Cognitive and neuropsychological outcomes at 5 years of age in preterm children born in the 2000s

Transcription

1 DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY ORIGINAL ARTICLE Cognitive and neuropsychological outcomes at 5 years of age in preterm children born in the 2000s ANNIKA LIND 1,2 MARIT KORKMAN 3 LIISA LEHTONEN 4 HELENA LAPINLEIMU 4 RIITTA PARKKOLA 5 JAAKKO MATOM¾KI 1 LEENA HAATAJA 6 THE PIPARI STUDY GROUP 1 Department of Pediatrics, Turku University Hospital, Turku, Finland. 2 Department of Psychology, Šbo Akademi University, Turku, Finland. 3 Department of Psychology, University of Helsinki, Helsinki, Finland. 4 Department of Pediatrics, University of Turku and Turku University Hospital, Turku, Finland. 5 Department of Radiology, Turku PET Center, University of Turku and Turku University Hospital, Turku, Finland. 6 Department of Pediatric Neurology, University of Turku and Turku University Hospital, Turku, Finland. Correspondence to Dr Leena Haataja at Turku University Hospital, Department of Pediatric Neurology, PO Box 52, Turku, Finland. leena.haataja@utu.fi This article is commented on Cooke on page 197 of this issue. PUBLICATION DATA Accepted for publication 22nd August Published online 17th December AIM The aim of this study was to assess the cognitive level and neuropsychological performance at 5 years of age in children with a very low birthweight (VLBW; birthweight <1501g) born in 2001 to METHOD A regional cohort of 97 children with a VLBW (mean gestational age 28wks [SD 3wks]; mean birthweight 1054g [SD 259g]; 50 male; 47 female) and a comparison group of 161 healthy children born at term (mean gestational age 40wks [SD 1wks]; mean birthweight 3644g [SD 446g]; 80 male; 81 female) were included in this study. At 5 years of age (SD 2mo) cognitive level was assessed with the Wechsler Preschool and Primary Scale of Intelligence revised and neuropsychological performance was assessed using NEPSY II. RESULTS The mean intelligence of the VLBW group corresponded to the normative mean of 100 but was lower than the mean of the low-risk comparison group (p<0.001). Neuropsychological performance was also significantly poorer than that of the comparison group. In NEPSY II, 25% of the VLBW group had a significantly impaired performance in one of 11 subtests and 33% had a significantly impaired performance in more than one, while 19% of the comparison group had a significantly impaired performance in one subtest and 10% had a significantly impaired performance in more than one. INTERPRETATION The mean cognitive capacity of the VLBW cohort corresponded to the normative mean. However, assessments of cognitive level only at preschool age do not provide the information on neuropsychological deficit that is necessary for planning adequate educational support. Preterm birth predisposes the infant to long-term developmental problems. As diverging developmental outcomes in different regions, 1 centres, 2 hospital levels, 3 and time periods 4,5 have been reported, continuous regional follow-up is needed for the evaluation of treatment and for the planning of adequate developmental support. There are some encouraging reports on the decreasing rate of cerebral palsy (CP) among preterm children, 4,6 as well as on the improved cognitive outcome of preterm children at 2 years of age, 7 but similar trends regarding the neurocognitive outcome at preschool age have not yet been reported. Thus, there is a need to learn more about the long-term cognitive and neuropsychological development of children with a very low birthweight (VLBW) born during the present decade. It has been shown that children who were born preterm achieve lower scores on intelligence tests 8,9 than their peers born at term. They also have more impairments in various domains of neuropsychological functioning than children born at term Some studies suggest that preterm birth is associated with general intellectual impairment rather than with specific dysfunctions, 13 whereas others report specific impairments despite cognitive levels within a normal range. 10 The aim of the present study was to assess cognitive and neuropsychological outcomes at 5 years of age in a regional cohort of children with a VLBW born in 2001 to 2003 and to analyse the effects of selected pre-, peri-, and postnatal data. Owing to improvements in neonatal care, it was expected that the outcome in this cohort would be favourable compared with outcomes reported in earlier publications. METHOD Participants This study is a part of the Finnish multidisciplinary, longitudinal research project, PIPARI (Development and Functioning of Very Low Birth Weight Infants from Infancy to School Age). When possible, all births less than 32 gestational weeks in south-west Finland are centralized to Turku University Hospital. All surviving preterm infants with a VLBW 256 DOI: /j x ª The Authors. Journal compilation ª Mac Keith Press 2010

2 (<1501g) born between January 2001 and December 2003 to Finnish-speaking families who lived inside the hospital catchment area were eligible for this study. Exclusion criteria were major congenital anomalies or genetic or chromosomal syndromes. A total of 103 children were eligible for the study but six were withdrawn by their families. Thus, the VLBW group consisted of 97 children in total, 50 male and 47 female. The comparison group consisted of 161 healthy, term infants (80 male; 81 female) born between November 2001 and December 2003 at Turku University Hospital. The comparison group was recruited by asking the parents of the first male and the first female born each Monday to take part in the study. If they refused, the parents of the next male or female were asked. Children were included in the study if they met the following criteria: their birthweight was more than 2SD below the mean according to age- and sex-specific Finnish growth charts; their gestational age at birth was equal to or greater than 37 weeks; they had not been admitted into neonatal care during the first week of life; and their family was Finnish speaking and lived inside the hospital catchment area. Exclusion criteria were major congenital anomalies or genetic or chromosomal syndromes and mother s use of illicit drugs or alcohol during the pregnancy. The collection of data and the testing of the children continued in the PIPARI study after the present paper was prepared. In this present study, we included all preterm children and all comparison children who had been tested by the date this paper was prepared. Written informed consent was obtained from both biological parents. The families were recruited when the children were newborn and very few refused to take part in the study. The PIPARI study protocol was approved by the Ethics Review Committee of the Hospital District of the South-West Finland in December Brain imaging Brain pathology classifications based separately on serial cranial ultrasound examinations and magnetic resonance imaging (MRI) of the brain were used as background variables. The ultrasound examinations were performed as described by Reiman et al. 14 on all the preterm infants at ages 3 to 5 days, 7 to 10 days, 1 month, and monthly thereafter until discharge from hospital and again at term-equivalent age. The infants were categorized into three groups (normal, mildly abnormal, severely abnormal) according to the most pathological finding in the ultrasound examinations and as described by Rademaker et al. 15 An MRI study was performed at term-equivalent age on all preterm infants. The design of this study and categorization into three groups (normal, minor brain abnormality, major brain abnormality) according to the most pathological finding on MRI have been described in Maunu et al. 16 What this paper adds This paper indicates that the prognosis for the long-term cognitive outcome of children with a very low birthweight born between 2001 and 2003 seems to have improved compared with that reported in earlier publications. Despite the improved cognitive outcome, this paper shows that children with a very low birthweight had a higher level of difficulty than the comparison group in all four neuropsychological domains. This paper suggests that cognitive outcome and neuropsychological performance did not systematically worsen with lower gestational age. Wechsler Preschool and Primary Scale of Intelligence revised The cognitive level of the children was evaluated at 5 years of chronological age (SD 2mo) by a psychologist using a short form of the Wechsler Preschool and Primary Scale of Intelligence revised (WPPSI-R 17 ) that included the following subtests: information, sentences, arithmetic, block design, geometric design, and picture completion. The psychologist was not blinded to group. Full-scale intelligence quotient (FSIQ), Verbal intelligence quotient (VIQ), and Performance intelligence quotient (PIQ) were estimated (normal group, mean=100, SD 15). A quotient of 85 or above (within )1SD) was considered as normal intelligence, a quotient of 70 to 84 ()1SD to )2SDs) as slightly below normal, and a quotient of 69 or below (less than )2SDs) as significantly below normal. NEPSY II At 5 years of chronological age (SD 2mo), the neuropsychological performance of the children was evaluated by a psychologist using NEPSY II (the standardization edition of the Finnish version). 18,19 Age-appropriate subtests were selected to form a set of tests that tap domains of neurocognitive development comprehensively. The selected subtests assess attention (subtests auditory attention and visual attention ), executive functioning (the subtest inhibition ), memory functions (subtests word list interference, narrative memory, and memory for designs ), visuomotor and visuospatial functions (subtests visuomotor precision and design copy ), and language (subtests comprehension of instructions, phonological processing, and speeded naming ). The scoring instructions for NEPSY II were under development at the time of the study. Later modifications in scoring and start - stop rules prevented our use of the later published norms. The subtests and the differences from the final version have been described in Lind et al. 20 Standard scores were based on the results of the comparison group; a standard score of 8 or above was considered as average performance, a score of 6 or 7 as slightly below average, and a score of 5 or below as significantly below average, in accordance with the manual. 18,19 Statistical analysis In the comparison of neuropsychological functioning between children born at different gestational ages, domains of the NEPSY II subtests were created in order to reduce the number of variables in NEPSY II. The subtests auditory attention, visual attention, and inhibition formed the domain executive functioning. The subtests narrative memory, memory for designs, and word list interference formed the domain memory. The subtests visuomotor precision and design copy formed the domain motor skills. Finally, subtests speeded naming, comprehension of instructions, and phonological processing formed the domain language. The mean value of the standard scores for the subtests in each Preterm Children Born in the 2000s Annika Lind et al. 257

3 domain was considered as a standard score for the domain. The domain names differed slightly from the domain names in the manual. The manual domain names were reworded in accordance with our previous publication. 20 The mean values of the test scores between the VLBW and comparison groups and between males and females with a VLBW were compared using an independent-sample t-test. Hierarchical multiple linear regression analyses were performed to study the effect of demographic and neonatal data on the outcome variables. The following variables were used as dependent variables: FSIQ, VIQ, PIQ, and the scores of 11 NEPSY II subtests. In the first stage of regression analysis, the length of mother s education in years was used as an independent variable. In the second stage, the use of prenatal corticosteroids was added to the equation. In the third stage, gestational age and sex were entered. In the fourth stage, neonatal conditions: (1) any inflammatory disease (chronic lung disease [need for supplementary oxygen at an age equivalent to 36 gestational weeks], sepsis or meningitis, and or intestinal perforations including those related to necrotizing enterocolitis and isolated perforations) and (2) minor and major brain abnormality, on either brain MRI or ultrasound images were added. First the significant changes in coefficient of determination (DR 2 ) are presented to describe how much variance in the dependent variable can be explained by one or a set of new independent variables over and above that explained by an earlier set. Then the significant associations of the individual variables are presented, taking into account the impact of all the other variables in the stage. The results are presented using estimated regression coefficients (b)and95%confidence intervals (CIs). Diagnostic plots were used to assess the assumptions of the regression models. Residuals were approximately normally distributed. Residual plots indicated neither heteroscedasticity nor non-linear relationships. As our data contained only a few children who did not receive prenatal corticosteroids, these children had the highest leverage values. The statistical analyses were performed using SAS (version 9.1; SAS Institute, Cary, NC, USA) and p-values below 0.05 were considered statistically significant. All the p-values were two-tailed. RESULTS One child was excluded from the VLBW cohort because of an anomaly syndrome (Goldenhar syndrome). Of the 103 children eligible for the study, six families withdrew before the child was 5 years old. Thus, the final VLBW group consisted of 97 children. However, six children could not be assessed with the WPPSI-R and the NEPSY II because of severe developmental problems, and five children failed to participate in the assessments within 2 months of their fifth birthday. Of the 169 comparison children whose families agreed to participate in the study, eight families withdrew before the child was 5 years old. Thus, the comparison group consisted of 161 children. Of these, 12 children failed to participate in the assessment at 5 years of age. Fifteen (15%) VLBW children had neurodevelopmental impairments: seven children had isolated CP, two had an isolated hearing impairment (defined as the need for a hearing aid), two of the children had a cognitive impairment (IQ <70; -2SD), one had both CP and a hearing impairment, one had both CP and a cognitive impairment, and two had both CP and an intellectual impairment. None of the children was blind. Of these 15 children with neurodevelopmental impairments, five could not be assessed with the WPPSI-R and NEPSY II and were excluded from the analyses (one had a hearing impairment, one had an intellectual impairment, one had CP and a hearing impairment, and two had CP and an intellectual impairment). In addition, one child could not be assessed with NEPSY II owing to severe attention deficits. The background variables are presented in Table I. The WPPSI-R and NEPSY II results are presented in Tables II and III. Analyses were performed using complete cases only. The number of IQs NEPSY subtests missing ranged from 0 to 6. The mean IQs of the VLBW group corresponded to the average for Finnish children according to the test norms, but were significantly lower than the comparison group mean. Scores for all the NEPSY II subtests except the narrative memory subtest were significantly lower in the VLBW group. Twenty-one (25%) children in the VLBW group had a standard score below 6 in one NEPSY II subtest, and 28 (33%) had a score below 6 in more than one. Twenty-seven (19%) of the comparison children had a standard score below 6 in one subtest and 15 (10%) had a standard score below 6 in more than one. When the WPPSI-R and NEPSY II test scores were compared between children with a VLBW without CP, a hearing impairment, or intellectual impairment and the comparison group, the differences remained significant except those regarding the subtests comprehension of instructions (p=0.071; VLBW group: mean 9.5; SD 2.8; comparison group: mean 10.2; SD 2.6) and visuomotor precision (p=0.069; VLBW group: mean 9.1; SD 3.2; comparison group: mean Table I: Background variables of the very-low-birthweight (VLBW) children and the comparison children VLBW children Comparison children Mother s education (y), mean (SD) 14.6 (3.2) 15.1 (2.6) Prenatal corticosteroids Yes, n (%) 91 (94) No, n (%) 6 (6) Gestational age (weeks+days), 28+2 (2+5) 40+0 (1+1) mean (SD) Sex Male, n (%) 50(52) 80(50) Female, n (%) 47(48) 81(50) Chronic lung disease, sepsis or meningitis and or intestinal perforation Yes, n (%) 40 (41) No, n (%) 57 (59) Brain pathology, cranial ultrasound Normal, n (%) 45 (46) Mildly abnormal, n (%) 43 (44) Severely abnormal, n (%) 9 (9) Brain pathology, magnetic resonance imaging Normal, n (%) 52 (54) Minor brain abnormality, n (%) 20 (21) Major brain abnormality, n (%) 24 (25) 258 Developmental Medicine & Child Neurology 2011, 53:

4 Table II: Test scores of the very-low-birthweight (VLBW) and comparison groups and the differences between them Test VLBW group, mean (SD) Comparison group, mean (SD) Difference, mean (95% CI) p-value WPPSI-R FSIQ (16.0) (14.8) 11.7 ( ) <0.001 VIQ (15.0) (13.9) 5.7 ( ) PIQ 98.2 (16.3) (13.7) 12.7 ( ) <0.001 NEPSY II Auditory attention 8.4 (3.0) 10.0 (2.6) 1.5 ( ) <0.001 Visual attention 9.2 (1.9) 10.0 (2.6) 0.8 ( ) Inhibition 8.3 (3.3) 9.9 (3.0) 1.5 ( ) <0.001 Narrative memory 9.8 (2.5) 10.0 (2.7) 0.3 () ) Memory for designs 8.3 (3.3) 10.0 (2.7) 1.7 ( ) <0.001 Word list interference 8.4 (4.0) 10.1 (2.8) 1.7 ( ) <0.001 Visuomotor precision 8.9 (3.3) 10.0 (3.4) 1.0 ( ) Design copy 7.6 (3.2) 9.8 (3.0) 2.2 ( ) <0.001 Speeded naming 9.0 (3.1) 9.8 (3.0) 0.8 ( ) Comprehension of instructions 9.3 (2.8) 10.2 (2.6) 0.8 ( ) Phonological processing 8.7 (2.3) 10.0 (2.6) 1.3 ( ) <0.001 CI, confidence interval; WPPSI-R, Wechsler Preschool and Primary Scale of Intelligence revised; FSIQ, Full-scale intelligence quotient; VIQ, Verbal intelligence quotient; PIQ, Performance intelligence quotient. Table III: Percentages of the very-low-birthweight (VLBW) and comparison children with an IQ between 70 and 84 standard score between 6 and 7, and with an IQ below 70 standard score below 6 VLBW group (%) Comparison group (%) <70 < <70 <6 WPPSI-R FSIQ VIQ PIQ NEPSY II Auditory attention Visual attention Inhibition Narrative memory Memory for designs Word list interference Visuomotor precision Design copy Speeded naming Comprehension of instructions Phonological processing WPPSI-R, Wechsler Preschool and Primary Scale of Intelligence revised; FSIQ, Full-scale intelligence quotient; VIQ, Verbal intelligence quotient; PIQ, Performance intelligence quotient. 10.0; SD 3.4). The only sex differences occurred in auditory attention (p=0.011; females: mean9.4; SD3.0; males: mean 7.5; SD 3.5) and visuomotor precision (p=0.023; females: mean 9.7; SD 3.3; males: mean 8.1; SD 3.1). Regarding intelligence, the hierarchical regression analysis showed that neonatal conditions (stage 4) were significantly associated both with FSIQ (p=0.005; DR 2 =0.21) and PIQ (p=0.006; DR 2 =0.20). The length of maternal education (stage 1) was significantly associated with VIQ (p=0.011; DR 2 =0.08). A 1-year increase in maternal education was associated with an increase in VIQ (p=0.001; b=1.33; 95% CI ). Regarding NEPSY II subtests, neonatal conditions (stage 4) were significantly associated with narrative memory (p=0.044; DR 2 =0.16) and design copy (p=0.044; DR 2 =0.15). Children with major pathology on MRI had lower scores in narrative memory (p=0.029; b=)1.89; 95% CI )3.59 to )0.20). In addition, gestational age and sex (stage 3) were significantly associated with the visuomotor precision subtest scores (p=0.007; DR 2 =0.13). A 1-day increase in gestational age was associated with increased scores (p=0.026; b=0.05; 95% CI ) and females scored better than males (p=0.045; b=1.55; 95% CI ). Finally, maternal education (stage 1) was significantly associated with the speeded naming subtest score (p=0.016; DR 2 =0.08). A 1-year increase in maternal education was significantly associated with higher scores (p=0.016; b=0.26; 95% CI ). The results of the regression analysis are presented in more detail in the Appendix SI (published online). The distributions of VIQ, PIQ, and FSIQ are shown in Fig. 1 and the distribution of scores in the neuropsychological domains is shown in Fig. 2.Pretermchildrenscoredlower than term children, but the scores were not systematically lowest in the children born at the lowest gestational weeks there were some trends towards lower scores in more mature but VLBW infants. All nine children in the group born beyond gestational week 31 were small for gestational age (birthweight 2SDs or more below the mean according to age- and sex-specificfinnishgrowthcharts).thesixchildrenwhocouldnot be assessed with the WPPSI-R and NEPSY II because of severe developmental impairments were born at 23, 23, 24, 26, 27, and 28 weeks of gestation. DISCUSSION To our knowledge, this is the first report on the cognitive and neuropsychological 5-year outcomes of preterm children born in the 2000s. The children in the VLBW group had an average cognitive capacity, even though it was lower than that of Preterm Children Born in the 2000s Annika Lind et al. 259

5 VIQ PIQ FSIQ Figure 1: Distribution of Verbal intelligence quotient (VIQ), Performance intelligence quotient (PIQ), and Full-scale intelligence quotient (FSIQ) in preterm children (n=86) born at gestational weeks 23 to 24 (n=6), 25 to 27 (n=32),28to31(n=39), and above 31 (n=9), and in control children born at term (n=146). The mean value is marked by +, the line inside the box represents the median, and the box represents the quartiles. the comparison group of healthy, term children. However, all the neuropsychological functions except verbal memory were significantly poorer in the VLBW group than in the comparison group. Thirty-three per cent of the children in the VLBW group had a significantly impaired performance in more than one subtest that assessed neuropsychological functions compared with 10% of the comparison group. Unexpectedly, IQs or scores in neuropsychological tests were not systematically worse in children born at a lower gestational age. In the EPIPAGE study, 21 68% of 5-year-old children achieved normal intelligence test scores (above 84). In our study, as many as 85% of the VLBW group achieved scores above 84. The children in the EPIPAGE study were born in 1997, and the children in our study between 2001 and 2003, which indicates an improved cognitive outcome for preterm children born in the 2000s. However, the preterm populations were not identical; for example, the EPIPAGE study included children born before gestational week 33, whereas our study used a birthweight limit (1501g). Furthermore, the assessment methods were different. However, indications of improved outcome can also be seen when our results are compared with those of VLBW Swedish children born between 1988 and 1993 and tested at 5 years 6 months of age with the Swedish versionoftheintelligencetestweused. 8 In that study, the mean FSIQ of the children at 5 years 6 months chronological age was 91.1, the mean VIQ was 96.5, and the mean PIQ was In preterm children, PIQ, representing difficulties with, for example, visual perception and eye hand coordination, appears to be more adversely affected than VIQ, 8,12 and visuomotor and visuospatial functions appear to be more adversely affected than other neuropsychological functions when several neuropsychological domains have been studied simultaneously. 12 Similarly, in our study, the difference in PIQ between the VLBW and comparison groups was larger than the difference in VIQ. Concordantly, the NEPSY II subtest assessing visuospatial functioning, showed the largest difference between the groups. Our study shows that assessment of neuropsychological functions provides significant additional information compared with using only cognitive-level assessment in the follow-up of children with a VLBW. As the mean FSIQ of our VLBW group was the same as that in Finnish WPPSI-R norms, 17 85% of the tested children would have been classified as normal if only intelligence had been assessed. Despite this, the VLBW group had poorer neuropsychological functions, with the exception of verbal memory, than the comparison group. Regarding working memory, as many as 44% of the children had scores below the average range. Performance was significantly impaired in more than one subtest assessing neuro- 260 Developmental Medicine & Child Neurology 2011, 53:

6 Executive functioning Memory Motor skills Language Figure 2: Distribution of scores in the neuropsychological domains in preterm children (n=86) born at gestational weeks 23 to 24 (n=6), 25 to 27 (n=32), 28 to 31 (n=39), and above 31 (n=9), and in control children born at term (n=146). The scores are standard scores in the domains in NEPSY II. The mean value is marked by +, the line inside the box represents the median, and the box represents the quartiles. psychological functions in 33% of the preterm children. High scores in the comparison group may be explained by the selection of the comparison group because, for example, all infants who were small for gestational age or who were admitted to the neonatal intensive care unit (over 10% of the newborn population) were excluded. Another explanation is that the WPPSI-R norms are 14 years old, and it is known that IQ tends to increase over time (Flynn effect). 22 Several studies have demonstrated that a lower gestational age is associated with poorer developmental outcome. 11,21 In our study, the effect of gestational age was not consistent, although a higher gestational age was associated with better visuomotor skills. In fact, the mean score in the domains executive functioning and memory was higher for the children born at 23 to 24 weeks of gestation than for infants who were small for gestational age born at 32 weeks or later. Therefore, growth retardation seems to be at least an equal risk factor for certain neuropsychological domains than an extremely preterm birth. Because of the small group sizes in each category,theconclusionsfromthisanalysismustbedrawnwith caution. Additionally, three of the six children who could not be assessed with WPPSI-R and NEPSY II because of severe developmental impairments were born at 23 to 24 weeks of gestation. Associations between parental education and development have been described previously, and it has even been suggested that children born to mothers with low education might benefit from early intervention. 23 The impact of sex 24 and neonatal illnesses 12 on the developmental outcome is also apparent from previous studies. Furthermore, severe brain pathology in MRI was associated with poor verbal memory. A limitation of this study is the exclusion of the most impaired children from the analyses of cognitive level and neuropsychological functioning. Children with severe disabilities cannot be appropriately examined with the tests used in this study, which might have skewed our results. An additional limitation is the large number of statistical analyses performed, which might result in significant associations emerging by chance. As there were a large number of results, we may have put too much emphasis on the p-values in the selection of the presented results in order to increase the readability. A further possible constraint is the relatively small population size. However, there are no other studies available yet on the long-term cognitive and neuropsychological outcomes of preterm children born during the 2000s. In conclusion, this first report on the long-term cognitive and neuropsychological outcomes of children with VLBW born in the 2000s gives a new, more positive perspective on the outcome of preterm survivors. The cognitive outcome of preterm children, as well as the outcome of children with the lowest gestational age, seems to have improved in this Preterm Children Born in the 2000s Annika Lind et al. 261

7 decade, when compared with earlier published studies, but are still poorer than the cognitive outcome of low-risk term children. ACKNOWLEDGEMENTS The PIPARI study group includes Satu Ekblad; Eeva Ekholm; Leena Haataja; Mira Huhtala; Pentti Kero; Riikka Korja; Harry Kujari; Helena Lapinleimu; Liisa Lehtonen; Hanna Manninen; Jaakko Matomäki; Jonna Maunu; Petriina Munck; Pekka Niemi; Pertti Palo; Riitta Parkkola; Jorma Piha; Annika Lind; Marika Leppänen; Liisi Rautava; Päivi Rautava; Milla Reiman; Hellevi Rikalainen; Katriina Saarinen; Elina Savonlahti; Matti Sillanpää; Suvi Stolt; Päivi Tuomikoski- Koiranen; Anniina Väliaho; Tuula Äärimaa. This study was supported by grants from the Foundation for Paediatric Research The South- Western Finnish Foundation of Neonatal Research and the Signe and Ane Gyllenberg Foundation. ONLINE SUPPLEMENTARY INFORMATION Additional material and supporting information may be found in the online version of this article. REFERENCES 1. Tommiska V, Heinonen K, Lehtonen L, et al. No improvement in outcome of nationwide extremely low birth weight infant populations between and Pediatrics 2007; 119: Vohr BR, Wright LL, Dusick AM, et al. Center differences and outcomes of extremely low birth weight infants. Pediatrics 2004; 113: Rautava L, Lehtonen L, Peltola M, et al. The effect of birth in secondary- or tertiary-level hospitals in finland on mortality in very preterm infants: a birth-register study. Pediatrics 2007; 119: e Wilson-Costello D, Friedman H, Minich N, et al. Improved neurodevelopmental outcomes for extremely low birth weight infants in Pediatrics 2007; 119: Vohr BR, Wright LL, Poole WK, McDonald SA. Neurodevelopmental outcomes of extremely low birth weight infants <32 weeks gestation between 1993 and Pediatrics 2005; 116: Robertson CM, Watt MJ, Yasui Y. Changes in the prevalence of cerebral palsy for children born very prematurely within a population-based program over 30 years. JAMA 2007; 297: Bode MM, D Eugenio DB, Forsyth N, Coleman J, Gross CR, Gross SJ. Outcome of extreme prematurity: a prospective comparison of 2 regional cohorts born 20 years apart. Pediatrics 2009; 124: Bohm B, Katz-Salamon M, Institute K, Smedler AC, Lagercrantz H, Forssberg H. Developmental risks and protective factors for influencing cognitive outcome at years of age in very-low-birthweight children. Dev Med Child Neurol 2002; 44: Caravale B, Tozzi C, Albino G, Vicari S. Cognitive development in low risk preterm infants at 3 4 years of life. Arch Dis Child Fetal Neonatal Ed 2005; 90: F Bayless S, Stevenson J. Executive functions in school-age children born very prematurely. Early Hum Dev 2007; 83: Foulder-Hughes LA, Cooke RW. Motor, cognitive, and behavioural disorders in children born very preterm. Dev Med Child Neurol 2003; 45: Mikkola K, Ritari N, Tommiska V, et al. Neurodevelopmental outcome at 5 years of age of a national cohort of extremely low birth weight infants who were born in Pediatrics 2005; 116: Hoff Esbjorn B, Hansen BM, Greisen G, Mortensen EL. Intellectual development in a Danish cohort of prematurely born preschool children: specific or general difficulties? J Dev Behav Pediatr 2006; 27: Reiman M, Kujari H, Maunu J, et al. Does placental inflammation relate to brain lesions and volume in preterm infants? J Pediatr 2008; 152: Rademaker KJ, Uiterwaal CS, Beek FJ, et al. Neonatal cranial ultrasound versus MRI and neurodevelopmental outcome at school age in children born preterm. Arch Dis Child Fetal Neonatal Ed 2005; 90: F Maunu J, Kirjavainen J, Korja R, et al. Relation of prematurity and brain injury to crying behavior in infancy. Pediatrics 2006; 118: e Wechsler D. Wechslerin älykkyystestistö esikouluikäisille, käsikirja. Helsinki: Psykologien Kustannus, Korkman M, Kirk U, Kemp SL. NEPSY II, 2nd edn. San Antonio, TX: PsychCorp Pearson Clinical Assessment, Korkman M, Kirk U, Kemp SL. NEPSY II. Helsinki: Psykologien Kustannus, Lind A, Haataja L, Rautava L, et al. Relations between brain volumes, neuropsychological assessment and parental questionnaire in prematurely born children. Eur Child Adolesc Psychiatry 2010; 19: Larroque B, Ancel PY, Marret S, et al. Neurodevelopmental disabilities and special care of 5-year-old children born before 33 weeks of gestation (the EPIPAGE study): a longitudinal cohort study. Lancet 2008; 371: Hiscock M. The Flynn effect and its relevance to neuropsychology. J Clin Exp Neuropsychol 2007; 29: Wang LW, Wang ST, Huang CC. Preterm infants of educated mothers have better outcome. Acta Paediatr 2008; 97: Wolke D, Samara M, Bracewell M, Marlow N, EPICure study group. Specific language difficulties and school achievement in children born at 25 weeks of gestation or less. J Pediatr 2008; 152: Developmental Medicine & Child Neurology 2011, 53: