- Email: [email protected]
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants Isabelle Guellec, MD1,2, Stephane Marret, MD, PhD3,4, Olivier Baud, MD, PhD5,6, Gilles Cambonie, MD, PhD7, Alexandre Lapillonne, MD, PhD8, Jean-Christophe Roze, MD, PhD9, Jeanne Fresson, MD1,10, Cyril Flamant, MD, PhD9, Marie-Laure Charkaluk, MD, PhD1, Catherine Arnaud, MD10, and Pierre-Yves Ancel, MD, PhD1,11 Objectives To determine whether small head circumference (HC) or birth weight (BW) or both are associated with neonatal and long-term neurologic outcome in very preterm infants. Study design All 2442 live births from the 1997 Epipage study between 26 and 32 weeks of gestational age in 9 regions of France were analyzed. A total of 1395 were tested at age 5 years for cognitive performance and 1315 with school performance reports at age 8 years. Symmetric growth restriction (SGR) was defined by HC and BW <20th percentile and in the same percentile range, and asymmetric growth restriction by at least 1 of HC and BW <20th percentile and the other in a higher decile range. There were 2 forms of asymmetric growth restriction: head growth restriction (HGR) and weight growth restriction (WGR). Appropriate for gestational age was defined by both BW and HC >20th percentile. Results Compared with appropriate for gestational age, SGR was significantly associated with neonatal mortality (aOR 2.99, 95% CI 1.78-5.03), moderate and severe cognitive deficiency (aOR 1.65, 95% CI 1.01-2.71 and aOR 2.61, 95% CI 1.46-4.68, respectively), and poor school performance (aOR 1.79; 95% CI 1.13-2.83). HGR was significantly associated with severe cognitive deficiency (aOR 2.07, 95% CI 1.15-3.74). WGR was not significantly associated with cognitive or school performance despite higher rates of neonatal morbidity. Conclusions SGR in preterm infants was associated with neonatal mortality and impaired cognitive and school performance. The outcome of asymmetric growth restriction differed according to HC. HGR was associated with impaired cognitive function; WGR was not. (J Pediatr 2015;-:---).
C
hanges in perinatal management have been associated with a substantial increase in the survival of infants at very low gestational ages, an increase that raises questions about their long-term neurologic outcomes.1 Preterm growthrestricted infants are a population of particular interest because they combine immaturity secondary to low gestational age with the consequences of growth restriction. Growth restriction remains a concept difficult to study especially in preterm infants. It intermixes mechanisms including placental insufficiency, with or without brain sparing,2 congenital abnormalities, toxic, environmental,3 and maternal diseases.4,5 It is difficult to distinguish their consequences from the separate consequences of immaturity. Because these mechanisms might affect various anthropometric measurements, such as head circumference (HC) and birth et de la Recherche From the Institut National de la Sante dicale (INSERM), UMR 1153, Equipe de recherche en Me weight (BW) differently, these measurements might also be associated differently miologie Obste tricale, Pe rinatale et Pe diatrique; Epide with specific outcomes. ^ pitaux de Paris, Department of Assistance Publique - Ho Neonatology and Intensive Care, A. Trousseau Hospital, Previous studies have demonstrated that small for gestational age birth is assoParis; Rouen University Hospital, Neonatal Medicine; 6-9 Institute of Biomedical Research, University, Inserm ciated with a high mortality rate and impaired cognitive development. Avenir Research Group, IFR 23, Rouen; INSERM UMR ^ pitaux de Paris, Although growth restriction is a dynamic process, it is commonly defined by a 1141; Assistance Publique - Ho Hospital, Paris; Department of Neonatology, R. Debre BW <10th percentile. Moreover, preterm growth reference curves underestimate Montpellier University Hospital Center, Neonatal and 10 Pediatric Intensive Care Unit, Montpellier; Assistance growth restriction. Growth restriction because of placental insufficiency ^ pitaux de Paris, Department of Publique - Ho Neonatology, Necker Hospital, Paris; Nantes University, should, thus, be studied as a dynamic process that reduces the fetus’s growth caDepartment of Neonatology, Maternite Regionale, pacity secondary to the failure of compensation mechanisms or the severity of the Clinical Epidemiology and Biostatistics Department France, Nantes University, INSERM CIC004; Clinical illness.11 It is thought that growth-restricted fetuses attempt to compensate for Epidemiology and Biostatistics Department, CHRU miologie et Nancy; and Centre de Recherche Epide the “substrate limitation associated with placental insufficiency by preferentially 1
2
3
4
5
6
7
8
9
10
11
Statistique Sorbonne Paris Cite DHU Risques et Paris Descartes, Paris, France grossesse Universite
AGA BW HC HGR SGR WGR
Appropriate for gestational age Birth weight Head circumference Head growth restriction Symmetric growth restriction Weight growth restriction
et de la ReFunded by Institut National de la Sante dicale (INSERM), Directorate General for cherche Me Health of the Ministry for Social Affairs, and “Hospital Program for Clinical Research 2001” (AOMO1117) of the French Department of Health. The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2015.08.025
1
THE JOURNAL OF PEDIATRICS
www.jpeds.com
perfusing the central nervous system.”12,13 If so, development would be limited first by a reduction in BW and only thereafter by smaller head size. If growth restriction occurs early in pregnancy or if no adaptive mechanism protects the fetus, growth restriction might be symmetric. If it occurs later, however, or is accompanied by adaptive phenomena, it might result in asymmetric growth restriction involving BW only. If this hypothesis is correct, head growth restriction (HGR) appears to be due to another mechanism and may be associated with a different outcome. Accordingly, we analyzed growth restriction in its different clinical forms by examining 2 specific anthropometric measurements (BW and HC) and their relations to 3 different outcomes in very preterm children: short-term mortality and morbidity and long-term neurodevelopmental outcomes.
Methods Our data come from the 1997 Epipage cohort study, which included all live births between 22 and 32 weeks of gestation in 1997 in 9 regions on France.14 Because 65% of those born at 22-25 weeks died before discharge, we limited our analysis to children born alive at 26-32 weeks (n = 2694) for whom HC and BW were available (n = 2442, 90.6%). Two regions with large samples included only 1 of every 2 infants for follow-up (70 infants not included). Parent refusal resulted in exclusion of 89 infants from follow-up. At 5 years of age, 1648 (80%) had medical examinations and 1395 (68%) cognitive assessments, all by trained physicians and psychologists. The parents of 1520 children (74%) also completed questionnaires. Around the children’s eighth birthdays, parents received a questionnaire about their school performance. In all, 1315 (64%) responses were available for analysis (Figure 1; available at www.jpeds.com). The Commission Nationale de l’Informatique et des Libertes (French Data Protection Authority) approved the study. Parents provided verbal consent. Ethics committee approval was not required because this was an observational study of usual care, with no intervention. Gestational age was defined by completed weeks of gestation, determined from the date of the last menstrual period and early ultrasound findings. Maternal and obstetric data were recorded on standardized questionnaires at birth in each maternity unit. Maternal data included nationality, age at birth, and parity. Family socioeconomic status was recorded according to the French classification of occupations and social position, defined by the higher-status parental occupation (or the mother’s, if she did not live with the father). Obstetric data included type of pregnancy (singleton or higher-order) and antenatal corticoid use. Infant Characteristics and Neonatal and Long-Term Outcomes Neonatal data were prospectively collected at each hospital. This study considers sex, BW, and HC, measured by the 2
Vol. -, No. maximum occipital-frontal HC at birth. Congenital abnormalities were also recorded. In-hospital mortality was defined as death in the delivery room or neonatal unit. Length of hospitalization was defined as the number of days until discharge home. Bronchopulmonary dysplasia was defined as oxygen dependency at 28 days. Duration of central line was defined as the total number of days during which a central line was maintained, regardless of the reason. Intraventricular hemorrhage and white matter damage were diagnosed from cranial ultrasonography, performed by qualified neonatologists or radiologists. Major brain lesions included intraventricular hemorrhage with ventricular dilatation (grade III) or intraparenchymal hemorrhage (grade IV), according to the Papile classification, cystic periventricular leukomalacia, or hyperechogenicity persisting more than 14 days without cystic formation.15 We used the European definition to define cerebral palsy, which requires at least 2 of the following: abnormal posture or movement, increased tone and hyperreflexia (spastic cerebral palsy), involuntary movements (dyskinetic cerebral palsy), or absence of coordination (ataxic cerebral palsy).16 The French version of the Kauffman assessment battery for children was used to assess cognitive function, expressed as a mental processing composite score (IQ equivalent), standardized with a mean of 100 and a SD of 15 in a French population born in the 1990s.17 Moderate cognitive deficiency was defined by a score between 70 and 84, and severe cognitive deficiency <70. Behavioral problems were assessed by the French versions of the strength and difficulties questionnaire,18 completed by parents. It includes 4 scales (inattention-hyperactivity, conduct, emotional, and peer problems) that were added together for a total behavioral difficulties score. The cut-off was defined by the 90th percentile of the scores observed in the reference group of term infants included in Epipage. School difficulties were assessed at age 8 years based on a parental questionnaire. Special schooling (institution, special school and special class in mainstream school, compared with mainstream class) or low grades were considered school difficulties.19 Growth Restriction Percentiles of BW and HC were determined by gestational age and sex from the data of this cohort of very preterm births (Figure 2). Our population was divided into 4 different categories according to the percentile of their HC and BW. Symmetric growth restriction (SGR) was defined by BW and HC percentiles both <10th percentile or both between 10th and 19th percentile. Two different types of asymmetric growth restriction were defined. HGR was defined by a HC <20th percentile, with BW in at least the next higher decile group, and weight growth restriction (WGR) by a BW <20th percentile, with HC in at least the next higher decile group. The group without growth restriction, that is, appropriate for gestational age (AGA), was defined by BW and HC both above the 20th percentile. Guellec et al
- 2015
ORIGINAL ARTICLES
Figure 2. Description of the population.
Statistical Analyses We studied the relations between maternal and obstetric characteristics and growth restriction at birth. We examined these different forms of growth restriction at birth in relation to neonatal and long-term outcomes. c2 tests were used, and a P value of <.05 was considered significant. We used the ANOVA test for continuous variables, with the same P value defining significance. Logistic regression models were used to analyze associations between growth restriction and each binary outcome (ie, mortality, major brain lesions, cerebral palsy, behavioral difficulties, and school difficulties). Multinomial logistic regression models were used for cognitive outcomes (ie, no cognitive deficiency, moderate cognitive deficiency, and severe cognitive deficiency). Associations were quantified by ORs and their 95% CIs. Covariates were included in logistic regression models if they were known risk factors and associated with the outcome with a P value of <.2 in the univariate analysis. All analyses were adjusted for gestational age and sex.6 For in-hospital mortality, factors included in the final model were antenatal corticosteroids (P < .01) and family social class (P = .06); for major brain abnormalities, nationality (P = .12), and antenatal corticosteroids (P < .01); and for cerebral palsy, the type of pregnancy (singleton vs higher-order) (P = .14). Finally, the same factors were included in the models for both behavioral problems and school difficulties: family social class (P < .01 for both), maternal age (P < .01 for both), parity (P = .05 and <.01, respectively), and type of pregnancy (P = .03 and .08, respectively). Associations between growth restriction and cognitive outcomes were studied after adjustment for so-
cial class (P < .01), age (P = .01), nationality (P = .01), and parity (P < .01). All analyses were performed with Stata 10 SE (StatCorp, College Station, Texas).
Results First, we compared the characteristics of children with missing data for BW, sex, or HC with children with complete data. They were more frequently male (P = .023), were born earlier (29 vs 30 weeks gestational age), and died more often in the hospital (40% vs 9%, P < .001) especially in the delivery room (18.7% vs 0%). Long-term outcomes did not differ between these 2 groups. Mean mental processing composite score was 94.2 for missing data group vs 93.7 for complete data (P = .827), cerebral palsy occurred in 7.7% vs 8.5% of these groups (P = 1.000), and total behavioral difficulties percentage were similar (20.8% vs 21.1%, P = .000). School difficulties also were similar (25.6% vs 22.6%, P = .578). Figure 2 summarizes the number of total, SGR, HGR, WGR, and AGA births by decile range of BW and HC. Of the 2442 live births available for analysis, 213 (8.7%) were SGR, 209 (8.6%) HGR, 201 (8.2%) WGR, and 1819 (74.5%) AGA. In the SGR group, 63% had both BW and HC <10th percentile, and 37% between the 10th and 19th percentiles. In the WGR group, 43% had a BW <10th percentile, and 57% between 10th and 19th percentiles. In the HGR group, 34% had a HC <10th percentile, and 66% between the 10th and 19th percentiles. Maternal and obstetric characteristics are reported by growth restriction (Table I; available at www.jpeds.com). Children of nulliparous women were growth-restricted
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
3
THE JOURNAL OF PEDIATRICS
Vol. -, No. -
www.jpeds.com
Table II. Neonatal outcome according to growth restriction category In hospital mortality n/N (%) OR [95% CI] aOR* [95% CI] Major brain pathology n/N (%) OR [95% CI] OR† [95% CI] Bronchopulmonary dysplasia n/N (%) OR [95% CI] ORz [95% CI]
SGR
HGR
WGR
AGA
P
30/213 (14.1%) 2.04 [1.34-3.12] 2.99 [1.78-5.03]
21/209 (10.0%) 1.39 [0.86-2.26] 1.68 [0.93-3.02]
21/201 (10.5%) 1.46 [0.90-2.36] 1.39 [0.74-2.63]
135/1819 (7.4%) 1 1
.006
41/211 (19.4%) 0.86 [0.60-1.23] 0.83 [0.55-1.26]
40/203 (19.7%) 0.88 [0.61-1.26] 1.04 [0.70-1.54]
42/196 (21.4%) 0.97 [0.68-1.39] 1.06 [0.72-1.57]
388/1773 (21.9%) 1 1
.807
42/206 (20.4%) 3.09 [2.11-4.52] 3.74 [2.48-6.65]
28/203 (13.8%) 1.93 [1.25-2.98] 2.09 [1.32-3.31]
33/194 (17.0%) 2.42 [1.63-3.74] 2.70 [1.74-4.20]
135/1763 (7.7%) 1 1
<.001
*Adjusted for gestational age, sex, family social class, and antenatal corticosteroids. †Adjusted for gestational age, sex, mother’s nationality, and antenatal corticosteroids. zAdjusted for gestational age, sex, mother’s age, type of pregnancy, and antenatal corticosteroids.
more frequently than those of multiparous women (P = .032). The SGR group received antenatal corticosteroids significantly more frequently than the others (P = .008). Less than 10% of the population had congenital abnormalities, a proportion that did not differ according to growth restriction (P = .302). Neonatal mortality and morbidity were significantly associated with growth restriction categories (Table II). Inhospital mortality was significantly higher in the SGR (aOR 2.99, 95% CI 1.78-5.03) than the AGA group. Hospitalization time increased significantly (P < .01) from 54 (54) days for AGA children to 65 (37), 71 (43), and 77 (0) days for HGR, WGR, and SGR children, respectively. Similarly, duration of central-line placement varied and was higher in the WGR (27 days [31]) and SGR groups (28 days [32]) than for HGR (21 days [19]), and AGA children (17 days [18]). The
bronchopulmonary dysplasia rate was higher in growthrestricted children than in the reference group. Major brain lesions did not vary by growth status. The prevalence of intellectual disability varied across growth restriction categories (Table III). SGR children had higher risks of both moderate (aOR 1.65, 95% CI 1.012.71) and severe (aOR 2.61, 95% CI 1.46-4.68) cognitive deficiency, and HGR children only of severe cognitive deficiency (aOR 2.07, 95% CI 1.15-3.74). Although WGR was initially associated with moderate cognitive deficiency (crude OR 1.66, 95% CI 1.08-2.56), this association was not significant after adjustment. The prevalence of cerebral palsy (P = .26) and behavioral problems (P = .24) did not vary significantly by growth-restriction categories before or after adjustment. School difficulties varied, although not significantly (P = .054): 21% in the AGA group, 24% in the WGR group,
Table III. Long-term outcome according to category of growth restriction Cerebral palsy* n/N (%) OR [95% CI] aOR* [95% CI] Moderate cognitive deficiency 70# MPC <85 n/N (%) OR [95% CI] aOR† [95% CI] Severe cognitive deficiency MPC <70 n/N (%) OR [95%CI] aOR† [95% CI] Behavioral problemsz n/N (%) OR [95%CI] aORz [95% CI] School difficultiesz n/N (%) OR [95% CI] aORz [95% CI]
SGR
HGR
WGR
AGA
P
8/133 (6.0%) 0.66 [0.30-1.31] 0.63 [0.30-1.34]
8/130 (5.2%) 0.64 [0.30-1.34] 0.63 [0.30-1.32]
8/140 (5.7%) 0.59 [0.28-1.23] 0.60 [0.28-1.26]
116/1245 (9.3%) 1 1
.262
27/112 (24.1%) 1.48 [0.92-2.38] 1.65 [1.01-2.71]
21/107 (19.6%) 1.10 [0.66-1.84] 1.20 [0.70-2.04]
35/126 (27.8%) 1.66 [1.08-2.56] 1.57 [0.99-2.47]
205/1050 (19.5%) 1 1
20/112 (17.9%) 1.99 [1.16-3.42] 2.61 [1.46-4.68]
18/107 (16.8%) 1.71 [0.98-2.99] 2.07 [1.15-3.74]
16/126 (12.7%) 1.38 [0.78-2.46] 1.61 [0.87-2.97]
113/1050 (10.8%) 1 1
28/121 (23.1%) 1.21 [0.77-1.89] 1.28 [0.81-2.03]
33/124 (26.6%) 1.46 [0.95-2.23] 1.42 [0.91-2.19]
31/131 (23.7%) 1.25 [0.81-1.91] 1.34 [0.86-2.10]
228/1144 (19.9%) 1 1
.242
35/115 (30.4%) 1.65 [1.08-2.52] 1.79 [1.13-2.83]
28/98 (28.6%) 1.51 [0.95-2.40] 1.48 [0.90-2.43]
26/110 (23.6%) 1.17 [0.73-1.86] 1.21 [0.74-1.99]
208/992 (21.0%) 1 1
.054
.021
MPC, mental processing composite. *Adjusted for gestational age, sex, and type of pregnancy. †Adjusted for gestational age, sex, family social class, mother’s age, nationality, and parity at birth. zAdjusted for gestational age, sex, family social class, mother’s age and parity, and type of pregnancy.
4
Guellec et al
- 2015
ORIGINAL ARTICLES
29% in the HGR group, and 30% in the SGR group (Table III). After adjustment, SGR children had a higher rate of school difficulties than AGA children (aOR 1.79, 95% CI 1.13-2.83).
Discussion We found that preterm children with SGR had the poorest outcomes in terms of mortality, neonatal morbidity, cognitive function, and school performance. The HGR group had risks of severe cognitive deficiency and school difficulties significantly higher than the AGA reference group. No similar associations were observed in children with WGR. Our study is based on data from the Epipage cohort, one of the largest population-based cohort studies of very preterm children. Its greatest strengths are its prospective design and long follow-up. Gestational age was assessed by early ultrasound scans. The geographical approach decreased the bias associated with studies in selected perinatal centers. We used validated tools to investigate motor deficiency (cerebral palsy), cognitive function (Kaufman assessment Battery for Children), and behavioral problems (Strengths and difficulties Questionnaire). The long-term follow-up at 8 years of age provided a reliable assessment of cognitive function. Missing data for BW or HC at birth prevented assessment of 9% of the cohort. Almost 40% died during the neonatal period. Thus, underestimation of the growth-restriction rate and its relation with in-hospital mortality cannot be excluded. In contrast, among survivors, outcomes did not differ between children with and without missing data. Thus, the impact of this selection on the associations between growth restriction and neurodevelopmental outcome is unclear. Loss to follow-up is an important issue in large prospective population-based cohort studies. Several studies have found that those lost to follow-up are more likely to have disabilities than those continuing to participate.20 In our study, children lost to follow-up more frequently had parents who were younger or had low socioeconomic status; both factors that influence cognitive and school outcomes (data not shown). We used an internal reference to define groups of growth restriction because of the absence of contemporary references for BW and HC growth in France for very preterm infants. Another difficulty is that we might have misclassified some preterm babies, especially in the SGR group. Nonetheless, if some of them were “constitutionally small for gestational age” instead of growth-restricted, this misclassification would have reduced rather than increased the association between SGR and cognitive disabilities. WGR was associated with neonatal complications (mortality, longer need for central line, length of hospitalization, and bronchopulmonary dysplasia), but not, or at least to a lesser extent, with long-term cognitive performance. Interestingly, despite these high rates of neonatal morbidities, children with normal HC had better cognitive outcomes than those with SGR or HGR. These results are consistent with previous
studies reporting that growth restriction defined by BW only is associated with a higher risk of mortality21,22 and of bronchopulmonary dysplasia.23 This difference may help explain the conflict in the literature about the effects of growth restriction on BW and cognitive performance24-27 in articles that do not distinguish between growth restriction with and without HC involvement. Another important result of our study is that HC was strongly related to cognitive function. Both SGR and HGR children were at risk of cognitive impairment and school difficulties. Mechanisms of HGR differ from those for WGR. HC appears to be a risk marker of future adverse neurologic outcome. We hypothesize that SGR involves exhaustion of the compensation enabled by the brainsparing response to placental insufficiency, and HGR might result from antenatal exposure to infections, drugs, chemicals, or high levels of ionizing radiation. All of these events are likely to be underdiagnosed. McElrath et al13 showed that fetal growth restriction and preeclampsia correlate strongly with microcephaly.13 Nonetheless, head size at birth is poorly related to intelligence,28,29 even though HC is related to brain volume.30 Businelli et al31 reported accelerated cortical development in growth-restricted fetuses, despite smaller brain volume, and Bocca-Tjeertes et al32 found developmental delay independent of HC at birth. Mechanisms of cognitive delay have not yet been completely elucidated. Fetal growth restriction may cause irreversible damage to the developing brain by disrupting gray matter. Tolsa et al33 used advanced magnetic resonance imaging techniques to show that children with growth restriction have significantly reduced intracranial volume and cortical gray and hippocampus matter compared with a matched control group. Disrupted network features have been observed in human babies with intrauterine growth restriction.34 Impairments of synaptic development and white matter have also been observed in animal models of intrauterine growth restriction.35 In our study, we observed that a not insignificant percentage of the Epipage population (8.2%) had a small HC without global growth restriction, associated with severe cognitive deficiency. We can hypothesize that growth restriction was limited to head growth and to a decrease in the number of endothelial and neural cells or connections between them, although adequate nutrition prevented underweight. In these cases, we suggest that maternal and/ or placental factors have specific deleterious effects on forebrain development. For example, maternal depression, or alcohol or cocaine use during pregnancy might act specifically on cerebral angiogenesis, neurotransmitter release, and synaptogenesis.36,37 Moreover some authors have described genetic particularities, such as the 1245 A-allele of the IGF1 promoter SNP, which is associated with a small head size and a reduced brain-sparing effect in small for gestational age infants.38 Even though we did not find more congenital abnormalities or chromosomal aberrations in our populations, genetics may be one path toward understanding cognitive deficiency.
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
5
THE JOURNAL OF PEDIATRICS
Vol. -, No. -
www.jpeds.com
In our population, 40%-66% of the preterm growthrestricted children were between the 10th and 19th percentiles. They are usually not recognized as growth-restricted because preterm children are smaller than their in-utero peers.10 The true proportion of growth-restricted children is, thus, underestimated when the 10th percentile of neonatal growth curves is used to define growth restriction.39 Accordingly, we found significant associations between neonatal mortality, various forms of neonatal morbidity (eg, bronchopulmonary dysplasia), cognitive functions, and growth restriction categories, including moderate growth restriction. Many studies do not consider these children to be at risk. Moreover, moderate HGR (ie, 10th-19th percentile) is not considered at risk, especially when the BW is normal. Our results provide support for the view that both low BW and HC are continuums, rather than the discrete categories. In very preterm babies, small head size at birth, either as part of SGR or isolated, is associated with long-term neurologic deficiencies, and isolated WGR is associated only with poorer neonatal outcome but not with long-term neurologic deficiencies. These factors aggravate the prognosis of preterm birth and may be useful for parental counseling and follow-up. n
10.
11.
12. 13.
14.
15.
16.
17. 18.
Submitted for publication May 22, 2015; last revision received Jul 10, 2015; accepted Aug 6, 2015.
19.
Reprint requests: Isabelle Guellec, MD, service de reanimation neonatale et diatrique, ho ^ pital Trousseau, APHP, boulevard A.Netter, 75012 Paris, pe France. E-mail: [email protected]
20.
References 1. Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet 2008;371:261-9. 2. Flood K, Unterscheider J, Daly S, Geary MP, Kennelly MM, McAuliffe FM, et al. The role of brain sparing in the prediction of adverse outcomes in intrauterine growth restriction: results of the multicenter PORTO Study. Am J Obstet Gynecol 2014;211:288.e1-5. 3. Hannam K, McNamee R, Baker P, Sibley C, Agius R. Air pollution exposure and adverse pregnancy outcomes in a large UK birth cohort: use of a novel spatio-temporal modelling technique. Scand J Work Environ Health 2014;40:518-30. 4. Nilsson PM, Li X, Sundquist J, Sundquist K. Maternal cardiovascular disease risk in relation to the number of offspring born small for gestational age: national, multi-generational study of 2.7 million births. Acta Paediatr 2009;98:985-9. 5. Boggess KA, Beck JD, Murtha AP, Moss K, Offenbacher S. Maternal periodontal disease in early pregnancy and risk for a small-forgestational-age infant. Am J Obstet Gynecol 2006;194:1316-22. 6. Morsing E, Asard M, Ley D, Stjernqvist K, Marsal K. Cognitive function after intrauterine growth restriction and very preterm birth. Pediatrics 2011;127:e874-82. 7. Monset-Couchard M, de Bethmann O, Relier JP. Long term outcome of small versus appropriate size for gestational age co-twins/triplets. Arch Dis Child Fetal Neonatal Ed 2004;89:F310-4. 8. Guellec I, Lapillonne A, Renolleau S, Charlaluk ML, Roze JC, Marret S, et al. Neurologic outcomes at school age in very preterm infants born with severe or mild growth restriction. Pediatrics 2011; 127:e883-91. 9. Kallankari H, Kaukola T, Olsen P, Ojaniemi M, Hallman M. Very preterm birth and foetal growth restriction are associated with specific 6
21.
22.
23.
24.
25.
26.
27.
28.
29.
cognitive deficits in children attending mainstream school. Acta Paediatr 2015;104:84-90. Hutcheon JA, Platt RW. The missing data problem in birth weight percentiles and thresholds for “small-for-gestational-age”. Am J Epidemiol 2008;167:786-92. Scherjon S, Briet J, Oosting H, Kok J. The discrepancy between maturation of visual-evoked potentials and cognitive outcome at five years in very preterm infants with and without hemodynamic signs of fetal brain-sparing. Pediatrics 2000;105:385-91. Baschat AA. Fetal responses to placental insufficiency: an update. BJOG 2004;111:1031-41. McElrath TF, Allred EN, Kuban K, Hecht JL, Onderdonk A, O’Shea TM, et al. Factors associated with small head circumference at birth among infants born before the 28th week. Am J Obstet Gynecol 2010;203:138. e1-8. Larroque B, Ancel PY, Marret S, Marchand L, Andre M, Arnaud C, 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;371:813-20. Ancel PY, Livinec F, Larroque B, Marret S, Arnaud C, Pierrat V, et al. Cerebral palsy among very preterm children in relation to gestational age and neonatal ultrasound abnormalities: the EPIPAGE cohort study. Pediatrics 2006;117:828-35. Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Surveillance of Cerebral Palsy in Europe (SCPE). Dev Med Child Neurol 2000;42:816-24. Kaufman A, Kaufman N. Kaufman assessment battery for children. Circle Pines, MN: American Guidance Service; 1983. Goodman R. The Strengths and Difficulties Questionnaire: a research note. J Child Psychol Psychiatry 1997;38:581-6. Larroque B, Ancel PY, Marchand-Martin L, Cambonie G, Fresson J, Pierrat V, et al. Special care and school difficulties in 8-year-old very preterm children: the Epipage cohort study. PLoS One 2011;6:e21361. Castro L, Yolton K, Haberman B, Roberto N, Hansen NI, Ambalavanan N, et al. Bias in reported neurodevelopmental outcomes among extremely low birth weight survivors. Pediatrics 2004;114:404-10. Katz J, Wu LA, Mullany LC, Coles CL, Lee AC, Kozuki N, et al. Prevalence of small-for-gestational-age and its mortality risk varies by choice of birth-weight-for-gestation reference population. PLoS One 2014;9: e92074. Sharma P, McKay K, Rosenkrantz TS, Hussain N. Comparisons of mortality and pre-discharge respiratory outcomes in small-for-gestationalage and appropriate-for-gestational-age premature infants. BMC Pediatr 2004;4:9. Soudee S, Vuillemin L, Alberti C, Mohamed D, Becquet O, Farnoux C, et al. Fetal growth restriction is worse than extreme prematurity for the developing lung. Neonatology 2014;106:304-10. Karagianni P, Kyriakidou M, Mitsiakos G, Chatzioanidis H, Koumbaras E, Evangeliou A, et al. Neurological outcome in preterm small for gestational age infants compared to appropriate for gestational age preterm at the age of 18 months: a prospective study. J Child Neurol 2010;25:165-70. Mu SC, Lin CH, Chen YL, Chang CH, Tsou KI. Relationship between perinatal and neonatal indices and intelligence quotient in very low birth weight infants at the age of 6 or 8 years. Pediatr Neonatol 2008;49:13-8. O’Keeffe MJ, O’Callaghan M, Williams GM, Najman JM, Bor W. Learning, cognitive, and attentional problems in adolescents born small for gestational age. Pediatrics 2003;112:301-7. Padilla N, Perapoch J, Carrascosa A, Acosta-Rojas R, Botet F, Gratacos E. Twelve-month neurodevelopmental outcome in preterm infants with and without intrauterine growth restriction. Acta Paediatr 2010;99: 1498-503. Brennan TL, Funk SG, Frothingham TE. Disproportionate intra-uterine head growth and developmental outcome. Dev Med Child Neurol 1985; 27:746-50. Kan E, Roberts G, Anderson PJ, Doyle LW. The association of growth impairment with neurodevelopmental outcome at eight years of age in very preterm children. Early Hum Dev 2008;84:409-16.
Guellec et al
- 2015 30. Cheong JL, Hunt RW, Anderson PJ, Howard K, Thompson DK, Wang HX, et al. Head growth in preterm infants: correlation with magnetic resonance imaging and neurodevelopmental outcome. Pediatrics 2008;121:e1534-40. 31. Businelli C, de Wit C, Visser GH, Pistorius LR. Ultrasound evaluation of cortical brain development in fetuses with intrauterine growth restriction. J Matern Fetal Neonatal Med 2014 Sep 10;1-6. [Epub ahead of print]. 32. Bocca-Tjeertes I, Bos A, Kerstjens J, de Winter A, Reijneveld S. Symmetrical and asymmetrical growth restriction in preterm-born children. Pediatrics 2014;133:e650-6. 33. Tolsa CB, Zimine S, Warfield SK, Freschi M, Sancho Rossignol A, Lazeyras F, et al. Early alteration of structural and functional brain development in premature infants born with intrauterine growth restriction. Pediatr Res 2004;56:132-8. 34. Batalle D, Eixarch E, Figueras F, Munoz-Moreno E, Bargallo N, Illa M, et al. Altered small-world topology of structural brain networks in infants with intrauterine growth restriction and its association with later neurodevelopmental outcome. Neuroimage 2012;60:1352-66.
ORIGINAL ARTICLES 35. Schober ME, McKnight RA, Yu X, Callaway CW, Ke X, Lane RH. Intrauterine growth restriction due to uteroplacental insufficiency decreased white matter and altered NMDAR subunit composition in juvenile rat hippocampi. Am J Physiol Regul Integr Comp Physiol 2009;296:R681-92. 36. Jegou S, El Ghazi F, de Lendeu PK, Marret S, Laudenbach V, Uguen A, et al. Prenatal alcohol exposure affects vasculature development in the neonatal brain. Ann Neurol 2012;72:952-60. 37. Montiel JF, Kaune H, Maliqueo M. Maternal-fetal unit interactions and eutherian neocortical development and evolution. Front Neuroanat 2013;7:22. 38. Ester WA, van Meurs JB, Arends NJ, Uitterlinden AG, de Ridder MA, Hokken-Koelega AC. The -G1245A IGF1 polymorphism is related with small head size and less brain sparing in small for gestational age born children. Eur J Endocrinol 2009;160:549-55. 39. Charkaluk ML, Marchand-Martin L, Ego A, Zeitlin J, Arnaud C, Burguet A, et al. The influence of fetal growth reference standards on assessment of cognitive and academic outcomes of very preterm children. J Pediatr 2012;161:1053-8.
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
7
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. -, No. -
Figure 1. Study population. MPC, mental processing composite; SDQ, Strengths and difficulties Questionnaire.
7.e1
Guellec et al
- 2015
ORIGINAL ARTICLES
Table I. Maternal and obstetric characteristics according to growth restriction category n Family SES Professional, manager Intermediate white-collar Administrative/public service, selfemployed, student Sales and service workers Manual worker or unemployed French nationality Maternal age at birth (y) <25 25-34 $35 Parity 0 1-2 $3 Type of pregnancy Singleton Antenatal corticosteroids Sex of child (male) Congenital abnormalities
SGR, n (%)
HGR, n (%)
WGR, n (%)
AGA, n (%)
P
309 533 521
27 (14.0) 45 (23.3) 50 (25.9)
19 (9.9) 55 (28.7) 37 (19.3)
32 (16.9) 37 (19.6) 42 (22.2)
231 (13.6) 396 (23.2) 392 (23.0)
.481
357 558 1968
29 (15.0) 42 (21.8) 171 (91.0)
31 (16.2) 50 (26.0) 169 (92.4)
25 (13.2) 53 (24.2) 169 (89.0)
272 (16.0) 413 (24.2) 1459 (87.4)
570 1467 391
45 (21.4) 121 (57.6) 44 (21.0)
54 (26.2) 118 (57.3) 34 (16.5)
38 (19.1) 121 (60.8) 40 (20.1)
433 (23.9) 1107 (61.1) 273 (15.1)
.133
1272 890 266
124 (59.1) 70 (33.3) 16 (7.6)
116 (56.6) 66 (32.2) 23 (11.2)
118 (59.0) 67 (33.5) 15 (7.5)
914 (50.4) 687 (37.9) 212 (11.7)
.032
1678 1782 1298 229
165 (77.5) 169 (82.8) 121 (56.8) 26/212 (12.3)
155 (74.2) 154 (75.9) 109 (52.2) 22/208 (10.6)
145 (72.1) 156 (80.0) 111 (55.2) 21/199 (10.6)
1213 (66.7) 1303 (75.5) 957 (52.6) 160/1813 (8.8)
.002 .008 .620 .302
.145
SES, socioeconomic status.
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
7.e2
C
hanges in perinatal management have been associated with a substantial increase in the survival of infants at very low gestational ages, an increase that raises questions about their long-term neurologic outcomes.1 Preterm growthrestricted infants are a population of particular interest because they combine immaturity secondary to low gestational age with the consequences of growth restriction. Growth restriction remains a concept difficult to study especially in preterm infants. It intermixes mechanisms including placental insufficiency, with or without brain sparing,2 congenital abnormalities, toxic, environmental,3 and maternal diseases.4,5 It is difficult to distinguish their consequences from the separate consequences of immaturity. Because these mechanisms might affect various anthropometric measurements, such as head circumference (HC) and birth et de la Recherche From the Institut National de la Sante dicale (INSERM), UMR 1153, Equipe de recherche en Me weight (BW) differently, these measurements might also be associated differently miologie Obste tricale, Pe rinatale et Pe diatrique; Epide with specific outcomes. ^ pitaux de Paris, Department of Assistance Publique - Ho Neonatology and Intensive Care, A. Trousseau Hospital, Previous studies have demonstrated that small for gestational age birth is assoParis; Rouen University Hospital, Neonatal Medicine; 6-9 Institute of Biomedical Research, University, Inserm ciated with a high mortality rate and impaired cognitive development. Avenir Research Group, IFR 23, Rouen; INSERM UMR ^ pitaux de Paris, Although growth restriction is a dynamic process, it is commonly defined by a 1141; Assistance Publique - Ho Hospital, Paris; Department of Neonatology, R. Debre BW <10th percentile. Moreover, preterm growth reference curves underestimate Montpellier University Hospital Center, Neonatal and 10 Pediatric Intensive Care Unit, Montpellier; Assistance growth restriction. Growth restriction because of placental insufficiency ^ pitaux de Paris, Department of Publique - Ho Neonatology, Necker Hospital, Paris; Nantes University, should, thus, be studied as a dynamic process that reduces the fetus’s growth caDepartment of Neonatology, Maternite Regionale, pacity secondary to the failure of compensation mechanisms or the severity of the Clinical Epidemiology and Biostatistics Department France, Nantes University, INSERM CIC004; Clinical illness.11 It is thought that growth-restricted fetuses attempt to compensate for Epidemiology and Biostatistics Department, CHRU miologie et Nancy; and Centre de Recherche Epide the “substrate limitation associated with placental insufficiency by preferentially 1
2
3
4
5
6
7
8
9
10
11
Statistique Sorbonne Paris Cite DHU Risques et Paris Descartes, Paris, France grossesse Universite
AGA BW HC HGR SGR WGR
Appropriate for gestational age Birth weight Head circumference Head growth restriction Symmetric growth restriction Weight growth restriction
et de la ReFunded by Institut National de la Sante dicale (INSERM), Directorate General for cherche Me Health of the Ministry for Social Affairs, and “Hospital Program for Clinical Research 2001” (AOMO1117) of the French Department of Health. The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2015.08.025
1
THE JOURNAL OF PEDIATRICS
www.jpeds.com
perfusing the central nervous system.”12,13 If so, development would be limited first by a reduction in BW and only thereafter by smaller head size. If growth restriction occurs early in pregnancy or if no adaptive mechanism protects the fetus, growth restriction might be symmetric. If it occurs later, however, or is accompanied by adaptive phenomena, it might result in asymmetric growth restriction involving BW only. If this hypothesis is correct, head growth restriction (HGR) appears to be due to another mechanism and may be associated with a different outcome. Accordingly, we analyzed growth restriction in its different clinical forms by examining 2 specific anthropometric measurements (BW and HC) and their relations to 3 different outcomes in very preterm children: short-term mortality and morbidity and long-term neurodevelopmental outcomes.
Methods Our data come from the 1997 Epipage cohort study, which included all live births between 22 and 32 weeks of gestation in 1997 in 9 regions on France.14 Because 65% of those born at 22-25 weeks died before discharge, we limited our analysis to children born alive at 26-32 weeks (n = 2694) for whom HC and BW were available (n = 2442, 90.6%). Two regions with large samples included only 1 of every 2 infants for follow-up (70 infants not included). Parent refusal resulted in exclusion of 89 infants from follow-up. At 5 years of age, 1648 (80%) had medical examinations and 1395 (68%) cognitive assessments, all by trained physicians and psychologists. The parents of 1520 children (74%) also completed questionnaires. Around the children’s eighth birthdays, parents received a questionnaire about their school performance. In all, 1315 (64%) responses were available for analysis (Figure 1; available at www.jpeds.com). The Commission Nationale de l’Informatique et des Libertes (French Data Protection Authority) approved the study. Parents provided verbal consent. Ethics committee approval was not required because this was an observational study of usual care, with no intervention. Gestational age was defined by completed weeks of gestation, determined from the date of the last menstrual period and early ultrasound findings. Maternal and obstetric data were recorded on standardized questionnaires at birth in each maternity unit. Maternal data included nationality, age at birth, and parity. Family socioeconomic status was recorded according to the French classification of occupations and social position, defined by the higher-status parental occupation (or the mother’s, if she did not live with the father). Obstetric data included type of pregnancy (singleton or higher-order) and antenatal corticoid use. Infant Characteristics and Neonatal and Long-Term Outcomes Neonatal data were prospectively collected at each hospital. This study considers sex, BW, and HC, measured by the 2
Vol. -, No. maximum occipital-frontal HC at birth. Congenital abnormalities were also recorded. In-hospital mortality was defined as death in the delivery room or neonatal unit. Length of hospitalization was defined as the number of days until discharge home. Bronchopulmonary dysplasia was defined as oxygen dependency at 28 days. Duration of central line was defined as the total number of days during which a central line was maintained, regardless of the reason. Intraventricular hemorrhage and white matter damage were diagnosed from cranial ultrasonography, performed by qualified neonatologists or radiologists. Major brain lesions included intraventricular hemorrhage with ventricular dilatation (grade III) or intraparenchymal hemorrhage (grade IV), according to the Papile classification, cystic periventricular leukomalacia, or hyperechogenicity persisting more than 14 days without cystic formation.15 We used the European definition to define cerebral palsy, which requires at least 2 of the following: abnormal posture or movement, increased tone and hyperreflexia (spastic cerebral palsy), involuntary movements (dyskinetic cerebral palsy), or absence of coordination (ataxic cerebral palsy).16 The French version of the Kauffman assessment battery for children was used to assess cognitive function, expressed as a mental processing composite score (IQ equivalent), standardized with a mean of 100 and a SD of 15 in a French population born in the 1990s.17 Moderate cognitive deficiency was defined by a score between 70 and 84, and severe cognitive deficiency <70. Behavioral problems were assessed by the French versions of the strength and difficulties questionnaire,18 completed by parents. It includes 4 scales (inattention-hyperactivity, conduct, emotional, and peer problems) that were added together for a total behavioral difficulties score. The cut-off was defined by the 90th percentile of the scores observed in the reference group of term infants included in Epipage. School difficulties were assessed at age 8 years based on a parental questionnaire. Special schooling (institution, special school and special class in mainstream school, compared with mainstream class) or low grades were considered school difficulties.19 Growth Restriction Percentiles of BW and HC were determined by gestational age and sex from the data of this cohort of very preterm births (Figure 2). Our population was divided into 4 different categories according to the percentile of their HC and BW. Symmetric growth restriction (SGR) was defined by BW and HC percentiles both <10th percentile or both between 10th and 19th percentile. Two different types of asymmetric growth restriction were defined. HGR was defined by a HC <20th percentile, with BW in at least the next higher decile group, and weight growth restriction (WGR) by a BW <20th percentile, with HC in at least the next higher decile group. The group without growth restriction, that is, appropriate for gestational age (AGA), was defined by BW and HC both above the 20th percentile. Guellec et al
- 2015
ORIGINAL ARTICLES
Figure 2. Description of the population.
Statistical Analyses We studied the relations between maternal and obstetric characteristics and growth restriction at birth. We examined these different forms of growth restriction at birth in relation to neonatal and long-term outcomes. c2 tests were used, and a P value of <.05 was considered significant. We used the ANOVA test for continuous variables, with the same P value defining significance. Logistic regression models were used to analyze associations between growth restriction and each binary outcome (ie, mortality, major brain lesions, cerebral palsy, behavioral difficulties, and school difficulties). Multinomial logistic regression models were used for cognitive outcomes (ie, no cognitive deficiency, moderate cognitive deficiency, and severe cognitive deficiency). Associations were quantified by ORs and their 95% CIs. Covariates were included in logistic regression models if they were known risk factors and associated with the outcome with a P value of <.2 in the univariate analysis. All analyses were adjusted for gestational age and sex.6 For in-hospital mortality, factors included in the final model were antenatal corticosteroids (P < .01) and family social class (P = .06); for major brain abnormalities, nationality (P = .12), and antenatal corticosteroids (P < .01); and for cerebral palsy, the type of pregnancy (singleton vs higher-order) (P = .14). Finally, the same factors were included in the models for both behavioral problems and school difficulties: family social class (P < .01 for both), maternal age (P < .01 for both), parity (P = .05 and <.01, respectively), and type of pregnancy (P = .03 and .08, respectively). Associations between growth restriction and cognitive outcomes were studied after adjustment for so-
cial class (P < .01), age (P = .01), nationality (P = .01), and parity (P < .01). All analyses were performed with Stata 10 SE (StatCorp, College Station, Texas).
Results First, we compared the characteristics of children with missing data for BW, sex, or HC with children with complete data. They were more frequently male (P = .023), were born earlier (29 vs 30 weeks gestational age), and died more often in the hospital (40% vs 9%, P < .001) especially in the delivery room (18.7% vs 0%). Long-term outcomes did not differ between these 2 groups. Mean mental processing composite score was 94.2 for missing data group vs 93.7 for complete data (P = .827), cerebral palsy occurred in 7.7% vs 8.5% of these groups (P = 1.000), and total behavioral difficulties percentage were similar (20.8% vs 21.1%, P = .000). School difficulties also were similar (25.6% vs 22.6%, P = .578). Figure 2 summarizes the number of total, SGR, HGR, WGR, and AGA births by decile range of BW and HC. Of the 2442 live births available for analysis, 213 (8.7%) were SGR, 209 (8.6%) HGR, 201 (8.2%) WGR, and 1819 (74.5%) AGA. In the SGR group, 63% had both BW and HC <10th percentile, and 37% between the 10th and 19th percentiles. In the WGR group, 43% had a BW <10th percentile, and 57% between 10th and 19th percentiles. In the HGR group, 34% had a HC <10th percentile, and 66% between the 10th and 19th percentiles. Maternal and obstetric characteristics are reported by growth restriction (Table I; available at www.jpeds.com). Children of nulliparous women were growth-restricted
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
3
THE JOURNAL OF PEDIATRICS
Vol. -, No. -
www.jpeds.com
Table II. Neonatal outcome according to growth restriction category In hospital mortality n/N (%) OR [95% CI] aOR* [95% CI] Major brain pathology n/N (%) OR [95% CI] OR† [95% CI] Bronchopulmonary dysplasia n/N (%) OR [95% CI] ORz [95% CI]
SGR
HGR
WGR
AGA
P
30/213 (14.1%) 2.04 [1.34-3.12] 2.99 [1.78-5.03]
21/209 (10.0%) 1.39 [0.86-2.26] 1.68 [0.93-3.02]
21/201 (10.5%) 1.46 [0.90-2.36] 1.39 [0.74-2.63]
135/1819 (7.4%) 1 1
.006
41/211 (19.4%) 0.86 [0.60-1.23] 0.83 [0.55-1.26]
40/203 (19.7%) 0.88 [0.61-1.26] 1.04 [0.70-1.54]
42/196 (21.4%) 0.97 [0.68-1.39] 1.06 [0.72-1.57]
388/1773 (21.9%) 1 1
.807
42/206 (20.4%) 3.09 [2.11-4.52] 3.74 [2.48-6.65]
28/203 (13.8%) 1.93 [1.25-2.98] 2.09 [1.32-3.31]
33/194 (17.0%) 2.42 [1.63-3.74] 2.70 [1.74-4.20]
135/1763 (7.7%) 1 1
<.001
*Adjusted for gestational age, sex, family social class, and antenatal corticosteroids. †Adjusted for gestational age, sex, mother’s nationality, and antenatal corticosteroids. zAdjusted for gestational age, sex, mother’s age, type of pregnancy, and antenatal corticosteroids.
more frequently than those of multiparous women (P = .032). The SGR group received antenatal corticosteroids significantly more frequently than the others (P = .008). Less than 10% of the population had congenital abnormalities, a proportion that did not differ according to growth restriction (P = .302). Neonatal mortality and morbidity were significantly associated with growth restriction categories (Table II). Inhospital mortality was significantly higher in the SGR (aOR 2.99, 95% CI 1.78-5.03) than the AGA group. Hospitalization time increased significantly (P < .01) from 54 (54) days for AGA children to 65 (37), 71 (43), and 77 (0) days for HGR, WGR, and SGR children, respectively. Similarly, duration of central-line placement varied and was higher in the WGR (27 days [31]) and SGR groups (28 days [32]) than for HGR (21 days [19]), and AGA children (17 days [18]). The
bronchopulmonary dysplasia rate was higher in growthrestricted children than in the reference group. Major brain lesions did not vary by growth status. The prevalence of intellectual disability varied across growth restriction categories (Table III). SGR children had higher risks of both moderate (aOR 1.65, 95% CI 1.012.71) and severe (aOR 2.61, 95% CI 1.46-4.68) cognitive deficiency, and HGR children only of severe cognitive deficiency (aOR 2.07, 95% CI 1.15-3.74). Although WGR was initially associated with moderate cognitive deficiency (crude OR 1.66, 95% CI 1.08-2.56), this association was not significant after adjustment. The prevalence of cerebral palsy (P = .26) and behavioral problems (P = .24) did not vary significantly by growth-restriction categories before or after adjustment. School difficulties varied, although not significantly (P = .054): 21% in the AGA group, 24% in the WGR group,
Table III. Long-term outcome according to category of growth restriction Cerebral palsy* n/N (%) OR [95% CI] aOR* [95% CI] Moderate cognitive deficiency 70# MPC <85 n/N (%) OR [95% CI] aOR† [95% CI] Severe cognitive deficiency MPC <70 n/N (%) OR [95%CI] aOR† [95% CI] Behavioral problemsz n/N (%) OR [95%CI] aORz [95% CI] School difficultiesz n/N (%) OR [95% CI] aORz [95% CI]
SGR
HGR
WGR
AGA
P
8/133 (6.0%) 0.66 [0.30-1.31] 0.63 [0.30-1.34]
8/130 (5.2%) 0.64 [0.30-1.34] 0.63 [0.30-1.32]
8/140 (5.7%) 0.59 [0.28-1.23] 0.60 [0.28-1.26]
116/1245 (9.3%) 1 1
.262
27/112 (24.1%) 1.48 [0.92-2.38] 1.65 [1.01-2.71]
21/107 (19.6%) 1.10 [0.66-1.84] 1.20 [0.70-2.04]
35/126 (27.8%) 1.66 [1.08-2.56] 1.57 [0.99-2.47]
205/1050 (19.5%) 1 1
20/112 (17.9%) 1.99 [1.16-3.42] 2.61 [1.46-4.68]
18/107 (16.8%) 1.71 [0.98-2.99] 2.07 [1.15-3.74]
16/126 (12.7%) 1.38 [0.78-2.46] 1.61 [0.87-2.97]
113/1050 (10.8%) 1 1
28/121 (23.1%) 1.21 [0.77-1.89] 1.28 [0.81-2.03]
33/124 (26.6%) 1.46 [0.95-2.23] 1.42 [0.91-2.19]
31/131 (23.7%) 1.25 [0.81-1.91] 1.34 [0.86-2.10]
228/1144 (19.9%) 1 1
.242
35/115 (30.4%) 1.65 [1.08-2.52] 1.79 [1.13-2.83]
28/98 (28.6%) 1.51 [0.95-2.40] 1.48 [0.90-2.43]
26/110 (23.6%) 1.17 [0.73-1.86] 1.21 [0.74-1.99]
208/992 (21.0%) 1 1
.054
.021
MPC, mental processing composite. *Adjusted for gestational age, sex, and type of pregnancy. †Adjusted for gestational age, sex, family social class, mother’s age, nationality, and parity at birth. zAdjusted for gestational age, sex, family social class, mother’s age and parity, and type of pregnancy.
4
Guellec et al
- 2015
ORIGINAL ARTICLES
29% in the HGR group, and 30% in the SGR group (Table III). After adjustment, SGR children had a higher rate of school difficulties than AGA children (aOR 1.79, 95% CI 1.13-2.83).
Discussion We found that preterm children with SGR had the poorest outcomes in terms of mortality, neonatal morbidity, cognitive function, and school performance. The HGR group had risks of severe cognitive deficiency and school difficulties significantly higher than the AGA reference group. No similar associations were observed in children with WGR. Our study is based on data from the Epipage cohort, one of the largest population-based cohort studies of very preterm children. Its greatest strengths are its prospective design and long follow-up. Gestational age was assessed by early ultrasound scans. The geographical approach decreased the bias associated with studies in selected perinatal centers. We used validated tools to investigate motor deficiency (cerebral palsy), cognitive function (Kaufman assessment Battery for Children), and behavioral problems (Strengths and difficulties Questionnaire). The long-term follow-up at 8 years of age provided a reliable assessment of cognitive function. Missing data for BW or HC at birth prevented assessment of 9% of the cohort. Almost 40% died during the neonatal period. Thus, underestimation of the growth-restriction rate and its relation with in-hospital mortality cannot be excluded. In contrast, among survivors, outcomes did not differ between children with and without missing data. Thus, the impact of this selection on the associations between growth restriction and neurodevelopmental outcome is unclear. Loss to follow-up is an important issue in large prospective population-based cohort studies. Several studies have found that those lost to follow-up are more likely to have disabilities than those continuing to participate.20 In our study, children lost to follow-up more frequently had parents who were younger or had low socioeconomic status; both factors that influence cognitive and school outcomes (data not shown). We used an internal reference to define groups of growth restriction because of the absence of contemporary references for BW and HC growth in France for very preterm infants. Another difficulty is that we might have misclassified some preterm babies, especially in the SGR group. Nonetheless, if some of them were “constitutionally small for gestational age” instead of growth-restricted, this misclassification would have reduced rather than increased the association between SGR and cognitive disabilities. WGR was associated with neonatal complications (mortality, longer need for central line, length of hospitalization, and bronchopulmonary dysplasia), but not, or at least to a lesser extent, with long-term cognitive performance. Interestingly, despite these high rates of neonatal morbidities, children with normal HC had better cognitive outcomes than those with SGR or HGR. These results are consistent with previous
studies reporting that growth restriction defined by BW only is associated with a higher risk of mortality21,22 and of bronchopulmonary dysplasia.23 This difference may help explain the conflict in the literature about the effects of growth restriction on BW and cognitive performance24-27 in articles that do not distinguish between growth restriction with and without HC involvement. Another important result of our study is that HC was strongly related to cognitive function. Both SGR and HGR children were at risk of cognitive impairment and school difficulties. Mechanisms of HGR differ from those for WGR. HC appears to be a risk marker of future adverse neurologic outcome. We hypothesize that SGR involves exhaustion of the compensation enabled by the brainsparing response to placental insufficiency, and HGR might result from antenatal exposure to infections, drugs, chemicals, or high levels of ionizing radiation. All of these events are likely to be underdiagnosed. McElrath et al13 showed that fetal growth restriction and preeclampsia correlate strongly with microcephaly.13 Nonetheless, head size at birth is poorly related to intelligence,28,29 even though HC is related to brain volume.30 Businelli et al31 reported accelerated cortical development in growth-restricted fetuses, despite smaller brain volume, and Bocca-Tjeertes et al32 found developmental delay independent of HC at birth. Mechanisms of cognitive delay have not yet been completely elucidated. Fetal growth restriction may cause irreversible damage to the developing brain by disrupting gray matter. Tolsa et al33 used advanced magnetic resonance imaging techniques to show that children with growth restriction have significantly reduced intracranial volume and cortical gray and hippocampus matter compared with a matched control group. Disrupted network features have been observed in human babies with intrauterine growth restriction.34 Impairments of synaptic development and white matter have also been observed in animal models of intrauterine growth restriction.35 In our study, we observed that a not insignificant percentage of the Epipage population (8.2%) had a small HC without global growth restriction, associated with severe cognitive deficiency. We can hypothesize that growth restriction was limited to head growth and to a decrease in the number of endothelial and neural cells or connections between them, although adequate nutrition prevented underweight. In these cases, we suggest that maternal and/ or placental factors have specific deleterious effects on forebrain development. For example, maternal depression, or alcohol or cocaine use during pregnancy might act specifically on cerebral angiogenesis, neurotransmitter release, and synaptogenesis.36,37 Moreover some authors have described genetic particularities, such as the 1245 A-allele of the IGF1 promoter SNP, which is associated with a small head size and a reduced brain-sparing effect in small for gestational age infants.38 Even though we did not find more congenital abnormalities or chromosomal aberrations in our populations, genetics may be one path toward understanding cognitive deficiency.
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
5
THE JOURNAL OF PEDIATRICS
Vol. -, No. -
www.jpeds.com
In our population, 40%-66% of the preterm growthrestricted children were between the 10th and 19th percentiles. They are usually not recognized as growth-restricted because preterm children are smaller than their in-utero peers.10 The true proportion of growth-restricted children is, thus, underestimated when the 10th percentile of neonatal growth curves is used to define growth restriction.39 Accordingly, we found significant associations between neonatal mortality, various forms of neonatal morbidity (eg, bronchopulmonary dysplasia), cognitive functions, and growth restriction categories, including moderate growth restriction. Many studies do not consider these children to be at risk. Moreover, moderate HGR (ie, 10th-19th percentile) is not considered at risk, especially when the BW is normal. Our results provide support for the view that both low BW and HC are continuums, rather than the discrete categories. In very preterm babies, small head size at birth, either as part of SGR or isolated, is associated with long-term neurologic deficiencies, and isolated WGR is associated only with poorer neonatal outcome but not with long-term neurologic deficiencies. These factors aggravate the prognosis of preterm birth and may be useful for parental counseling and follow-up. n
10.
11.
12. 13.
14.
15.
16.
17. 18.
Submitted for publication May 22, 2015; last revision received Jul 10, 2015; accepted Aug 6, 2015.
19.
Reprint requests: Isabelle Guellec, MD, service de reanimation neonatale et diatrique, ho ^ pital Trousseau, APHP, boulevard A.Netter, 75012 Paris, pe France. E-mail: [email protected]
20.
References 1. Saigal S, Doyle LW. An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet 2008;371:261-9. 2. Flood K, Unterscheider J, Daly S, Geary MP, Kennelly MM, McAuliffe FM, et al. The role of brain sparing in the prediction of adverse outcomes in intrauterine growth restriction: results of the multicenter PORTO Study. Am J Obstet Gynecol 2014;211:288.e1-5. 3. Hannam K, McNamee R, Baker P, Sibley C, Agius R. Air pollution exposure and adverse pregnancy outcomes in a large UK birth cohort: use of a novel spatio-temporal modelling technique. Scand J Work Environ Health 2014;40:518-30. 4. Nilsson PM, Li X, Sundquist J, Sundquist K. Maternal cardiovascular disease risk in relation to the number of offspring born small for gestational age: national, multi-generational study of 2.7 million births. Acta Paediatr 2009;98:985-9. 5. Boggess KA, Beck JD, Murtha AP, Moss K, Offenbacher S. Maternal periodontal disease in early pregnancy and risk for a small-forgestational-age infant. Am J Obstet Gynecol 2006;194:1316-22. 6. Morsing E, Asard M, Ley D, Stjernqvist K, Marsal K. Cognitive function after intrauterine growth restriction and very preterm birth. Pediatrics 2011;127:e874-82. 7. Monset-Couchard M, de Bethmann O, Relier JP. Long term outcome of small versus appropriate size for gestational age co-twins/triplets. Arch Dis Child Fetal Neonatal Ed 2004;89:F310-4. 8. Guellec I, Lapillonne A, Renolleau S, Charlaluk ML, Roze JC, Marret S, et al. Neurologic outcomes at school age in very preterm infants born with severe or mild growth restriction. Pediatrics 2011; 127:e883-91. 9. Kallankari H, Kaukola T, Olsen P, Ojaniemi M, Hallman M. Very preterm birth and foetal growth restriction are associated with specific 6
21.
22.
23.
24.
25.
26.
27.
28.
29.
cognitive deficits in children attending mainstream school. Acta Paediatr 2015;104:84-90. Hutcheon JA, Platt RW. The missing data problem in birth weight percentiles and thresholds for “small-for-gestational-age”. Am J Epidemiol 2008;167:786-92. Scherjon S, Briet J, Oosting H, Kok J. The discrepancy between maturation of visual-evoked potentials and cognitive outcome at five years in very preterm infants with and without hemodynamic signs of fetal brain-sparing. Pediatrics 2000;105:385-91. Baschat AA. Fetal responses to placental insufficiency: an update. BJOG 2004;111:1031-41. McElrath TF, Allred EN, Kuban K, Hecht JL, Onderdonk A, O’Shea TM, et al. Factors associated with small head circumference at birth among infants born before the 28th week. Am J Obstet Gynecol 2010;203:138. e1-8. Larroque B, Ancel PY, Marret S, Marchand L, Andre M, Arnaud C, 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;371:813-20. Ancel PY, Livinec F, Larroque B, Marret S, Arnaud C, Pierrat V, et al. Cerebral palsy among very preterm children in relation to gestational age and neonatal ultrasound abnormalities: the EPIPAGE cohort study. Pediatrics 2006;117:828-35. Surveillance of cerebral palsy in Europe: a collaboration of cerebral palsy surveys and registers. Surveillance of Cerebral Palsy in Europe (SCPE). Dev Med Child Neurol 2000;42:816-24. Kaufman A, Kaufman N. Kaufman assessment battery for children. Circle Pines, MN: American Guidance Service; 1983. Goodman R. The Strengths and Difficulties Questionnaire: a research note. J Child Psychol Psychiatry 1997;38:581-6. Larroque B, Ancel PY, Marchand-Martin L, Cambonie G, Fresson J, Pierrat V, et al. Special care and school difficulties in 8-year-old very preterm children: the Epipage cohort study. PLoS One 2011;6:e21361. Castro L, Yolton K, Haberman B, Roberto N, Hansen NI, Ambalavanan N, et al. Bias in reported neurodevelopmental outcomes among extremely low birth weight survivors. Pediatrics 2004;114:404-10. Katz J, Wu LA, Mullany LC, Coles CL, Lee AC, Kozuki N, et al. Prevalence of small-for-gestational-age and its mortality risk varies by choice of birth-weight-for-gestation reference population. PLoS One 2014;9: e92074. Sharma P, McKay K, Rosenkrantz TS, Hussain N. Comparisons of mortality and pre-discharge respiratory outcomes in small-for-gestationalage and appropriate-for-gestational-age premature infants. BMC Pediatr 2004;4:9. Soudee S, Vuillemin L, Alberti C, Mohamed D, Becquet O, Farnoux C, et al. Fetal growth restriction is worse than extreme prematurity for the developing lung. Neonatology 2014;106:304-10. Karagianni P, Kyriakidou M, Mitsiakos G, Chatzioanidis H, Koumbaras E, Evangeliou A, et al. Neurological outcome in preterm small for gestational age infants compared to appropriate for gestational age preterm at the age of 18 months: a prospective study. J Child Neurol 2010;25:165-70. Mu SC, Lin CH, Chen YL, Chang CH, Tsou KI. Relationship between perinatal and neonatal indices and intelligence quotient in very low birth weight infants at the age of 6 or 8 years. Pediatr Neonatol 2008;49:13-8. O’Keeffe MJ, O’Callaghan M, Williams GM, Najman JM, Bor W. Learning, cognitive, and attentional problems in adolescents born small for gestational age. Pediatrics 2003;112:301-7. Padilla N, Perapoch J, Carrascosa A, Acosta-Rojas R, Botet F, Gratacos E. Twelve-month neurodevelopmental outcome in preterm infants with and without intrauterine growth restriction. Acta Paediatr 2010;99: 1498-503. Brennan TL, Funk SG, Frothingham TE. Disproportionate intra-uterine head growth and developmental outcome. Dev Med Child Neurol 1985; 27:746-50. Kan E, Roberts G, Anderson PJ, Doyle LW. The association of growth impairment with neurodevelopmental outcome at eight years of age in very preterm children. Early Hum Dev 2008;84:409-16.
Guellec et al
- 2015 30. Cheong JL, Hunt RW, Anderson PJ, Howard K, Thompson DK, Wang HX, et al. Head growth in preterm infants: correlation with magnetic resonance imaging and neurodevelopmental outcome. Pediatrics 2008;121:e1534-40. 31. Businelli C, de Wit C, Visser GH, Pistorius LR. Ultrasound evaluation of cortical brain development in fetuses with intrauterine growth restriction. J Matern Fetal Neonatal Med 2014 Sep 10;1-6. [Epub ahead of print]. 32. Bocca-Tjeertes I, Bos A, Kerstjens J, de Winter A, Reijneveld S. Symmetrical and asymmetrical growth restriction in preterm-born children. Pediatrics 2014;133:e650-6. 33. Tolsa CB, Zimine S, Warfield SK, Freschi M, Sancho Rossignol A, Lazeyras F, et al. Early alteration of structural and functional brain development in premature infants born with intrauterine growth restriction. Pediatr Res 2004;56:132-8. 34. Batalle D, Eixarch E, Figueras F, Munoz-Moreno E, Bargallo N, Illa M, et al. Altered small-world topology of structural brain networks in infants with intrauterine growth restriction and its association with later neurodevelopmental outcome. Neuroimage 2012;60:1352-66.
ORIGINAL ARTICLES 35. Schober ME, McKnight RA, Yu X, Callaway CW, Ke X, Lane RH. Intrauterine growth restriction due to uteroplacental insufficiency decreased white matter and altered NMDAR subunit composition in juvenile rat hippocampi. Am J Physiol Regul Integr Comp Physiol 2009;296:R681-92. 36. Jegou S, El Ghazi F, de Lendeu PK, Marret S, Laudenbach V, Uguen A, et al. Prenatal alcohol exposure affects vasculature development in the neonatal brain. Ann Neurol 2012;72:952-60. 37. Montiel JF, Kaune H, Maliqueo M. Maternal-fetal unit interactions and eutherian neocortical development and evolution. Front Neuroanat 2013;7:22. 38. Ester WA, van Meurs JB, Arends NJ, Uitterlinden AG, de Ridder MA, Hokken-Koelega AC. The -G1245A IGF1 polymorphism is related with small head size and less brain sparing in small for gestational age born children. Eur J Endocrinol 2009;160:549-55. 39. Charkaluk ML, Marchand-Martin L, Ego A, Zeitlin J, Arnaud C, Burguet A, et al. The influence of fetal growth reference standards on assessment of cognitive and academic outcomes of very preterm children. J Pediatr 2012;161:1053-8.
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
7
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Vol. -, No. -
Figure 1. Study population. MPC, mental processing composite; SDQ, Strengths and difficulties Questionnaire.
7.e1
Guellec et al
- 2015
ORIGINAL ARTICLES
Table I. Maternal and obstetric characteristics according to growth restriction category n Family SES Professional, manager Intermediate white-collar Administrative/public service, selfemployed, student Sales and service workers Manual worker or unemployed French nationality Maternal age at birth (y) <25 25-34 $35 Parity 0 1-2 $3 Type of pregnancy Singleton Antenatal corticosteroids Sex of child (male) Congenital abnormalities
SGR, n (%)
HGR, n (%)
WGR, n (%)
AGA, n (%)
P
309 533 521
27 (14.0) 45 (23.3) 50 (25.9)
19 (9.9) 55 (28.7) 37 (19.3)
32 (16.9) 37 (19.6) 42 (22.2)
231 (13.6) 396 (23.2) 392 (23.0)
.481
357 558 1968
29 (15.0) 42 (21.8) 171 (91.0)
31 (16.2) 50 (26.0) 169 (92.4)
25 (13.2) 53 (24.2) 169 (89.0)
272 (16.0) 413 (24.2) 1459 (87.4)
570 1467 391
45 (21.4) 121 (57.6) 44 (21.0)
54 (26.2) 118 (57.3) 34 (16.5)
38 (19.1) 121 (60.8) 40 (20.1)
433 (23.9) 1107 (61.1) 273 (15.1)
.133
1272 890 266
124 (59.1) 70 (33.3) 16 (7.6)
116 (56.6) 66 (32.2) 23 (11.2)
118 (59.0) 67 (33.5) 15 (7.5)
914 (50.4) 687 (37.9) 212 (11.7)
.032
1678 1782 1298 229
165 (77.5) 169 (82.8) 121 (56.8) 26/212 (12.3)
155 (74.2) 154 (75.9) 109 (52.2) 22/208 (10.6)
145 (72.1) 156 (80.0) 111 (55.2) 21/199 (10.6)
1213 (66.7) 1303 (75.5) 957 (52.6) 160/1813 (8.8)
.002 .008 .620 .302
.145
SES, socioeconomic status.
Intrauterine Growth Restriction, Head Size at Birth, and Outcome in Very Preterm Infants
7.e2