Does ADHD worsen inhibitory control in preschool children born very premature and/or with very low birth weight?

Abstract Introduction Deficits in executive functioning, especially in inhibitory control, are present in children born very premature and/or with very low birth weight (VP/VLBW) and in children with attention-deficit/hyperactivity disorder (ADHD). Objective To evaluate whether ADHD imposes additional inhibitory control (IC) deficits in preschoolers born VP/VLBW. Methods 79 VP/VLBW (4 to 7 years) children were assessed for ADHD using the Schedule for Affective Disorders and Schizophrenia for School Aged Children – Present and Lifetime Version (K-SADS-PL). IC was measured with Conners’ Kiddie Continuous Performance Test (K-CPT 2) and the Behavior Rating Inventory of Executive Function – Preschool Version (BRIEF-P).Results: No significant differences were found between ADHD (n = 24) and non-ADHD children (n = 55) for any of the measures (p = 0.062 to p = 0.903). Both groups had deficits in most K-CPT 2 scores compared to normative samples, indicating poor IC and inconsistent reaction times. Conclusions ADHD does not aggravate IC deficits in VP/VLBW children. Either neuropsychological tasks and parent reports of executive functions (EFs) may not be sensitive enough to differentiate VP/VLBW preschoolers with and without ADHD, or these children’s EFs are already so impaired that there is not much room for additional impairments imposed by ADHD.


Introduction
Preterm birth is defined as birth occurring before 37 full weeks of gestation. Around the world, its prevalence ranges from 5% to 18%. 1 There are sub-categories of preterm birth, based on gestational age: extremely premature (EP): < 28 weeks; very premature (VP): 28 to < 32 weeks, and moderate to late preterm infants (MLP): 32 to < 37 weeks of gestation. Prematurity can also be classified according to birth weight: extremely low birth weight (ELBW): < 1,000 grams and very low birth weight (VLBW): 1,000 to < 1,500 grams. 1,2 Our study will focus on very preterm and/or very low birth weight (VP/VLBW) children.
Although the survival rates for VP and VLBW babies have increased over recent years due to advances in perinatal and neonatal care, 3 these groups are still at high risk of death and disability. 2 For instance, several structures of the central nervous system may be compromised in VP/VLBW children. 4 Abnormalities of brain structure may, in turn, translate to neuropsychological deficits in executive functions. 3 Prematurity is frequently associated with neurodevelopmental disorders, such as ADHD. 5 This disorder is characterized by inattention and/or hyperactivity/impulsivity interfering with functioning. 6 The validity of ADHD diagnosis in preschool children has been established. 7 Overall, neuropsychological studies of preschoolers with ADHD show similar results to studies of school-aged children, 8 including deficits in response inhibition, delay aversion, working memory, and sustained attention. 9,10 Inhibitory control is an executive control function that enables an ongoing response to be stopped. It enables inhibition and alteration of inappropriate strategies, such as detecting and correcting an error in an academic task. 11 Preschool children with ADHD tend to perform significantly worse in inhibitory control tasks and inventories, when compared to those without the disorder. [12][13][14] Although ADHD is a prevalent disorder in children born prematurely and/or at very low birth weight, 3 Figure 1).

Data collection and diagnostic procedures
Eligible participants were selected from the neonatology service's records. They were contacted  = large effect size. 22 The frequency of cases with worse performance (> 1 standard deviation from normative data) was also calculated, and between-group (ADHD vs. non-ADHD) differences were investigated using chi-square tests. Normative scores for both tests were obtained respectively from a Brazilian sample 23 and the American sample used for the original BRIEF-P manual. 21

Results
Data from 79 participants were analyzed for BRIEF-P scores. Nine participants refused to finish the K-CPT 2, or their assessments were considered invalid, resulting in a total sample of 70 children with K-CPT 2 data.
Demographic and clinical data can be found in Table 1.
ADHD and non-ADHD groups did not differ in terms of age, gender, IQ, gestational weeks, or socioeconomic status. The frequency of anxiety disorders was also similar between groups, but the ADHD group had significantly higher rates of ODD, which is in line with previous literature. 24 Findings emerging from ANCOVAs did not reveal significant differences between ADHD and non-ADHD groups, either for any of the BRIEF-P scores or for K-CPT 2 variables (p = 0.062 to p = 0.903). All effect sizes were small (Table 2). Correlations between K-CPT 2 and BRIEF-P scores were all non-significant and ranged from -0.197 to 0.121.
The frequency of cases with worse performance (> 1 standard deviation [SD] from the normative mean) on both instruments was also analyzed. Regarding BRIEF-P scores, 12 children (15.2%) satisfied this condition for the ISCI. For the FI and the EMI, 6 (7.6%) and 17 (21.5%) children respectively were included in   the groups with the worst performance. No association was found between ADHD and worse performance in BRIEF-P indexes (see Table 3).  6%). We also did not find an association between ADHD and worse performance in the K-CPT 2 scores (see Table 3).

Discussion
Consistent findings of behavioral and/or cognitive profiles in preterm born children have provided evidence of a "preterm phenotype". Psychiatric disorders, especially ADHD, anxiety, and autism spectrum disorder, are the most prevalent among subjects born preterm.
This preterm behavioral phenotype is usually found in association with cognitive impairment. 25 Investigations with adults born prematurely at VLBW support these findings and show that preterm birth has long-term consequences, 26 including neurocognitive deficits. 27 Specifically with relation to inhibitory control, current evidence shows that a deficit is present both in ADHD, 8,10,13,14 and in children born very prematurely. 3,28 However, to our knowledge, only one study has compared inhibitory control in VP/VLBW children with and without ADHD. Contrary to our original hypothesis, our results revealed no significant differences in inhibitory control between VP/VLBW children with and without ADHD.
In general, weak correlations are found between tests of executive function performance and behavior scales in the literature. 29 The type of measurement selected (behavioral reports or laboratory measures) can also influence the results, either by their intercorrelations or by their associations with ADHD. 30 This is why we used both a performance measure (K-CPT 2) and a parent-rated behavior scale (BRIEF-P). We found consistent results across instruments, indicating no differences between ADHD and non-ADHD groups. Also, as described in the literature, correlations between the two types of measures were low.
Given this finding, we also analyzed whether the groups' results were 1 SD or more from the normative mean, indicating a deficit in comparison to the normative samples for K-CPT2 and BRIEF-P. We chose a lenient threshold, of one standard deviation or 15.9% from the normal distribution mean, to identify cases that showed even a mild difficulty. The BRIEF-P results indicated that the ADHD group had a slightly higher frequency than that predicted by a normal distribution in the inhibitory self-control and emergent metacognition scales. This was only observed for emergent metacognition in the non-ADHD group.
Nevertheless, the differences were not significant, and at least 70.83% of the children had performance considered to be in the normal range in both groups.  34 Although this construct still remains understudied in prematurity, it is well established that a higher than usual RTV is present in ADHD and is also a marker of general psychopathology. [33][34][35] The pathophysiology of RTV is usually associated with abnormal frontal lobe volume and/or activation. 33 Cortical maturation of the frontal lobe is delayed in ADHD children, 36 and a meta-analysis showed an overall reduction in brain volume, including white and gray matter in VP/VLBW school-aged children. 37 Specifically in preschool children, a delay in normal cortical and surface development is associated with prematurity. 38 Another critical aspect to discuss is that our sample was characterized by a below-average mean IQ, irrespective of group. However, this is not an unexpected finding since low birth weight is also associated with below-average IQ. The high degree of nervous system immaturity and the greater susceptibility to neonatal complications may lead to cognitive impairment. The cerebral networks of children with low birth weight are less connected brain networks, with lower brain volumes and lower cortical surface area, which might result in impaired cognitive functions. 2,39 Moreover, results of a meta-analysis also showed that ADHD is associated with lower overall cognitive ability when compared to healthy controls. 40 The nature of the relationship between intelligence and executive functions is still controversial. 41 Although many variables are included in the broader concept of executive functions, they are separable to a certain degree and differently correlated to each other as well as to the many areas of the brain. 42 Among all other executive functions, inhibition has been found to have a low correlation to IQ. 43 The same is valid for RTV: although higher inconsistency is related to lower IQ, the variance explained by IQ is usually small. 44  in both groups might have decreased the power for detection of between-group differences. However, it is essential to note that all between-group ES were small for both BRIEF-P and K-CPT 2 scores.
Furthermore, the absence of a full-term control sample imposes limitations on our analyses, making us rely on normative data for comparisons. Still, regarding K-CPT 2 and BRIEF-P, validity studies for the Brazilian population were not available. Finally, another aspect we did not adequately control was use of medications.
Although none of the participants were using ADHD medication, we do not have information accurately recorded in our dataset about other clinical medications that might interfere with inhibitory control.

Conclusions
This study provides evidence of performance deficits on inhibitory control tests in preschool children born VP and/or with VLBW. Presence of ADHD did not impose any additional burden on the children's test performance.
Parents' reports seem to be less sensitive for capturing executive deficits in this population. Impairment in RTV, a more basic cognitive process, was found in most participants, independent of group status. Our findings suggest either that assessment of EFs using neuropsychological tasks and parent reports may not be sufficiently sensitive to differentiate between ADHD and non-ADHD in VP/VLBW preschool children, or that these children already have a level of impairment in executive functions that does not leave much room for additional impairment. More large-scale studies investigating the nature and long-term effects of inhibitory control deficits in these vulnerable populations are needed.