News and Notes about Scientific Research on Autism and other Developmental and Behavioral Disorders

Editor: Bill Ahearn, Ph.D., BCBA
Director of Research, New England Center for Children

Combating Autism Bill

The United States Congress has passed and the President signed into law the Combating Autism Act of 2006.  This legislation earmarks nearly $1 billion over the next five years for research, early detection, and early intervention for autism spectrum disorders.  It is thought that this could increase government funding for autism by 50%. In commenting on the bill, White House press secretary Tony Snow said:

“This bill will increase public awareness about autism and provide enhanced federal support for autism research and treatment by creating a national education program for doctors and the public about autism. The legislation will help more people recognize the symptoms of autism. This will lead to early identification and intervention, which is critical for children who struggle with the disorder. The President is confident that the legislation will serve as an important foundation for our nation's efforts to find a cure for autism.”

New evidence implies ASDs are not caused by vaccines

The prevalence of autism has risen dramatically in the past two decades.  Prevalence is an estimate of the number of affected persons at a point in time.  The last issue of the NECC Research Newsletter discussed the first comprehensive prevalence studies conducted in the United States.  These studies show that the prevalence of ASDs in 2003-2004 is somewhere between 1 in 139 and 1 in 181 children between the ages of 4 and 17.  This implies that there are around 300,000 children diagnosed with an ASD.  Most scientists feel the rise in prevalence is due to a combination of changes in the diagnostic criteria for autism spectrum disorders (ASDs) and increased awareness of the disorder (e.g., Wing and Potter, 2002). 

However, some have suggested that the increased prevalence of ASDs corresponds with an increase in the number of vaccinations recommended for children.  Correlation of two events is not sufficient evidence to assert that one caused the other as the two events could be unrelated.  Therefore, further study of such correlations is necessary to reveal evidence to either support or disconfirm a causal hypothesis.  One specific hypothesis of vaccines being linked to autism suggests that thimerosal, a preservative previously used in childhood vaccines that was removed from vaccines manufactured in the US in 1999, can cause autism.  Thimerosal is still present in some versions of the flu vaccine.  Several versions of this theory target different mechanisms for how thimerosal damages the child and causes autism.  They all, however, state that some damage occurs to the developing child after vaccination.  Advocates of the “thimerosal causes ASDs” hypothesis have also suggested that the prevalence of ASDs will substantially decrease subsequent to thimerosal being removed from childhood vaccines. 

A recent study published in Pediatrics by Fombonne and colleagues (2006) has shown that decreased thimerosal in childhood vaccines is not correlated with decreased prevalence of ASDs.  Fombonne and colleagues have collected data on the prevalence of ASDs in children in Montreal from 1987 through 1998.  Thimerosal was removed from vaccines in Canada by 1996.  Fombonne et al. found that there was a statistically significant increase in the prevalence of ASDs for children who received thimerosal-free versions of the recommended childhood vaccines.  Similar findings have been obtained in Denmark and Sweden and will likely be revealed in the US in the not too distant future.  Informal review of several educational data systems suggests that the prevalence of ASDs is continuing to increase in the US. 

Another hypothesis suggests that the measles-mumps-rubella (MMR) vaccine causes autism.  ASDs clearly have a genetic origin but one environmental trigger has been identified.  If a pregnant mother contracts rubella, there is an increased chance of the child having an ASD (Chess et al., 1971/74/77).  As the MMR vaccine decreases the chance of rubella infections, its use should decrease the likelihood of this environmental triggering of autism.  However, some have posited that the MMR vaccine triggers regression that is sometimes part of the course autism (i.e., some children with ASDs lose skills that they previously acquired in early development).  Most of the popularization of this theory has been spurred by Andrew Wakefield.  He suggested that the MMR injures a child’s gut in a specific manner producing problems similar to those caused by irritable bowel syndrome.  He also stated that measles was the cause because he and his colleagues purportedly detected components of the measles virus in the gastrointestinal tracts and blood of children with autism that were not present in typically developing children. 

Another study recently published in Pediatrics attempted to determine whether measles was more likely to be found in the bodies of children with ASDs than in typically developing children.  D’Souza and colleagues (2006) collected the largest sample of subjects for this type of study and used the same technique, polymerase chain reaction assays, that had purportedly detected measles in children with ASDs.  They found that this technique produced many positive reactions in both children with ASDs and typical children.  However, these reactions were further analyzed and found to be false positives for all subjects.  The products of the reactions were cloned and genetically sequenced and none of these sequences contained the components of the measles virus.  That is, neither the children with ASDs nor the typical children showed any evidence of measles virus in their bodies.  Furthermore, there were no differences found in anti-measles antibodies across the study groups of children. 

This, taken in combination with numerous other studies showing no relation between the MMR vaccine and ASDs, provides fairly definitive evidence against the “MMR causes autism” hypothesis.  It should also be noted that Andrew Wakefield was found to have been paid over $150,000 by a group seeking to pursue litigation against vaccine manufacturers in the United Kingdom.  Once this severe conflict of interest was revealed, ten of Wakefield’s co-authors requested that their names be withdrawn from the original publication used as support for this hypothesis.  More information on Andrew Wakefield can be found at BrianDeer.com, a website maintained by the investigative journalist who revealed the conflict of interest

D’Souza et al., (2006).  No Evidence of Persisting Measles Virus in Peripheral Blood Mononuclear Cells From Children With Autism Spectrum Disorder.  Pediatrics, 118(4), 1744-1745.

Fombonne et al., (2006).  Pervasive Developmental Disorders in Montreal: Prevalence and Links with Immunizations.  Pediatrics, 118(1), 139-150. 37.

Wing, L. & Potter, D. (2002).  The epidemiology of autistic spectrum disorders: Is the prevalence rising?  Mental Retardation and Developmental Disabilities Research Reviews, 8(3), 151-161

Genetics research identifies gene linked to autism

It has long been thought that autism has a genetic origin.  Early studies of twins showed that identical twins were more likely than fraternal twins to both have autism.  More recent studies suggest that there is a nearly 90% probability of an identical twin having an autism spectrum disorder (ASD) when his sibling also has one.  In fact, twins and other siblings are much more likely to have a diagnosis of an ASD if there is an immediate family member with autism.  There are a number of other variables that suggest a genetic origin to ASDs such as the increased prevalence of seizures in this population and the correlations between autism and other disorders (i.e., Fragile X syndrome and tuberous sclerosis) of known genetic origin.  However, there has been little success in identifying the specific gene or genes involved in autism. 

Campbell and colleagues (2006) at Vanderbilt University have uncovered a component of a gene sequence that seems to be strongly correlated with autism and that results in suppression of that gene’s function.  They studied genetic material made available to them through the Autism Genetic Resource Exchange (AGRE) which has been formed to aid in biomedical autism research. The affected gene is involved in producing a protein, MET.  The MET protein is involved in development and problems with it have previously been linked to metastases of cells that occur with cancer.  The C variant of the MET gene was much more likely to be found in persons with an ASD.  It was also found that the C variant was often present in parents of children with an ASD if those parents had more than one child diagnosed with an ASD. 

These researchers have also closely studied how the MET gene may affect development.  They have found that the MET gene is involved in neural development and have suggested that because the C version of the MET gene does not function properly, it might be involved in the abnormal brain growth thought to occur in persons with ASDs.  This correlation between the presence of the C variant of the MET gene and ASDs must be replicated to confirm this link but it is a promising finding that corresponds to our current understanding of how autism affects brain development.  Acosta and Pearl (2003) have reviewed neuro-developmental research which indicates that autism likely involves abnormal brain growth development that is prenatally determined.  Though this finding is important for understanding the origin of autism, potential applications of this finding are not immediately imminent.

For more information on this work, go to:

http://www.nih.gov/news/pr/oct2006/nimh-17.htm

Acosta, M.T. & Pearl, P.L. (2003).  The neurobiology of autism: New pieces of the puzzle. Current Neurology and Neuroscience Reports, 3, 149-156.

Campbell et al. (2006).  A genetic variant that disrupts MET transcription is associated with autism.  Proceedings of the National Academy of Science, 103(45), 16834-16839.

Research at NECC

Editor’s note: Staff from NECC have recently attended the 27th annual meeting of the Berkshire Association for Behavior Analysis and Therapy in Amherst, MA. In this issue of the newsletter, we’re including the abstracts for four of the over 30 presentations by NECC staff.  The four abstracts were presented in a symposium chaired by NECC Program Director, Rick Graff, detailing work that he and his colleagues have been conducting on preference and reinforcement.  This line of work has consisted of modifications and extensions of methods for identifying preferred stimuli and reinforcers in individuals with developmental disabilities.  Identifying effective reinforcers is a necessary first step in developing effective educational programs for children diagnosed with autism.  For this reason, enhancing current methods for assessing preferred items is critical for promoting skill acquisition and independence for students at NECC.  The first abstract is for Theresa Cerrone’s presentation. This talk detailed some work that has been conducted with preschool children at NECC.  This work has consisted of comparing outcomes from verbal preference assessment (asking children what they want) and tangible preference assessments (presenting multiple items) to determine whether the former method is as effective as the latter in identifying reinforcers for children diagnosed with autism.

Using preference assessment technology to assess the correspondence between “saying” and “doing” in preschoolers with ASD.  THERESA CERRONE, (New England Center for Children), Amy Lipcon (New England Center for Children), Danielle Vigeant (New England Center for Children, Candice Colón (New England Center for Children), and Rick Graff (New England Center for Children)

Although a great deal of research has been published on identifying reinforcers for individuals with developmental disabilities, the extent to which preschool children with autism can identify their own reinforcers has rarely been explored.  In this study, verbal and tangible paired-stimulus preference assessments were compared in 4 preschoolers with autism.  In the tangible assessment, on each trial 2 stimuli were placed in front of the participant.  In the verbal assessment, on each trial participants were asked, “Do you want x or y”.  For both assessments, the percentage of opportunities each stimulus was chosen or named was calculated, and preference hierarchies were developed.  A second observer independently collected data on 50% of trials on both assessments.  For all participants, interobserver agreement was above 98% for the stimulus chosen or named.  The two assessments identified the same most- and least-preferred item for only 2 of 4 participants; for these participants, reinforcer assessments demonstrated that high-preference items functioned more effectively as reinforcers than low-preference items.  These results suggest that verbal preference assessments may be reliable predictors of reinforcers for some, but not all, young children with autism.

Editor’s note: The next abstract is a study by Jennifer Keras and Rick Graff. This study involved providing parents of preschool children diagnosed with autism with a 90-min training on how to conduct stimulus preference assessments.   They showed that a training involving lecture, demonstration, and practice, was effective in teaching parents to accurately conduct and record data on stimulus preference assessment procedures in their home with their child.  

Teaching staff and parents to conduct paired-stimulus preference assessments.  JENNIFER KERAS (New England Center for Children) and Rick Graff (New England Center for Children)

Over the past 20 years, a great deal of literature has been published on assessing preferences and identifying reinforcers with individuals with disabilities.  It is not clear how well this research has translated to practice, however, and it seems unlikely that parents frequently conduct systematic preference assessments with their own children with disabilities.  In this study, 14 parents of preschool children with autism were taught to conduct paired stimulus preference assessments.  Prior to training, a pretest was given to assess parents’ knowledge of preference and reinforcement; mean accuracy was 52.6% (range, 21% to 87%).  Parents then received 90 min of training on conducting preference assessments that involved lecture, demonstration, and practice.  Following the training, mean accuracy on posttests increased to 80.6% (range, 64% to 100%).  The following week, parents conducted a paired-stimulus assessment in their home with their child, and a trained observer was present to collect interobserver agreement (IOA) and procedural integrity (PI) data on at least 50% of trials.  Interobserver agreement and PI data indicated that parents accurately conducted preference assessment procedures and accurately recorded data on 98% of trials, suggesting that parents can learn some aspects of behavior analytic technology in relatively short periods of time.

Editor’s note: The next abstract is a study by Tracey (Toran) Galiatsatos, Rebecca Maxfield, and Rick Graff.  This study involved conducting stimulus preference assessments for community-based activities by providing pictures that represented various activities.  Although the activities chosen where not immediately provided, results showed that items frequently selected on the preference assessment were also often selected when trading in their tokens.  These results suggest that stimulus preference assessments can show participants’ preference for activities that cannot be immediately delivered (e.g., going to the movies).  

Assessing preferences for community-based activities.  TRACEY (TORAN) GALIATSATOS (New England Center for Children), Rebecca Maxfield (New England Center for Children), ELISE Battle (New England Center for Children), and Rick Graff (New England Center for Children)

Although pictorial paired-stimulus (PPS) preference assessments have been used to successfully identify edible and sensory reinforcers, their utility in accurately identifying preferences for community activities has not been explored.  In this study, 2 individuals with developmental disabilities, ages 34-36, participated.  Both participants had token reinforcement programs that allowed them access to additional community activities contingent on the absence of challenging behavior.  Seven sets of PPS assessments were conducted with 6 community activities.  During the PPS assessment, pictures of two community activities were randomly selected and placed in front of the participant, who was asked, “Where do you want to go?”  No consequence was provided for pointing to one of the pictures.   The percentage of opportunities each picture was touched was calculated, and preference hierarchies were developed.  Interobserver agreement (IOA) data were recorded in 75% of sessions across participants and assessments; mean IOA was 100%.  Immediately after completing the PPS assessment, the participants traded in their tokens, and went to the community location of their choice.  On 5 of 7 occasions, the participant traded in their tokens to access the item that ranked first on the PPS assessment, suggesting that PPS assessments may accurately identify preferences for community-based activities.

Editor’s note: The last abstract is of a paper presented by Rick Graff describing two studies. In the first, the effects of including various items in the preference assessment array were evaluated.  In the second study, the effects of using easy relative to more difficult tasks when conducting reinforcer assessments was explored.  Results showed that outcomes obtained from preference assessments may more accurately predict absolute reinforcement effects when a range of items are included in the stimulus array and when difficult tasks are included in the reinforcer assessment.

Graff, R. B., Larsen, J., Leonard, M A., & Kelly, E. (October, 2006). Toward an analysis of variables that affect preference and reinforcer assessment outcomes.  RICK GRAFF (New England Center for Children), Jill Larsen (New England Center for Children, Michelle Leonard (New England Center for Children), and Erin Kelly (New England Center for Children)

In Study 1, an initial paired-stimulus assessment (PS-1) used items presumed to be high-preference.  Distinct preference hierarchies were generated, but subsequent single-operant reinforcer assessments (RA-1) indicated that high- and low-preference stimuli were associated with similar response rates.  A second PS assessment (PS-2) was then conducted, using the least-preferred item from PS-1, plus 7 new stimuli.  The items ranked as least preferred on PS-1 were now ranked as most preferred on PS-2.  On RA-2, the top-ranked items from PS-2 were associated with high response rates, while the items ranked as least preferred did not function as reinforcers.   These results indicate that PS assessments do not suggest the absolute reinforcing value of stimuli, but instead only provide information related to the relative reinforcing value of stimuli.  Study 2 examined how the results of reinforcer assessments were influenced by task difficulty.   High- and low-preference stimuli were identified using paired-stimulus preference assessments for 3 individuals with developmental disabilities.  Reinforcer assessments were conducted using easy tasks versus more difficult tasks.  Results showed that both high and low preference stimuli functioned as reinforcers when the easy task was used and only high preference items functioned as reinforcers when the difficult task was used for two of the three participants.

Web Resources

• For information about the New England Center or to access and electronic version of this or previous NECC Research Newsletters, visit our Web site www.necc.org.
• For information about autism, visit the National Library of Medicine’s autism site www.nlm.nih.gov/medlineplus/autism.html.
• For information about applied behavior analysis in the treatment for autism visit www.behavior.org.
• For science-based information on biomedical treatments and theories in autism visit www.autism-biomed.org.
• For information on health issues in general visit the World Health Organization www.who.int.
• For professionally-screened information on health care (including some treatments for autism and other developmental disabilities), visit www.quackwatch.com.
• For information on the Berkshire Association for Behavior Analysis and Therapy, visit http://www.babat.org.
• For information on the Association for Behavior Analysis, visit http://www.abainternational.org/.

Back to Research Newsletter index page