Advances in pediatric cardiac care have resulted in an increasing number of adults with congenital heart disease (CHD) being followed in tertiary care centers. This has generated interest in adult CHD as a new subspecialty of cardiology; there are now more adults than children in the world population affected with CHD.1,2 The increasing longevity of this population has led to increases in lifelong medical, psychosocial and behavioral problems, raising concerns about these patients’ well-being and perceived quality of life.3,4 Many of them also have deficits in neurocognitive development that lead to greater difficulty in adapting to educational and occupational tasks.

There are many well-known reasons for neurocognitive impairment in CHD, all related to hypoxia. Recent studies indicate that cerebral damage may occur during fetal life in certain forms of severe CHD related to reduced delivery of nutrients and oxygen, responsible for cerebral insult, mostly during the third trimester of pregnancy, due to abnormal cerebral circulation resulting from altered flow patterns in the aortic arch, and others have studied the parameters of cardiac and circulatory dysfunction that influence fetal brain development.5–8

In 2012, after reviewing the literature on surveillance, screening, evaluation, and management strategies for children with CHD, the American Heart Association published a scientific statement9 which made recommendations to optimize neurodevelopmental outcomes in the pediatric CHD population.

It is therefore important to study the neurocognitive performance of this patient population, as well as their psychosocial and emotional status and educational and occupational achievements.

The main objectives of this study were to assess the neuropsychological performance of adolescents and young adults with CHD, comparing them with a group of healthy controls, to determine whether there are different neurocognitive phenotypes in CHD through the establishment of clusters of cognitive function, and to identify their relation to sociodemographic, neonatal, clinical and psychological adjustment variables.

Statistically significant differences were found between patients with CHD and the control group of healthy participants in all areas of neurocognitive performance, with better results for the latter. Table 4 shows the results of neuropsychological testing and comparisons between controls and CHD patients.

Solutions of two (r2=0.368), three (r2=0.508) and four (r2=0.575) clusters were tested. The three-cluster solution was considered optimal since the determination coefficient (r2) was only slightly lower than in the four-cluster solution, while the latter had a small number of participants (n=9) in one of the clusters. Therefore, the three-cluster analysis provided groups of CHD patients that could be analyzed with suitable sample sizes for further statistical procedures.

Table 5 shows the mean neurocognitive performance for the three clusters. Their neurocognitive profiles are represented in Figure 1. Univariate effects were significant for all neurocognitive domains: memory (F=76.61; p<0.001); processing speed (F=65.12; p<0.001); attention (F=44.12; p<0.001); executive function (F=21.9; p<0.001); visuoconstructive ability (F=240.56; p<0.001); and global cognitive function (F=159.08; p<0.001).

Figure 1.

Mean performance of the three clusters in different neurocognitive domains. A: attention; EF: executive function; GF: global cognitive function; GI: globally impaired cluster; M: memory; MI: moderately impaired cluster; NI: non-impaired cluster; PS: processing speed; VC: visuoconstructive.

(0.04MB).

The NI cluster (non-impaired) included 114 patients. Their neurocognitive performance was significantly superior to that of the moderately impaired (MI) and globally impaired (GI) clusters. No significant differences were found between this patient group and controls across the neurocognitive domains: memory (p=0.997); processing speed (p=0.557); attention (p=0.306); executive function (p=0.099); visuoconstructive ability (p=0.1); and global cognitive function (p=0.063).

The MI cluster included 76 patients. Compared to the NI cluster, significantly poorer results were found in all cognitive domains and in global cognitive function.

The GI cluster included 27 patients. Significant differences were found in comparison to the MI cluster, with poorer performance in memory, processing speed, visuoconstructive ability and global cognitive function. There are no significant differences in attention or executive function.

The aims of this study were to assess the neuropsychological performance of adolescents and young adults with CHD, comparing them with a group of healthy controls, to determine whether there are different neurocognitive phenotypes in CHD, and to identify their relation to sociodemographic, neonatal, clinical and psychological adjustment variables.

Our study demonstrated that CHD patients have consistently poorer neurocognitive performance than healthy controls, confirming data from several other published reports.28–32 There is growing evidence that, although they score within the average range on intelligence testing, CHD patients present poor performance in various neurocognitive abilities, with a consistent pattern of sequelae in gross and fine motor skills, attention, visuospatial ability, speech and language, executive function, social cognition and impulsive behavior. This pattern was confirmed in our study.

A major strength of our study was that it enabled us to identify three neurocognitive phenotypes (non-impaired, moderately impaired and globally impaired) in our patients and the features associated with each. The MI cluster, compared to the NI cluster, presented deficiencies in memory, processing speed, attention and executive function, while the GI cluster showed striking deficiencies in memory, processing speed, executive function and visuoconstructive ability compared to the MI cluster. The NI cluster had larger head circumference and higher birth weight than the other two clusters, which is consistent with magnetic resonance imaging (MRI) findings in fetuses and infants. They also presented higher 1- and 5-min Apgar scores. As is to be expected, participants in the NI and MI clusters had more years of schooling than those in the GI cluster.

Interestingly, compared to the NI and MI clusters, the GI phenotype presented more aggressive behaviors on self-report measures, and their caregivers reported higher rates of attention and social problems on observational measures.

Furthermore, we found that patients with more severe impairment in neurocognitive function (MI and GI clusters) had a higher proportion of cyanotic forms of disease than the NI cluster. Our findings are also consistent with other published data showing that although all children with CHD are at increased risk for developmental disorders or delay, cyanotic heart lesions are associated with increased vulnerability for these conditions.9,30,33,34

Recent studies35,36 focusing on neurological sequelae in adult survivors of CHD reveal that severe forms of disease are associated with significant deficits in several cognitive domains, including attention, processing speed, and executive function. However, cognitive function in moderate and mild forms did not differ significantly from normal. These findings are consistent with data in our study showing that severe forms of disease, such as tetralogy of Fallot and transposition of the great arteries, are more frequent among patients with more severe cognitive impairment.

Many authors report that birth weight and head circumference are good predictors of neurodevelopment in children with severe CHD.37,38 Consistently, we found that head circumference, weight and length measured at birth and 1-min Apgar score in our patients were able to differentiate between the NI, MI and GI clusters, associating with different neurocognitive abilities in later life. This finding highlights the impact of the antenatal impairment in brain growth and development that occurs in CHD fetuses.

In previous studies, several mechanisms found in CHD implicated in poor neurocognitive development have been investigated, and their relative contributions analyzed, some arising in fetal life, others occurring after birth and especially during or after surgery (anoxia and perioperative ischemia). MRI scans and Doppler studies in fetuses have shown that some forms of CHD lead to derangements of fetal blood flow and decreased oxygen and nutrient delivery, resulting in impaired brain development and increased likelihood of sequelae.37–39 Data from a review38 of the mechanisms involved in the delay and impairment of brain maturation in complex CHD, unique to each form of illness, are consistent with the smaller head circumference found in neonates with complex CHD compared to normal babies.

Our findings are important because they may contribute to a new generation of studies focusing on better understanding of the ongoing processes of poor neurocognitive development and injury, that could enable decisions on neuroprotective strategies, early detection and rehabilitation from injury and disability.

In summary, our study analyzed various parameters in CHD patients, and generally confirmed findings of previously published reports. However, it is the first to examine so many different clinical, neurocognitive and psychosocial variables together, and this, we suggest, constitutes the originality and social importance of our research.

Our study also has limitations. Firstly, the number of participants was small for some of the comparisons between groups, and we were thus unable to draw robust conclusions. This resulted from the lack of data on some variables, such as fetal and neonatal measures, for some participants. Furthermore, the study's retrospective design, with all the measurements obtained in a single session, may have biased patient recruitment.

Another limitation of the study is that it does not assess the neurocognitive profile of babies born with low birth weight who do not have CHD, which would shed further light on the nature of such disabilities.

This study is of clinical value, helping to predict which children are at increased risk for neurocognitive disabilities and therefore to plan prevention and neuroprotective strategies. To our knowledge, it is the first study to provide data on all these variables in a Latin population.

In conclusion, we gathered information on the impact of various clinical and psychosocial variables in CHD, providing evidence that is useful for prevention, neuroprotection and rehabilitation.

In our opinion, future research should compare CHD patients with low birth weight children who do not have CHD, and those with intrauterine growth restriction resulting in low birth weight, to improve understanding of the mechanisms that give rise to impaired neurocognitive development.

The authors have no conflicts of interest to declare.