European Heart Journal (1998) 19, 647651

Psychopathology in young adults with congenital heart disease

Follow-up results
E. M. Utens*, H. J. Versluis-Den Bieman*, F. C. Verhulst*, F. J. Meijboom, R. A. Erdman and J. Hess
*Department of Child and Adolescent Psychiatry; Department of Paediatric Cardiology, Sophia Childrens Hospital, Erasmus University Rotterdam; Department of Medical Psychology and Psychotherapy; Department of Cardiology, Erasmus University Rotterdam, Rotterdam, The Netherlands

Aims The aim of the present study was to assess the occurrence of a wide range of behavioural and emotional problems long-term (>9 years) after surgical correction for congenital heart disease in infancy and childhood. Methods The problem scores on the Young Adult Self-Report of 166 1925-year-old adults with congenital heart disease were compared with those of subjects of similar age from the general population. Results On most Young Adult Self-Report scales no dierences were found between the mean problem scores of the congenital heart disease-adults and reference peers. On only two Young Adult Self-Report scales (i.e. Somatic Complaints and Strange) and the total problem score were signicant though small dierences found between the

mean problem scores of the congenital heart disease adults and reference peers. No signicant relationship was found between cardiac diagnosis and problem behaviours in congenital heart disease adults. No relationship was found between IQ scores and problem behaviours in congenital heart disease adults. Conclusion Overall, it can be concluded that the results of the congenital heart disease adults were not unfavourable, since the dierences in mean (total) problem scores between the congenital heart disease sample and reference group were small and limited to only two specic problem areas. (Eur Heart J 1998; 19: 647651) Key Words: Psychopathology, congenital heart disease, adults, follow-up.

Introduction
Advances in cardiovascular diagnostic and surgical techniques have resulted in greatly increased survival in patients with congenital heart disease over the last three decades. However, little is known about the long-term psychosocial eects of surgery for congenital heart disease in adulthood. The few follow-up studies conducted in this eld have yielded contradictory results. Furthermore, it is dicult to draw rm conclusions from existing studies, since they often
Revision submitted 30 September 1997, and accepted 13 October 1997. Correspondence: E. M. Utens, Ph.D., Department of Child and Adolescent Psychiatry, Sophias Children Hospital, Erasmus University Rotterdam, Dr. Molewaterplein 60, 3015 GJ Rotterdam, The Netherlands. 0195-668X/98/040647+05 $18.00/0 hj970824

contain methodological weaknesses, such as small sample size, heterogeneous sample composition, low response rates and non-standardized assessment procedures[1,2]. The present study is part of a larger descriptive and explorative follow-up study concerning long-term medical and psychosocial outcome in children, adolescents and young adults who underwent surgery for congenital heart disease[1,2]. The aims of the present study were: (1) to compare the level of behavioural/ emotional problems of 1925-year-old congenital heart disease patients with that of a reference group of similar age, (2) to determine the relationship between cardiac diagnosis and problem behaviours, (3) to determine the relationship between intellectual functioning and problem behaviours. The strengths of the present study are the use of a large consecutive series of patients and a standardized assessment procedure.
 1998 The European Society of Cardiology

648

E. M. Utens et al.

Methods Description of the samples

Inclusion criteria The present study population was derived from the larger follow-up study mentioned above. In this larger follow-up study all consecutive patients who underwent their rst open heart surgery for congenital heart disease between 1 January 1968 and 1 January 1980 in the University Hospital of Rotterdam, and who were younger than 15 years at the time of surgery, were eligible for a follow-up, which took place in the same hospital (19891991) (for detailed description[1,2]). For the present study, congenital heart disease patients who were 1925 years old at follow-up were selected, since for this age-range Dutch normative reference data were available. Patient sample At follow-up, of the eligible 1925-year-old congenital heart disease-adults (n=250) 29 patients had died, 16 were lost to follow-up and 32 refused to participate. Two mentally retarded patients were unable to complete the self-report assessment instrument. Five patients, who preferred to complete the self-report at home, did not return it (see assessment procedure). The present patient sample consisted of the remaining 166 congenital heart disease adults (mean age: 217 years). The response rate, corrected for deceased patients and persons lost to follow-up, was 809%. Within the congenital heart disease sample a dierentiation was made between ve diagnostic groups: atrial septal secundum defect, ventricular septal defect, tetralogy of Fallot, transposition of the great arteries and pulmonary stenosis (numbers and mean ages respectively: n=50, n=40, n=28, n=13, n=15 and 217, 216, 220, 208 and 222 years of age). The results of 20 patients (mean age: 218 years) were excluded from the statistical analyses because they belonged to a miscellaneous diagnostic group, consisting of small numbers of patients with a variety of congenital heart defects. The Dutch normative reference group (mean age: 216 years), from which data were obtained in 1991, is described in detail elsewhere[3].

Table 1 Percentages of adults in the congenital heart disease sample and reference group scoring in the Young Adult Self-Report psychopathological range
% >Cut-o* Reference (n) Males 1921 2225 Females 1921 2225 ConHD (n) Binomial test P

year** year** year year

90 (156) 90 (166) 97 (195) 95 (189) 96 (706)

163 (43) 220 (41) 125 (40) 167 (42) 157 (166)

ns <002 ns ns <002

Total sample

*The 90th percentile of the cumulative frequency distribution of the Young Adult Self-Report total problem scores obtained from the reference group was chosen as the cut-o point to distinguish problem from non-problem adults. Total problem scores exceeding this cut-o fall in the psychopathological range. Due to gaps between rank-ordered scores, the percentages of reference adults scoring above the 90th percentile were not exactly 10%. **The age categories were based on the median split of the patient sample. ConHD=congenital heart disease.

neither to Internalizing nor to Externalizing. To assess intelligence, the Groninger Intelligence Test-short form[9], a standardized Dutch intelligence test, was used.

Assessment procedure
All patients were traced, approached uniformly and signed an informed consent before participating. Patients were tested by a clinical psychologist (E.U.), who was blind to their cardiac diagnosis. Thirteen patients preferred to complete the Young Adult SelfReport at home and returned it by mail. In these cases, intelligence tests could not be performed.

Results Proportion of problem adults

In Table 1, the proportions of problem adults in the congenital heart disease sample and reference group are shown. Problem adults are dened here as persons scoring in the psychopathological range. Scores in this range reect levels of problem behaviour similar to those of young adults typically referred to mental health departments.

Instruments
The Young Adult Self-Report[4], a reliable and valid[58] standardized self-report for ages 1830, was used to assess the occurrence of a broad range of psychopathology. It contains 110 problem items and measures two broad areas: Internalizing problems (consisting of the scales: Anxious/Depressed, Withdrawn and Somatic Complaints) and Externalizing problems (consisting of: Shows O, Delinquent Behaviour and Aggressive Behaviour). Further, the Young Adult Self-Report contains the scale: Strange (concerning obsessions, hallucinations, strange behaviour and ideas), which belongs
Eur Heart J, Vol. 19, April 1998

Mean problem scores

To test dierences in mean total problem scores and mean scale scores of both groups, analyses of variance (ANOVAs) were computed in a group (congenital heart disease vs reference) sex (males vs females) age

Psychopathology congenital heart disease

649

Table 2 Dierences in mean problem scores for Young Adult Self-Report scales and total problem scores of congenital heart disease and reference young adults
Mean problem scores Syndromes ConHD (n=166) 30 18 81 27 35 30 27 137 83 359 Reference (n=706) 30 18 73 20 31 29 20 122 79 302 Group* (%)

Aggressive behaviour Delinquent behaviour Anxious/depressed Somatic complaints Shows o Withdrawn Strange Internalizing Externalizing Total problem score***

18 21 06** 17

An ANOVA computed on the total problem score and all scale scores of the congenital heart disease adults, with sex and age as independent variables, showed signicant main eects of sex on the scales Delinquent Behaviour (congenital heart disease males scoring higher than congenital heart disease females) and Somatic Complaints (congenital heart disease females scoring higher than congenital heart disease males), accounting for, respectively, 120% and 72% of variance. No main eects of age on the mean (total) problem scores were found within the congenital heart disease sample. No two-way interaction was found.

Diagnostic groups
ANOVAs computed with cardiac diagnosis, sex and age as independent variables showed that cardiac diagnosis had no main eect on any of the scales, or on the total problem score (Table 3). This indicates that no signicant dierences were found between the mean (total) problem scores of dierent cardiac diagnostic groups. No signicant two- or three-way interactions were found.

*Percentage of variance accounted for by signicant group eect (congenital heart disease vs reference), (P<0005, applying Bonferroni correction for 10 comparisons). **Not signicant after Bonferroni correction. ***The total problem score is derived by summing the scores for all 110 problem items. The response format is 0=not true, 1=somewhat or sometimes true, and 2=very true or often true. ConHD=congenital heart disease.

(1921 vs 2225 years) factorial design. To correct for chance ndings, a Bonferroni correction (for 10 comparisons) was applied on all analyses of variance executed. In Table 2, the magnitude of signicant group eects (congenital heart disease vs reference) is indicated in terms of percentage of variance accounted for. According to Cohen[10], eects accounting for 1059% of variance are considered small, 59138% medium and >138% large. On the total problem score and on only Somatic Complaints and Strange were signicant though small group eects found. No signicant two- or three-way interactions were found.

Intellectual functioning
Five mentally retarded patients could not complete the IQ test and 16 could not complete it for practical reasons (see procedure). The mean IQ score of the congenital heart disease-sample was: x=105, SD=15, n=145. Of the total congenital heart disease sample, 21 patients (127%; n=166) were mentally retarded or had borderline intellectual functioning (IQ scores, respectively, <70 and 7085). No signicant (Pearson product moment) correlation was found between Young Adult SelfReport total problem scores and IQ scores.

Table 3 Dierences in mean problem scores for Young Adult Self-Report syndromes and total problem scores across cardiac diagnostic groups in ANOVAs
Cardiac diagnosis (n=146) Syndromes (n) ASD (50) 28 18 83 28 34 31 28 142 80 364 VSD (40) 27 17 71 26 38 33 25 128 82 355 F4 (28) 33 18 96 30 32 26 34 150 88 395 TGA (13) 31 22 76 29 43 31 39 132 96 377 PS (15) 31 17 77 26 28 29 17 133 76 311 Main eects

Aggressive behaviour Delinquent behaviour Anxious/depressed Somatic complaints Shows o Withdrawn Strange Internalizing Externalizing Total problem score

ns ns ns ns ns ns ns ns ns ns

ASD-atrial septal secundum defect; VSD=ventricular septal defect; F4=tetralogy of Fallot; TGA=transposition of the great arteries; PS=pulmonary stenosis. Eur Heart J, Vol. 19, April 1998

650

E. M. Utens et al. congenital heart disease and reference adults in reporting somatic complaints. Further, our nding, that a signicant dierence was found on the syndrome Strange, is in line with Brandhagen et al.s ndings on the scale obsessivecompulsive. The response rate in Brandhagen et al.s study was low: 363%. We found no relationship between problem behaviour and dierent cardiac diagnoses. Our results conrm those of previous studies[1,2,12,13]. To examine further to what extent the dierence between the total problem scores of the congenital heart disease and reference group could be attributed to somatic problems, an ANOVA was again computed, but now 15 items with a somatic content were excluded. Again a signicant group eect was found (23% of variance accounted for, P<001), which indicated that the higher Young Adult Self-Report total problem score of the congenital heart disease adults could only partly be explained by somatic complaints. (The percentage of variance accounted for increased by 06%, which can be attributed to the fact that the mean total problem scores and corresponding variance diered when somatic items were included or excluded). In several follow-up studies[12,14,15], the educational level of congenital heart disease adults was above average. In none of these studies, were IQ tests performed. In the present study, the IQ scores of the reference group were not available. However, since the reference group consisted of a random general population sample, we presumed that its distribution of IQ scores was normal (mean IQ=100, 165% scoring <86). Since the mean IQ score of our congenital heart disease sample was 105 (n=145) and 127% showed borderline intellectual functioning or mental retardation, we conclude that our overall results regarding intellectual functioning are favourable. However, the results should be interpreted with caution since a relatively large proportion of IQ scores is missing and patients with Downs syndrome were not included, since it was the policy of the cardiac unit not to operate on such patients in those years. No relationship was found between problem behaviours and IQ scores of congenital heart disease adults, which conrms the results of our previous study[1]. The dierence in the IQ scores of the dierent diagnostic groups in the present study should be carefully interpreted due to missing observations and consequently small cells. The strengths of the present study were: the high response rate, the use of a normative reference group, a psychometrically sound assessment procedure and the dierentiation according to diagnostic groups. However, one should refrain from generalizing our results to congenital heart disease children operated upon nowadays, since the medical and surgical treatment of children with a congenital heart defect has been strongly improved over the past 20 years. Future research should take potential risk factors[1] for developing later adjustment problems (e.g. number of operations, age at surgical repair and present health status) into consideration so that adult patients

The mean IQ scores for the dierent cardiac diagnostic groups were: atrial septal secundum defect: mean IQ=107 12 (3); ventricular septal defect: mean IQ=106 18 (7); tetralogy of Fallot: mean IQ=96 18 (4); transposition of the great arteries: mean IQ=96 17 (2), pulmonary stenosis: mean IQ=113 8 (3) (numbers in parentheses=missing scores). An ANOVA showed a signicant (P<001) dierence between the IQ scores of dierent cardiac diagnostic groups; patients with cyanotic heart defects (tetralogy of Fallot, transposition of the great arteries) obtained the lowest mean IQ-scores.

Discussion
In this study, the occurrence of a broad range of clinically relevant problem behaviours was assessed in a standardized manner. Overall, it can be concluded that our results reect quite satisfactory psychosocial functioning of congenital heart disease adults since the dierences in mean (total) problem scores between the congenital heart disease sample and the reference group were small and limited to only two specic problem areas, Somatic Complaints and Strange. The nding that more young congenital heart disease adults scored in the psychopathological range than reference peers (157% vs 96%) appeared to be determined mainly by problem reporting of the 2225-year-old congenital heart disease males. The small dierence between the congenital heart disease sample and reference group on Somatic Complaints can be explained by assuming that congenital heart disease adults, due to alertness to bodily symptoms, have a tendency to report these symptoms more often. Besides, some somatic complaints (such as heart pounding) could be caused by the cardiac defect itself. As to the scale designated as Strange, further statistical analysis showed that the congenital heart disease sample scored signicantly higher on only three items reecting obsessive and strange thoughts. Healthrelated worries (scar and memory of the cardiac operation) were reported on these items several times. We reported earlier[2] that congenital heart disease adolescents (1118 years old) showed signicantly more problems than reference peers of a similar age on all scales of the Youth Self-Report[11], the adolescent version of the Young Adult Self-Report. Although caution should be adopted when comparing the results of these congenital heart disease adolescents with the present congenital heart disease adults, since the data are obtained by cross-sectional and not by longitudinal studies, it can be concluded that behavioural and emotional problems seem to decrease as patients grow older. The present studys results contrast with those of Brandhagen et al.[12], who reported a broad range of neurotic symptoms in congenital heart disease adults, whereas we found higher problem-reporting in only two specic areas. In both Brandhagen et al.s study and ours, a signicant dierence was found between
Eur Heart J, Vol. 19, April 1998

Psychopathology congenital heart disease who are at risk for developing psychopathological problems can be identied and preventive interventions be developed.
The present Rotterdam Quality of Life Study was supported by grants from the Netherlands Heart Foundation (no. JH 8802). The normative data were gathered by means of grants from the Sophia Foundation for Medical Research, and the Dutch National Fund for Mental Health. We thank Dr Robert Ferdinand for his advice concerning data analysis.

651

[7]

[8]

[9]

References
[1] Utens EMWJ, Verhulst FC, Erdman RAM et al. Psychosocial functioning of young adults after surgical correction for congenital heart disease in childhood: A follow-up study. J Psychosom Res 1994; 38: 74558. [2] Utens EMWJ, Verhulst FC, Meijboom FJ et al. Behavioural and emotional problems in children and adolescents with congenital heart disease. Psychol Med 1993; 23: 41524. [3] Ferdinand RF, Verhulst FC. The prevalence of self-reported problems in young adults from the general population. Soc Psychiatry Psychiatric Epidemiol 1996; 31: 1020. [4] Achenbach TM. The Young Adult Self-Report. Burlington, VT: University of Vermont, Department of Psychiatry, 1990. [5] Achenbach TM, Howell CT, McConaughy SH, Stanger C. Six-year predictors of problems in a national sample: III. Transitions to young adults syndromes. J Am Acad Child Adolesc Psychiatry 1995; 34: 65869. [6] Ferdinand RF, Verhulst FC. The prediction of poor outcome in young adults: Comparison of the Young Adult Self-Report,

[10] [11] [12] [13] [14]

[15]

the General Health Questionnaire, and the Symptom Checklist. Acta Psychiatr Scand 1994; 89: 40510. Ferdinand RF, Verhulst FC, Wiznitzer M. Continuity and change of self-reported problem behaviors from adolescence into young adulthood. J Am Acad Child Adolesc Psychiatry 1995; 34: 68090. Wiznitzer M, Verhulst FC, Van Den Brink W et al. Detecting psychopathology in young adults: a comparison of the Young Adult Self-Report, the General Health Questionnaire and the Symptom Checklist as screening instruments. Acta Psychiatr Scand 1992; 86: 327. Kooreman A, Luteijn F. Groninger Intelligentie Test. GIT. Schriftelijke Verkorte Vorm. Lisse, the Netherlands: Swets & Zeitlinger, 1987. Cohen J. Statistical Power for the Behavioural Sciences. New York: Academic Press, 1977. Achenbach TM. Integrative guide to the 1991 CBCL 4-18, YSR and TRF Proles. Burlington, VT: University of Vermont, Department of Psychiatry, 1991. Brandhagen DJ, Feldt RH, Williams DE. Long-term psychologic implications of congenital heart disease: A 25-year follow-up. Mayo Clinic Proc 1991; 66: 4749. DeMaso DR, Beardslee WR, Silbert AR, Fyler DC. Psychological functioning in children with cyanotic heart defects. J Devel Behav Pediatrics 1990; 11: 28994. Kaemmerer H, Tintner H, Konig U, Fritsch J, Sechtem U, Hopp HW. Psychosoziale Probleme Jugendlicher und Erwachsener mit angeborenen Herzfehlern. Zeitschrift fur Kardiologie 1994; 83: 194200. Otterstad JE, Tjore I, Sundby P. Social function of adults with isolated ventricular septal defects. Scand J Soc Med 1986; 14: 1523.

Eur Heart J, Vol. 19, April 1998