Autism, originally described by Kanner (1943), is among the most severe of neuropsychiatric disorders. It is a pervasive developmental disorder affecting social, communicative, and compulsive/repetitive behaviors characterized by stereotypic complex hand and body movements, craving for sameness, and narrow repetitive interests. Individuals with autism spectrum disorders (ASD) are characteristically heterogeneous and show marked variability in their response to interventions. Studies of behavioral and psychopharmacological interventions document approximately 1/3 of ASD participants fail to respond to targeted treatments. Efforts to evaluate the specificity of treatment effects are important to inform conceptualizations about the disorder, identify behavioral phenotypes, and to aide clinical decision making.

The goal of this study is to evaluate the use of clinical behavioral pharmacology methods, functional behavioral assessments (FBA), in assessing the treatment effects of pediatric medications in children with ASD. The present study of FBA procedures in pharmacological treatment will be conducted as a separate, but parallel study within IRB approved, federally funded, double-masked, placebo controlled medication trials of citalopram (GCO # 01-1295 PS*), an SSRI hypothesized to reduce stereotyped and repetitive behaviors in ASD and divalproex sodium (GCO # 01-0294), a medication recently found to reduce repetitive behaviors in ASD (Hollander et al., in press). This study will focus on the use of FBAs in distinguishing responders vs. nonresponders on the basis of behavior function, in evaluating functional patterns for stereotypy, aggression, and impulsivity, and in using descriptive FBAs as outcome measures in clinical trials.

FBAs are behavioral assessment methods used to hypothesize about the function of maladaptive behaviors. FBAs are conducted either through experimental manipulations known as functional analyses or through descriptive analyses procedures, which involve structured observations and parent/caregiver interviews. Descriptive analyses will be conducted with all participants (n=24). The more rigorous, functional analyses will be conducted with a sub-set of the sample (n=6) to corroborate the findings of the descriptive analyses. Data from the FBAs will be collected using videotaped recordings of behavior and coded by trained raters for both the descriptive and experimental analyses.

Our pilot data and other published data suggest that certain medications such as citalopram (celexa) and divalproex sodium (Depakote) may improve global functioning in autistic patients and repetitive/compulsive behaviors and social deficits. The addition of FBA methods to evaluate outcome are an important step in extending the research and knowledge of the conditions associated with good and poor treatment response to pediatric medications in children with autism.

Research geared towards increasing knowledge of the variables contributing to treatment response is critical to advancing intervention studies in ASD. Literature in the growing field of clinical behavioral pharmacology suggests functional behavioral assessments (FBA), when integrated into clinical trials of medication, can be important tools in elucidating the specificity of treatment effects (Mace & Mauk, 1995, Northup & Gully, 2001). The goal of this study is to evaluate the use of clinical behavioral pharmacology methods, primarily descriptive functional behavioral assessments (FBA), in a double-masked, placebo controlled trial of medications used to treat children with autism. These medications are hypothesized to reduce the stereotyped and repetitive behaviors in ASD. This study will focus on the use of descriptive FBAs in distinguishing responders vs. non-responders on the basis of behavior function, in evaluating functional patterns for stereotypy, and in using descriptive FBAs as outcome measures in clinical trials.

This study is accordance with CAN research priorities in its focus on improving methodological approaches in clinical trials (Hollander, et al., 2004). This study is also consistent with recent NIMH recommendations for clinical research. The recommendations identify the integration of behavioral research methods with traditional methods of clinical research (i.e. group designs) as important aims for ASD treatment research (CPEA/STAART working group, 2004). As such, this study fills significant gaps in two research areas: treatments for repetitive behaviors and research methodology in treatment trials for individuals with ASD.

The bio-behavioral methods proposed for use in this evaluation have been previously used to stratify and match treatments for self-injurious behavior in adults with developmental disabilities (Mace & Mauk, 1995), identify specific treatment effects of methylphenidate in children with ADHD (Northup & Gully, 2001), and more recently in evaluating the specificity of risperidone's effects on destructive behaviors in individuals with autism (Crosland, et al, 2003).

Despite that stereotyped and repetitive behaviors are a core symptom domain of ASD, behavioral pharmacology approaches to date have only evaluated the correlated symptoms of aggression, self-injurious, and destructive behaviors in ASD. Repetitive behaviors in ASD include stereotypic movements, repetition of routine behaviors, repetitive play, perseverative speech, and over-focus on restricted interests. These repetitive movements may be associated with emotional distress, tension, or dysphoria and have significant impacts on daily life. It is believed that the methods used to elucidate and specify treatment effects in other medications and disorders will also be valuable to the understanding of SSRI effects in autism.

Both behavioral and pharmacological treatments have been used to treat repetitive and stereotyped behaviors in autism. Behavioral treatments include the use of noncontingent reinforcement (Britton, et al., 2002) and alternative response training to maintain homeostasis (Johnson, Laarhoven, & Repp, 2002). Psychopharmacological approaches include the use of SSRIs given the similarity between repetitive behaviors in autism and OCD (McDougle, 1995) and evidence for serotonin dysfunction in ASD (Hollander, 2000). To date, there are few controlled studies with any of the SSRIs in pediatric populations with ASD. Hollander & colleagues (Hollander, et al., in press) at Mt. Sinai recently completed a study suggesting the efficacy of low-dose liquid fluoxetine on reducing repetitive behaviors in a double-blind, placebo controlled crossover study of 45 children with ASD. This proposed FBA evaluation would occur within another SSRI trial at Mt. Sinai, which aims to extend previous findings in a large scale, multisite investigation of the safety and efficacy of liquid citalopram in children with ASD.

Behavioral pharmacology methods Functional behavioral assessments (FBA) are the cornerstone of clinical behavioral pharmacology approaches. The goals of FBAs are to hypothesize and experimentally test the functions of maladaptive behaviors. Within the field of behavior analysis, determining the function of the maladaptive behavior is primary to developing effective interventions.

FBAs involve both experimental and descriptive assessments of behavior functions. Experimental methods, also known as functional analyses (Iwata, et al., 1982/1994), have the advantage of using experimental manipulations and direct, continuous observations of behavior. The procedures outlined by Iwata, et al. (1982/1994) are considered the gold-standard for functional analyses. The procedure involves the use of multi-element, single-subject experimental design. During the functional analysis, consequences presented for a maladaptive behavior are varied. The experimental conditions outlined by Iwata, et al. (1982/1994) are labeled by the presumed function tested in a condition and include: escape/avoidance, attention, automatic reinforcement/self-stimulation, and control (i.e. preferred toys + social attention).

Experimental methods provide valuable and reliable data on behavior functions, but involve significant time and expense for patients and experimenters. In contrast, descriptive methods are based on parent/caregiver-report, direct behavioral observations, and provide information on both consequent and antecedent predictors of behavior. Descriptive methods are strongly correlated with experimental analyses and have advantages because of the information provided on antecedent functions and in their relative cost-effectiveness (Horner, et al., 1997). This study will pilot the use of descriptive assessment methods as an outcome measure and use experimental FBAs to corroborate the results of the descriptive analyses.

The methods and outcome of this investigation will be used to design an independent clinical trial using a multi-method approach, including descriptive and experimental FBAs as outcome measures, in medication studies of repetitive behaviors in ASD. Given this research has limited funding options outside of CAN, funding this project may be a critical, preliminary step in improving our treatment-specificity research and clinical research methods in ASD.

Specific Aims

The primary aim of this study is to determine the usability of functional behavioral assessment (FBA) procedures in predicting and measuring treatment response within a clinical trial of citalopram, an SSRI hypothesized to reduce repetitive behaviors in ASD. The specific aims of the study are as follows:

1. To predict treatment response through results from descriptive FBAs.
2. To evaluate functional patterns for stereotypy/repetitive behaviors in ASD.
3. To evaluate the use of descriptive FBAs as outcome measures in SSRI trials in ASD and to correlate results from descriptive FBAs with standard global rating measures such as the Child Yale-Brown Obsessive Compulsive Scale (CYBOCS) and the Aberrant Behavior Checklist (ABC) for repetitive behaviors and stereotypy.

Inclusion Criteria:

• Male or female of any race or ethnicity.
• Ambulatory status (outpatient or day-treatment) at time of randomization.
• Age 5-17 years inclusive at the time of consent
• Subjects will meet criteria for the diagnosis of Autistic Disorder, Asperger's Disorder, or PDD-NOS as determined by ADI-R administered by raters who are trained to research reliability, and confirmed by an experienced and reliable clinician using DSM-IV-TR criteria.
• Subjects must also have a score greater than 7 on the first 3 items of the Compulsions Subscale of the Revised CYBOCS.
• Subjects must have a rating of at least moderate (ratings of 4 or greater) behavioral disturbance based on the modified Clinical Global Impression-Severity of Illness score (CGI-S) at the time of screening (See description below).
• Subject must demonstrate a mental age >18 months as determined by the Vineland Adaptive Behavior Scales.
• Subjects must be free of psychotropic medication for at least one month for fluoxetine, two weeks for other SSRIs and neuroleptics, and for 5 days for stimulants prior to baseline ratings.
• Subjects and their parents (guardians) must be judged reliable for medication compliance and must agree to keep appointments for study visits and tests as outlined in the protocol.
• Subjects enrolled in an applied behavior analytic educational program for children with autism will be included in the Experimental FBA study only. **

Exclusion Criteria:

• Subjects who have medical contraindications to therapy with SSRIs as determined by medical history or known allergy.

• Concomitant medication that would interfere with participation in the study.

• Subjects with a prior history of treatment failure to a clinically adequate trial of two selective serotonin reuptake inhibitors (SSRI).

• Subjects with a previous diagnosis of Rett's Disorder or Childhood Disintegrative Disorder.

• Subjects with documented need for other ongoing psychotropic medications besides study medication (with the exception of stable dose (at least 3 month) anti-convulsants for seizures).

• Subjects with uncontrolled epilepsy (seizure within past 6 months).

• Presence of chronic medical conditions that might interfere with study participation or where study participation would be contraindicated, or clinically significant abnormal baseline laboratory testing.

• Subject with known personal history of bipolar disorder or prior manic episode induced by antidepressant exposure.

• Subjects with a prior history of an exposure to citalopram of sufficient dose or duration to determine response status.

• Subjects that are not enrollment in an applied behavior analytic educational program for children with autism will be excluded from the Experimental FBA study only.

  • Cases which do not meet exact inclusion or exclusion criteria but questioned as appropriate by the site PI will be reviewed by the multi-site STAART protocol committee for final decision about study eligibility. The protocol committee will take care to insure that the exception 1) is consistent with the subject's welfare and 2) does not compromise the scientific purpose of the study. Such exceptions will be coded as an approved protocol deviation and will be communicated to the local Institutional Review Board in a timely manner.

Subjects who are enrolled in the approved double-blind placebo controlled trial of citalopram in young children with autism.