Determinants of symptom profile and severity of conduct disorder in a tertiary level pediatric care set up: A pilot study.

BACKGROUND: Conduct disorders (CDs) are one of the most common causes for referral to child and adolescent mental health centers. CD varies in its environmental factors, symptom profile, severity, co-morbidity, and functional impairment. AIMS: The aim was to analyze the determinants of symptom profile and severity among childhood and adolescent onset CD. SETTINGS AND
DESIGN: Clinic based study with 60 consecutive children between 6 and 18 years of age satisfying International Classification of Disease-10 Development Control Rules guidelines for CD, attending behavioral pediatrics unit outpatient.
MATERIALS AND METHODS: The family psychopathology, symptom severity, and functional level were assessed using parent interview schedule, revised behavioral problem checklist and Children's Global Assessment Scale. STATISTICAL ANALYSIS: The correlation and predictive power of the variables were analyzed using SPSS 16.0 version.
RESULTS: There was significant male dominance (88.3%) with boy girl ratio 7.5:1. Most common comorbidity noticed was hyperkinetic disorders (45%). Childhood onset group was more predominant (70%). Prevalence of comorbidity was more among early onset group (66.7%) than the late-onset group (33.3%). The family psychopathology, symptom severity, and the functional impairment were significantly higher in the childhood onset group.
CONCLUSION: The determinants of symptom profile and severity are early onset (childhood onset CD), nature, and quantity of family psychopathology, prevalence, and type of comorbidity and nature of symptom profile itself. The family psychopathology is positively correlated with the symptom severity and negatively correlated with the functional level of the children with CD. The symptom severity was negatively correlated with the functional level of the child with CD.

INTRODUCTION

Behavioral problems are one of the most common referral causes to child guidance clinics. Usually they are subjected to all kinds of severe punishments. They will be labeled as “problem child” and alienated from all spheres. It aggravates situation and creates a vicious cycle. All those who come in proximity with them are under stress.

As per International Classification of Disease-10 (ICD-10)[1] conduct disorder (CD) is characterized by repetitive and persistent pattern of dissocial, aggressive or defiant conduct. At its extreme form, behavior should amount to major violation of age appropriate social expectations with an enduring pattern of 6 months or longer. If behavioral problems are characterized by argumentativeness, but absence of more severe dissocial or aggressive acts, below age of 9 or 10 years, it is called oppositional defiant disorder (ODD).[1]

Conduct disorder classified as childhood and adolescence onset type if behavioral problems occur before or after age of 10 years.[23]

Aim and objectives

To define the symptom profile of CD

To compare the symptom profile and co-morbidity pattern among childhood and adolescent onset CD

To analyze factors affecting the symptom profile and its severity among childhood and adolescent onset CD.

MATERIALS AND METHODS

Study design

Descriptive outpatient (OP) clinic based pilot study.

Study population

A total of 60 consecutive children between ages of 6 and 18 years attending the behavioral pediatrics unit OP who satisfied the ICD-10 Development Control Rules (DCR) diagnostic guidelines for CD were included in the study population.

Behavioral pediatrics unit is an exclusive child and adolescent mental health service unit, headed by a senior consultant who qualified in both pediatrics and psychiatry under Department of Child Health, SAT Hospital, Government Medical College, Trivandrum, Kerala, and South India.

Study period

2012 January to 2012 December.

Exclusion criteria

Children with neurological illness, including head injury, seizure disorder, cerebral palsy, mental retardation, psychosis, mania, and depressive disorder were excluded.

Instruments

International Classification of Disease-10 DCR, 1993 (WHO 23-item symptom list of operational diagnostic guideline for CD for research).

Parent interview schedule (PIS)[4] is a 9-item scale to assess and quantify the abnormal psycho social situation (Axis-V) or the family psychopathology.

Revised behavioral problem checklist (RBPC)[5] to quantify the behavioral symptoms among the children with CD.

Children's Global Assessment Scale (CGAS)[6] to assess the functional level, which ranges from 0 to 100.

Revised behavioral problem checklist use weighted scoring. Each item is given a scoring of 0 (not a problem), 1 (mild problem), and 2 (severe problem). It contains 89-item, and six scales namely CD having 22-item; socialized aggression (SA) 17-item; attention problems 16-item; anxiety-withdrawal 11-item; psychotic behavior 6-item, and motor tension-excess 5-item. Twelve-items which are named under not relevant category because factor analysis showed these items did not load significantly on identified RBPC factors or they loaded on more than one factor.

Procedure

Consent was taken from the parent and older children. Detailed history taking, observation and evaluation were done using the intake proforma. Mother was the key informant. Child, mother, father, and other key players in the family were interviewed individually and together. School teachers were also interviewed in indicated cases. The diagnosis was done based on the ICD-10 DCR guidelines. The 23 symptom list of ICD-10, DCR were either marked as self-reported, elicited, and not complained. Family psychopathology including abnormal psycho social situation assessed using the PIS. IQ assessment was done for children with the clinical diagnosis of specific learning disability. Behavioral symptom rating was done using the RBPC. Children global assessment function was assessed using the CGAS. All the data were entered in to the coded intake proforma.

In order to bring about cultural correlation, the tool RBPC was revalidated as part of this study before using it for assessing the symptom severity (the average intra class correlation coefficient for the inter rater reliability was 0.9981 with 95% confidence interval (CI) 0.9967-0.9997. The average intra class correlation coefficient for the test retest reliability was 0.9118 with 95% CI, 0.8452-0.9597. Correlation of RBPC with CGAS and PIS were analyzed using Spearman's rho; with CGAS the Spearman's rho is −0.578 with P - 0.008 at 0.05 level and with PIS the Spearman's rho are 0.517 with P - 0.019 at 0.05 levels.

Following the baseline rating of symptom severity and global assessment of functional level the comprehensive intervention package with the cognitive problem solving skill therapy and parent management training to the children and their parents respectively, were administered. Pharmacotherapy was given to children based on the symptoms of violence and comorbidity. Comorbid conditions were managed accordingly.

Statistical analysis was done using Statistical Package for Social Sciences. SPSS Inc. Released 2007. SPSS for Windows, Version 16.0. Chicago, SPSS Inc.

RESULTS

Among the study population, there was boys’ domination with 88.3% and girls 11.7%.

Regression equations

Prediction of revised behavioral problem checklist (T) score on parent interview schedule

Revised behavioral problem checklist (T) score (y) =36.870 + 4.038 × PIS score (t = 7.868, P = 0.000) r2 = 0.516, that is, 51.6% prediction.

Regression analysis shows that predictive power of family psychopathology on symptom severity of children with CD is 51.6%.

Prediction of children's global assessment scale on parent interview schedule

Children's Global Assessment Scale score (y) = 57.405-0.935 × PIS score (t = −4.506, P = 0.000) r2 = 0.259, that is, 25.9% prediction.

The regression analysis shows that the predictive power of family psychopathology on functional level of the children with CD is 25.9%.

Prediction of children's global assessment scale on revised behavioral problem checklist (t) score

Children's Global Assessment Scale score (y) = 60.713-0.164 × RBPC (Total) score (t = −4.408, P = 0.000) r2 = 0.251, that is. 25.1% prediction.

The regression analysis shows that the predictive power of symptom severity of children with CD on functional impairment is 25.1%.

DISCUSSION

Prevalence of CD is between 5% and 10% in the industrialized western world.[7] Boys affected more than girls.

The prevalence varies among the Indian studies, Deivasigamani (11.13%),[8] and Sarkar et al. (7.1%.).[9] Malhothra et al.[10] had reported a prevalence of 4.94% in a retrospective clinical study. Srinath et al.[11] had given a low prevalence of 0.2% in an epidemiological study. Sarkhel et al.[12] had reported a prevalence rate of 4.58% school going population.

In the present study, children with CD were more commonly seen among the primary and upper primary age group than the later childhood in the present study. The CD will usually become manifest and problematic in family, school, and peer group during this developmental period. About 28% of the study population belongs to 6-8 years of age group since it was conducted in the pediatric tertiary care referral center. There was significant male dominance (88.3%) in the present study [Table 1] with boy girl ratio is 7.5:1. The predominance of boys were a little higher than in other studies.[712] The reason for this could be due to the difference in the clinic and community population and the service seeking attitude.

Table 1

Age distribution of the study group (n=60)

Among the study population, the rural (60%), and semi urban (36.7%) were the major group. The majority of the parents was having only primary education (father 68.3% and mother 53.3%) and mainly engages in the manual labor (father 40% and mother 16.7%). Mother was housewife among 61.7% of the population. Both income groups of above poverty line (APL) (53.3%) and below poverty line (46.7%) category were more or less equally present in the population. But the studies show that children with CD, while present in all economic levels it is more common among urban low income communities.[1314] This is due to the fact that most of our manual laborers now come under the definition of APL category. The predominant living arrangement was the nuclear type of family (78.3%), which reflects the current social change. The primary care giver was the biological parents in 71.7%. However, inconsistent parental availability and discipline were the common feature noticed western studies.[1415]

The family history of mental health problems was present in the 55% of the study population. The main problems noticed were alcoholism (25%), mental illness (16.7%), and domestic violence (11.67%).

Symptoms profile analysis revealed that among the self-reported group of symptoms the most frequently repeated five were frequent or severe temper tantrums (91.7%), often argues with adults (90.0%), actively refuses adults’ requests or defies rules (88.3%), frequently initiates physical fights (76.7%), and exhibits physical cruelty to other people (66.7%). Among the elicited groups of symptoms the most frequently repeated five were “touchy” or easily annoyed by others (73.3%), blames others for his or her own mistakes or misbehavior (51.7%), deliberately destroys the property of others (50%), often angry or resentful (36.7%), and frequently bullies others (36.7%). The less frequently repeated five were as follows; forces another person into sexual activity (0% - not complained at all), breaks into someone else's house, building, or car (3.3%), commits a crime involving confrontation with the victim (5%), deliberately sets fires with a risk or intention of causing serious damage (8.3%), and has run away from parental or parental surrogate home at least twice or has run away once for more than a single night (10%). The symptom “forces another person in to sexual activity” was not there in any children in the previous Indian study also.[12] The symptom profile in the present study clearly delineates the nature of CD from the western literature where more aggression, violence, forced sexual activity, and problems with law enforcement authority are seen. The majority of the children with CD in the present study came under the moderate degree of severity (76.7%) with problems in the family, school and peer group, followed by mild degree among 13.3%, and severe among 10% cases. There were only three cases that were registered as children in conflict with law in the present study. The degree of severity in the present study with clinic population is higher compared with the previous study.[12]

Among the family psychopathology (PIS) the most frequently noticed five problems were as follows; physical child abuse (75%), intra-familial discord among adults (61.7%), inadequate parental supervision/control (58.3%), inadequate or distorted intra-familial communication (56.7%), and discordant relationship with peers (51.7%). It is quantified [Table 2]. The influence of environmental factors like problems in the child rearing practices such as poor parental supervision, harsh discipline, broken family, single parent, physical, and sexual abuse, poverty, alcohol and violent behavior of the parent will produce mal adaptive behavior in children and adolescents.[16] The extensive body of research has documented a large number of dispositional and contextual risk factors that can play a role in the etiology of this disorder. A developmental theory model was proposed.[17] The nature and quantity of family psychopathology influenced the symptom severity of CD.[1819] Nature of the family psychopathology in the present study reveals that our issues are different from that of the western part of the world as mentioned earlier.[1415]

Table 2

Abnormal psychosocial situation (Axis-V) (PIS) and functional impairment (CGAS) among the study group (n=60)

The symptom profile of CD reflects the heterogeneity (multiple behavioral domains) in its presentation and its severity [Table 3]. In one child it is aggression and violence where as in another, the problem is lying, stealing, and SA. Apart from this there were children with mixed symptoms in the multiple behavioral domains of CD. This complicates the clinical picture, intervention, and prognosis. The variations in these behavioral domains in its presentation and severity have to be considered for planning the intervention and predicting the prognosis.

Table 3

Symptom profile and severity of CD among the study group (RBPC score, n=60)

The mean level of functional impairment [Table 2] is moderate which reflects that there is interference in functioning with most of the social areas or severe impairment in one area.

Among the subtypes commonest was the CD confined to family (28.3%) followed by ODD (20%). This pattern is unique to Indian situation.

The most common comorbidity noticed in the present study is hyperkinetic disorders (45%) the prevalence of which stands in between the previous studies.[121819] The comorbidity with attention-deficit hyperactivity disorder (ADHD) increases the severity of the CD.[16202122] The ADHD, ODD, and CD are together described as disruptive behavioral disorders. ODD is a relatively benign disorder with good prognosis.[23] Other important comorbidities were specific learning disability (16.7%) and mixed disorders of emotion and conduct (10%). The conditions like somatoform disorder and sibling rivalry were also noted. Multiple comorbidity was present in some cases.

Among CD, childhood onset group was more predominant (70%) than adolescent onset ones, which were reported similarly in the previous study.[12] The male dominance continues the same pattern in early onset (childhood) and late onset (adolescent) groups [Table 4].

Table 4

Sex distribution among the diagnostic types the study group (n=60)

Comparison of the childhood and adolescent onset group in the present study reveals that there was marked difference in the centers from which they are referred and personnel by whom they were referred, subtypes, prevalence and types of comorbidity, family psychopathology, symptom profile including various behavioral domains, its severity, and functional impairment.

The common centers from where the early onset group were referred are tertiary care center (52.4%) and school (23.8%) where as it is the school in adolescent onset group (61%). The common referral personnel were pediatrician (42.8%) in the first group and it was the teacher (61%) in the second group. These observations necessitates the need and significance for giving training to the primary care personnel for identifying the behavioral problem among the children and improving the facility and bringing the interventional program in the primary care centers. Every other school teachers have to be trained in school mental health program and primary counseling.

The common subtypes noticed among the childhood onset were those confined to family (38.1%), ODD (28.6%), and un socialized (14.3%) whereas among the adolescent onset they were un specified (44.4%) and socialized (38.9%).

The prevalence of comorbidity among the early onset group (66.7%) was more than the other group (33.3%). The prevalence of hyperkinetic disorder among the early onset group is (59.3%) whereas in the later onset group it was less (11.1%). The prevalence and type of co-morbidity will influence the symptom severity and prognosis.[14] Studies show that CD became more severe and persistent when children also exhibit ADHD.[2021] The comorbid ADHD is found to influence the development, course, and severity of CD. CD with comorbid ADHD have a much earlier age of onset of disruptive behavior than CD alone.[22] Boys who meet criteria for CD with an age of onset of less than 10 years are 8.7 times more likely to show at least one aggressive symptom than the youths who qualify for CD at a later age.[24]

The family psychopathology, symptom severity, and symptom profile were significantly higher in the childhood onset [Tables 5 and 6] except the SA which is significantly higher in adolescent onset group. This shows that moving with friends and going for stealing, alcoholism, substance abuse, gang war etc., usually starts in the adolescent age in our culture. The prevalence of more psychopathology among the childhood onset group might have leads to severity of behavioral symptoms. The relationship between family psychopathology and symptom severity were documented in previous studies.[142526] Favorable parenting may be protective factor.[27]

Table 5

Comparison of PIS, RBPC and CGAS between child hood and adolescent onset CD (PIS, n=60)

Table 6

Comparison of symptom profile and its severity of CD in various domains between childhood and adolescent onset CD (RBPC score, n=60)

Functional impairment was also more among childhood onset group [Table 5]. But the difference is not statistically significant. This could be due to the difference in the severity of SA of childhood and adolescent onset groups. Among adolescent onset group, though they had low total score of symptom [Table 5] they had significantly higher score on subscale SA [Table 6]; which might have impaired their global functional level in family, school, and peer group. This might be the reason for relatively higher functional impairment among adolescent group with a low score of total symptom severity (with no statistically significant difference) in comparison with childhood onset group. However, these observations need further explorations.

There is a significant correlation [Table 7] among the variables, family psychopathology (PIS), quantity of symptom (RBPC), and functional impairment (CGAS). The family psychopathology is positively correlated with the symptom severity and negatively correlated with the functional level of the children with CD. The symptom severity was negatively correlated with the functional level of the child with CD. The regression analysis (ANOVA) showed that the predictive power of family psychopathology on symptom severity was 51.6% [Figure 1] and on functional level was 25.9% [Figure 2]. The predictive power of symptom severity on functional level was 25.1% [Figure 3]. This supports the developmental model of CD proposed in various studies.[1617]

Table 7

Correlation between abnormal psychosocial situations, symptom severity and functional level of children (n=60)

Figure 1

Regression plot showing the relationship between parent interview schedule score and revised behavioral problem checklist total score

Figure 2

Regression plot showing the relationship between parent interview schedule score and Children's Global Assessment Scale total score

Figure 3

Scatter plot showing the relationship between revised behavioral problem checklist total score and Children's Global Assessment Scale total score

To summarize, the determinants of symptom profile and its severity are early onset (childhood onset CD), nature, and quality of family psychopathology, prevalence, and type of comorbidity and nature of symptom profile itself apart from the genetic vulnerability.

CONCLUSION

The children with CD are not homogenous population, but heterogeneous in nature with variations in the age of onset, subtypes, prevalence of comorbidity, abnormal psycho social situations, symptom profile, symptom severity, and functional level of impairment

There was significant male dominance (88.3%) with boy girl ratio is 7.5:1. The common comorbidity noted was hyperkinetic disorders (45%). Majority of the children with CD came under moderate degree of severity (76.7%) with problems in the family, school, and peer group. There were only a few cases that were registered as children in conflict with law, which differs from the western literature where more severe cases of children with CD and those in conflict with law

The childhood onset group was 70% with continuation of the male dominance (85.7%), having more family psychopathology, symptom profile, symptom severity, functional impairment, and comorbidity

The determinants of symptom profile and severity were early onset (childhood onset CD), nature and quantity of family psychopathology, prevalence, and type of comorbidity and nature of symptom profile itself apart from the genetic vulnerability

Family psychopathology was positively correlated with the symptom severity and negatively correlated with functional level of the children with CD. The symptom severity was negatively correlated with functional level

The predictive power of family psychopathology on symptom severity was 51.6% and on functional level was 25.9% among the children with CD. The predictive power of symptom severity on functional level was 25.1%.

Limitations of the study

Small sample size.