The purpose of this study is to learn more about autism and its subtypes. Autism is a developmental disorder in which children have problems with communication and social skills and display restricted interests and repetitive behaviors.

This study has several goals. One aim is to look at types of autism that are known, such as the regressive subtype, (where skills are lost). We will explore whether there is a unique change in immune functioning related to this subtype. Another aim is to serve as one of the sites that will pilot a larger natural history study, entitled Autism Phenome Project. The goal is to further understand autism by identifying other subtypes.

We will look at several types of medical issues that may be related to autism, including immunologic problems. Children will be followed over the course of several years. We aim to capture medical problems that may be related to autism as they develop, and study outcomes in areas such as behavior and language, in order to explore known and new subtypes of autism.

Normally developing children (aged 1-6) and children with autism (age 1-6), developmental delays other than autism (age 2-6), or Rett's disorder (age 1-4) may be eligible for this study.

Depending on each child's study group and age, participants may undergo the following tests and procedures:

Baseline Visit

• Medical and developmental history, physical examination, psychological, cognitive and medical tests to assess symptoms of autism or other developmental disorders, photographs of the child's face, collection of hair, urine and baby teeth samples. If available, hair samples from the baby's first haircut and from the biological mother's hair are also collected.
• Overnight electroencephalogram (EEG; for children autism developmental delays and Rett Syndrome): A special cap with electrodes is placed on the child's head to measure brain waves (brain electrical activity) while the child sleeps in the hospital overnight. Healthy volunteers do not undergo this procedure.
• Magnetic resonance imaging (MRI) scan: The child stays in the scanner, lying still for 10 to 15 minutes at a time. Since it may be difficult for the child to lie still, the test may be scheduled for a time when the child is likely to be sleepy, or the child may be sedated.
• Lumbar puncture (for some children in the autism, developmental delay and Rett Syndrome groups). This test may be done under sedation.

Follow-Up Visits

Follow-up visits are scheduled at different intervals, depending on study group, age and aspect of the study the child is enrolled in. The visits include a short interview session with the child's caregiver and assessment of the child's development and behavior. Some of the assessment measures used during the baseline examination are repeated, including symptoms ratings, behavioral tests and a blood test. For some children, the final visit will include repeats of the medical procedures.

The current investigation will focus on the interrelationship between regression in autism and immune dysfunction. We propose to conduct a longitudinal natural history study that will result in a sample of 50 children with autism and strictly defined regression, 50 children with autism and no history of regression, 50 typically developing children, and 25 children who meet criteria for Rett's Disorder (a genetic disorder in which children have autistic features with clear regression). Subjects will be 12-48 months of age at study entry in order to capture regression as close to its onset as possible.

Systematic prospective evaluations will be utilized in order to provide validation to the phenomenon of regression in autism, determine diagnostic and functional outcomes, and evaluate behavioral, medical and immunologic functioning. These evaluations will include comprehensive medical history, behavioral assessment, physical and neurologic examination, sleep EEG, structural MRI, blood work for laboratory assays, and for some children, lumbar puncture. In an investigation of potential immunologic abnormalities and associated neuroinflammation, repeated assessments are necessary to determine whether both behavioral and/or immunologic parameters are due to state versus trait alterations: elements of these assessments will be conducted every 6 to 12 months, depending on the child's age.

Our objective is to determine if there is a unique alteration in immune function among autistic children with a regressive clinical course. Our primary hypothesis is that at least some children with regressive autism will have abnormalities in immune function. These abnormalities will not be found among autistic children without a regressive course nor will they be found among children without an autistic spectrum disorder. Our specific aims are:

Specific Aim #1: To characterize and validate the regressive phenotype in children with autism.

Working hypothesis: Children with the diagnosis of autism who had a regressive clinical course can be reliably distinguished from children with non-regressive autism, Rett's Disorder and healthy children. When compared to children with non-regressive autism, the history of children in the regressive group will consist of similar or higher skill levels prior to regression, and lower skill levels following regression.

Specific Aim #2: To characterize the immune response in children with autism.

Working Hypothesis 2a: Distinct immunologic responses and cytokine abnormalities will be evident in children in the regressive autism sample, but not in the non-regressive autism, Rett's Disorder, or typically developing samples.

Working Hypothesis 2b: MRI scans will reveal regional neuroinflammation among children in the regressive autism sample, but not in the non-regressive autism, Rett's Disorder, or typically developing samples.

Specific Aim #3: To identify neurobiologic markers for autism through the techniques of metabolomics, proteomics and genomics (e.g. gene expression profiles).

Working hypothesis: These neurobiologic markers will reliably differentiate between regressive autism, non-regressive autism, Rett's Disorder, and typically developing children.

In addition, although progress has been made in identifying children affected by autism, little attention has been paid to the different manifestations of autism (or autisms). As has been demonstrated in "diabetes" and "cancer", it is expected that the differences between individuals with autism may be as informative as their behavioral similarities. Complementing this study that investigates one possible subtype (regression), we propose to pilot a phenomenologic investigation to identify other biological and behavioral subgroups of autistic individuals, in order to facilitate investigations of etiology, pathophysiology, treatment and prevention. The present investigation thus includes a pilot study to evaluate feasibility of conducting a large-scale, multi-site longitudinal investigation of 1,800 children (600 children with autism, 600 healthy controls, and 600 children with developmental delays). The protocol has been approved by the UC-Davis IRB (see attachments) and the study is currently underway at the M.I.N.D.

Institute in Sacramento, California. NIMH is an ideal second site for the protocol, because of the overlap between the design of this regression study with their Autism Phenome Project. Children participating in 06-M-0102 will be invited to participate in this study; if interested, their data will be entered into the Phenome study. The age at study entry for the Phenome study (2 - 6 years) is slightly older than the original regression study, so we will recruit up to 40 children with autism and 25 age-/sex-matched healthy controls to this investigation. In addition, as many as 25 children (ages 2 - 6 years) with developmental delays (without autism) will be recruited as contrast subjects.

• Ahlsen G, Rosengren L, Belfrage M, Palm A, Haglid K, Hamberger A, Gillberg C. Glial fibrillary acidic protein in the cerebrospinal fluid of children with autism and other neuropsychiatric disorders. Biol Psychiatry. 1993 May 15;33(10):734-43.
• Ashwood P, Anthony A, Pellicer AA, Torrente F, Walker-Smith JA, Wakefield AJ. Intestinal lymphocyte populations in children with regressive autism: evidence for extensive mucosal immunopathology. J Clin Immunol. 2003 Nov;23(6):504-17.
• Ashwood P, Van de Water J. Is autism an autoimmune disease? Autoimmun Rev. 2004 Nov;3(7-8):557-62. Review.

• INCLUSION CRITERIA:

The sample at study endpoint will contain 50 children diagnosed with idiopathic autism and regression (AUT-R), 50 children diagnosed with idiopathic autism (and no history of regression; AUT-NR), 75 typically developing children (TYP), matched on ethnicity, sex, chronological age and daycare enrollment. Matching for daycare setting (home versus setting with other children) will be attempted due to the potentially greater immunologic challenge experienced by children in daycare settings. In addition, as many as 40 children with autism (with or without a history of regression) will be included in the evaluations and 25 children with developmental delays (without autism) will be recruited as contrast subjects. The age range of all groups will be between 12 and 60 months at first visit. We will over-recruit for children in the 12-36 month age range, in order to maximize participants of the age when regression associated with autism diagnosis is most likely.

Regression will be defined as:

Language loss: Loss of at least 3 spontaneously meaningful words (excluding mama or dada) used for at least a month, and lost for at least a month.

AND/OR

Nonverbal communication/social loss: Loss of more than one nonverbal communicative behavior (e.g. gestures, joint attention, eye contact, imagination, pretend play, sharing, showing, watching children, orienting to name, social smiling, social games, spontaneous imitation of actions, response to social overtures).

Autism groups (AUT-R & AUT-NR): Children with a diagnosis of autism or referral from professional with concerns of a diagnosis of autism will be evaluated with the ADI-R and ADOS. Those meeting research criteria for autism will be included.

Although there is overlap in the age period for regression specified in this protocol (regression between 15-30 months) and that described for childhood disintegrative disorder (CDD) (the DSM-IV criteria for CDD indicate apparently normal development for the first 2 years after birth), the criteria for CDD also indicate a diagnosis can only be made if symptoms are not better accounted for by another pervasive developmental disorder such as autism. Symptoms of CDD that separate it from autism include loss of acquired skills in areas such as motor skills and bowel or bladder control, while the focus of the current study is on regression in core symptoms of autism (i.e. socio-communicative skills). Thus, the current protocol will include children with regression over 2 who meet criteria for autism, but not those who only meet criteria for CDD.

Typically Developing control group (TYP): No diagnosis of developmental delay and no first-degree relatives with history of developmental problems.

Developmental Delay group (DD): Developmental scores (Performance Quotient and Verbal Quotient) greater than 1.5 standard deviations below mean on Mullen Scales of Early Learning

EXCLUSION CRITERIA:

1. Diagnosis of cerebral palsy. For the typical controls only, a history of extremely low birthweight (due to prematurity or intrauterine growth failure).
2. If behavioral management issues (e.g. self-injury, aggressiveness) are severe to the extent that the screening protocol was aborted.