Autism is a complex neurodevelopmental disorder that is thought to involve an interaction between multiple and variable susceptibility genes, environmental factors, and epigenetic effects. Great concern has been raised about the marked increase in the prevalence of autism spectrum disorders in the last decade. Risperidone, the most studied atypical antipsychotic used in children, has been shown to improve severe behavioral difficulties in over half of children with autism who have these difficulties. However, not all children with autism and severe behavioral problems respond to risperidone, and for a few, it has significant side effects.

Two controlled studies and numerous open-label and long term studies in children with autism spectrum disorders using the atypical antipsychotic risperidone show a significant decrease of associated serious behavioral problems. The use of atypical antipsychotics is of great concern, however, because of their significant side effects and the fact that only two-thirds of children positively respond. Ways to predict response, appropriate dosage and serious side effects are needed.

For this study, we will identify 40 children (4 to 18 years old) with autism who also have serious behavioral problems. We will then treat them with risperidone. Blood samples will be obtained prior to treatment and at eight weeks of treatment or study exit. At that time, efficacy will be assessed using the Clinical Global Impression-Improvement scale (CGI-I) and the Irritability subscale of the Aberrant Behavior Checklist (ABC). Blood genomic profiles before and after risperidone treatment will be determined using Affymetrix oligonucleotide microarrays combined with RT-PCR.

Blood genomic profiles are shown to predict medication response for disorders such as cancer and epilepsy. This exploratory or discovery study will use blood genomic profiles before and after risperidone treatment in children with autism and severe behavioral difficulties to determine if the profiles can predict response to treatment. The ultimate goal of this line of research is to develop methods to predict which medications work for which child before initiating treatment, to predict which child might develop particular side effects, and to identify new treatment targets for future medication development.

Risperidone will be started at 0.5 mg at bedtime for 4 days and, if the current dosage is tolerated as evidenced by no more than mild sedation, no EPS or other moderate to severe AEs, and if there are continued behavioral symptoms, the dose will be increased to 1 mg at bedtime for an additional 4 days. If tolerated and indicated, 0.5 mg will be added in the AM for a daily total of 1.5 mg. After that, dosages may be increased if there does not appear to be an adequate clinical response. Dosage will not be increased if there are side effects (e.g. excessive sedation, salivation, EPS, lactation) and may be decreased if it is not tolerated. If the investigator determines that a significant adverse reaction occurs or if the subject or his or her family wants to stop the study, the medication will be tapered or stopped depending on the dose and reason for stopping and the subject will be offered alternative treatment at the M.I.N.D. Institute Clinic or referred elsewhere. This dosing schedule mirrors that used in the two recent positive trials of risperidone for treating severe behavioral problems in autism (McCracken et al., 2002; Shea et al., 2004).

Inclusion Criteria:

1. Willingness to participate and written informed consent obtained from parent and when indicated by subject
2. Confirmed DSM-IV-TR diagnosis of Autistic Disorder or Asperger's Disorder using the ADOS and the ADI-R, the current gold standards for diagnosing autism spectrum disorders, subject history, and clinical consensus with PI. The ADI-R and ADOS will be administered by research-reliable clinicians, including a clinical psychologist, a certified trainer on the ADOS and the ADI-R
3. rated by study clinician as at least "moderate" on the CGI-Severity scale (a rating of ≥ 4) and greater than 18 on the ABC Irritability subscale

Males or females of any race between 4 and 18 years of age (5) A nonverbal IQ greater than or equal to 55 on the Stanford-Binet:V (6) Women of childbearing potential must use an adequate method of contraception throughout the study.

Exclusion Criteria:

1. Primary diagnosis of bipolar disorder, schizophrenia, or autism spectrum disorder other than Autistic and Asperger's Disorders
2. Nonverbal IQ lower than 55 (Stanford-Binet:V)
3. History of seizure activity in the past year (active seizures might confuse efficacy ratings)
4. fever, infection, metabolic disturbance or any severe medical illness in the past year
5. typical or atypical antipsychotic use within 8 weeks of study entry
6. Inability of parents or care takers to give informed consent, travel to the visits, administer medication, or arrange for completion of rating scales. Other non-antipsychotic medications and non-pharmacological treatments will be allowed if started at least 2 months prior to the initial screening and must remain constant for the 8 weeks of this study. These treatments will be recorded as will a history of past trials of medications. Prohibiting ongoing treatment would be difficult to justify to parents and to the IRB and would make adequate recruitment for this pilot study difficult.