Antenatal Steroids and Neurodevelopment in 12-Year-Old Children Born Extremely Preterm

16512227, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apa.16140 by Nicaragua Hinari access, Wiley Online Library on [30/11/2022].
See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
| |
Received: 25 May 2021 Revised: 19 September 2021 Accepted: 5 October 2021
DOI: 10.1111/apa.16140
REGULAR ARTICLE
Antenatal steroids and neurodevelopment in 12-year-old

children born extremely preterm
Olga Kochukhova1,3 | Ylva Fredriksson Kaul1 | Martin Johansson1 |

Cecilia Montgomery1 | Gerd Holmström2 | Katarina Strand Brodd1,4 |
Lena Hellström-Westas1
1
Departments of Women’s and Children’s
Health, Uppsala University, Sweden Abstract
Aim: To investigate neurodevelopmental outcome in 12-year-old children born very
2
Departments of Neuroscience/
Ophthamology, Uppsala University,
Sweden
preterm in relation to perinatal, neonatal and socioeconomic variables. To examine
3
Departments of Psychology, Uppsala whether previously described positive effects of antenatal steroids on cognition per-
University, Sweden sist at 12 years.
4
Centre for Clinical Research Sörmland,
Methods: Prospective cohort, 78 children with gestational ages 22.7–31.9 weeks,
Uppsala University, Sweden
born in 2004–2007 and examined at 12 years of age with cognitive, motor and visual
Correspondence
motor integration tasks and compared to an age-matched control group (n = 50). Two
Olga Kochukhova, Department of
Psychology, Uppsala University, Box 1225, preterm subgroups were studied: very preterm children (28–31 gestational weeks, n
75142, Uppsala, Sweden,
= 53) and extremely preterm children (22–27 gestational weeks, n = 25).
Email: olga.kochukhova@psyk.uu.se
Results: The preterm children had significantly lower scores on all cognitive, motor
Funding information
and visual motor integration tasks than the controls. Gestational age and maternal
Financial support was obtained from
the Swedish Research Council (2016- education influenced associations differently in the two preterm subgroups. Also, se-
03109), the Swedish Brain Foundation
vere retinopathy of prematurity demonstrated strong associations to outcome. In the
(FO2016-0293), the Linnéa and Josef
Carlsson Foundation, the Promobilia extremely preterm group, administration of antenatal steroids was associated with
Foundation, the Foundation Sunnerdahls
better cognition, basic attention, word generation and motor skills.
Handikappsfond, the Sven Jerrings Fond
Conclusion: At 12 years of age, very preterm children born in the 2000s still have
deficits across several neurodevelopmental domains compared to term-born peers.
Administration of antenatal steroids has long-lasting associations to cognition and
motor skills in extremely preterm-born children.
KEYWORDS
attention, long-term follow-up, verbal fluency, very preterm, visual motor integration
Abbreviations: ADHD, attention deficit disorder; Beery-VMI, Beery-Buktenica Developmental Test of Visual-Motor Integration –Sixth Edition; EXPRESS, extremely preterm infant
study in Sweden; Full-scale IQ, full-scale intelligence quotient; LOVIS, LOngitudinal study of VISuomotor capacity in very preterm infants; MABC-2 , Movement Assessment Battery for
Children, Second edition; NEPSY-II, Developmental NEuroPSYchological Assessment, Second Edition; ROP, Retinopathy of prematurity stage 3 or higher; TEA-Ch, The Test of Everyday
Attention for Children; WISC-V, Wechsler Intelligence Scales for Children, Fifth Edition.
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in
any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2021 The Authors. Acta Paediatrica published by John Wiley & Sons Ltd on behalf of Foundation Acta Paediatrica.
|
314
wileyonlinelibrary.com/journal/apa Acta Paediatrica. 2022;111:314–322.
|
16512227, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apa.16140 by Nicaragua Hinari access, Wiley Online Library on [30/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
KOCHUKHOVA et al. 315
1 | I NTRO D U C TI O N
Key notes
The long-term risks of being born very preterm are well-known and
• At 12 years of age, both children born very and ex-
include various neurodevelopmental sequels such as cognitive defi-
tremely preterm performed lower on cognitive, motor
cits, cerebral palsy, developmental coordination disorder, attention
and visual motor integration abilities compared to full
deficit disorders (ADHD) and autism.1 Although outcomes have im-
term born peers.
proved over the last decades, 2 these children are still at increased
• The level of maternal education influenced cognitive
risk, and differences in test results relative to age-matched peers
3 outcome at 12 years in the children born very preterm.
seem to increase over time. There are few reports of long-term
• The administration of antenatal steroids was associated
outcomes in very preterm children born in the 2000s.4,5 We have
with cognitive and motor outcome in extremely preterm
previously reported outcome at 6.5 years in a regional cohort of very
12-year-old children.
preterm children born 2004–2007. In line with other studies, the
extremely preterm infants, with gestational ages below 28 weeks,
had the highest risks of neurodevelopmental sequels, but also the
children born at 28–31 gestational weeks in the study cohort had in- i.e. below −2 standard deviations of the expected birth weight for
creased risks as compared to term-born children.6,7 In the extremely gestational age), and presence of severe neonatal complications
preterm children, the risks of cognitive deficits were significantly such as intraventricular haemorrhage grade 3–4, cystic periventricu-
related to neonatal morbidities and to administration of antenatal lar leukomalacia, medically and/or surgically treated persistent duc-
steroids, whilst these associations were not present in the children tus arteriosus, neonatal sepsis, retinopathy of prematurity stage 3
born at 28–31 gestational weeks.6 These results are in line with a or higher (ROP) and bronchopulmonary dysplasia. No child received
recent Cochrane review that could see probable effects of antenatal postnatal steroids and no child developed necrotising enterocolitis.
8
steroids for reducing risk of developmental delay in childhood and A flow chart describing the study population is shown in Figure 1.
brain injury.8,9 At 12 years, the children were invited to a new evaluation in-
There are, so far, relatively few reports of long-term outcomes in cluding psychological tests and assessment of motor function. We
very preterm children born in the 2000s.4,5 The aim of this study was also collected data on the mothers’ educational level. The LOVIS was
consequently to further investigate long-term outcome at 12 years performed in parallel with the national EXPRESS study (Extremely
in the same cohort as described above. We targeted functions that Preterm infants in Sweden Study), and the assessment age, as well
belong to developmental domains where impairments have previ- as parts of the protocols were identical. Data from the age-matched
ously been detected in children born preterm. A secondary aim was Uppsala regional EXPRESS controls, (with inclusion criteria: singleton
to further investigate if the specific effects on cognition of antenatal birth, GA 37–41 weeks, APGAR at 5 min ≥7, as described in11), were
6
corticosteroids that were present at 6.5 years remained at 12 years. included with permission from the EXPRESS steering committee.
We specifically asked the following questions: Do 12-year-old chil- Individual consent to participate as controls in both the EXPRESS
dren born very preterm still show overall signs of neurodevelop- and the LOVIS 12-year follow-up was obtained, and 50 (75% of the
mental sequels in cognitive, motor and visual motor domains, when controls that participated at the 6.5-year assessment6) controls liv-
compared to age-matched peers? If so, how is neurodevelopment ing in the Uppsala region were included. All children's legal guardians
related to neonatal and socioeconomic background factors? Are provided written informed consent, and Uppsala regional research
there differences in outcomes related to administration of antena- ethics committee approved the study (2016/400).
tal steroids? If so, do any differences extend to more finely grained
cognitive functions?
2.2 | Study protocol
2 | M E TH O D S The children were assessed by clinical psychologists and physiother-

apists at the Uppsala University Hospital.
2.1 | Participants General intellectual ability was measured using the Swedish ver-
sion of the Wechsler Intelligence Scales for Children, Fifth Edition
The LOngitudinal study of VISuomotor capacity in very preterm in- (WISC-V ),12 giving scores mean and standard deviation (SD) 10013
fants (LOVIS) is a population-based prospective cohort study that for Full-Scale Intelligence Quotient (Full scale-IQ), as well as for the
included children born very preterm (gestational age <32 weeks) in five indices: Verbal Comprehension Index, Visual Spatial Index, Fluid
Uppsala county, Sweden, between 2004 and 2007. Perinatal and Reasoning Index, Working Memory Index and Processing Speed
neonatal characteristics were prospectively collected, as previously Index. Visuomotor integration (eye-hand coordination), mean (SD)
described10: administration of antenatal steroids (betamethasone, 10013 was measured with the Beery-Buktenica Developmental Test
at least one dose), mode of delivery, mother's age, infant sex, birth of Visual-Motor Integration –Sixth Edition (Beery-VMI),14 and motor
weight (including birth weight classified as small for gestational age, competence was measured by the Total score (range from 15 to 113
|
316 KOCHUKHOVA et al.
F I G U R E 1 Flow chart of the LOVIS

(Longitudinal study of Visuomotor
capacity) study group
points) of the Movement Assessment Battery for Children, Second to name the pictures of squares as ‘circle’ and vice versa as quickly as
15
edition (MABC-2), age span 11–16 years. possible); (3) switching in which automatic responding and inhibition
Basic visual attention was measured by the subtest Sky-Search are altered according to predefined cues as quickly as possible. The
from The Test of Everyday Attention for Children (TEA-Ch). The conditions 2 and 3 represent different levels of executive control of
task aims to measure selective attention by identifying pairs of attention as well as cognitive flexibility.16,17 The naming condition
identical objects amongst distractors as quickly as possible (mean serves to distinguish if a poor performance on the inhibition and at-
(SD) 10 (3)).13 tention switching conditions can be attributed to executive control
In addition, two subtests (Word generation and Inhibition) from over attention, or if it is more likely to be a consequence of basic
the test battery A Developmental NEuroPSYchological Assessment, naming ability and sustained attention.
Second Edition (NEPSY-II) were administered. The subtests (mean
(SD) 10 (3)) contain different sets of tasks, analysed relative each
other in order to separate if a deficit relies on more basic or higher 2.3 | Statistical analysis
cognitive functions. First, the subtest Word Generation measures
verbal fluency and includes: (1) Semantic category task—measures Drop out analysis was performed between participating and non-
if word production is helped by a provided category, e.g. ‘Name as participating/excluded children using Fisher's exact test, χ2 for
many animals as possible’. (2) Initial letter task—measures phonolog- categorical variables and t test or Mann–Whitney for continuous
ical word production according to a provided first letter only, e.g. variables. The Mann–Whitney U test was used to explore group
‘Name as many words beginning with the letter “s” as possible’. The differences in WISC-V, Beery VMI and MABC-2, and χ2 tests were
initial letter task is more executively demanding as semantic associ- used to explore differences in categorised outcomes between the
ations provided by given categories cannot be used. The Inhibition groups. Cutoffs were derived from the control group's mean and
from (NEPSY-II) explores attention using three conditions: (1) nam- standard deviation (SD) for the WISC-V and the Beery VMI scores,
ing condition, measuring basic ability to attend and integrate visual and percentile distributions were used for the MABC-2. The 12-year
information processing with verbal answers (the child is e. g. asked outcomes were classified as: no impairment, i.e. results at or above
to name pictures of circles and squares as quickly as possible); (2) in- −1 SD for the cognitive scores and scores >16th percentile for the
hibition of automatic responses to visual cues (the child is e.g. asked MABC-2; the category impairment included any cognitive test result
|
below −1 SD, or MABC-2 scores at or below the 16th percentiles, 3.1 | Relation between neonatal risk factors and
respectively, and severe impairment was defined as any cognitive 12-year cognitive and motor outcomes
test result below −2 SD, or MABC-2 scores equal to or below the
5th percentile. For the whole LOVIS group, the univariate regressions demonstrated
In a different set of analyses, we used the continuous test significant associations between Full-Scale IQ and antenatal steroids,
scores as outcome variables. We divided factors that could poten- persistent ductus arteriosus, ROP, and mother's education, respec-
tially affect outcome in three categories: prenatal risks (antenatal tively. Beery VMI scores were related to gestational age, persistent
steroids, caesarean section, gestational age, small for gestational ductus arteriosus and ROP and MABC-2 scores were associated with
age), neonatal risks (the number of severe neonatal complications) gestational age, periventricular leukomalacia, intraventricular haem-
and maternal characteristics (age at delivery, years of education). orrhage grade 3–4, ROP, and age of mother. In the extremely preterm
Next, a set of univariate linear regressions were used to test associ- born group, Full-Scale IQ was significantly related to antenatal ster-
ations between these variables and the outcome scores. In order to oids, ROP and mother's education. Beery VMI scores were associ-
further explore which factors that were most important for the 12- ated to gestational age and ROP, and MABC-2 to antenatal steroids,
year outcomes, we included all factors that significantly (p < 0.05) gestational age and ROP. In the very preterm born group, Full-Scale
or near-significantly (p = 0.05–0.10) affected the test results (no IQ was related to mother's education only, and MABC-2 to perive-
impairment, impairment, and severe impairment) were added in ntricular leukomalacia. The explained variance (R 2) in these regres-
consecutive steps in hierarchical regression analyses: prenatal sion models varied from 5 to 45% and beta values (degree of change
risks at Step 1, neonatal risks at Step 2 and maternal characteris- in the outcome variable for every 1-unit of change in the predictor)
tics at Step 3. Hierarchical regressions were performed separately varied from (0.23 to 0.67). In general, children of mothers with longer
for the subgroups very preterm and extremely preterm infants. education scored higher on Full-Scale IQ. Severe ROP was associated
These analyses aimed to answer whether neonatal and maternal with impairment children's performance in all explored developmen-
characteristics, in addition to prenatal risks, improved the regres- tal domains. Administration of antenatal steroids was associated with
sion models’ ability to predict neurodevelopmental performance at higher scores on Full-scale IQ and MABC-2, but only in the extremely
12 years of age. preterm children. Mode of delivery, small for gestational age, the
To address the secondary aim of the study, to investigate the child's sex, neonatal sepsis or bronchopulmonary dysplasia, respec-
effect of administered antenatal steroids on neurodevelopment, tively, were not associated to cognitive or motor scores at 12 years.
univariate and hierarchical regressions were performed as described The results of the first step in the hierarchical regression for ex-
above for all five WISC-V indices, the Sky-Search subtest (TEA-Ch), plaining Full-Scale IQ indicated that prenatal risk factors significantly
and the two subtests of NEPSY-II, separately. explained data variance in the whole population and in the extremely
In order to explore whether we had enough statistical power to preterm subgroup. Administration of antenatal steroids was the only
perform the regression analyses with the given group sizes (n = 25– significant predictor of Full-Scale IQ (Table 3). In the second step, post-
78), statistical power for all the regressions was calculated post hoc natal risk factors were entered, but only ROP accounted for a significant
(alpha level =0.05, two-t ailed; number of predictors in the regres- rise of the explained variance, and only for the whole population. In the
2
sions from 2 to 6; effect sizes (f ) varied from medium (0.27) to large third step, the mothers’ level of education made significant contribu-
(1.7).18 The analyses resulted in strong statistical power ranging from tions to the predictability of the model in both the whole cohort and
0.95 to 0.99 for all multiple regression models. in the two subgroups. The same hierarchical regression procedure was
performed for the Beery-VMI and the MABC-2 results, Table 3. For the
Beery-VMI, only gestational age (for the whole population and the two
3 | R E S U LT S subgroups) and ROP made significant contributions to the predictive
ability of the regression model. Gestational age (whole population and
The baseline characteristics of the LOVIS cohort in infancy and for the extremely preterm group) and antenatal steroids (only extremely
those participating at 12 years of age were comparable, Table 1. preterm group) played a significant role for motor performance, whilst
Table 2 shows the categorised results of the Full-Scale IQ, the periventricular leukomalacia (whole population and very preterm
Beery-VMI index and the MABC-2. Overall, the LOVIS group had group) and ROP accounted for additional explained variances for the
significantly higher rates of impairment than the control group. MABC-2 percentile scores. Maternal characteristics did not contribute
Extremely preterm subgroup demonstrated significantly more mild to these models, neither for the Beery-VMI nor for the MABC-2 scores.
and severe impairments compared to the reference group. However,
the rates of severe impairments in the preterm children born at 28–
31 gestational weeks did not differ statistically from the control 3.2 | Antenatal steroids use:
group. At the same time, the very preterm subgroup significantly neurodevelopment and fine grain cognitive functions
differed from the reference group first and foremost at the amount
of mild impairments, and children with severe impairments were less Antenatal steroids were only associated to 12-year outcome in the
frequent in this group compared to the extremely preterm group. extremely preterm subgroup. Univariate linear regression models
|
TA B L E 1 Characteristics of the LOVIS cohort and the two subgroups (extremely preterm, EPT and very preterm, VPT). (A) Characteristics
from infancy; (B) 12-year outcome and maternal characteristics. p-values were obtained using t-test, Mann Whitney U-test, χ2 and Fisher
exact test, as appropriate. p-values for drop out analysis are presented in section A where included LOVIS children are compared to the drop
outs at 12 years. In section B the p-values correspond to comparisons between participating 12-year LOVIS children (with sub groups) and
the reference group. ***p < 0.001, **p < 0.01, *p < 0.05, †0.05 < p < 0.10
Original cohort in infancy Children at 12 years of age
LOVIS EPT VPT Drop out b LOVIS EPTc VPTc Reference
A Child N = 113 n = 35 n = 78 n = 31 N = 78 n = 25 n = 53 n = 50
Deceased, n (%) 4 (3.5) 3 (8) 1 (1)
Antenatal steroids, n (%) 78 (69) 23 (66) 55 (71) 20 (64) 56 (72) 18 (72) 38 (72)
Cesarian Section, n (%) 68 (60) 22 (63) 46 (59) 21 (68) 44 (56) 15 (60) 29 (55)
Female, n (%) 52 (46) 16 (46) 36 (46) 12 (38.7) 35 (45) 13 (52) 30 (57) 32 (64)
GA weeks, mean (SD) 29 (2) 26 (2) 30 (1) 29.2 (2.4) 29(2) 26 (2) 30 (1)
Birth weight grams, 1198 (353) 838 1359 1240 (365) 1204 859 1365
mean (SD) (220) (272) (343) (222) (260)
SGA, n (%) 22 (20) 5 (14) 17 (22) 7 (22.6) 13 (17) 2 (8) 11 (21)
IVH ≥grade 3, n (%) 5 (4) 2 (6) 3 (4) 0 (0) 4 (5) 2 (8) 2 (4)
PVL, n (%) 7 (6) 0 7(9) 0 (0) 7 (9) † 0 (0) 7 (12)
PDA, n (%) 25 (22) 18 (51) 7(9) 3 (9.7) 18 (23) 12 (48) 6 (11)
Sepsis, n (%) 22 (20) 15 (43) 7 (9) 3 (9.7) 17 (22) 11 (44) 6 (11)
ROP ≥stage 3, n (%) 9 (8) 9 (26) 0 4 (12.9) 4 (5) 4 (16) 0
BPD, n (%) 25 (22) 19 (54) 6 (8) 4 (12.9) 17 (22) 12 (48) 5 (9)
a *
B Cerebral Palsy, n (%) - 8 (10) 2 (8) 6 (11)* 0
a
Autism, n (%) - 5 (6) 2 (8) 3 (6) 1 (2)
** ** *
Full-Scale IQ, mean (SD) - 94 (17) 90 (13) 96 (16) 102 (11)
Beery VMI index, mean - 80 (15)*** 77 (18)*** 82 (13)*** 92 (12)
(SD)
MABC−2 total score, - 64 (18)*** 61(20)*** 66 (17)*** 79 (11)
mean (SD)
Mother
Age at child birth, years, 32 (5) 32 (5) 32 (5) - 32 (5) 32 (5) 32 (5)
mean (SD)
Education
≤9 years, n (%) 8 (11)** 2 (12) 6 (10)* 1 (2)
**
10–12 years, n (%) 36 (47) 8 (47) 28 (47)* 14 (29)
** *
≥ 13 years, n (%) 32 (42) 7 (41) 25 (42) 34 (69)
Note: Abbreviations: BPD, bronchopulmonary dysplasia; GA, gestational age; IVH, intraventricular haemorrhage; PDA, medically and surgically
treated persistent ductus arteriosus; PVL, cystic periventricular leukomalacia; ROP, retinopathy of prematurity for ROP stage 3 and higher; SGA,
small for gestational age.
a
2 VPT children had both Autism and CP diagnoses.
b
Drop out analysis performed between included LOVIS (n = 78) and not participating/excluded LOVIS participants (n = 31) at 12 years of age.
c
Percentages are calculated from the sub groups.
demonstrated that antenatal steroids significantly predicted higher 17% to 33% and the beta values from 0.41 to 0.58. However, the ex-
Full-Scale IQ predominantly through better Verbal Comprehension ecutive aspects of attention (inhibitory control and switching) were
and Fluid Reasoning Indices. Antenatal steroids also accounted for not significantly related to antenatal steroids. Hierarchical multivari-
improved basic visual attention (TEA-Ch) and ability to use visual ate regression models, confirmed findings of independent effects of
information for quick verbal retrieval, e.g. naming of different geo- antenatal steroids, also when controlling for other neonatal variables
metrical shapes (NEPSY II, Inhibition –Naming condition). Antenatal and maternal characteristics, Table 4.
steroids also accounted for better verbal fluency, both phonologi- In addition, logistic regression analyses were performed on all
cally and semantically (NEPSY II, Word Generation) and the ex- five WISC-V Indices, subtests of the NEPSY II and TEA-Ch and the
plained variance (R 2) in the significant regression models varied from total score of the MABC-2. Administration of antenatal steroids
|
TA B L E 2 Categorised results of the Full-Scale IQ, Beery VMI Index and MABC-2 total score in the LOVIS study in relation to the term-
born control group. The number of available test results varied slightly between the different tests. Statistical differences between the
LOVIS cohort, and the two subgroups very preterm (VPT) and extremely preterm (EPT) children, respectively, compared to the reference
group were obtained by χ2 test and are denoted: ***p < 0.001, **p < 0.01, *p < 0.05, †0.05 < p < 0.10
Proportion (%) of children with results below −1 SD Proportion (%) of children with results below −2 SD
LOVIS EPT VPT Reference LOVIS EPT VPT Reference
N of participants 71–78 24–25 52–53 48–50 71–78 24–25 52–53 48–50

** ** ** * ** †
Full-Scale IQ 40 44 37 14 19 28 15 4
Beery VMI Index 39* 46* 35† 18 21** 33*** 15† 4
th th
Proportion (%) on or below the 16 percentile Proportion (%) on or below the 5 percentile
MABC−2 45*** 56*** 39** 15 26* 36* 22† 9
only significantly reduced the risk of cognitive impairment, i.e. a association with cognition. Motor function was related to both an-
Full-Scale IQ below −1 SD, and this effect was mainly mediated via tenatal steroids and gestational age, whilst visual motor integration
higher scores for Verbal Comprehension and Visual Spatial Indices was only associated with gestational age. Amongst the postnatal
(for details see Table S1). risks, severe ROP showed most associations to outcome, which con-
firms findings from the 6.5-year follow-up.6 These findings are also
in line with a recent study by Petursdottir et al, reporting signifi-
4 | D I S C U S S I O N cantly lower scores of visuomotor functions in young, prematurely
born adults who had been treated for ROP. 21 Severe ROP, which
The present study demonstrated that children born in the 2000 still only affected the extremely preterm infants, was the only neonatal
exhibit adverse long-term consequences after very preterm birth, risk factor consistently associated to all outcome measures in the
especially children born extremely preterm. At 12 years, a signifi- present study cohort. The subjects with severe ROP are all in the
cantly larger proportion of the very preterm children performed at extremely preterm group, but this subgroup was not large enough
the level of impairment across all degrees of very preterm birth and to achieve statistical power. However, as gestational age was con-
in all the investigated domains, i.e. cognition, visual motor integra- trolled for in the hierarchical analysis of the whole cohort, ROP ex-
tion and motor function. The proportion of children performing at plains additional variance in data over and above gestational age.
the level of impairment and severe impairment was 2–3 times higher The level of maternal education influenced Full-Scale IQ across both
amongst the children born at 28–31 gestational weeks than in the subgroups. Overall, associations between antenatal and postnatal
control group, although some differences failed to reach statistical factors, respectively, and cognition, attention and motor function
significance. The category severe impairment reflects sequels that were scarce in the very preterm group. Periventricular leukomalacia
affect daily life. Difficulties existed in several areas, including basic was the only factor that had an association with motor outcome in
abilities needed to solve everyday tasks, manoeuvring in the envi- the very preterm subgroup.
ronment and acquiring new skills. The functions targeted by the tests The influence of neonatal risks on 12-year outcomes seemed to
involve the broader domains of intelligence, motor ability, attention weaken up to 12 years of age, but the environmental factors related
and executive function. The affected skills are likely to influence not to maternal education prevailed. In the more vulnerable, extremely
19
only academic achievements, but also other aspects of day-to-day preterm subgroup; however, maternal education did not override the
activities.1 Poor motor skills are known to affect participation, and importance of antenatal steroids. General intelligence is partly inher-
physical and mental well-being. 20 Visual motor integration is de- ited22 and tends to correlate strongly to level of education. 23 With
pendent on perception and motor skills, but also requires planning all likelihood, preterm birth alters the trajectory of optimal develop-
and organisation. Each of these skills may function adequately in ment. Children born to highly educated mothers, however, would
isolation, but the integration requires a higher level of functioning. have had higher test results if not exposed to preterm birth because
Compared to the LOVIS cohort's results at 6.5 years, the prev- of genetic traits. 24 The children born before 28 gestational weeks
alence of cognitive impairment and severe cognitive impairment in might be neurologically so vulnerable that the effects of antenatal
the children born very preterm remained higher than in the full term steroids are not outweighed by genetics and family socioeconomics.
born controls at 12 years,6 as did the poorer motor performance that The present study results clearly indicate that antenatal ste-
was present in infancy7 and at 2.5 years.6 All children with cerebral roids have a salutogenic importance for neurodevelopment in the
palsy were detected before the age of 4 years, and rates of autism extremely preterm subgroup associated with a decrease in clinically
remained similar between 6.5 and 12 years. relevant impairments. For the extremely preterm subgroup, the
In the present cohort, the prenatal characteristics influenced out- positive relation was found both in categorical analysis (Table S1),
come differently in the two gestational age groups. In the extremely indicating decline in clinically relevant impairments, and even
preterm subgroup, antenatal steroids had a major independent clearer in the hierarchical regression and in light of other predictors.
|
TA B L E 3 Summary of hierarchical regression analyses for variables predicting performance on Full-Scale IQ, Beery-VMI Index and
MABC-2 total scores. Different regressions include different predictors, as univariate dependencies in the children groups varied. All
significant results are depicted. ΔR 2—depicts change in explained data variance on each additional regression step, but not for independent
variables. ***p < 0.001, **p < 0.01, *p < 0.05, †0.05 < p < 0.10
Full-Scale IQ
Predictors LOVIS EPT VPT
ΔR2 β ΔR2 β ΔR2 β

* **
Step 1: Prenatal risk factors .11 .37 .09
Antenatal steroids .26* .50** no
†
Gestational Age .22 .27 .40†
SGA no no −.22
*
Step 2: Postnatal risk factors .08 .05 no
PDA −.11 no no
*
ROP −.30 −.28 no
Step 3: Mother's characteristics .13*** .11* .15*
Age no no .19
Education .37*** .37* .33*
Beery-VMI Index
Step 1: Prenatal risk factors .08* .38*** .00
* ***
Gestational Age .29 .61 .04
Step 2: Postnatal risk factors .13** .10† no
PDA −.05 no no
ROP −.39** −.41† no
Step 3: Mother's characteristics .05 no
Education no .23 no
MABC−2 total scores
Step 1: Prenatal risk factors .10* .51*** .06
*
Antenatal steroids no .33 no
Cesarian Section .20† no no
Gestational Age .25* .58*** −.04
SGA no no .23
*** † *
Step 2: Postnatal risk factors .18 .12 .14
IVH −.04 no −.11
*
PVL −.26 no −.31*
ROP −.39** −.34† no
†
Step 3: Mother's characteristics .04 no .04
Age −.20† no −.20
Note: no—variable not entered into the hierarchical regression because of non-significant association in univariate analysis.
Abbreviations: EPT, extremely preterm subgroup, gestational age <28 weeks; IQ, intelligence quotient; IVH, intraventricular haemorrhage grade
3–4; LOVIS, LOngitudinal study of VISuomotor capacity in very preterm children; MABC-2, Movement ABC-second edition; PDA, medically and/or
surgically treated persistent ductus arteriosus; PVL, cystic periventricular leukomalacia; ROP, retinopathy of prematurity stage 3 and higher; SGA,
small for gestational age; VMI, visual motor integration; VPT, very preterm subgroup, gestational age 28–31 weeks.
Additional to the positive relationship with Full-Scale IQ, the Verbal subtest (NEPSY-II), both the semantic category and more executively
Comprehension (concept formation and vocabulary) and Fluid demanding, Initial letter task, were equally associated to administra-
Reasoning (visual logic) were better in children who received ante- tion of antenatal steroids, indicating that basic language production
natal steroids. rather than higher order (executive) functions was affected. The
There was also a statistical effect on the basic ability to attend same pattern was found for attention (TEA-Ch; NEPSY-II, Inhibition),
and integrate visual information processing with verbal answers. where basic focussed visual attention was affected, whilst atten-
Regarding the verbal fluency measured by the Word generation tional control and inhibition were not. Also, antenatal steroids were
|
TA B L E 4 Summary of hierarchical regression analyses for antenatal steroids in relation to 12-year cognitive outcome in the extremely
preterm subgroup (gestational age <28 weeks) of the LOVIS cohort. Different regressions include different predictors, as univariate
associations varied. ΔR 2—depicts change in explained data variance on each additional regression step, but not for independent variables.
***
p < 0.001, **p < 0.01, *p < 0.05, †0.05 < p < 0.10
Predictors WISC-V NEPSY-II
Naming
Fluid Reasoning Index Verbal Comprehension Index condition
ΔR2 β ΔR2 β ΔR2 β

* * *
Step 1: Prenatal risk factors .34 .24 .23
* *
Antenatal steroids .44 .49 .48*
†
Gestational Age .32 no no
Step 2: Postnatal risk factors .08 .05
ROP −.36 −.25 no
**
Step 3: Mother's characteristics .23 .09 .02
Education .54** .31 .13
TEA-Ch NEPSY-II
Sky-Search task Semantic category Initial letter task
**
Step 1: Prenatal risk factors .33 .20* .20*
Antenatal steroids .57** .45* .45*
*
Step 2: Postnatal risk factors .08 .13 .08
ROP −.28 −.41* −.32
Note: no—variable not entered into the hierarchical regression because of non-significant association in univariate analysis.
Abbreviations: LOVIS, LOngitudinal study of VISuomotor capacity in very preterm children; NEPSY-II, A Developmental NEuroPSYchological
Assessment 2nd ed; ROP, retinopathy of prematurity stage 3 and higher; TEA-Ch, test of everyday attention for children; WISC-V-Wechsler
Intelligence Scale for Children 5th ed.
associated with better motor function but no associations were The limitations of the current study include a relatively small
found with visual motor functions. sample, where the prevalence of some neonatal risk factors as well
Previous studies have demonstrated that antenatal steroids in- as severe impairment was relatively low. This might have resulted in a
crease survival in infants born preterm, 25,26 and decreases risks for loss of statistical power. However, the results were still robust, con-
neonatal morbidities in very preterm infants. However, the potential firming larger prevalence of impairments in children born preterm
effects of antenatal steroids on neurodevelopment have been more as compared to full-term controls. Also, the investigators could not
uncertain, although no adverse effects have been documented. 27,28 be blinded, as the assessments were carried out in a clinical setting.
In a large observational cohort study a dose-dependent relation Concerning the analysis of antenatal steroids, the sample was not
between antenatal steroids and survival, as well as with neurode- large enough to investigate possible dose dependent effects. 29 The
velopment (cognition and cerebral palsy), was present in extremely strengths of the study include the prospectively collected, popula-
preterm children at 18–22 months. That was in part mediated through tion based regional cohort, extensively followed up into early ad-
a lower rate of intraventricular haemorrhages and periventricular olescence with measures on several aspects of development. Few
leukomalacia in children receiving antenatal steroids. 26 However, other studies of long-term follow-up of children born in the 2000s
in the present study including only survivors, no obvious mediating are available, and to our knowledge, no studies have investigated
effects were discernible. The positive association between antena- antenatal steroids over such a large age span.
tal steroids and Full-scale IQ as well as verbal comprehension were
identified in the extremely preterm children already at 6.5 years.6
The dropout rate for this study was 28%, which is comparable to 5 | CO N C LU S I O N S
other studies. All children with cerebral palsy, intraventricular haem-
orrhage grade 3–4 and periventricular leukomalacia participated at Sequels after very preterm birth in several neurodevelopmental
12 years, which potentially could skew results. Amongst the children domains persist into early adolescence also for children born in the
not participating in the 12-year follow-up, the reasons ranged from 2000s, and in several neurodevelopmental domains. Maternal ed-
not responding to the invitation, moved to another county, no pos- ucation is associated to outcomes, but is difficult to influence for
sibility to take time off school/work. In one case, parents deemed the clinician. In the extremely preterm group, antenatal steroids
the functional level of their child as too low to be able to participate. were important for 12-year cognitive, basic attentional, and word
|
14. Bna BKE. The Beery-Buktenica developmental test of visual motor

generation skills as well as for motor development. This emphasises integration: administration, scoring and teaching manual, 5th ed.
the salutogenic properties of antenatal steroids, and the importance NCS Pearson Inc; 2006.
of including them in standard praxis. 15. Henderson SE, Sugden DA, Barnett A. Movement assessment bat-
tery for children, 2nd ed. Psychological Corporation; 2007.
16. Korkman M, Kirk U, Kemp SL. NEPSY II – Clinical and Interpretive
AC K N OW L E D G E M E N T S
Manual. NCS Pearson Inc; 2007.
We thank the children and parents participating in the study and 17. Korkman M, Kirk U, Kemp SL. NEPSY II, Swedish version. NCS
acknowledge the contribution of Lina Degréus, Cecilia Ewald, Sirkku Pearson Inc; 2011.
Setänen and Jonina Hreinsdottir. 18. Selya AS, Rose JS, Dierker LC, Hedeker D, Mermelstein RJ. A
Practical Guide to Calculating Cohen's f(2), a Measure of Local
Effect Size, from PROC MIXED. Front Psychol. 2012;3:111.
C O N FL I C T O F I N T E R E S T 19. Johnson S, Hennessy E, Smith R, Trikic R, Wolke D, Marlow N.
The author declares that there is no conflict of interest that could be Academic attainment and special educational needs in extremely
perceived as prejudicing the impartiality of the research reported. preterm children at 11 years of age: the EPICure study. Arch Dis
Child Fetal Neonatal Ed. 2009;94:F283-289.
20. Kaul YF, Johansson M, Månsson J, et al. Cognitive profiles of ex-
ORCID tremely preterm children: Full-Scale IQ hides strengths and weak-
Olga Kochukhova https://orcid.org/0000-0002-0268-0877 nesses. Acta Paediatr. 2021;110(6):1817-1826 10.1111/apa.15776.
Cecilia Montgomery https://orcid.org/0000-0002-5290-4090 21. Petursdottir D, Holmström G, Larsson E. Böhm B Visual-motor
functions are affected in young adults who were born prema-
ture and screened for retinopathy of prematurity. Acta Paeddiatr.
REFERENCES 2021;110:127-133.
1. Allotey J, Zamora J, Cheong-See F, et al. Cognitive, motor, be- 22. Nisbett RE, Aronson J, Blair C, et al. Intelligence: New findings and
havioural and academic performances of children born preterm: theoretical developments. Am Psychol. 2012;67:130-159.
a meta-analysis and systematic review involving 64 061 children. 23. Ritchie SJ, Tucker-Drob EM. How much does education improve
BJOG. 2018;125:16-25. intelligence? A Meta-Analysis. Psychol Sci. 2018;29:1358-1369.
2. Stoll BJ, Hansen NI, Bell EF, et al. Trends in care practices, mor- 24. Van Houdt CA, van Wassenaer-Leemhuis AG, Oosterlaan J, van
bidity, and mortality of extremely preterm neonates, 1993–2012. Kaam AH, Aarnoudse-Moens CSH. Developmental outcomes of
JAMA. 2015;314:1039-1051. very preterm children with high parental education level. Early
3. O’Reilly H, Johnson S, Ni Y, Wolke D, Marlow N. Neuropsychological Hum Dev. 2019;133:11-17.
Outcomes at 19 Years of Age Following Extremely Preterm Birth. 25. Glass HC, Costarino AT, Stayer SA, Brett CM, Cladis F, Davis
Pediatrics. 2020;145:e20192087. PJ. Outcomes for extremely premature infants. Anesth Analg.
4. Nyman A, Korhonen T, Lehtonen L, Haataja L, PIPARI Study 2015;120:1337-1351.
Group. School performance is age appropriate with support ser- 26. Chawla S, Natarajan G, Shankaran S, et al. national institute of
vices in very preterm children at 11 years of age. Acta Paediatr. child health and human development neonatal research network.
2019;108:1669-1676. Association of neurodevelopmental outcomes and neonatal mor-
5. Ruys CA, Bröring T, Van Schie PEM, et al. Neurodevelopment of chil- bidities of extremely premature infants with differential exposure
dren born very preterm and/or with a very low birth weight: 8-Year to antenatal steroids. JAMA Pediatr. 2016;170:1164-1172.
follow-up of a nutritional RCT. Clin Nutr ESPEN. 2019;30:190-198. 27. Roberts D, Brown J, Medley N, Dalziel SR. Antenatal corticoste-
6. Kaul YF, Naseh N, Strand Brodd K, Böhm B, Holmström G, roids for accelerating fetal lung maturation for women at risk of
Hellström-Westas L. Average 2.5-year neurodevelopmental test preterm birth. Cochrane Database Syst Rev. 2017;21:CD004454.
results in children born very preterm did not rule out cognitive defi- 28. Cartwright RD, Crowther CA, Anderson PJ, Harding JE, Doyle LW,
cits at 6.5 years of age. Acta Paediatr. 2021;110:846-854. McKinlay CJD. Association of fetal growth restriction with neuro-
7. Montgomery C, Kaul YF, Strand Brodd K, Persson K, Hellström- cognitive function after repeated antenatal betamethasone treat-
Westas L. Structured observation of motor performance in infants: ment vs placebo: secondary analysis of the ACTORDS randomized
Level and quality associated with later motor development. Acta clinical trial. JAMA Netw Open. 2019;2:e187636.
Paediatr. 2021;110:307-313. 29. Chawla S, Bapat R, Pappas A, Bara R, Zidan M, Natarajan G.
8. McGoldrick E, Stewart F, Parker R, Dalziel SR. Antenatal cortico- Neurodevelopmental outcome of extremely premature infants ex-
steroids for accelerating fetal lung maturation for women at risk of posed to incomplete, no or complete antenatal steroids. J Matern
preterm birth. Cochrane Database Syst Rev. 2020;12:CD004454. Fetal Neonatal Med. 2013;26:1542-1547.
9. Hershkovich Shporen C, Reichman B, Zaslavsky-Paltiel I, Lerner-Geva
L, Flidel-Rimon O. Antenatal corticosteroid therapy is associated
with a lower risk of cystic periventricular leukomalacia. Acta Paediatr. S U P P O R T I N G I N FO R M AT I O N
2021;110(6):1795-1802. https://doi.org/10.1111/apa.15772. Additional supporting information may be found in the online ver-
10. Strand Brodd K, Grönqvist H, Holmström G, Grönqvist E, Rosander sion of the article at the publisher’s website.
K, Ewald U. Development of smooth pursuit eye movements in very
preterm born infants: 3. Association with perinatal risk factors.
Acta Paediatr. 2012;101:164-171.
How to cite this article: Kochukhova O, Fredriksson Kaul Y,
11. Serenius F, Kallen K, Blennow M, et al. Neurodevelopmental
Outcome in Extremely Preterm Infants at 2.5 Years After Active Johansson M, et al. Antenatal steroids and neurodevelopment
Perinatal Care in Sweden. :11. in 12-year-old children born extremely preterm. Acta Paediatr.
12. Wechsler D. Wechsler intelligence scale for children, 5th ed. NCS 2022;111:314–322. https://doi.org/10.1111/apa.16140
Pearson, Inc; 2016.
13. Manly T, Robertson IH, Anderson V, Nimmo-Smith I. Test of every-
day attention for children (TEA-Ch). Harcourt Assessment; 1999.

Antenatal Steroids and Neurodevelopment in 12-Year-Old Children Born Extremely Preterm

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Antenatal Steroids and Neurodevelopment in 12-Year-Old Children Born Extremely Preterm

Uploaded by

Copyright:

Available Formats

16512227, 2022, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/apa.16140 by Nicaragua Hinari access, Wiley Online Library on [30/11/2022].