1. Introduction

There is a remarkable consistency in outcomes over time and between countries, cultures, and healthcare systems, which provides evidence for a universal “preterm phenotype” associated with the neurodevelopmental immaturity conferred by a VP birth [1]. Relative to those born at term, very preterm (VP) children are at an increased risk for cognitive impairments, attention deficits, and social–emotional problems. However, there is no increased risk for disruptive or oppositional behavioral problems [2]. Moreover, developmental coordination disorders (DCD) are associated with behavioral, cognitive, and/or attention disorders [3,4] in EPT school-aged children who are free from severe deficiencies. In addition, two meta-analyses confirm a high prevalence of psychiatric disorders in children and adults born prematurely as compared to a population born at term. These VP children have a high risk of attention disorders with or without hyperactivity, autism spectrum, and anxiety disorders [1,5]. However, these disorders are always accompanied by cognitive, neurological, and learning comorbidities [6].

As traditionally seen in the literature in cohort populations, the cognitive fate of premature children is usually classified into four groups: (1) absence of incapacity, (2) minor disability, (3) moderate disability, and (4) severe disability, based on data from the total Full Scale Intelligent Quotient (FSIQ) scores. A threshold of ≥89 is considered normal, which is a low average [4,7]. However, the performances observed with a total FSIQ are the result of a complex processes involving multiple intellectual and non-intellectual characteristics (attention, emotions, motivation, movement planning, etc.) [8]. Furthermore, for infants born prematurely, the FSIQ is most often calculated on a dissociated subtest value that is not necessarily reflective of the child’s cognitive function and thus constitutes a false methodology [8]. The analysis of the subtests’ dispersions aims to highlight those that deserve a thorough interpretation. Indeed, a child considered “normal” is likely to have impairments such as a dysexecutive syndrome and/or impaired behavior and/or DCD that could disrupt brain function.

Therefore, most longitudinal studies focus on general mental health classifications rather than individual clinical diagnostic criteria and comorbidity profiles (motor and behavior). Since some outcomes may be deteriorating rather than improving over time, knowledge of all comorbidities may be useful in improving an early identification of the highest risk children in order to address their clinical needs. Inclusion of behavioral difficulties and developmental coordination disorders with the definition of mild neurodevelopmental disabilities describes, in more depth, the complexity of those difficulties faced by these preterm children [9].

In fact, there have been few studies directed on individual clinical descriptions in which discrete or non-specific signs of psychiatric disorders and/or subtle impairments of executive functions occur together and/or are associated with cognitive and/or neurological impairments in school-aged VP children. To our knowledge, only one study of these children who share similar profiles within a group on measurements of intelligence and executive functions does not take into account behavior [8]. The assessment of a true neurobehavioral “phenotype” (NBP) of prematurity, which takes into account not only the FSIQ but also executive functions, neurologic examinations, and behavior should be a principal objective of epidemiological research on prematurity [8]. Finally, no studies have measured this neurobehavioral profile and its social consequences particularly on the QoL of these children.

2. Analysis on Research Results

2.1. Population

We categorized 231 children into three clusters (Figure 1) with a mean age of 8.46 (±0.75) years. The FSIQ mean was 91.62 (14.93) (low mean average), and 106 (59.55%) had a dysexecutive syndrome. The mean emotional symptoms were slightly elevated (3.48, 2.42). The comparison between children who had all examinations (psychometric tests of WISC-IV and NEPSY-2, behavioral assessment, and QoL) and those children who had at least one missing data of these examinations did not show any statistical differences (Table S1—Online supporting information).

Figure 1. Study population’s flow chart. Note: Quality of Life (QoL); Cluster 1: Moderate Impaired NeuroBehavior (MODINB); Cluster 2: Minor Impaired NeuroBehavior (MINB); Cluster 3: UnImpaired NeuroBehavior (UINB).

2.2. Cognitive/Behavioral Impairments Profiles from New Classification

The 231 children were classified according to the severity of their neurobehavioral assessment into three neurobehavioral impairment groups:

• Cluster 1 Moderate Impaired NeuroBehavior (MODINB) with multiple impairments and behavior troubles (n = 46 (20%));
• Cluster 2 Minor Impaired NeuroBehavior (MINB) with less severe multiple impairments and behavioral troubles, (n = 99 (43%)); and
• Cluster 3 (n = 86 (37%)), UnImpaired NeuroBehavior (UINB), with only slight emotional symptoms (mean score: 3.5 (SDs) = 2.56: slightly below average, therefore pathological), with a normal neurocognitive assessment outcome.

Behavior and anxiety were significantly different between each different cluster and with a gradient of severity (Table 1).

2.3. Neurobehavioral Impairment Profiles: Behavioral and Neurocognitive Comorbidities

In Cluster 1 (MODINB), emotional troubles had high scores (4.05 (2.62)), and hyperactivity was slightly below the mean (5.32 (2.45)). The anxiety score was (35.30 (8.28)), 30 children (65.22%) had language delays, and 44 (95.65%) had a dysexecutive syndrome. The other impairments (visuo-spatial integration delay, attention deficit disorder, and ideomotor dyspraxia) were presents in 15 to 17% of the cases (Table 1).

In Cluster 3 (UINB), emotional symptoms were always below the mean and only 26 (33%) of the children had a dysexecutive syndrome. Dysexecutive disorders were a composite of performance across four subtests: working memory index (WMI) < 85 and/or planning score < 8 and/or the mental flexibility score < 8 and/or inhibition score < 10th. It may be that some children had slightly lower performances (just less than −1 SD) on just one of these tests over −1 SD (Table 1).

2.4. Neurobehavioral Cognitive Assessment by Cluster

In Cluster 1 (MODINB), the WMI and the PRI were close to −2 SDs, 74.93 (SD = 11.07), and 77.26 (SD = 8.99) respectively, the other indices were close to −1 SD (Table 1 and Figure 2).

Figure 2. Neuropsychological evaluation of the four WISC IV indices by cluster. Note: Verbal comprehension index (VCI), perceptual reasoning index (PRI), working memory (WMI), and processing speed index (PSI). Average = 100, 1 SD = 15.

In Cluster 2 (MINB), the PSI and the PRI were close to −1 SD, respectively 88.49 (SD = 10.75) and 85.41(SD = 9.24); the other indices (WMI and VCI) were close to the low average.

In Cluster 3 (UINB), the WMI was close to the average as was the PSI, respectively 101.15 (SD = 11.41) and 102.90 (SD = 12.61). The other subtests VCI and PRI were above average with a difference of 0.5 SDs.

Mental flexibility was below average (9.48 (SD = 2.68)) in Cluster 3 (UINB), less than −1 SDs (7.76 (SD = 2.72)) in Cluster 2 (MINB), and less than −2 SDs (6.50 (SD = 2.27)) in Cluster 1 (MODINB) (Table 1 and Figure 3).

Figure 3. Neuropsychological assessment: planning, mental flexibility, auditory, and visual attention according to the clusters. Note: Average: 10, deviation: 3. A <8 score reflected a disorder in the category concerned: <−1 SD).

All clusters had a behavioral score slightly below average for the variable “emotional symptoms”. Cluster 2 (MINB) had a slightly below average “social relationship disorder” score. Cluster 1 (MODINB) had a slightly below average scores for all variables in the Goodman test, except for the variable “prosocial behavior” (Figure 4).

Figure 4. Anxiety assessment by STAI questionnaire and behavior assessments by Goodman-SDQ test by clusters. Note: Goodman-SDQ (Strengths and Difficulties Questionnaire): Scores were correlated with Achenbach’s Child Behavior Checklist (CBCL), including 25 items targeted for the parents. The questionnaire reflected a total assessment score of the difficulties and defined five subcategories composed of five items each: emotional disorders: average: 0–3, slightly below average: 4, high: 5–6, very high 7–10; behaviors: average: 0–2, slightly below average: 3, high: 4–5, very high 6–10; hyperactivity: average: 0–5, slightly below average: 6–7, high: 4–5, very high 6–10; social relationship disorder: average: 0–2, slightly below average: 3, high: 4–6, very high 7–10. A total difficulty score was the sum of the first four items, with each one-point increase corresponding to an increase in the risk of developing a mental disorder. The categories assessed was in the average 0–13, slightly below the average 14–16, high 17–19, or very high 20–40. “Prosocial” behavior is counted separately and varies in the opposite direction: average: 8–10, slightly below average: 7, high: 6, very high 0–5. STAIC: State–Trait Anxiety Inventory for children. Schematically: standard anxiety state: Normal: 20–40, High Anxiety: higher than 40.

2.5. Correlation of Population Study and Quality of Life

The QoL was reduced for all children when compared to the reference population. The exceptions to this were the relationships with the family and the teacher, as evaluated by the parents. For the children, the areas impacted were ranked as (1) Self-esteem, (2) Relationship with friends, and (3) Hobbies. The parents cited the areas most impacted for their children were (1) Psychological well-being, (2) School-work, and (3) General well-being (Table 2).

Table 2. Quality of life comparisons of studied (n = 178) between cluster and reference populations.

2.5.1. Quality of Life Comparisons between the Neurobehavior Impairment Clusters

There are significant QoL differences between the clusters (as indicated by the total VSP-A index), in self- and hetero-evaluations. The QoL deterioration is determined by the severity of the three neurobehavioral “phenotypes”, which are defined as well as by dysexecutive and/or behavioral disorders. The areas most specifically concerned were the school-work and self-esteem categories (Table 2).

2.5.2. Quality of Life in Cluster 3

UINB as compared to the reference population (Table 2). There was a significant reduction in the QoL in Cluster 3, as compared to the reference population. The most impacted areas for the children were (1) Relationships with friends, (2) Leisure, and (3) General well-being. From a parental point of view, the areas most affecting their children were (1) Psychological well-being, (2) General well-being, and (3) Vitality. The area of school-work was not impacted.

This entry is adapted from the peer-reviewed paper 10.3390/children8110943