Stein, D., Ring, Al, Shulman, C., Meir, D., Holan, A., Weizman, A., & Barak, Y. (2003).  Brief report:  Children with autism as they grow up - description of adult inpatients with severe autism.  JADD, 31, 355-360.

• most follow-up studies - show improvement over the years?  THEY DO?
• 2/3 have poor outcomes
  • 10-20% good by adolescence or adulthood
  • 1-2% live independently
  • predictor - IQ and speech
    • siezures seem to have an effect as well
• clinical features seem to be stable over time
• three clinical subgroups
  • aloof and indifferent
  • passive and friendly
  • active but odd
    • more stable clinically if lower IQ
    • most stable characteristics
      • social impairment
      •  steroetyped movements
      • concrete thinking
      • odd language
      • flat affect
      • anxiety
• present study - looks at adults diagnosed with autism as children
  • will the clinical features still be there?
  • what are the most prominent ones?
  • will they still meet the criterion?
  • IQ = more severe outcome?
• 28 inpatient individuals with autism
  • seems to be like an institution in the US - both of these are in Jerusalem
  • all are 20 or older
  • all diagnosed with autism in Early childhood
  • diagnosed by old DSM standards, ICD standards
• Instruments
  • CARS
  • Structured questionnaire based on the DSM-IV
  • Structured interview for other Axis I disorders
  • nonverbal IQ - Raven's Colored Progressinve matrices test - used figures to test:
    • analogy
    • comparisons
    • perceptual parts of a whole
  • 5 participants did not test,  one was blink
• Procedures
  • CARS - done by 2 physicians
  • DSM questionairre done by independent psychiatrists
    • 2 day observation and interview of staff and family
  • third team - interviewed parents and staff for other Axis I impairments
  • review of medical records
  • Clinical psychologist - Raven's test
• Stats - two-tailed t-test, chi square
  • Pearson and Spearman's RHO for CARS and DSM questionnaire scores
• Results
  • no between groups differences (from two different hospitals) - N size is somewhat low
    • 20 males
    • 8 females
    • mean - 26.9 years old
      • hospitalized mean - 12.5 years of age
  • 26 patients - have full ASD, 2 autistic disorder, "residual type"?
  • 82% social interaction
  • 61% communication
  • 50% stereotypy
  • 96% lack of peer development
  • 86% lack of social reciprocity
  • 58% lack of spoken language
  • CARS - all scored higher than 30
    • significant correlation between the CARS and the DSM structured interview
  • mean nonverbal IQ 45.1
    • correlated with the CARS but not the DSM structured interview
    • did not correlate with the amount of time hospitalized
  • no one had any other Axis I diagnoses
  • 46% siezure disorder - more likely to have a lower IQ score
  • 75% taking antipsychotics
  • 57% antiseizure
    • 39% taking both
    • no differences in medication between those with and without siezure disorders (use and dosage)
• Discussion
  • almost all the participants continued to have classic autism
  • did not change, did not develop different psychiatric diagnosies
  • most had severe MR as well
  • impairments in social and communication, as well as repetitive activities
  • 2/3 - no spoken language
  • CARS and IQ - worse outcome - not sure what this means in terms of the data collected?
  • can't speculate about relationships between language and outcome - interaction effects possible
  • often receiving high doses of medication for problem behavior
    • can cause adverse effects
    • should consider lowering the doses
  • more adults with siezure disorders - more severe the disorder, more siezures?
    • could this also affect the IQ?
  • limitations
    • selective group and retrospectie design
    • reliance on informant assessment and checklists
    • no diagnosis Axis I might have been an artifact of the autism itself
    • cohort effects
    • possible observer bias (even though the raters were blind to the hypothesis)
  • advantages
    • independent teams
    • structured ratings
    • evaluations - many informants, thorough evaluations
    • homogeneous group
• more longitudinal research is needed to make conclusions
  •