The authors have declared that no competing interests exist.
Contributed to text manuscript and supervised project: JBVG. Conceived and designed the experiments: CSHA HJD JO. Performed the experiments: CSHA. Analyzed the data: CSHA HJD JO. Wrote the paper: CSHA NW JO.
Objective of this study was to examine the impact of executive function (EF) on mathematical and attention problems in very preterm (gestational age ≤ 30 weeks) children. Participants were 200 very preterm (mean age 8.2 ± 2.5 years) and 230 term children (mean age 8.3 ± 2.3 years) without severe disabilities, born between 1996 and 2004. EFs assessed included verbal fluency, verbal working memory, visuospatial span, planning, and impulse control. Mathematics was assessed with the Dutch Pupil Monitoring System and parents and teachers rated attention problems using standardized behavior questionnaires. The impact of EF was calculated over and above processing speed indices and IQ. Interactions with group (very preterm versus term birth status) were examined. Analyses were conducted separately for two subsamples: children in preschool and children in primary school. Very preterm children performed poorer on tests for mathematics and had more parent and teacher rated attention problems than term controls (ßs>.11, Ps<.01). IQ contributed unique variance to mathematics in preschool and in primary school (ßs>.16, Ps<.007). A significant interaction of group with IQ (ß = −. 24, P = .02) showed that IQ contributed unique variance to attention problems as rated by teachers, but that effects were stronger for very preterm than for term infants. Over and above IQ, EF contributed unique variance to mathematics in primary school (ß = .13, P<.001), to parent rated inattention in preschool and in primary school (ßs>−.16, Ps<.04), and to teacher rated inattention in primary school (ß = −.19; ß = .19, Ps<.009). In conclusion, impaired EF is, over and above impaired IQ, an important predictor for poor mathematics and attention problems following very preterm birth.
Most very preterm (gestational age ≤ 30 weeks) infants survive without major disabilities.
A large body of literature on term children has demonstrated that higher-order neurocognitive processes, the so-called executive functions (EF) are crucial in explaining academic difficulties and behavior problems.
Research has consistently described that very preterm birth affects EF.
Given the ongoing debate on whether IQ and EF are related or distinct concepts,
Aim of this study was to capture the unique contribution of EF to mathematical and attention problems in very preterm children over and above that of processing speed indices and IQ. EF’s were selected on which our very preterm sample had deficits [>.3 standardized mean difference [SMD]) compared to term control children.
The sample of 200 very preterm (gestational age ≤ 30 weeks) children was derived from all (n = 1260) very preterm infants admitted between 1996–2004 to the neonatal intensive care unit (NICU) of the Erasmus University Medical Centre, Sophia Children's Hospital Rotterdam, The Netherlands. 252 infants died. Twins (n = 302) were excluded as inclusion of these children would violate the assumption of independence of data. Very preterm children with severe disabilities not being able to perform tests as employed in the present study were also excluded (n = 77). Such severe disabilities were classified according to Wood et al,
Minor neurosensory dysfunctions as observed in participating children are presented in
Groups | ||||||
Very Preterm (n = 200) | Term (n = 230) | |||||
Age |
8.2 | 2.5 | 4.0–12.0 | 8.3 | 2.3 | 4.0–12.0 |
Gestational Age, mean, SD, range, wk | 28.1 | 1.4 | 24.5–30.0 | 39.9 | 1.2 | 37.0–43.0 |
<28 wk, n, % | 87.0 | 43.5 | 0.0 | 0.0 | ||
Birthweight, mean, SD, range, g | 1013.0 | 287.0 | 460.0–1900.0 | 3578.0 | 482.0 | 2500.0–5025.0 |
<1500 g, n, % | 191.0 | 95.5 | 0.0 | 0.0 | ||
Boys, n, % | 106.0 | 53.0 | 106.0 | 46.1 | ||
Estimated IQb | 93.3 | 15.8 | 70.0–138.0 | 105.0 | 13.4 | 70.0–141.0 |
Parental Educationc, n, % | ||||||
High | 45.0 | 23.1 | 109.0 | 47.3 | ||
Intermediate | 75.0 | 38.2 | 79.0 | 34.3 | ||
Low | 80.0 | 38.7 | 33.0 | 14.3 | ||
Minor Neurosensory Dysfunction, n, % | 37.0 | 18.5 | 13.0 | 5.6 | ||
Minor Vision Loss or Corrected with Contact Lenses or Glasses | 26.0 | 13.0 | 13.0 | 5.6 | ||
Minor Hearing Loss or Corrected with Hearing Aids | 5.0 | 2.5 | 0.0 | 0.0 | ||
Spastic Unilateral Cerebral Palsy | 6.0 | 3.0 | 0.0 | 0.0 |
Age of the very preterm children is not corrected for prematurity.bAdjusted for parental education. cHighest of two parents. Low = primary education only or prevocational secondary education; intermediate = 3-year secondary education or middle vocational education; high = higher professional, university training or PhD.
In the Netherlands, preschool starts at the child's fourth birthday and constitutes at least two years. Primary school starts with grade 1 in August for children who turn 6 years of age between October of the previous year and the following September. Mathematics in preschool and primary school were assessed using standardized tests that are part of the Dutch National Pupil Monitoring System.
Inattention in preschool children was rated by parents and teachers using the Attention Problems scale of the Child Behavior Checklist-Preschool [CBCL/1-5),
To enhance reliability for these primary school questionnaires, we averaged scores on the parent DBD and CBCL attention scales. The same was done for the teacher DBD and TRF attention scales. Average scores were calculated for parent and teacher ratings separately, since interrater correlations were moderate (
EF’s that were measured included: 1) verbal fluency, measured in a test that required children to name as many examples of two specific categories: “animals” and “things you can eat or drink” within a 40-second time frame.
Speed and inconsistency in speed were assessed using the correctly executed go-trials of the Stop Signal test,
IQ was measured with the subtests Vocabulary and Block Design of the WISC-III
Mathematical skills in preschool and in primary school were individually assessed by trained school staff. For very preterm children, completion of behavior questionnaires and assessment of EF and IQ took place at the Erasmus University Medical Centre Rotterdam Sophia Children's Hospital Rotterdam. Term children were assessed at their schools.
Data on mathematics were available for 75.3% (n = 311) of the participating children. For the remaining children, data on mathematics were not available because they were either in special education (n = 24), or their school used a different pupil monitoring system (n = 30), or they were too young (n = 55) to be assessed with the mathematics test at the time of participation in our study.
Preschool parent rated attention was available for all children. Preschool teacher rated attention was available for 70.0% of the children. Gestational age, birthweight, parental education, and gender, did not differ between children with and without these teacher ratings (Fs<.68, Ps>.41), although the latter group was on average 3 months younger (F = 5.03, P = .03). Primary school parent rated attention was available for 80.70% of the children and teacher rated attention was available for 74.10% of the children. Gestational age, birthweight, parental education, and gender, did not differ between children with and without these ratings (Fs<2.90, Ps>.09), except that parents of children with teacher ratings available had a higher level of education, (F = 5.99, P = .02).
For dependent variables derived from the Verbal Fluency, Digit Span, and Stop Signal test, there was missing data (< 7.0%) which resulted from either examiner error or child noncompliance. These missing values were replaced by means of maximum likelihood estimation (Expectation Maximization).
Whether poor mathematics and attention problems could be predicted from EF was examined using hierarchal linear regression analyses. These analyses were conducted separately for children in preschool and primary school. Raw scores were used in all analyses and P-values of <.05 (two-tailed) were considered statistically significant. R-square change values (δR2) of each step in the analyses were evaluated. Steps that did not reach the threshold for significance (P<.05) were not incorporated in further analyses. In step 1, the predictor group (very preterm versus term birth status) was entered, adjusted for grade (in case of analyses on mathematics
Perinatal Characteristics | n | % |
Intra Uterine Growth Retardation | 47.0 | 23.3 |
Caesarian Section | 120.0 | 60.0 |
Preeclampsia | 65.0 | 32.5 |
Patent Ductus Arteriosus | 84.0 | 42.0 |
Septicaemia | 109.0 | 54.5 |
Necrotizing Enterocolitis Grade II/III | 5.0 | 2.5 |
Respiratory Distress with Surfactant Treatment | 131.0 | 65.5 |
Retinopathy of Prematurity Grade I/II/III | 21.0/16.0/2.0 | 10.5/8.0/1.0 |
Intra-Ventricular Hemorrhage Grade I/II/III/IV | 17.0/25.0/8.0/2.0 | 8.5/12.5/4.0/1.0 |
Oxygen Dependence at 6 Weeks Corrected Age | 11.0 | 5.4 |
Prenatal Steroids (Celestone) | 141.0 | 70.5 |
Postnatal Steroids (Dexamethasone) | 35.0 | 17.3 |
Dopram | 62.0 | 31.0 |
Duration of Assisted Ventilation, mean, SD, days | 9.1 | 0.2 |
Duration of Stay on Neonatal Intensive Care, mean, SD, days | 43.0 | 36.8 |
Intra uterine growth retardation is defined as an SDS score of -2.0
Step | Predictor | ΔR2 | P | β | P |
|
|||||
|
|||||
1 | Groupa |
|
|
|
|
2 | Groupa |
|
|
−.17 | .21 |
IQ |
|
|
|||
|
|
||||
1 | Groupa |
|
|
|
|
4 | Groupa |
|
|
|
|
Impulse Control |
|
|
|||
|
|
||||
1 | Groupa |
|
|
|
|
|
|||||
|
|||||
1 | Groupa |
|
|
|
|
4 | Groupa |
|
|
−.04 | .16 |
Speed Indices | −.05 | .10 | |||
IQ |
|
|
|||
Visuospatial Span |
|
|
|||
|
|
||||
1 | Groupa |
|
|
|
|
4 | Groupa |
|
|
||
IQ | −.14 | .06 | |||
Visuospatial Span |
|
|
|||
|
|
||||
1 | Groupb |
|
|
|
|
Group*Gender |
|
|
|||
4 | Groupb |
|
|
|
|
Group*Gender |
|
|
|||
IQ | −.005 | .96 | |||
Group*IQ |
|
|
|||
Visuospatial Span |
|
|
|||
Impulse Control |
|
|
Significant associations (
In preschool, very preterm versus term birth status (i.e. group) significantly predicted poor mathematics and parent and teacher rated inattention. Step 2 and 3 yielded significant effects of IQ on mathematics. Processing speed indices were not significantly predictive for mathematical nor for attention problems (ΔR2<.04, βs<−.21, Ps>.15). In Step 4, impulse control significantly predicted parent rated attention problems over and above speed and IQ, but EFs did not significantly predict mathematical problems nor teacher rated attention problems (βs<−.29, Ps>.10).
In primary school, group significantly predicted poor mathematics. Step 2 and 3 yielded significant effects of processing speed and IQ on mathematics, although effects of processing speed disappeared (β = −.05, P = .10), with EFs entered into the analyses in Step 4. Of the EFs entered in Step 4, poor visuospatial span significantly predicted poor mathematics. Group also significantly predicted parent rated and teacher rated inattention, and a significant interaction between group and gender (ΔR2 = .03, β = −.23, P = .005) indicated that very preterm boys had the highest teacher ratings of inattention. Step 2 and 3 yielded significant effects of low IQ on parent and teacher rated inattention (βs>−.18, Ps<.02). A significant interaction with group (ΔR2 = .03, β = −.26, P = .01) indicated that the association between IQ and teacher rated inattention was significantly stronger for very preterm than for term children. Processing speed indices did not significantly predict parent nor teacher rated inattention (ΔR2<.02, βs<.13, Ps>.08). Of the EFs, entered in Step 4, poor verbal working memory and visuospatial span significantly predicted parent rated attention problems and poor visuospatial span and impulse control significantly predicted teacher rated attention problems. None of the remaining interactions with group were significant.
This study compared 200 very preterm children to 230 term control children on measures of mathematics and parent and teacher ratings of attention with the main hypothesis of the study being that poor EF would account, over and above response speed indices and IQ, for problems in mathematics and symptoms of inattention in very preterm children. Since such mathematical and attention problems are two of the most commonly reported problem areas in this population,
Results confirmed that very preterm children performed worse than term peers on all measures of mathematics and attention. Poor mathematics in preschool was fully accounted for by IQ. In primary school both IQ and EF accounted for group differences in mathematics, suggesting that in primary school, mathematical problems become increasingly complex and demanding and appeal to a higher level of neurocognitive abilities. The strong impact of IQ on mathematics reflects visuospatial as well as verbal abilities being important requisites for mathematical achievement.
The EF impulse control explained unique variance in parent rated inattention in preschool, which corresponds with the view that impulse control deficits underlie inattentiveness in young children,
In primary school, both IQ as well as EF explained unique variance in parent and teacher rated inattention. In addition, gender interacted with group indicating that the excess in attention problems as rated by teachers was mainly found in very preterm boys. IQ also interacted with group in the prediction of teacher rated attention problems in primary school, such that IQ showed stronger effects on inattention in very preterm than in term infants. Visuospatial span explained unique variance in parent as well as in teacher rated inattention, though impulse control only explained unique variance in teacher rated inattention, findings that converge with the few earlier studies on this issue.
Processing speed indices did not uniquely contribute to mathematics nor to attention problems. Though,‘lower-order’ speed is impaired in very preterm children, it does not capture contributions of ‘higher-order’ executive processes.
In our study, effects of EF were calculated while adjusting for IQ, whereas in earlier studies, e.g.
Very preterm birth is associated with deficits in mathematics and symptoms of inattention. This study showed that impaired EF was, over and above impaired IQ, an important predictor for these adverse outcomes. Given the increasing body of research
(DOC)