Risk Factors of Acute Behavioral Regression in Psychiatrically Hospitalized Adolescents with Autism

J Can Acad Child Adolesc Psychiatry. 2010 May; 19(2): 100–108.

Copyright © 2010 Canadian Academy of Child and Adolescent Psychiatry


During adolescence, some individuals with autism engage in severe disruptive behaviors, such as violence, agitation, tantrums, or self-injurious behaviors. We aimed to assess risk factors associated with very acute states and regression in adolescents with autism in an inpatient population.
Between 2001 and 2005, we reviewed the charts of all adolescents with autism (N=29, mean age=14.8 years, 79% male) hospitalized for severe disruptive behaviors in a psychiatric intensive care unit. We systematically collected data describing socio-demographic characteristics, clinical variables (severity, presence of language, cognitive level), associated organic conditions, etiologic diagnosis of the episode, and treatments.
All patients exhibited severe autistic symptoms and intellectual disability, and two-thirds had no functional verbal language. Fifteen subjects exhibited epilepsy, including three cases in which epilepsy was unknown before the acute episode. For six (21%) of the subjects, uncontrolled seizures were considered the main cause of the disruptive behaviors. Other suspected risk factors associated with disruptive behavior disorders included adjustment disorder (N=7), lack of adequate therapeutic or educational management (N=6), depression (N=2), catatonia (N=2), and painful comorbid organic conditions (N=3).
Disruptive behaviors among adolescents with autism may stem from diverse risk factors, including environmental problems, comorbid acute psychiatric conditions, or somatic diseases such as epilepsy. The management of these behavioral changes requires a multidisciplinary functional approach.
Keywords: autism, adolescence, acute behavioral state, regression, intellectual disability
Adolescence is a particularly important period in the course of development. Regarding neuropsychiatric disorders, adolescence is the period of onset for schizophrenia, bipolar disorder, and catatonia (Cohen et al., 2005). Adolescence is also one of the peak periods of seizure occurrence (Tuchman and Rapin, 2002). In the field of autism, some authors have observed that the onset of puberty is temporally associated with clinical deterioration (Gillberg and Schaumann, 1981; Billstedt et al., 2005). However, most children with autism grow up and become adolescents and then adults, without manifesting major adult psychiatric disorders. Indeed, the core symptoms of autism – deficiencies in social interaction, language delay and communication disabilities, and restricted and stereotyped behavior – tend to show improvement over time (Rutter and Lockyer, 1967; Rutter et al., 1967; Kobayashi and Murata, 1998; Seltzer et al., 2003), and only a few individuals with autism experience severe disruptive disorders during adolescence.
The literature on autism during adolescence and adulthood is sparser than that concerning autism during childhood despite some challenging behaviors that persist throughout adulthood. In a review of studies published before 1996, Nordin and Gillberg (1998) observed that cognitive or behavioral regression occurred in 12 to 22% of adolescents with autism. The largest study of this phenomenon, a survey conducted on 201 young adults with autism born in Japan, indicated that 32% of them showed marked deterioration during adolescence (Kobayashi et al., 1992). More recently, a prospective study conducted on 120 Swedish autistic subjects showed that behavioral and cognitive regression, catatonia, and “adult psychosis” occurred during adolescence in 16%, 12%, and 8% of those studied, respectively (Billstedt et al., 2005). Finally, Mouridsen et al. (2008) showed using a case-control method that adults with autism had a higher frequency of additional treatable psychiatric disorders than controls, in particular psychotic disorders and affective disorders.
Normal intellectual abilities seem to be protective: among individuals with autism, those without intellectual disability experience less deterioration than those with mental disabilities (Venter et al., 1992; Ballaban-Gil et al., 1996). Similarly, among individuals with intellectual disability, those with autism experience more deterioration than those without autism. From an epidemiological sample of subjects with intellectual disabilities in Ontario [Canada], Bradley and Bolton (2006) compared the prevalence of episodic psychiatric disorders in adolescents with intellectual disability with (N=36) and without (N=36) autism. They found that autism was significantly associated with the total number of episodes and with psychotropic prescriptions. At least one episode of psychiatric illness occurred in the 17 individuals with autism. These episodic illnesses were depressive disorders (N=8), bipolar disorders (N=2), adjustment disorders (N=3), and persistent and unclassifiable disorders (N=4) (Bradley and Bolton, 2006).
In summary, severe behavioral changes and mental health problems in adolescents with autism are poorly investigated and inadequately understood. In particular, no empirical guidelines are available regarding etiology and treatment, as there are very few studies concerning inpatient treatment of subjects with autism (Shattuck et al., 2007) and, to our knowledge, none focusing on adolescence. The present study aimed to screen all patients with autism hospitalized in a psychiatric intensive care unit for severe behavioral disturbances occurring during adolescence. Some guidelines for diagnosis and treatment are suggested based on this clinical experience.
Methods and Results
The current study provides a detailed clinical picture of a population of adolescents with autism who were hospitalized in a psychiatric setting for acute behavioral impairment or cognitive regression and documents the factors influencing the outcome.
The main limitations of this study are (a) the retrospective design, (b) the lack of standardized instruments to assess patients’ level of intellectual disability or language, and (c) the use of a limited number of clinical instruments [GAS, CARS]. The strengths of this study are (a) the multidisciplinary approach, including the use of experts in epilepsy and genetic/metabolic disease, (b) the free access to inpatient care regardless of patients’ socioeconomic background, and (c) the use of long inpatient stays in an intensive care unit to monitor outcomes of multiple therapeutic techniques for behavioral improvement.
It is noteworthy that the population studied is unique in terms of the severity of the problematic behaviors observed. Indeed, psychiatric hospitalization is quite unusual among people with autism and generally occurs after the failure of first-line interventions, such as behavioral management or an outpatient trial of a psychotropic drug.
Low IQ, absence of communication skills, and comorbid epilepsy are well-known factors leading to a poor prognosis, including an impaired course of development during adolescence (Gillberg and Steffenburg, 1987; Venter et al., 1992; Ballaban-Gil et al., 1996; Rapin, 1997; Howlin et al., 2004). In the current sample, all 29 individuals were severely impaired and had multiple challenging behaviors. This point is not surprising, as it has been shown in adults that these behaviors often co-occur (Matson et al., 2008). All had intellectual disability; all but one had severe autism as measured by the CARS; only a third had functional language; half of them had “syndromal or complex autism” (Cohen et al., 2005; Miles et al., 2005). However, this study also demonstrates that other factors, including specific care, family support, and socio-economic factors, appear to be crucial in the occurrence of these acute states. The male-female ratio was 3.8: 1 in the sample. Given the fact that all of the patients exhibited significant cognitive delay and severe autistic syndrome, the expected sex ratio should be closer to 2: 1 (Fombonne 2003, Amiet et al, 2008). This discrepancy can be explained by the fact that aggressivity and conduct disorders are more frequent among boys. Moreover, challenging behaviors are less tolerable among boys, leading to more frequent hospitalization requirement.
Absence of care was the retained diagnosis for the acute state in about 25% of the cases. This finding highlights two facts. First, adolescence is often a period when a new institution is needed, resulting in a higher risk of a break in the continuity of care (Fombonne et al., 1997). Individuals become too old for children’s institutions, but no adolescent institution exists. Second, as demonstrated in a large naturalistic study that described the follow-up of a cohort of 495 children with PDD, the major risk factor for the absence of appropriate care in France is a high degree of behavioral impairment (Thevenot et al., 2008). Therefore, despite free access to care in France, both adolescence and behavioral impairment may serve as exclusion factors in individuals with PDD. As a result, a vicious circle occurs as the absence of care may increase behavioral impairment.
In addition, the importance of family support is demonstrated by cases in which consequences occurred after major family events. In six cases, a diagnosis of adjustment disorder was retained for the acute state. Poor adaptive skills are a classical feature of both intellectual disability and autism. However, it is noteworthy that the life events implicated in our cases were very severe in nature (death, hospitalization of a parent for psychiatric illness or cancer, and maltreatment) and would also have affected ordinary adolescents.
In four cases, a psychiatric illness was diagnosed. Two adolescents exhibited a severe form of catatonia with akinesia, catalepsy, and mutism. Classically associated with schizophrenia when it occurs in youths (Cohen et al., 2005; Cornic et al., 2009), catatonia can be associated with bipolar disorders (Brunelle et al., 2009), somatic illness (Lahutte et al., 2008), and also autism (Wing and Shah, 2000; Ohta et al., 2006; Cohen et al., 2009). In a clinical follow-up study, Wing and Shah (2000) reported a 6% rate of catatonic episodes in a group of 506 adolescents and adults, especially when subjects had intellectual disability. Only one case of major depressive disorder was reported in our study, but this statistic could be underestimated. There is emerging evidence that depression is probably the most common psychiatric disorder that occurs in autistic people, but this disorder can be difficult to recognize in subjects with autism with intellectual disability and poor communication skills (Kobayashi et al., 1992; Ghaziuddin et al., 2002). It could be hypothesized that some subjects with a diagnosis of adjustment disorder were, in fact, depressed. Finally, one patient had bipolar disorder, a comorbidity that has been reported in association with complex differential diagnosis issues (Atals and Gerbino-Rosen, 1995; Brunelle et al., 2009).
The importance of somatic illness is probably the most significant result in this study. For one-third of the subjects (9 out 29), the acute behavioral state was a consequence of a treatable medical illness. A systematic approach is warranted, given the variety of possible conditions (see below). Among organic conditions, uncontrolled epilepsy should be the first-line hypothesis. Epilepsy was found to be responsible for the acute behavioral deterioration in six of our cases. Three of these received a diagnosis of epilepsy for the first time during the hospitalization. The high prevalence rate in this group is not surprising, given that epilepsy in autism is associated with intellectual disability (Amiet et al., 2008) and that early childhood and adolescence have been reported to be the peak periods for seizure onset (Tuchman and Rapin, 2002). However, the relationship between autism and epilepsy is complex, and their association may have different origins. Various seizure types and epileptic syndromes have been described in association with autism. Moreover, epileptic anomalies are frequently observed on the EEGs of autistic patients despite an absence of seizures, suggesting at least a low epileptic threshold (Tuchman and Rapin, 2002). It may also be true that autism and epilepsy share a genetic and/or neurodevelopmental cause, at least in the case of secondary autism (e.g., tuberous sclerosis). Epilepsy by itself may induce the development of autistic symptoms (Jambaqué et al., 1998). Two examples illustrate this point: (a) autistic regression may occur when the epileptic focus is located in a critical brain area, mainly in temporo-frontal locations, with substantial improvement after medication or even surgical treatment (Nass et al., 1999; Neville et al., 1997); (b) several epileptic encephalopathies are associated with intellectual disability and/or autistic traits, probably through a specific developmental impact. This association is especially true in West syndrome (Saemundsen et al., 2007). Finally, a coincidental association between autism and epilepsy cannot be ruled out in some cases, considering the high frequency of epilepsy in the general population. In terms of the diagnostic and therapeutic approach, we, like others, consider collaboration with an experienced neurologist to be crucial (Tuchman and Rapin, 2002).
To formulate rational guidelines for the diagnosis and treatment of acute behavioral states or regression in adolescents with autism, one cannot remain at the level of a symptomatic approach. Given the diversity of psychopathologies found in this study, we recommend a systematic integrative multidisciplinary approach that should include (a) a careful social and family evaluation, (b) a systematic search for frequent painful organic conditions, based on a clinical evaluation and a minimal screening [common blood tests, dental X-ray, urinalysis], (c) a neurological examination, with a specific focus on possible seizures, and an EEG, despite the technical issues in this population, and (d) a psychiatric evaluation, taking into account the particular profile of these adolescents [poor language skills, ID] and using adapted rating tools when available. This process should lead to a functional evaluation of each individual case and to the formulation of a principal hypothesis regarding the cause of the acute behavioral state. This step is crucial given the tendency to use too many medications in this field despite the minimal evidence available regarding their effectiveness for challenging behavior associated with autism (Matson and Neal, 2009). Considering alternative psychosocial-based interventions and careful functional assessments appear to be advisable. Figure 2 summarizes our view. Although this view is not a scientific fact and is essentially experience-based, this integrative approach can assist during the process of making treatment decisions.
Adolescents with autism who present with acute behavioral regression that compromises safety need to be examined with a multidisciplinary approach that includes organic, social, and psychiatric investigations. Given the complexity of these situations, hospitalization in psychiatric settings that collaborate with other disciplines is warranted. The adolescent psychiatrist plays a key role in coordinating investigations, giving a proper diagnosis, and treating the acute state.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s