Gastrointestinal pathologies in autism: Did Mayo's Ibrahim and colleagues err?

Teresa Binstock
Researcher in Developmental & Behavioral Neuroanatomy
January 31, 2010


In 2009, Samar H. Ibrahim and colleagues at Mayo Clinic published a study summarizing the researchers' glimpse at gastrointestinal pathologies in children with and without autism (1). Question arise: Did the study's methodology have one or more major weaknesses? How valid are the group's findings and interpretations? Is the researchers' article generating misleading impressions?

The Ibrahim et al study was published online on July 27, 2009, in the journal "Pediatrics". On July 28th, 2009, an initial critique was shared by autism parent and physician Bryan Jepson, M.D., author of a fully citationed but readable text Changing the Course of Autism.

In regard to the Ibrahim et al study, Dr. Jepson wrote:

There are several problems with the Mayo study.

First, it is retrospective which means that it relies on how well the physicians record the symptoms (even if the access to records is great).  It has been shown in reviews of other studies in autism comparing retrospective to prospective designs, that many physicians simply ignore the GI symptoms in children with autism.  There may be many reasons for this, including that they don’t believe that their GI symptoms are relevant to the presenting complaint, there is difficulty getting a history from a non-verbal child (i.e. the kids may not complain), the symptoms present differently from other kids (they don’t describe heartburn for reflex; or the stools are only mushy, not watery so they don’t consider it “diarrhea”; or their stools are infrequent and hard to pass but they come out loose so they don’t consider this constipation; or they don’t look at or measure the abdomen and since the kids don’t complain, they don’t record abdominal bloating).    When studies have been done that look prospectively at GI symptoms in autism with specific targeted questionnaires, they have always shown a difference between kids with autism and controls.


Second and more importantly, they [Ibrahim et al] measure a cumulative incidence.  That means that if any of the kids had diarrhea recorded in their chart, even if only once, it would be counted.  That is why they have such high incidence in both kids with autism and normal kids.  It is very common to have one or two bouts with constipation or diarrhea over your entire childhood.  The important difference is how chronic it is, not if you’ve ever had it.  So having a viral illness with diarrhea that last for a week or two was counted the same as children who may have had chronic loose stools every day of their life.  They do not distinguish these issues in their study data.

The first and second concerns delineated by Dr. Jepson are very important. Indeed, a Medscape reviewer, William T. Basco, Jr., MD, FAAP, called attention to the weakness of tracking cumulative incidence and doing so without evaluating duration or severity of a child's gastrointestinal pathologies, writing  "The authors admit that a limitation of the study is the fact that the analyses do not account for duration or severity of symptoms, both of which would have required more extensive, primary data collection during each episode of gastrointestinal symptoms compared with chart review." In other words, in the Ibrahim study, a child who had diarrhea once for three days would be counted the same as a child who had diarrhea for 6 months.  As physicians Jepson and Basco suggest, the child with a brief bout of diarrhea or constipation may be physiologically very different from a child whose intestinal pathology is prolonged.

In his review of July 28, 2009, Dr. Jepson presented three additional concerns:

Third, they [Ibrahim et al] did find differences in constipation and food selectivity but they interpret this as a behavior consequence not a physiological problem.  That is their opinion.  There is nothing in their study design that would show if that is true or not true.

Fourth, they [Ibrahim et al] mentioned that very few of the autistic children had documented Crohn’s disease, Celiac disease or enzyme deficiency.  But they don’t mention how many of the kids were worked up for these things.  If you don’t look, you won’t find it.  Why not report the number of scopes, biopsies or celiac panels?  My guess is that it is because very few of these kids were actually worked up appropriately.

Fifth, assessing the prevalence of some of the GI symptoms in autism is very challenging without a further workup because many of the symptoms are subjective and require patient cooperation to get at the diagnosis.  For example, abdominal pain.  If a child cannot tell you that they are having abdominal pain, how are you going to identify it or record it.  Especially when many of the potential manifestations of abdominal pain in autism (irritability, aggressiveness, self-injurious behaviors, etc) are often and easily explained away as typical “autism” behavior and a pain etiology is not further explored.  How does a child with autism who is having reflux describe that to a doctor.  Unless it is severe enough to be causing vomiting, asthma, etc , it will be missed.  Yet, neurotypical kids can describe heartburn or epigastric discomfort and are more likely to have it diagnosed.  So mild cases will be diagnosed in neurotypical kids and only severe cased in autistic children.  This will change the prevalence numbers between the cases and controls  and it all comes down to how easy it is to obtain the history, not the true prevalence.  I suspect that if you took all of those children, both neurotypical and controls, and scoped them, that the level of pathology would be much, much higher in autism.


Oversights such as those inherent in the methodology, findings, and interpretations offered by Ibrahim and colleagues - especially given their Mayo Clinic affiliation - can be unintentionally presumed to be authoritative and may have subserved a recent U.S. News & World Report article, wherein obviously caring correspondent Judith Palfrey, M.D., cited Harvard's William Barbaresi, M.D., and offered, "Recent studies show that children with autism are at no higher risk of having gastrointestinal problems than are children without autism." (5)  That's a major point in the critiques by doctors Jepson and Basco. Studying incidence without including severity and duration as co-factors can lead to misleading oversimplifications.

Dr. Palfrey also mentioned diets and autism, "I'm wondering about the recent news showing little evidence that special diets do any good." Here it's important to offer a mild rebuttal. Evidence cited in reviews by Timothy Buie, M.D., and colleagues (6-7) is consistent with at least one other study (8) and with long-term data compiled by the Autism Research Institute (9): according to parents of children with autism, special diets help in ~60% of autistic children who were placed on a restricted diet (reviewed in 10).

Conclusion: That gastrointestinal pathologies in children with autism and other autism-spectrum disorders (ASDs) can be contemplated, evaluated, and treated is a major step forward and, to some extent, the study by Mayo Clinic's Samar Ibrahim and colleagues is helpful even as its methodology was seriously flawed and its findings and conclusions leave much to be desired. As months turn into years, the full and possibly etiologic significance of gastrointestinal pathologies in children with autism or one of the other ASDs will be more fully appreciated, including a direct gut-brain-language connection (11) and the increasingly described intestinal lymphoid hyperplasia (12-13).

Evaluating and treating autistic children's intestinal pathology has been decreed a legitimate concern (6-7).


References
:

1.
Incidence of Gastrointestinal Symptoms in Children With Autism: A Population-Based Study
Samar H. Ibrahim, Robert G. Voigt, Slavica K. Katusic, Amy L. Weaver, William Barbaresi.
Departments of Pediatric and Adolescent Medicine and Health Sciences Research
Mayo Clinic, Rochester, Minnesota
Pediatrics 2009;124;680-686. Epub 2009 Jul 27.
$ http://pediatrics.aappublications.org/cgi/content/full/124/2/680

2.
Changing the Course of Autism
Bryan Jepson, M.D., with Jane Johnson.
http://www.amazon.com/Changing-Course-Autism-Scientific-Physicians/dp/1591810612

3.
Critique of Ibrahim et al 2009{cite-1 herein}
Bryan Jepson, M.D.
July 28, 2009.
Personal communication, excerpts shared with permission.

4.
Do Children With Autism Experience High Rates of Gastrointestinal Problems?
{Review of Ibrahim et al, cite-1 herein}
William T. Basco, Jr., MD, FAAP
http://www.medscape.com/viewarticle/709656

5.
What's the Story About Gastrointestinal Problems in Kids With Autism?
Judith Palfrey, M.D.
http://www.usnews.com/health/blogs/health-advice/2010/01/29/whats-the-story-about-gastrointestinal-problems-in-kids-with-autism

6.
Evaluation, Diagnosis, and Treatment of Gastrointestinal Disorders in Individuals With ASDs: A Consensus Report
Buie T et al.
Pediatrics 2010;125:S1–S18.
http://pediatrics.aappublications.org/cgi/content/full/125/Supplement_1/S1

7.
Recommendations for Evaluation and Treatment of Common Gastrointestinal Problems in Children With ASDs
Buie T et al.
Pediatrics 2010;125:S19–S29
http://pediatrics.aappublications.org/cgi/content/full/125/Supplement_1/S19

8.
A randomised, controlled study of dietary intervention in autistic syndromes
Knivsberg AM, Reichelt KL, Høien T, Nødland M.
Nutr Neurosci. 2002 Sep;5(4):251-61.

9.
ARI Parent Ratings: For several decades, the Autism Research Institute has been creating a database comprised of parental reports of what treatments worked (for a specific child), what treatments didn't work; which set the child back, which didn't seem to have any effect. The data include pharmaceuticals, supplements, and diets. In each line-item within the data presentation, consider the ratio (per treatment) of got-better/got-worse.
Parent Ratings of Behavorial Effects of Biomedical Interventions
http://www.autism.com/treatable/form34qr.htm


10.
Gastrointestinal pathologies in ASD children: new autism-gastro stance by AAP
Teresa Binstock; January 05, 2010

11.
Anterior insular cortex: linking intestinal pathology and brain function in autism-spectrum subgroups.
Binstock T.
Med Hypotheses. 2001 Dec;57(6):714-7.

12.
Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children
Wakefield AJ et al.
Lancet. 1998 Feb 28;351(9103):637-41.

13.
Clinical Presentation and Histologic Findings at Ileocolonoscopy in Children with Autistic Spectrum Disorder and Chronic Gastrointestinal Symptoms
Arthur Krigsman, Marvin Boris, Allan Goldblatt and Carol Stott
Autism Insights 2010:2 1-11; 27 Jan 2010
http://www.la-press.com/clinical-presentation-and-histologic-findings-at-ileocolonoscopy-in-ch-a1816


Contact Teresa Binstock by email


Return to index of essays by Teresa Binstock