timepoint. We considered the following variations in modeling the covariance structure: (1) the variance of the random intercept independent of or depending on the GA group, (2) the variance of the error term independent of or depending on the GA group, and (3) the variance of the error term differing or equal across timepoints. We considered all eight possible combinations of the above choices and chose the most parsimonious model that could not be improved upon in a statistically significant manner at the 0.10 level by using a more complex model. Statistical comparisons between different covariance structures were carried out using the −2 log likelihood ratio chi-square test. This test takes the difference in the number of parameters between the larger and smaller models into account. AIC and BIC values are also reported for the different models. Gender, maternal education, and RNNA scores were then added to the fixed effects part of the model to determine, using backward selection, whether these characteristics independently predict ELC scores beyond GA group assignment. Maternal education was expressed as a binary variable, in terms of whether or not the mother had a bachelor's degree. The RNNA was handled as a quantitative variable. For each predictor, we examined the predictor’s main effect and the predictor x time interaction (with time again handled as a factor variable); in doing so we imposed the constraint that inclusion of a given predictor x time effect required inclusion of the corresponding predictor term. The statistical significance of the various terms was assessed using the Wald test, and we decided to include terms if they were “marginally statistically significant” (p-value between 0.05 and 0.10) as well as terms statistically significant at the conventional p < 0.05 level. 3. Results 3.1. Model fit An examination of the various choices of covariance structure indicated that for all of the dependent variables considered (i.e., the ELC and all subscales), no statistical evidence of heterogeneity of variance across the GA groups exists in the random intercept or error term. The inclusion of unequal error variance over time significantly improved the model fit for most of the scales and for consistency was therefore included for all scales (see Table 3 for the tests of model fit). 3.2. Mullen scales of early learning: mixed model analyses Descriptive statistics for the ELC scores are presented in Table 2 and Fig. 1. Overall, the pattern of change in the ELC scores was Table 3 Model fit for Early Learning Composite and subscales scores. ELC

Visual reception

Fine Motor

Gross Motor

Receptive Language

Expressive Language

Model with unequal error over time AIC BIC -2loglikelihood

5088.4 5164.9 5036.4

4690.7 4767.2 4638.7

4681.1 4757.6 4629.1

4841.7 4918.2 4789.7

4947.9 5024.4 4895.9

4738.8 4740.9 4686.8

Model with equal error over time AIC BIC -2loglikelihood

5092.2 5157.0 5048.2

4755.4 4820.1 4711.4

4736.4 4737.9 4692.4

4872.4 4873.9 4828.4

4993.8 5058.6 4949.8

4735.0 4736.6 4691.0

Model comparison Chi-square value of model comparisona (df = 14) p–value for Chi-square test

11.8 0.622

72.7 < 0.0001

63.3 < 0.0001

38.7 0.0004

53.9 < 0.0001

4.2 0.99

a note: Chi-square test were applied to compare models with equal and equal error across time-points, in terms of the differences in −2loglikelihood value of the models. Models with unequal error over time were applied for all scales in order to maintain consistency. For the ELC and Expressive Language scales, in which addition of this term did not improve model fit, analyses with equal error over time yielded similar results as with unequal error over time.

5

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

Fig. 1. Mullen Early Learning Composite scores.

irregular over time, with an initial decrease in the early period that was followed by an increase at 12 months and a decrease again at 18 months (see Table 2). As can be seen in Fig. 1, this unstable pattern was more prominent among VPT and EPT infants. The results of the analyses of the ELC and subscales are summarized in Table 4. The analyses generally reveal an association between the GA and MSEL scores, with the EPT infants scoring the lowest over time and presenting as the most disadvantaged group in all domains. As reflected in the significant main group effects in the ELC scores and across all subscales, differences in MSEL scores were observed among groups. Moreover, as reflected in the significant main time effects in the ELC scores and across all subscales reveal, changes in MSEL scores over time were also observed. Finally, as reflected in the significant group x time interactions in the ELC scores and most subscales, different patterns of change over time were observed among groups, with the patterns differing among the subscales (see Figs. 1 and 2). To further understand the different trajectories among GA groups, differences between groups in terms of both overall scores (i.e., averaged across time) and changes in scores over time (i.e., between 1 and 18 months) were examined. The results of pairwise comparisons for the ELC and all subscale scores are reported in Table 4. Table 5 presents pairwise ELC score comparisons between groups at all timepoints. The EPT infants’ overall ELC scores were significantly lower than those of the FT infants; the difference between the groups was significant at all timepoints. Comparing the groups’ scores at 1 vs. 18 months reveals that the gap between the groups significantly increased over time: the scores of the EPT infants declined, whereas those of the FT infants did not. These differences are also reflected in the subscales, with EPT infants’ overall scores significantly lower in all subscales. Moreover, the gaps between the EPT and FT infants significantly increased over time in the Visual Reception, Gross Motor, and Fine Motor subscales. Similar results emerged for the VPT infants, whose overall ELC scores were significantly lower than those of the FT infants; significant differences emerged from 4 months onward and the gap significantly increased over time. These differences are reflected in all subscales, in which the VPT infants’ overall scores were significantly lower and/or showed a significant decline compared to those of the FT infants. The overall average ELC score was significantly lower among MPT infants than among FT infants. However, examining individual timepoints reveals that a significant difference existed only at 8 months; no significant difference between groups was found regarding changes in scores between 1 and 18 months. Comparing the MPT and FT infants’ subscale scores yields different results for the subscales. The MPT infants' overall Visual Reception, Fine Motor, and Gross Motor scores were significantly lower than those of the FT infants. Moreover, a significantly increasing gap emerged for the Visual Reception and Gross Motor score. No significant differences emerged in the Receptive and Expressive Language subscales. Examining differences among PT groups reveals that the EPT infants’ overall ELC and subscale scores were significantly lower than those of the VPT infants, with no significant between-group differences in the gap between 1 month and 18 months. Similarly, the EPT infants’ overall ELC and subscale scores were significantly lower than those of MPT infants; however, the ELC and the Fine Motor, Gross Motor, and Expressive Language subscales featured a between-group gap that was significantly larger at 18 months than at 1 month. The VPT infants’ overall ELC and subscales scores were significantly lower than those of the MPT infants. Furthermore, a between-group gap that was significantly larger at 18 months than at 1 month was observed for the ELC scores as well as for the Fine Motor, Receptive Language, and Expressive Language subscales. 3.3. Predictors of ELC scores We initially considered the full model with all predictor and predictor x time terms, including gender, maternal education, and RNNA scores. As the predictor x time terms were significant at the 0.10 level (Wald test) for all three predictors, we took the full model as the final model (see Table 6). Thus, the ELC score predictions over time were more accurate when gender, maternal education, and RNNA scores were included in addition to the GA groups. As Fig. 3 demonstrates, the ELC scores of boys were significantly lower than those of girls, with an increasing gap over time. Similarly, the ELC scores of infants whose mothers had less 6

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

Table 4 Regression results and pairwise comparisons for Mullen composite scores and subscales. ELC

Visual reception

Fine motor

Gross motor

Receptive language

Expressive language

< 0.0001 < 0001 0.003

< 0.0001 < 0.0001 0.0002

< 0.0001 < 0.0001 0.0003

< 0.0001 < 0.0001 < 0.0001

0.0002 0.0093 0.1326

< 0.0001 < 0.0001 0.0159

−9.32 0.99 < 0.0001 −5.15 0.82 < 0.0001 −2.22 0.73 0.0027

−9.50 1.46 < 0.0001 −4.61 1.22 0.0002 −2.19 1.09 0.0455

−4.66 1.32 0.0005 −1.90 1.10 0.0853 0.62 0.97 0.5240

−7.32 1.18 < 0.0001 −4.44 0.99 < 0.0001 −1.61 0.88 0.0680

Pairwise comparisons among preterm groups – Average Level Across Time EPT vs VPT Estimate −7.22 −5.17 −4.17 SD 1.84 1.20 1.01 p-value 0.000‘ < 0.0001 < 0.0001 EPT vs MPT Estimate −12.53 −7.80 −7.10 SD 1.70 1.11 0.94 p-value < 0.0001 < 0.0001 < 0.0001 VPT vs MPT Estimate −5.30 −2.63 −2.93 SD 1.39 0.90 0.77 p-value 0.0002 0.004 0.0002

−4.89 1.50 0.0013 −7.31 1.39 < 0.0001 −2.42 1.14 0.0351

−2.76 1.35 0.0429 −5.28 1.25 < 0.0001 −2.52 1.02 0.0144

−2.88 1.21 0.0190 −5.70 1.12 < 0.0001 −2.83 0.92 0.0025

Pairwise comparisons with FT group – Difference Between 18 Months and 1 Month EPT vs FT Estimate −9.73 −9.15 −6.73 SD 3.69 2.70 2.56 p-value 0.009 0.0008 0.0091 VPT vs FT Estimate −9.14 −7.46 −1.90 SD 3.02 2.19 2.09 p-value 0.003 0.0008 0.3658 MPT vs FT Estimate −0.88 −5.06 2.09 SD 2.68 1.95 1.86 p-value 0.744 0.0103 0.262

−13.72 3.03 < 0.0001 −11.06 2.49 < 0.0001 −6.60 2.23 0.0036

1.41 3.71 0.7048 −6.04 3.02 0.0467 −0.01 2.68 0.9976

−5.31 2.84 0.0624 −3.32 2.33 0.1564 1.15 2.08 0.579

7.44 3.80 0.0513 1.41 3.54 0.6898 −6.03 2.80 0.0323

−1.99 2.91 0.4932 −6.47 2.70 0.0176 −4.47 2.17 0.0403

Main effects group effect time effect group*time

p-value p-value p-value

Pairwise comparisons with FT group – Average Level Across Time EPT vs FT Estimate −15.31 −10.14 SD 1.79 1.64 p-value < 0.0001 < 0.0001 VPT vs FT Estimate −8.08 −4.98 SD 1.50 0.97 p-value < 0.0001 < 0.0001 MPT vs FT Estimate −2.78 −2.34 SD 1.33 0.86 p-value 0.039 0.0073

Pairwise comparisons among preterm groups – Difference Between 18 Months and 1 Month EPT vs VPT Estimate −0.59 −1.69 −4.83 −2.66 SD 3.78 2.76 2.62 3.12 p-value 0.876 0.540 0.0663 0.3952 EPT vs MPT Estimate −8.86 −4.09 −8.82 −7.12 SD 3.52 2.57 2.44 2.92 p-value 0.0125 0.113 0.0004 0.0160 VPT vs MPT Estimate −8.26 −2.40 −3.99 −4.45 SD 2.80 2.04 1.95 2.37 p-value 0.0035 0.240 0.0414 0.0620

ELC- early learning composite; EPT – extremely preterm; VPT – very preterm; MPT – moderately preterm; FT – full-term.

academic education were significantly lower than those of infants whose mothers had attained more academic education, with an increasing gap over time (see Fig. 3). The ELC trajectories over time by RNNA score are shown in Fig. 3, where the colors are on a redyellow-green spectrum, red corresponding to low RNNA values and green representing high values. Higher (i.e., more abnormal) RNNA scores were associated with lower overall ELC scores over time. Moreover, the trajectories of the infants with more abnormal RNNA scores exhibited less stability over time. 4. Discussion The MSEL, which was administered at five timepoints in this study, provides a unique opportunity for an in-depth longitudinal assessment of preterm infants’ early development. To the best of our knowledge, this is one of the first studies to examine differences in early developmental trajectories longitudinally within preterm cohorts, with an emphasis on MPT infants for whom longitudinal data are less extensive. The EPT and VPT infants have been reported in the literature as consistently scoring lower than FT infants in cognitive, language, and motor assessments; the differences emerge early in life and persist or even increase over time (Barre, Morgan, Doyle, & Anderson, 2011; Erikson, Allert, Carlberg, & Katz-Salamon, 2003; Sansavini, Savini, Guarini, Broccoli, and Alessandroni, 2011; Sansavini et al., 2014; Van Haastert, De Vries, Helders, & Jongmans, 2006). Similar results regarding EPT and VPT infants were obtained in the 7

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

Fig. 2. Mullen subscale scores.

current study. Interestingly, although the EPT infants' scores were lower than those of VPT infants, the gap remained consistent and did not increase or decrease over time. Regarding MPT infants, findings from cross-sectional studies indicate that MPT infants have lower scores than FT infants but higher scores compared to EPT infants (Boyle et al., 2014; Johnson et al., 2015; Kerstjens et al., 2011; Potijk, Kerstjens, Bos, Reijneveld, & De Winter, 2013). This pattern of differences was mostly confirmed by our findings, as MPT infants scored below FT and above EPT infants in the cognitive and motor domains. However, the MPT infants' trajectories did not significantly differ from those of FT infants in the language domain. In line with previous studies (Janssen et al., 2011; Sansavini, Savini et al., 2011; Van Haastert et al., 2006), male gender and lower maternal education were associated with both lower ELC scores and a decline in scores over time. These effects were significant 8

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

Table 5 Pairwise Comparisons over time for Mullen Early Learning composite score. 1 month

4 months

8 months

12 months

18 months

EPT vs FT

Estimate SD p-value

−9.07 2.70 001.

−11.86 2.45 < 0.0001

−20.47 2.82 < 0.0001

−16.33 2.78 < 0.0001

−18.81 3.23 < 0.0001

VPT vs FT

Estimate SD p-value

−3.32 2.27 144.

−5.84 2.07 0054.

−13.17 2.37 < 0.0001

−5.60 2.35 0182.

−12.47 2.62 < 0.0001

MPT vs FT

Estimate SD p-value

−1.94 2.01 3365.

−2.04 1.82 2656.

−5.29 2.11 0128.

−1.81 2.07 3821.

−2.82 2.33 2292.

EPT vs VPT

Estimate SD p-value

−5.75 2.76 0395.

−6.01 2.52 0183.

−7.29 2.88 0124.

−10.75 2.86 0002.

−6.34 3.31 057.

EPT vs MPT

Estimate SD p-value

−7.14 2.56 006.

−9.82 2.32 < 0.0001

−15.17 2.68 < 0.0001

−14.52 2.63 < 0.0001

−15.99 2.08 < 0.0001

VPT vs MPT

Estimate SD p-value

−1.39 2.10 5094.

−3.80 1.92 0495.

−7.88 2.20 0004.

−3.79 2.18 0845.

−9.65 2.44 0001.

EPT – extremely preterm; VPT – very preterm; MPT – moderately preterm; FT – full-term.

Table 6 Final model with predictors for Early Learning Composite Scores. Effect

Num DF

Den DF

f Value

P value

Group time Group × Time Gender Gender × Time Maternal education Maternal education × Time RNNA RNNA × Time

3 4 12 1 4 1 4 1 4

138 187 188 138 189 138 189 137 187

19.81 6.95 2.68 6.75 2.01 3.83 3.18 1.84 2.24

< 0.0001 < 0.0001 0.0024 0.0104 0.0942 0.0523 0.0149 0.1772 0.0667

RNNA – Rapid neonatal neurobehavioral assessment.

beyond GA, which suggests that male infants whose mothers are less educated may benefit from attentive follow-up and parental guidance regarding appropriate developmental stimulation. The rate of maternal academic education was somewhat high in the current sample, namely 67% compared to the 56% overall level of academic education in Israel (OECD, 2014); as such, generalization of the results to other populations may be limited. Neonatal neurobehavior, as assessed by RNNA at one-month CA, was also examined as a potential behavioral predictor of ELC scores over time. Previous studies have yielded inconsistent results regarding the potential of neonatal neurobehavioral assessment scores to predict preterm infants' developmental outcome (Harijan et al., 2012; Picciolini et al., 2016; Stephens et al., 2010). In the current study, abnormal neurobehavior at one-month CA was associated with lower ELC scores over time. Notably, it was also associated with less stability in scores over time − which may explain the inconsistency in the predictability of neonatal neurobehavioral assessment for long-term outcomes. For some infants, it may be indicative of a sustained injury that may result in persistent, long-term difficulties. For others, the presence of early neurobehavioral abnormalities may reflect neurological maturational processes that may be transient and naturally resolved over time and therefore not necessarily predict later outcomes or mandate further prevention and intervention efforts. Nevertheless, this association and the large variability and changes in scores over time among preterm infants emphasize the importance of longitudinal studies and repeated assessments. Several limitations of this study should be noted. First, although the follow-up was extensive, the endpoint of 18 months CA is rather early given the subject matter. Catch-up gains, which are seen in long-term follow-up studies (Fitzpatrick, Carter, & Quigley, 2016; Samuelsson et al., 2006), were not seen in this study. Moreover, developmental difficulties not yet apparent in the current study may emerge over time with the increasing language, academic, and social demands during late childhood. Second, the sample size and low incidence of neonatal complications did not allow for the examination of medical conditions that may be associated with various developmental outcomes. Moreover, there is an issue of confounding: The groups were defined in terms of GA, but GA is highly correlated with birthweight (the correlation in the sample was 0.94), so it is difficult to tell whether the differences between 9

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

Fig. 3. Mullen Early Learning Composite scores by gender, maternal education and RNNA scores.

groups are due to gestational age or to birthweight. This point should be taken into account in the interpretation of the results. Third, it should be taken into account that the MSEL has not been validated in Israel and therefor the MSEL the age-norms may be somewhat different for this country. For this study, the reference for comparing the preterm infants' scores is therefore the full-term group and not only the age-norms. Moreover, the patterns of change in the scores in the whole sample should be interpreted with caution. Fourth, the sample included both mono- and bilingual children. A more in-depth investigation may be useful to address the possible association between bilingualism and language development among preterm infants. Finally, this study did not address parental mental health or parenting styles, which may also affect developmental outcomes.

5. Conclusions In this study, the EPT and VPT infants displayed significant delays in all areas that emerged very early in life and increased over time. The MPT infants had more favorable trajectories than the EPT and VPT infants but achieved lower scores than the FT infants. Male infants, infants born to mothers with fewer educational qualifications, and infants with more neonatal neurobehavioral abnormalities were found to be at elevated risk. Future research may shed more light on the mechanisms that underlie the associations among infant gender, maternal education, and developmental outcomes. Biological factors, including genetic susceptibility, specific neurological injuries, and medical complications may be linked to specific deficiencies. Social factors such as early parent-child interactions may modify biological predisposition and should be addressed. Parents’ interaction style and verbal input may also affect preterm infants' development and should thus be further examined as well. Finally, long-term follow-up is important to examine the long-lasting effects of preterm birth.

Acknowledgement We wish to thank the participating families for their time and cooperation. 10

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

References Aarnoudse-Moens, C. S. H., Weisglas-Kuperus, N., Van Goudoever, J. B., & Oosterlaan, J. (2009). Meta-analysis of neurobehavioral outcomes in very preterm and/or very low birth weight children. Pediatrics, 124(2), 717–728. http://dx.doi.org/10.1542/peds.2008-2816. Als, H., Duffy, F. H., McAnulty, G. B., Rivkin, M. J., Vajapeyam, S., Mulkern, R., et al. (2004). Early experience alters brain function and structure. Pediatrics, 113(4), 846–857. http://dx.doi.org/10.1542/peds.113.4.846. Arpino, C., Compagnone, E., Montanaro, M. L., Cacciatore, D., De Luca, A., Cerulli, A., et al. (2010). Preterm birth and neurodevelopmental outcome: A review. Child’s Nervous System, 26(9), 1139–1149. http://dx.doi.org/10.1007/s00381-010-1125-y. Aylward, G. P. (2005). Neurodevelopmental outcomes of infants born prematurely. Journal of Developmental and Behavioral Pediatrics, 26(6), 427–440. Barre, N., Morgan, A., Doyle, L. W., & Anderson, P. J. (2011). Language abilities in children who were very preterm and/or very low birth weight: A meta-analysis. The Journal of Pediatrics, 158(5), 766–774. http://dx.doi.org/10.1016/j.jpeds.2010.10.032 e1. Beck, S., Wojdyla, D., Say, L., Betran, A. P., Merialdi, M., Requejo, J. H., et al. (2010). The worldwide incidence of preterm birth: A systematic review of maternal mortality and morbidity. Bulletin of the World Health Organization, 88(1), 31–38. http://dx.doi.org/10.1590/s0042-96862010000100012. Bhutta, A. T., Cleves, M. A., Casey, P. H., Cradock, M. M., & Anand, K. J. S. (2002). Cognitive and behavioral outcomes of school-aged children who were born preterm. JAMA − Journal of the American Medical Association, 288(6), 728. http://dx.doi.org/10.1001/jama.288.6.728. Blauw-Hospers, C. H., & Hadders-Algra, M. (2005). A systematic review of the effects of early intervention on motor development. Developmental Medicine & Child Neurology, 47(6), 421–432. http://dx.doi.org/10.1017/S0012162205000824. Boyle, E., Draper, E., Evans, T., Field, D., Manktelow, B., Marlow, N., et al. (2014). Infant neurodevelopmental outcomes following late and moderately preterm birth. Archives of Disease in Childhood. Fetal and Neonatal Edition, 99, A3. http://dx.doi.org/10.1136/archdischild-2014-306576.7. Brown, N. C., Doyle, L. W., Bear, M. J., & Inder, T. E. (2006). Alterations in neurobehavior at term reflect differing perinatal exposures in very preterm infants. Pediatrics, 118(6), 2461–2471. http://dx.doi.org/10.1542/peds.2006-0880. Burns, T. G., King, T. Z., & Spencer, K. S. (2013). Mullen scales of early learning: The utility in assessing children diagnosed with autism spectrum disorders, cerebral palsy, and epilepsy. Applied Neuropsychology: Child, 2(1), 33–42. http://dx.doi.org/10.1080/21622965.2012.682852. Carbajal, R., Rousset, A., Danan, C., Coquery, S., Nolent, P., Ducrocq, S., et al. (2008). Epidemiology and treatment of painful procedures in neonates in intensive care units. JAMA − Journal of the American Medical Association, 300(1), 60–70. http://dx.doi.org/10.1001/jama.300.1.60. Chyi, L. J., Lee, H. C., Hintz, S. R., Gould, J. B., & Sutcliffe, T. L. (2008). School outcomes of late preterm infants: Special needs and challenges for infants born at 32–36 weeks gestation. The Journal of Pediatrics, 153(1), 25–31. http://dx.doi.org/10.1016/j.jpeds.2008.01.027. De Kieviet, J. F., Piek, J. P., Aarnoudse-Moens, C. S., & Oosterlaan, J. (2009). Motor development in very preterm and very low-birth-weight children from birth to adolescence: A meta-analysis. JAMA- Journal of the American Medical Association, 302(20), 2235–2242. http://dx.doi.org/10.1001/jama.2009.1708. De Schuymer, L., De Groote, I., Striano, T., Stahl, D., & Roeyers, H. (2011). Dyadic and triadic skills in preterm and full term infants: A longitudinal study in the first year. Infant Behavior and Development, 34(1), 179–188. http://dx.doi.org/10.1016/j.infbeh.2010.12.007. Erikson, C., Allert, C., Carlberg, E. B., & Katz-Salamon, M. (2003). Stability of longitudinal motor development in very low birthweight infants from 5 months to 5.5 years. Acta Paediatrica, 92(2), 197–203. Fitzpatrick, A., Carter, J., & Quigley, M. A. (2016). Association of gestational age with verbal ability and spatial working memory at age 11. Pediatrics, 138(6), http:// dx.doi.org/10.1542/peds.2016-0578. Gardner, J., Karmel, B., & Freedland, R. (2001). Determining functional integrity in neonates: A rapid neurobehavior assessment tool. In T. Singer, & P. S. Zeskind (Eds.). Biobehavioral assessment of the infant (pp. 398–422). New York, NY: Guilford Press. Gardner, J. M., Karmel, B. Z., Freedland, R. L., Lennon, E. M., Flory, M. J., Miroshnichenko, I., et al. (2006). Arousal, attention, and neurobehavioral assessment in the neonatal period: Implications for intervention and policy. Journal of Policy and Practice in Intellectual Disabilities, 3(1), 22–32. http://dx.doi.org/10.1111/j.17411130.2006.00049. Goldenberg, R. L., Culhane, J. F., Iams, J. D., & Romero, R. (2008). Epidemiology and causes of preterm birth. The Lancet, 371(9606), 75–84. http://dx.doi.org/10. 1016/S0140-6736. Harijan, P., Beer, C., Glazebrook, C., Israel, C., Marlow, N., Whitelaw, A., & Johnson, S. (2012). Predicting developmental outcomes in very preterm infants: Validity of a neonatal neurobehavioral assessment. Acta Paediatrica, 101(7), 275–281. http://dx.doi.org/10.1111/j.1651-2227.2012.02663.x. Janssen, A. J., Akkermans, R. P., Steiner, K., De Haes, O. A., Oostendorp, R. A., Kollée, L. A., & Nijhuis-van der Sanden, M. W. (2011). Unstable longitudinal motor performance in preterm infants from 6 to 24 months on the Bayley Scales of Infant Development–Second edition. Research in Developmental Disabilities, 32(5), 1902–1909. http://dx.doi.org/10.1016/j.ridd.2011.03.026. Johnson, S., & Marlow, N. (2014). Growing up after extremely preterm birth: Lifespan mental health outcomes. Seminars in Fetal and Neonatal Medicine, 19(2), 97–104. http://dx.doi.org/10.1016/j.siny.2013.11.004. Johnson, S., Evans, T. A., Draper, E. S., Field, D. J., Manktelow, B. N., Marlow, N., et al. (2015). Neurodevelopmental outcomes following late and moderate prematurity: A population-based cohort study. Archives of Disease in Childhood – Fetal and Neonatal Edition, 100(4), F301–F308. http://dx.doi.org/10.1136/ archdischild-2014-307684. Kerstjens, J. M., De Winter, A. F., Bocca-Tjeertes, I. F., Ten Vergert, E. M. J., Reijneveld, S. A., & Bos, A. F. (2011). Developmental delay in moderately preterm-born children at school entry. The Journal of Pediatrics, 159(1), 92–98. http://dx.doi.org/10.1016/j.jpeds.2010.12.041. Larroque, B., Ancel, P. Y., Marret, S., Marchand, L., André, M., Arnaud, C., et al. (2008). Neurodevelopmental disabilities and special care of 5-year-old children born before 33 weeks of gestation (the EPIPAGE study): A longitudinal cohort study. The Lancet, 371(9615), 813–820. http://dx.doi.org/10.1016/S0140-6736(08) 60380-3. Luu, T. M., Vohr, B. R., Allan, W., Schneider, K. C., & Ment, L. R. (2011). Evidence for catch-up in cognition and receptive vocabulary among adolescents born very preterm. Pediatrics, 128(2), 313–322. http://dx.doi.org/10.1542/peds.2010-2655. Moster, D., Lie, R. T., & Markestad, T. (2008). Long-term medical and social consequences of preterm birth. The New England Journal of Medicine, 359(3), 262–273. http://dx.doi.org/10.1056/NEJMoa0706475. Mullen, E. M. (1995). Mullen scales of early learning, AGS edition: Manual and item administrative books. Circle Pines, MN: American Guidance Service. Nordhov, S. M., Rønning, J. A., Dahl, L. B., Ulvund, S. E., Tunby, J., Kaaresen, P. I., et al. (2010). Early intervention improves cognitive outcomes for preterm infants: Randomized controlled trial. Pediatrics, 126(5), e1088–94. http://dx.doi.org/10.1542/peds.2010-0778. Population with tertiary education. http://dx.doi.org/10.1787/0b8f90e9-en. Picciolini, O., Montirosso, R., Porro, M., Giann, M. L., Mosca, F., Porro, C. M., et al. (2016). Neurofunctional assessment at term equivalent age can predict 3-year neurodevelopmental outcomes in very low birth weight infants. Acta Paediatrica, 105(2), e47–53. http://dx.doi.org/10.1111/apa.13248. Platt, M. J., Cans, C., Johnson, A., Surman, G., Topp, M., Torrioli, M. G., & Krageloh-Mann, I. (2007). Trends in cerebral palsy among infants of very low birthweight (< 1500 g) or born prematurely (< 32 weeks) in 16 European centres: A database study. The Lancet, 369(9555), 43–50. http://dx.doi.org/10.1016/S0140-6736. Potijk, M. R., Kerstjens, J. M., Bos, A. F., Reijneveld, S. A., & De Winter, A. F. (2013). Developmental delay in moderately preterm-born children with low socioeconomic status: Risks multiply. The Journal of Pediatrics, 163(5), 1289–1295. http://dx.doi.org/10.1016/j.jpeds.2013.07.001. Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods. California, USA: Sage Publications. Saigal, S., & Doyle, L. W. (2008). An overview of mortality and sequelae of preterm birth from infancy to adulthood. The Lancet, 371(9608), 261–269. http://dx.doi. org/10.1016/S0140-6736. Samuelsson, S., Finnström, O., Flodmark, O., Gäddlin, P. O., Leijon, I., & Wadsby, M. (2006). A longitudinal study of reading skills among very-low-birthweight children: Is there a catch-up. Journal of Pediatric Psychology, 31(9), 967–977. http://dx.doi.org/10.1093/jpepsy/jsj108. Sansavini, A., Guarini, A., & Caselli, M. C. (2011). Preterm birth: Neuropsychological profiles and atypical developmental pathways. Developmental Disabilities Research Reviews, 17(2), 102–113. https://doi.org/10.1002/ddrr.1105. Sansavini, A., Guarini, A., Savini, S., Broccoli, S., Justice, L., Alessandroni, R., & Faldella, G. (2011). Longitudinal trajectories of gestural and linguistic abilities in very

11

Research in Developmental Disabilities xxx (xxxx) xxx–xxx

M. Yaari et al.

preterm infants in the second year of life. Neuropsychologia, 49(13), 3677–3688. http://dx.doi.org/10.1016/j.neuropsychologia.2011.09.023. Sansavini, A., Savini, S., Guarini, A., Broccoli, S., Alessandroni, R., & Faldella, G. (2011). The effect of gestational age on developmental outcomes: A longitudinal study in the first 2 years of life Child: Care. Health and Development, 37(1), 26–36. http://dx.doi.org/10.1111/j.1365-2214.2010.01143.x. Sansavini, A., Pentimonti, J., Justice, L., Guarini, A., Savini, S., Alessandroni, R., & Faldella, G. (2014). Language, motor and cognitive development of extremely preterm children: Modeling individual growth trajectories over the first three years of life. Journal of Communication Disorders, 49, 55–68. http://dx.doi.org/10. 1016/j.jcomdis.2014.02.005. Stephens, B. E., Liu, J., Lester, B., Lagasse, L., Shankaran, S., Bada, H., et al. (2010). Neurobehavioral assessment predicts motor outcome in preterm infants. The Journal of Pediatrics, 156(3), 366–371. http://dx.doi.org/10.1016/j.jpeds.2009.09.042. Van Haastert, I. C., De Vries, L. S., Helders, P. J. M., & Jongmans, M. J. (2006). Early gross motor development of preterm infants according to the Alberta Infant Motor Scale. The Journal of Pediatrics, 149(5), 617–622. http://dx.doi.org/10.1016/j.jpeds.2006.07.025. Van Noort-van der Spek, I. L., Franken, M. C. J. P., & Weisglas-Kuperus, N. (2012). Language functions in preterm-born children: A systematic review and metaanalysis. Pediatrics, 129(4), 745–754. http://dx.doi.org/10.1542/peds.2011-1728. Volpe, J. J. (2009). Brain injury in premature infants: A complex amalgam of destructive and developmental disturbances. The Lancet Neurology, 8(1), 110–124. http:// dx.doi.org/10.1016/S1474-4422. Woodward, L. J., Moor, S., Hood, K. M., Champion, P. R., Foster-Cohen, S., Inder, T. E., & Austin, N. C. (2009). Very preterm children show impairments across multiple neurodevelopmental domains by age 4 years. Archives of Disease in Childhood. Fetal and Neonatal Edition, 94(5), F339–44. http://dx.doi.org/10.1136/adc. 2008.146282.

12