Introduction
Approximately 1% of newborns are born with a CHD. Reference Khoshnood, Loane and Garne1 The survival rate has improved due to better treatment and diagnostics, revealing an increased lifetime risk of neurological, endocrinological, and pulmonary co-morbidities. Reference Warnes, Liberthson and Danielson2–Reference El-Chouli, Meddis and Christensen4 In recent years, complex and simple CHD research has further revealed neurocognitive, social, and psychiatric difficulties. Reference Asschenfeldt, Evald and Heiberg5–Reference Siciliano, Prussien and Lee9 Psychiatric disorders have a negative impact throughout life with a larger risk for lower educational attainments, unemployment, and divorce, Reference Mojtabai, Stuart, Hwang, Eaton, Sampson and Kessler10–Reference Mojtabai, Stuart and Hwang12 and early recognition and treatment may have the potential to change these negative impacts. However, studies investigating psychiatric co-morbidities in children and adolescents with CHD are sparse and show conflicting results. One study from 2014 found that children with a CHD had an increased risk of both attention deficit hyperactivity disorder and autism diagnoses. Reference Razzaghi, Oster and Reefhuis13 Another study reported parental-reported symptoms related to attention deficit hyperactivity disorder, depression, and anxiety and found no difference between children and adolescents with CHD and a heart-healthy control group. Reference Eichler, Köhler-Jonas and Stonawski14 Two previous systematic reviews have investigated the evidence for psychiatric symptoms in patients with CHD. Reference Clancy, Jordan, de Weerth and Muscara15,Reference Latal, Helfricht, Fischer, Bauersfeld and Landolt16 The first review reported that young children aged 1–6 years old, who underwent early cardiac surgery, had a higher likelihood of experiencing challenges across larger symptom clusters (internal and external behaviour problems). Reference Clancy, Jordan, de Weerth and Muscara15 The second review observed an increased risk of psychological maladjustment, including internal and external symptoms combined, among children and adolescents (aged 2 to under 17 years old) with various CHDs, Reference Latal, Helfricht, Fischer, Bauersfeld and Landolt16 but to the best of our knowledge, there is no systematic review investigating the most common child psychiatric psychopathology, that is, autism, attention deficit hyperactivity disorder, and emotional disorders, both on a diagnostic and symptomatic level in children and adolescents with CHD. The conflicting results on specific psychiatric diagnoses and symptoms (autism, attention deficit hyperactivity disorder, and emotional disorders) emphasise the need for a systematic review with this as the focus.
The aim of this study is to review the current literature on psychiatric co-morbidities, both symptoms and diagnoses, in children and adolescents with CHD.
Materials and methods
The systematic review was conducted according to the 24-step guide by Muka et Al, European Journal of Epidemiology. Reference Muka, Glisic and Milic17 The search strategies were defined in collaboration with a medical librarian from the Danish Royal Library. Searches were performed in three electronic databases (PubMed, Embase, and PsychINFO) using subject headings (MeSH, EMTREE, and PsycINFO thesaurus) and free text searches related to CHDs (CHD), psychiatric disorders and symptoms, and children and adolescents. The search from PubMed is available in Supplementary Material (S1). The searches in the other databases were made in a similar way. Two searches were performed the first on November 26, 2021, and a second follow-up search on February 6, 2023. All articles from the searches were screened based on the following in- and exclusion criteria: Studies needed to involve (1) children and adolescents aged 2–18 years who were born with a structural CHD; (2) outcomes related to diagnoses and symptoms of attention deficit hyperactivity disorder, depression, anxiety, or autism; (3) a heart-healthy control group or a normative sample. Studies that only examined patients with a known syndrome (e.g. Trisomy21, 22q11), heart/lung transplant recipients, or were not written in English were excluded. Additionally, the reference lists of all selected articles were manually searched to identify additional studies (CB). The screening process of titles and abstracts was performed independently in pairs by three of the authors (CB: 100% of articles, SLJ: 50% of articles, JH: 50% of articles), followed by a second screening on full-text articles (CB: 100% articles, SLJ: 50% of articles, KB: 50% of articles).
Any conflicts that arose during the screening process were resolved by consensus. The researchers were not blinded to journal titles, authors, or authors’ affiliations during the screening procedure. Data extraction was performed using a specific extraction template (based on PICO) and an online tool (Covidence) by two researchers (CB, SLJ). The outcome of the studies had to be a number (%) of participants (CHD and controls) with a psychiatric diagnosis or a mean (SD) of psychiatric symptoms of participants (CHD and controls). The severity of CHD was grouped as simple, moderate, complex, and a mixed group using the 2018 AHH/ACA guideline published in the Journal of the American College of Cardiology. Reference Stout, Daniels and Aboulhosn18
To assess the risk of bias, we utilised the Newcastle-Ottawa scale, which uses a star-based rating system to evaluate articles. The domains of Selection, Comparability, and Outcome are each assigned stars, with a maximum of 8 stars attainable. A study receiving 1–3 stars is deemed unsatisfactory, 4 stars are deemed satisfactory, 5–6 stars are considered good, and 7–8 stars are considered very good.
Due to substantial differences in outcome measures, comparative data, and study design the studies were deemed too heterogeneous for a meaningful meta-analysis of effect. Consequently, a narrative synthesis of the results was conducted instead.
Results
We included 19 articles Reference Razzaghi, Oster and Reefhuis13,Reference Eichler, Köhler-Jonas and Stonawski14,Reference DeMaso, Labella and Taylor19–Reference So, Li and Ho34 from the first search and 1 extra from the second search Reference Lang, Gerlach and Plank35 (Fig 1). Two articles were based on the same study. Reference Eichler, Köhler-Jonas and Stonawski14,Reference Lang, Gerlach and Plank35 Together, the included articles encompassed a total of 8035 unique patients with CHD. Fourteen articles reported on symptoms, four reported on psychiatric diagnoses, and two reported on both symptoms and diagnosis (Supplementary Material S2). In the articles (n = 13) that investigated proxy-reported symptoms, all but one article had parents as proxies. Reference Eichler, Köhler-Jonas and Stonawski14,Reference DeMaso, Labella and Taylor19–Reference Holland, Cassidy and Stopp21,Reference DeMaso24,Reference Schaefer, von Rhein and Knirsch26–Reference Miatton, De Wolf, François, Thiery and Vingerhoets31,Reference Lang, Gerlach and Plank35 The last articles had both teachers and parents as proxies. Reference Yamada, Porter and Conway32 The majority of the articles used control groups from another study Reference Razzaghi, Oster and Reefhuis13,Reference DeMaso, Labella and Taylor19,Reference Holland, Cassidy and Stopp21,Reference Sigmon, Kelleman, Susi, Nylund and Oster22,Reference Fredriksen, Diseth and Thaulow25,Reference Lang, Gerlach and Plank35 or healthy matched controls (sex, age) recruited from a local community sample. Reference DeMaso24,Reference Guan, Liu, Wang, Han and Jin30–Reference So, Li and Ho34 Four articles compared with normative data, Reference Gupta, Mitchell, Michael Giuffre and Crawford20,Reference Schaefer, von Rhein and Knirsch26–Reference Hansen, Poole and Nguyen28 one article included controls with one inpatient or emergency department record, Reference Gonzalez, Kimbro and Cutitta36 one article used a National Health Insurance Research Database, Reference Tsao, Lee and Jeng23 one article used a sibling control, Reference Mccusker, Armstrong, Mullen, Doherty and Casey29 and one article did not specify how the control group was recruited. Reference Eichler, Köhler-Jonas and Stonawski14
Figure 1. PRISMA flow diagram.
Three articles reported on simple CHD (repaired ventricular septal defect), 3 on complex, and the remaining 14 reported on a mix of simple, moderate, and complex CHD.
Of the 20 included articles, six had psychiatric diagnoses as the outcome Reference Razzaghi, Oster and Reefhuis13,Reference DeMaso, Labella and Taylor19,Reference Holland, Cassidy and Stopp21–Reference Tsao, Lee and Jeng23,Reference Gonzalez, Kimbro and Cutitta36 and 15 had psychiatric symptoms measured by different questionnaires as the outcome. Reference Eichler, Köhler-Jonas and Stonawski14,Reference DeMaso, Labella and Taylor19,Reference Holland, Cassidy and Stopp21,Reference DeMaso24–Reference Lang, Gerlach and Plank35 Of the six articles that had a psychiatric diagnosis as the outcome, five looked at attention deficit hyperactivity disorder Reference Razzaghi, Oster and Reefhuis13,Reference DeMaso, Labella and Taylor19,Reference Holland, Cassidy and Stopp21,Reference Tsao, Lee and Jeng23,Reference Gonzalez, Kimbro and Cutitta36 three looked at autism Reference Razzaghi, Oster and Reefhuis13,Reference Sigmon, Kelleman, Susi, Nylund and Oster22,Reference Tsao, Lee and Jeng23, and three looked at emotional disorders. Reference DeMaso, Labella and Taylor19,Reference Holland, Cassidy and Stopp21,Reference Gonzalez, Kimbro and Cutitta36 Of the 15 articles that had psychiatric symptoms as the outcome, 13 looked at attention deficit hyperactivity disorder Reference Eichler, Köhler-Jonas and Stonawski14,Reference DeMaso, Labella and Taylor19,Reference Holland, Cassidy and Stopp21,Reference DeMaso24–Reference Yamada, Porter and Conway32,Reference Lang, Gerlach and Plank35 and 12 looked at emotional disorders. Reference Eichler, Köhler-Jonas and Stonawski14,Reference DeMaso, Labella and Taylor19–Reference Holland, Cassidy and Stopp21,Reference DeMaso24,Reference Fredriksen, Diseth and Thaulow25,Reference Mccusker, Armstrong, Mullen, Doherty and Casey29–Reference Miatton, De Wolf, François, Thiery and Vingerhoets31,Reference Luyckx, Rassart, Goossens, Apers, Oris and Moons33–Reference Lang, Gerlach and Plank35
A total of 10 articles were conducted in the United States of America or Canada, 7 articles were conducted in Europe, and the remaining 3 articles in Asia.
No studies scored below 4 stars on the adapted NOS for risk of bias assessment; therefore, all studies are deemed satisfactory or above. The average score was 6 corresponding to a good study (Supplementary Material S2).
Psychiatric co-morbidities
Attention Deficit Hyperactivity Disorder
All of the included five attention deficit hyperactivity disorder articles found that the children and adolescents with CHD had a higher prevalence of attention deficit hyperactivity disorder than a matched (sex and age) control group (Table 1). The CHD group was mixed in three articles and complex in two. In three of the five articles, the attention deficit hyperactivity disorder diagnosis was made by a trained physician. One article used diagnoses based on registries, and one used parent information about a formerly established attention deficit hyperactivity disorder diagnosis.
Table 1. Diagnosed psychiatric disorder
Autism
All three articles concerning a diagnosis of autism documented a higher prevalence in children and adolescents with CHD compared to controls (Table 1). All studies included mixed groups of CHD. Two studies used diagnoses based on registries and one study used autism diagnosis collected from a questionnaire answered by a knowledgeable adult to the child.
Emotional disorders
Two articles investigated both depression and anxiety separately. They found no difference in depression rates or anxiety rates between children and adolescents with CHD (complex types) and controls (Table 1). Both depression and anxiety were diagnosed from an evaluation made by a trained physician. One article investigated depression and anxiety together as one metric and found that children and adolescents with CHD (mixed types) had a higher prevalence of emotional disorders (depression/anxiety) compared to controls (Table 1). This article found the diagnosis of depression/anxiety through hospital registers and/or prescription of medication.
Psychiatric symptoms
Attention deficit hyperactivity disorder
Self-reported
More attention deficit hyperactivity disorder symptoms were reported in three out of six articles. Two articles found no difference, and one article described no difference in boys but found that girls with CHD had more self-reported attention deficit hyperactivity disorder symptoms (Table 2).
Table 2. Symptoms of related psychiatric disorder
CADS: Conners’ ADHD Rating Scale, YRS: Youth Report Scale, SDQ: Strengths and Difficulties Questionnaire, DYPSIPS-III: ICD-10/DSM-IV forms CADS: Conners’ ADHD Rating Scale, SDQ: Strengths and Difficulties Questionnaire, SNAP-IV: Swanson, Nolan and Pelham Questionnaire, DYPSIPS-II/III: ICD-10/DSM-IV forms CKI: ADS: Children’s Depression Inventory, CES-D: Center for Epidemiological Studies Depression Scale, DYPSIPS-III: ICD-10/DSM-IV forms CBCL: Child Behavior Checklist, DYPSIPS-II/III: ICD-10/DSM-IV forms RCMAS: The Revised Children’s Manifest Anxiety Scale, YRS: Youth Report Scale, DYPSIPS-III: ICD-10/DSM-IV forms CBCL: Child Behavior Checklist, STAI: The State-Trait Anxiety Inventory, DYPSIPS-II/III: ICD-10/DSM-IV forms.
Proxy-reported
Twelve articles investigated proxy-reported attention deficit hyperactivity disorder symptoms. Four of these articles also investigated self-reported attention deficit hyperactivity disorder symptoms. Eight of the twelve articles found that the proxies reported the CHD children to have more attention deficit hyperactivity disorder symptoms than the proxies of the heart-healthy peers. Four articles found no difference in attention deficit hyperactivity disorder symptoms between the CHD groups and the controls (all of the studies investigated simple CHD) (Table 2).
Depression
Self-reported
Six of the included articles investigated symptoms of depression through self-evaluation.
Three of the six articles found more symptoms of depression in the CHD compared to the controls. One article found no difference, and two articles found fewer symptoms of depression in the CHD group (Table 2).
Proxy-reported
Only one out of four articles found more symptoms of depression amongst CHD patients compared to controls (Table 2). Two of the articles that found no difference between the two groups are based on the same study. Reference Eichler, Köhler-Jonas and Stonawski14,Reference Lang, Gerlach and Plank35
Anxiety
Self-reported
Three of the five articles found that the CHD group self-reported more anxiety symptoms than the control group (3/3 of the studies on complex CHD). One article found that girls with CHD had fewer anxiety symptoms than girls without CHD, but found no difference in anxiety symptoms between boys with CHD and boys without CHD (mixed CHD). The article reporting on simple CHD found no difference between the two groups in self-reported anxiety symptoms (Table 2).
Proxy-reported
Five articles investigated proxy-reported symptoms of anxiety. All articles found that there was no difference in proxy-reported anxiety symptoms between CHD and controls, but the one study that used both The State-Trait Anxiety Inventory and Child Behavior Checklist found that the CHD group have more anxiety/depression symptoms than the control group in the CBCL questionnaire (Table 2). Two of the articles that found no difference between the two groups are based on the same study. Reference Eichler, Köhler-Jonas and Stonawski14,Reference Lang, Gerlach and Plank35
Discussion
In this systematic review, we found a consensus in the included articles for a higher prevalence of diagnosed attention deficit hyperactivity disorder (ranging between 1.4 and 9 times higher than controls) and autism (ranging between 1.8 and 5 times higher than controls) in children and adolescents with CHD, but inconsistent results regarding depression and anxiety. All the articles that investigated symptoms of attention deficit hyperactivity disorder in children and adolescents with complex CHD found that the children and adolescents with CHD (both by self-report and proxy-report) had more attention deficit hyperactivity disorder symptoms than controls. The evidence is less strong in children and adolescents with mixed CHD. We found no difference in attention deficit hyperactivity disorder symptoms between children and adolescents with simple CHD and controls. We found no consensus in the articles in this review investigating symptoms of depression (both self- and proxy-reports). Two of the three articles that found no difference between proxy-reported depression symptoms are based on the same study. Reference Eichler, Köhler-Jonas and Stonawski14,Reference Lang, Gerlach and Plank35 This should be taken into account when interpreting the results of course, but we still did not find any consensus on depression symptoms. Children and adolescents with complex CHD self-reported more anxiety symptoms than controls. Mixed and simple CHD groups had the same amount of anxiety symptoms (both self- and proxy-reported) as controls.
There could be different explanations as to why we found that some children and adolescents with CHD had a higher prevalence of attention deficit hyperactivity disorder (mixed/complex CHD) and autism (mixed) and not depression and anxiety compared to peers. One explanation could be that neurodevelopmental diseases, such as attention deficit hyperactivity disorder and autism, are characterised by symptoms and impaired functioning that often appears in early childhood, and are often diagnosed at an early age. Depression and anxiety disorders, on the other hand, are emotional disorders that often have a debut later in life. Reference Solmi, Radua and Olivola37 Since we only included ages up to 18 years, depression and anxiety may not have fully surfaced yet. A prevalence study from 2019 found that attention deficit hyperactivity disorder and autism were more prevalent in adolescents in general, but depression and anxiety grew in prevalence in adulthood. Reference Khanna, Duca, Kay, Shore, Kelly and Crume38
Another explanation could be that these two diagnoses, attention deficit hyperactivity disorder and autism, arise from atypical brain development. Reference Hernandez, Rudie, Green, Bookheimer and Dapretto39 Brain development is influenced by a variety of factors both before and after birth, which can impact the risk of neurodevelopmental disorders such as autism and attention deficit hyperactivity disorder. Reference Baumer and Spence40,Reference Kavanagh, Rafiq and Iqbal41 Genetics, immunological dysregulation, metabolic disturbances, disturbances in the blood-gut axis, and early brain injury (both developmental and structural lesions) are suggested aetiological pathways for neurodevelopmental disorders. Some of the same mechanisms are found to be affected in patients with CHD. Reference Nielsen, Ellesøe, Larsen, Hjortdal and Nyboe42–Reference Omann, Nyboe and Kristensen47 As many of the above-mentioned factors are potentially more disturbed in complex CHD both prenatal with desaturated blood in the developing brain Reference Donofrio, Bremer and Schieken48,Reference Arduini, Rosati and Caforio49 and postnatal with early, difficult, and long-lasting surgeries, it is not surprising that it is in this patient group that we found an increased risk for attention deficit hyperactivity disorder. The advancements in the treatment and clinical care of complex CHD have led to a decrease in post-operative brain injuries, injuries which are possible contributors to neurodevelopmental co-morbidities. Reference Peyvandi, Xu and Barkovich50 It will be interesting to see if this has an effect on the increased prevalence of attention deficit hyperactivity disorder found in this patient group. It is also in this patient group with complex CHD, that most of the psychiatry research has been done so far. Psychiatric co-morbidities in children with simple CHD are not investigated to the same extent. As studies in adults with simple CHD have found a greater risk for neurodevelopmental challenges, social challenges, and a greater risk for psychiatric co-morbidities, in general, Reference Asschenfeldt, Evald and Heiberg5,Reference Nyboe, Udholm, Larsen, Rask, Redington and Hjortdal7,Reference Nyboe, Fonager, Larsen, Andreasen, Lundbye-Christensen and Hjortdal8,Reference Olsen, Sørensen, Hjortdal, Christensen and Pedersen51 further and larger studies in this group of children and adolescents are important.
Children and adolescents with complex CHD self-reported more symptoms of anxiety.
Anxiety is not always notable for the outside world; this could explain why it is only in the self-reported data that we find this difference. We have found the same pattern looking at adults with simple CHD where symptoms like inattention are difficult for a proxy to recognise. Reference Lau-Jensen, Asschenfeldt, Evald and Hjortdal6,Reference El Dabagh, Asschenfeldt, Kelly, Evald and Hjortdal52 This has also been found in a study investigating other chronic diseases in childhood. Reference Samuels, Menand and Mauer53
As far as we know, this systematic review is the first that looked at diagnosed psychiatric disorders and more specific psychiatric symptoms in children and adolescents born with a CHD. We included 20 articles (19 studies) from Europe, Asia, and North America. All studies compared the study group with a healthy control group or with normative data. All studies were scored satisfactory or above in a risk of bias assessment. A limitation of our study is the lack of a meta-analysis of the studies included. The included studies in this research vary significantly in their outcome measures, study design, and comparative data. This heterogeneity made it difficult to conduct a meaningful meta-analysis of the effects; therefore, a meta-analysis was not deemed feasible. The studies focused on different types of CHD patients, ranging from simple to complex, with some including both types of CHD in a mixed group. CHD is a highly heterogeneous congenital malformation. The studies also used different reference groups to compare the CHD group, including community-based control groups, children with one emergency/inpatient record, control groups from prior studies, siblings, insurance database controls, and normative data. The age range of participants across the studies was also wide, spanning from 2 to 18 years old. The outcome of interest was obtained through different methods across the studies, including psychiatric diagnoses through psychiatric evaluations, registry data, a combination of registered diagnoses and prescription medicine, and parents’ reports of psychiatric diagnoses. Studies that used psychiatric symptoms as the outcome used different survey measures.
Conclusion
In this systematic review, we found that children and adolescents with a CHD had a higher prevalence of attention deficit hyperactivity disorder and autism than controls, but inconsistent results regarding depression and anxiety. Focus on and subsequent interventions on psychiatric symptoms may have the potential to decrease the known risks of long-term social problems found in this patient group. A large and important aspect of further research in this area would be to make more comparable research. This could be done by not mixing very different CHD groups, but by looking at simple, moderate, or complex CHD separately and by making reasonable age intervals, for example preschool-age, early school-age, late school-age, and later adolescents.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1047951123003013
Acknowledgements
None.
Financial support
This work was supported by the Danish Heart Foundation (S.H.L and J.L.H 20-R143-A9431-22160).
Competing interests
None.
Open access
