Abstract
There is increasing recognition of the role of psychopathology as a risk factor for involvement in the abuse of psychoactive substances [1]. On the other hand, there is strong evidence to suggest that substance abuse can be a risk factor for psychiatric disorder [2], and as a factor that significantly influences the course, treatment and outcome of psychiatric illness [3].
Both emotional problems and conduct disorders have long been recognized as co-occurring with substance use disorders [4, 5]. Substance abuse is also linked with increased suicidal ideation and attempted suicide [6] as well as completed suicide [7]. Comorbidity between substance use disorders, attention deficit hyperactivity disorders [8] and eating disorders [9] has also been documented. Furthermore, recent reports link psychoactive substance dependence with a significantly increased risk of psychosis [10].
The aim of this study is to determine the rate and nature of substance use disorders (other than alcohol) comorbidity among inpatient youth with a severe psychiatric disorder, as there may be significant clinical, training and resource implications for our service and for similar services.
Method
Population and setting
Data were systematically gathered prospectively on all admissions as part of an ongoing outcomes project at the Christchurch Youth Inpatient Unit (YIU). The YIU is an 8-bed tertiary mental health facility for the assessment and treatment of youth aged 16–18 years, who have a severe psychiatric disorder and who cannot be effectively treated or managed in other mental health youth services. It is part of a comprehensive and well-resourced service with outpatient, day facilities and inpatient services. The Unit is a non-secure treatment facility and mainly accepts referrals from the Canterbury region, which has a population of 500 000, but is also a regional service for the whole of the South Island of New Zealand, which has a population of 800 000. It provides short-term intensive treatment for patients with severe mental illness with an objective of achieving enough stability and remission for early discharge to outpatient management and follow-up. The Unit does not accept patients with pure conduct disorder or substance use disorder. All youth admitted during the Unit's first 12 months of operation (since March 2001) were included.
Diagnostic categories
The diagnosis of Substance Use Disorder (which includes Substance Abuse Disorder (SAD) and Substance Dependence), and other Axis I diagnoses were based on the DSM-IV criteria, as is the clinical policy at the Unit. A Substance Use Disorder diagnosis was based on detailed history from the subjects, parents/caregivers, and other professionals that have been involved in the subject's care, as well as a detailed evaluation of the subject's psychiatric records. As the vast majority of subjects have been in the outpatient service for significant periods of time, it was not difficult to verify statements regarding psychoactive substance use. In cases where there was suspicion, urine analysis was done. Urine screening for psychoactive substances is normally carried out for all patients with psychotic symptoms, and for other patients where there was suspicion of substance use on admission and as a matter of policy at the Unit. We believe that it was highly unlikely that we missed a diagnosis of Substance Use Disorder in our sample.
Sociodemographic variables
The investigation also gathered the following data pertaining to substance use: age, gender, living situation and school attendance at the time of admission, and history of suicidal and criminal behaviour.
Data collection and analysis
Substance use, sociodemographic and clinical information was gathered by searching subjects' electronic clinical files on the mental health patient information system of the Canterbury District Health Board. Occasionally, an adolescent's psychiatrist, case manager, or primary nurse at YIU was consulted for clarification of the obtained information. The information was then entered into the Statistics Package for the Social Sciences (SPSS 10.05).
Statistical methods
Sociodemographic information was coded into the following binary (0 = no, 1 = yes) variables: living with parent(s), living with friends, living alone, inconsistent living arrangements, other care arrangement including foster care, attending school, contact with the law because of criminal behaviour, at least one suicide attempt and multiple suicide attempts. χ2 analyses were performed to examine differences in sociodemographic variables between substance abusing and non-substance abusing youth and to examine any effects of gender. The Student's t-test was performed to examine differences in continuous variables.
Results
In the first year of operation, the YIU admitted 62 adolescents, 37 females and 25 males, with a mean age of 16.35 years. Of these, 53 were of New Zealand European/Caucasian descent (85%), including 2 English-born and 1 Canadian-born, and 9 were indigenous New Zealand Maori (15%). Using DSM-IV criteria, the adolescents' primary diagnoses largely fell into one of three categories: Mood Disorder (60%), Major Psychotic Disorder (16%), or Anxiety Disorder (16%), predominantly Posttraumatic Stress Disorder (PTSD).
Forty (64.5%) of those admitted also had DSM-IV diagnosis of SAD (other than alcohol), the majority of which involved cannabis, although stimulants and hallucinogenic substances were also involved. None of the subjects had a diagnosis of Substance Dependence Disorder. Substance abusing and non-substance abusing adolescents presented with similar psychosocial characteristics. At the time of admission, 60% of all youth admitted lived with one or more of their parents, 55% were not attending school, and 60% had made at least one suicide attempt. Over one-third (34%) had made several lifetime suicide attempts, and 13% had contact with the law because of criminal behaviour. A summary of the psychosocial information is presented in Table 1.
Demographic and psychosocial characteristics of substance abusing and non-substance abusing youth inpatients
χ2 analyses for differences between groups with more than five persons were conducted for the variables: living with parents, attending school and attempted suicide. No differences were found between substance abusing and non-substance abusing youth or substance abusing males and females. Significantly, about one in four substance abusing youth had inconsistent living arrangements compared to one in 11 in non-substance abusing youth.
The rates of comorbidity among different diagnostic groups shows the same pattern as shown in Table 2.
Primary DSM-IV Axis I diagnosis and comorbid substance abuse disorder in youth inpatients
Furthermore, more than half of those with a SAD had a mood disorder and one in five had a psychotic disorder.
Discussion
The published literature on comorbidity shows huge variations in methodology (retrospective vs prospective, parent-report vs self-report), population (inpatient vs outpatient) and focus (general psychiatry vs drug use treatment programs vs correctional facility). It is therefore very hard to make exact comparisons between their results. However, the trends are consistent and our results are not very different.
The overall rate of comorbidity we found (64.5%) is consistent with those reported elsewhere. Grilo et al. [5] found a comorbidity rate of 65% (excluding conduct disorder) among adolescents in a general psychiatric hospital. Similarly, Grella et al. [11] in a sample of adolescents from drug treatment programmes, found that 64% of the sample had at least one comorbid mental disorder. Unlike previous reports, we did not find the same levels of comorbidity between conduct disorder and SAD simply because our unit excludes young people with conduct disorder from admission unless they had another Axis 1 diagnosis (other than SAD). This is also reflected in our finding of the remarkably low levels of criminal involvement among our subjects.
Some studies have reported family disruption and dysfunction to be significantly associated with varying degrees of substance use, both in inpatient and community samples [11–13]. Consistent with these reports, our comorbid youth were more likely to have problems within the family setting leading to inconsistent living arrangements and school problems.
In view of the Unit's admission criteria, and as would be expected, the general pattern of morbidity in our sample seems to show that internalizing psychopathology is the predominant feature. Among the internalizing disorders, mood disorders, especially bipolar disorder (BPD) seems to be particularly associated with SAD especially among younger patients. We found a SAD comorbidity rate of 60% among those with a mood-related disorder in line with the rates of 65% [2] and 61% [14] in previous reports. Wilens et al. [15] found a rate of SAD comorbidity of 39% among their adolescent population with BPD.
The association between substance use disorders and suicidal behaviour is well documented and significant. In fact, substance use disorders have even been reported to distinguish between those who attempt suicide or have repeated attempts and those who only think about suicide [4, 6, 16]. Given the nature of the data on suicidal behaviour that we collected, we cannot confirm or deny those findings. However, similar to others [13], we found that an almost equal proportion of our sample of youth with comorbid SAD (57.5%) and those without SAD (64%) had a suicide attempt. Furthermore, 30% and 41%, respectively, reported multiple attempts. This would seem to indicate that at the more severe end of depression (which is a requirement for admission) the risk of suicidal behaviour is high and is independent of coexisting SAD.
Confirming earlier evidence that posttraumatic stress symptoms are significantly associated with problematic drug and alcohol use [17], we found significant SAD comorbidity among our subjects with PTSD related anxiety. About two-thirds had a SAD. Others [14] reported a corresponding rate of 43%. We also found that among those with SAD, 13% (and 14% in non-SAD) had an anxiety disorder compared to other reports of 14.5% [4]. Jacobsen et al. [18] view the high rates of comorbidity as suggesting that PTSD and substance use disorders are functionally related to one another. They indicate that most published data support a pathway whereby PTSD precedes substance abuse or dependence. Substances are initially used to modify PTSD symptoms. With the development of dependence, physiologic arousal resulting from substance withdrawal may exacerbate PTSD symptoms, thereby contributing to a relapse of substance use.
Recent reports indicate that among adult patients with schizophrenia, 41% had a lifetime history of substance abuse or dependence, including 30% on a drug other than alcohol [19]. An English study [20] found a rate of 16% in a similar population. Our subjects show much higher rates (80%), which could be explained by the fact that they were drawn from a more acutely ill population. Grilo et al. [4] found a rate of 5.8% of psychosis comorbidity among those with substance use disorders, our rate of 20% is significantly higher, probably reflecting the population differences.
The role of substance use disorder (particularly cannabis and amphetamines) comorbidity among patients with schizophrenia and related psychoses is generating considerable interest. At least part of the interest is related to the well established psychotic effects of cannabis and amphetamines. The effect of amphetamines is related to its dopaminergic properties which may precipitate a psychotic state difficult to distinguish from schizophrenia but primarily characterized by delusional thinking and hallucinations [21]. Similarly, psychotic symptoms following cannabis use are commonly reported and may present as toxic confusional states or functional psychotic like states. It is widely accepted that cannabisrelated psychotic symptoms are dose-related. Despite that, there is considerable doubt as to the existence of a specific clinical syndrome identifiable as a ‘cannabis psychosis’ [22].
A considerable part of the interest, however, is closely related to the continuous debate about the legal status of cannabis. Central to many of the arguments is the question of whether cannabis (known to cause psychotic symptoms) is in any way linked with schizophrenia. Recent epidemiological studies confirmed the view that there is strong evidence to suggest that cannabis use is a significant risk factor for schizophrenia. Reports linked it with an increased risk of developing schizophrenia in a dose-dependent fashion among a very large Swedish cohort [23], a New Zealand cohort [24] and a large cohort of prisoners in the UK [10]. The question of causality, or self-medication is hard to answer, given that the evidence is based on epidemiological studies. The explanation most accepted is that cannabis triggers the onset or relapse of schizophrenia in predisposed people and also generally exacerbates the symptoms [10, 25, 26]. Our clinical experience is consistent with this view.
Because of its objectives and design, our study did not tell us much about causality and the direction of the relationship between psychopathology and comorbid substance use disorders. The available evidence is inconsistent. Some studies, for example, have found that the onset of psychopathology preceded the onset of a substance use disorder, and that substance use disorders are associated with more dysfunction and the need for hospitalization [27]. Rao et al.[28] reported that depressed adolescents have earlier onset of substance use disorders than controls. Yet, a recent Australian cohort study found that frequent cannabis use in teenage girls predicts later higher rates of depression and anxiety, and that depression and anxiety in teenagers did not predict cannabis use [29]. It is possible that there is a bi-directional causal relationship, the direction of which depends on each person's own constitutional, and personal characteristics as well as environment and circumstances.
Conclusions
Our findings suggest that the level of SAD comorbidity among inpatient youth with severe psychiatric disorders is relatively high, in line with previous reports. There are clear clinical and resource implications for our service and probably for services similar to ours. To start with, screening is essential to identify those patients who need a further detailed specialist assessment for substance use disorders. If and when a substance use disorder is confirmed, its possible impact on the presenting psychopathology and the treatment plan should be determined. Moreover, clinicians should take full account of and consideration for, comorbid substance use disorder in planning for discharge and follow-up. Above all, it should be treated in its own right regardless of whether it is ‘primary’ or ‘secondary’. The resource and training implications for the field of child and adolescent psychiatry are obvious and should be considered and planned for. Training programmes in child and adolescent psychiatry have largely overlooked the uniqueness of many aspects of substance use disorder in youth, particularly the issues of comorbidity with psychopathology and the developmental aspects of adolescence. Substance use disorders in young people have many significantly different aspects to those in adulthood [30]. These differences should be reflected and emphasized in psychiatric training programmes.