Autism and Gastrointestinal symptoms: two new studies

28 Jul

Autism and poop. You hear those two words in the same sentence a lot on the net. People have been asking for studies on whether autistics have a higher incidence of gastrointestinal (GI) problems for a long time. Well, two papers came out in the last week with answers…and many parents are not happy.

The two papers are:


The early stool patterns of young children with autistic spectrum disorder

by B Sandhu, C Steer, J Golding, A Emond of the University of Bristol

and

Incidence of Gastrointestinal Symptoms in Children With Autism: A
Population-Based Study

by Samar H. Ibrahim of the Mayo Clinic.

The Bristol group’s study came out last week. Given that the Mayo Clinic study was on the way, I figured I’d wait and blog them both at the same time. Actually, I considered not blogging them at all. These papers are more nails in the coffin for Andrew Wakefield’s hypothesis that MMR causes “autistic enterocolitis” and the belief by many that this drove much of the “autism epidemic”. But, tired as that story is, the question of whether autistics have GI problems at a higher rate is important and worth discussing.

The Bristol study has free pdf access. Not so the Mayo Clinic study: abstract only, but I have a copy. Rather than go through the studies in detail (if you are that interested you will likely read the paper for yourself), let’s just look at the results and conclusions sections of the abstracts:

Bristol group:

Results: Comparison of the ASD and control group during the first 3.5 years of life showed no major differences in stool colour or consistency, or in frequency of diarrhoea, constipation, bloody stools or abdominal pain. The ASD children had similar stool frequency up to 18 months, but there was a trend for ASD children to pass more stools at 30 months (OR 3.73, 95% CI 1.11 to 12.6; p=0.004) and at 42 months (OR 6.46, 95% CI 1.83 to 22.7; p,0.001), although only three children passed more than 4 stools/day. Repeating the analysis on only those cases diagnosed as having classical childhood autism resulted in very similar findings.

Conclusions: During the first 42 months of life, ASD children had a stool pattern that was very similar to that of other children, apart from a slight increase in stool frequency at 30 and 42 months. There were no symptoms to support the hypothesis that ASD children had enterocolitis.

Mayo Clinic:

RESULTS: Subjects were followed to median ages of 18.2 (case subjects) and 18.7 (control subjects) years. Significant differences between autism case and control subjects were identified in the cumulative incidence of constipation (33.9% vs 17.6%) and feeding issues/food selectivity (24.5% vs 16.1). No significant associations were found between autism case status and overall incidence of gastrointestinal symptoms or any other gastrointestinal symptom category.

CONCLUSIONS: As constipation and feeding issues/food selectivity often have a behavioral etiology, data suggest that a neurobehavioral rather than a primary organic gastrointestinal etiology may account for the higher incidence of these gastrointestinal symptoms in children with autism.

Or, to put in a single sentence: there is no evidence that children with autism have GI problems at a greater rate than the general public.

How about repeating that with emphasis: there is no evidence that children with autism have GI problems at a greater rate than the general public. They are not saying that there are no children with autism and GI issues. Quite the contrary. You wouldn’t know that to read some comments on the internet about these studies.

I’m a little surprised by these results. No, I don’t think that Wakefield was right. But, I wouldn’t be surprised if children with autism have other medical concerns at higher rates. Also, there were two abstracts from IMFAR 2008 that stuck in the back of my mind.

In the first, a team from the University of Connecticut presented a study suggesting that GI issues may be more common in children with children with ASD’s (but at a similar rate to children with other developmental delays).

No evidence for higher rates of gastrointestinal problems in young children with ASDs versus those with other developmental delays

Conclusions: In this sample of young community-based children with ASDs and other developmental delays, no significant group differences in parentally reported feeding problems and gastrointestinal symptoms were found at age two or at age four. Most published research has been conducted at specialty GI or DD/ASD clinics with older children. The results of this study suggest that their findings may not be applicable to young children or to children evaluated in community settings. While GI problems may be increased in children with developmental disorders, we found no evidence that they were specific to autism spectrum disorders.

The second abstract (which later became a paper that was discussed on this blog): David Mandell’s group presented a paper suggesting that a significant fraction of adults hospitalized with schizophrenia diagnoses might actually have autism:

Evidence of autism in a psychiatrically hospitalized sample

Their IMFAR presentation (and later published paper) showed an increased number of GI problems in their adult group. 36% of their adults had GI problems vs. 23% of the general psychiatric hospital population.

Unfortunately, these latest studies are getting the usual “online-autism-parents” community welcome. It follows the same pattern as vaccine/autism research:

a) Ask for studies to be done
b) Studies are done
c) Disagree with the data
d) try to slime the authors

Is it a surprise to anyone that some researchers have opted out of working on autism?

(note: minor edits were made shortly after publishing this article)

Advertisements

16 Responses to “Autism and Gastrointestinal symptoms: two new studies”

  1. RAJ July 29, 2009 at 13:00 #

    The same problem exists in twin and genetic studies. An hypothesis is proposed only to be refuted by subsequent reports. Bailey et al. (1995) published a twin study that is the most referenced article published in the history of autism. The article offered the following conclusions:

    http://www.ncbi.nlm.nih.gov/pubmed/7792363?

    Autism is a genetic disorder. The timing of onset is early in pregnancy based on their sample which reported minor congenital anomolies present in their sample, therefore the consistent reports of unfavorable events in the pre, peri and neonatal period were the consequence of genetically influenced abnormal development and were not related to etiology. Congenital anomolies are consistently found in genetic mental retardation syndromes (Fragile X, Downs Syndrome) as well as early (first trimester) environmental insults (thalidomide, rubella, Valproate Acid Syndrome)all of which are associated with autism.

    Subsequent reports did not support this conjecture and have reported major and minor congenital anomolies are present in a small subgroup as well as in healthy controls:

    http://www.ncbi.nlm.nih.gov/pubmed/16700944

    http://www.umfcv.ro/anomalii-congenitale-minore-in-autism

    http://www3.interscience.wiley.com/journal/120029261/abstract

    These studies show that congenital anomolies are present in a small subgroup (Less than 15% in these studies), not the majority of cases as Baily et al conjectured, but are also present (less than 15%) in healthy controls. These studies do not support the conjecture published by Baily et al (1995) yet this study has driven the research priorities in autism for decades, especially in genetic research.

    Remember the hype of the study a year ago that found inherited and denovo copy number variations in autism:

    http://www.sciencemag.org/cgi/content/abstract/1138659v1

    A more recent study found that Copy number variations in autism is associated with mental retardation and congenital anomolies:

    http://www.pubmedcentral.nih.gov/articlerender.fcgi?tool=pubmed&pubmedid=16840569

    All of the sample with copy number variations (deletions and additions) were mentally retarded and all had congenital anomolies.

    http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2563185&rendertype=table&id=T3

    http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2563185&rendertype=figure&id=F2

    Copy number variations are associated with mental retardation and congenital anomolies demonstrating the problem the research community has with differentiating mental retardation from an unambigous ASD diagnosis. Through parent support groups I have met perhaps several dozen autistic children and only observed a single case with congenital anomolies (A boy diagnosed with Fragile X) all othe other children were not only normal looking but many were very attractive toddlers.

    Another overhyped genetic paper associated ‘Autism’ with 16P11.2 copy number variations (deletions):

    http://www.ncbi.nlm.nih.gov/pubmed/18156158

    Other studies do not support the Boston Children’s Hospital study:

    http://www.ncbi.nlm.nih.gov/pubmed/19306953?

    Here again the association is with mental retardation.

    Autism is ill-defined, ambigous and highly subjective.

    Hakonarnon recently reported replicating a finding in a few families present in the AGRE data base with copy number variations in the region of chromosome 15Q11-13 and the mutations in the UBE3A gene associated with ‘Autism’.

    That region is associated with Angelman’s Syndrome which feature severe mental retardation, ataxia and seizures. Many autism researchers have suggested that Angelman’s Syndrome is a genetic autism sydrome.

    However the Angelman’s Syndrome Foundation on its home page states that Angelman’s Syndrome is often misdiagnosed as cerebral palsy or autism:

    http://www.angelman.org/stay-informed/

  2. RobinHausmanMorris July 29, 2009 at 14:07 #

    Hi Kev,
    I agree, initially I did not even want to write about the “diet” controversy. Perhaps the word controversy implies debate, which in fact remains dispelled by science.
    However, I did think it was important to share Dr. Margaret Bauman’s thoughts about “listening” to physical symptoms of our children.
    http://www.examiner.com/examiner/x-3565-Autism–Parenting-Examiner~y2009m7d28-Autism-101-Diet-therapy-for-stomach-problems

  3. Tom July 30, 2009 at 17:37 #

    Either RAJ doesn’t understand genetics or he purposely misrepresents findings. From the damn abstract he claims invalidates autism micro deletions at 16p11.2:

    “The assembled evidence indicates that recurrent 16p11.2 deletions are associated with variable clinical outcome, most likely arising from haploinsufficiency of one or more genes. The phenotypical spectrum ranges from MR and/or MCA, autism, learning and speech problems, to a normal phenotype.”

    RAJ, your slip is showing.

  4. Sullivan July 30, 2009 at 18:17 #

    Tom,

    RAJ’s comment has no relationship to the blog post so I didn’t see the need to discuss it. Not that I caught the glaring mistake he made, and not that I’m not glad you pointed it out.

  5. Tom July 30, 2009 at 18:43 #

    Hi Sullivan,

    Yeah, it was a non-sequitar. I really think that posting on LBRB is a privilege not a right. Repeated attempts to misrepresent deserve banishment. Maybe this would keep certain contributors on the straight and narrow.

  6. Prometheus July 31, 2009 at 00:26 #

    I initially thought that the comment (#3 above) by RAJ was simply “pasted” into the wrong post. Then I read the first two sentences:

    The same problem exists in twin and genetic studies. An hypothesis is proposed only to be refuted by subsequent reports.

    This appears to be a “linker”, allowing RAJ to post an otherwise completely unrelated comment about genetics in a post about gut problems and autism.

    Normally I would have let this non sequitur go unremarked. Unfortunately, RAJ’s understanding of genetics appears a bit thin in this comment and needs to be answered.

    For example:

    [1] While the Baley et al (1995) study may be the most cited twin study of autism, that doesn’t mean that it is the only or even the best study showing heritability of autism. Taniai et al (2008) showed a very high level of heritability in autism (more for females than males, oddly), as did Ronald et al (2006) and Hallmayer et al (2002). I could go on, but it would be merely beating the already dead horse. Suffice it to say that autism has been shown to be highly heritable.

    [2] The Wier et al (2006), Stoicanescu (2007) [note: published in a Romanian journal] and Ohdo et al (2007) studies do not actually contradict Bailey et al (1995). In fact, their findings disagree somewhat with each other. I’m not sure what point RAJ was driving at, but it appears to have been something about physical anomalies, which are slightly more common in autistic children than is the norm (but – according to Ohdo et al, less common than in Trisomy 21).

    [3] The Jacquemont et al (2006) citation used a 1 Mbp-resolution micro-array for their study. This means that they would have been able to detect only relatively large deletions and duplications. Since the “average” (there is a huge variation) gene size is about 10 to 15 Kbp, a 1 Mbp deletion/duplication could involve upwards of 100 genes. There is little wonder that the subjects they found were more dramatically affected.

    [4] The “overhyped” Kumar et al (2008) article on 16p11.2 in a number of subjects with autism was not contradicted by the subsequent Bijlsma et al (2009) paper, which found that the 16p11.2 deletion was not invariably associated with autism, as the authors stated even in their abstract.

    I’m not sure what RAJ is thinking with this set of citations. Is it an attempt to show that autism isn’t genetic? Who knows? Certainly not me.

    What is clear is that RAJ is fixated on twin studies at the moment and doesn’t appear to like the ones that show autism to be highly heritable. While this is all very interesting, I don’t see what that has to do with the (lack of) connection between autism and gut problems, which was the topic of this post.

    I apologize for using up space to rebut RAJ’s comment, but I couldn’t let it stand unanwered, even if it was not germane to the topic at hand.

    Prometheus

    • Sullivan July 31, 2009 at 00:39 #

      Prometheus,

      thanks for taking the time (and space) to add your comment.

  7. RAJ July 31, 2009 at 12:44 #

    “Tom,

    RAJ’s comment has no relationship to the blog post so I didn’t see the need to discuss it. Not that I caught the glaring mistake he made, and not that I’m not glad you pointed it out”.

    It as everything to do with your post. You cherrypick two studies you agree with as disproving all other studies with different results eg:

    http://www.ncbi.nlm.nih.gov/pubmed/10547242?

    The Bailey et al 1995 twin study does the same thing. The study does a lot of conjecturing, dismissing the consistent evidence over many decades of an increase in unfavorable events in the pre peri and neonatal period as being unrelated to autism etiology because they are the consequence of earlier genetically influenced abnormal development, and that the presence of minor congenital anomolies in their sample proves their conjecture. Subsequent studies have shown that the presence of minor and major congenital anomolies in autism is low, between 10 and 14% and minor and major congenital anomolies are also present in 6 to 14% of healthy controls in the three studies I referenced.

    Tom wrote:
    “Either RAJ doesn’t understand genetics or he purposely misrepresents findings. From the damn abstract he claims invalidates autism micro deletions at 16p11.2:

    “The assembled evidence indicates that recurrent 16p11.2 deletions are associated with variable clinical outcome, most likely arising from haploinsufficiency of one or more genes. The phenotypical spectrum ranges from MR and/or MCA, autism, learning and speech problems, to a normal phenotype.”

    RAJ, your slip is showing”

    Tom, that study showed no association between autism and 16P11.2 microdeletions, they were referring to earlier studies that made such a claim.

    See Hakonarnon’s most recent publication based on AGRE data:

    http://www.plosgenetics.org/article/info:doi/10.1371/journal.pgen.1000536

    Hakonarnon examining genetic data present in the AGRE multiple incident families reported a number of rare genetic variants contained in the AGRE dataset. The problem is that the same rare genetic variants he claimed were autism susecptbility genes have also been reported in studies that found the same genetic variants in mental retardation without autism.

    Since the introduction of DSM-IV and ICD-10 in 1994 and all the ‘Gold Standard’ diagnostic tools based on the DSM-IV and ICD-10 diagnostic criteria the definition of autism is so ill-defined that researchers involved in autism see autism where researchers who specialize in specific mental retardation syndromes do not see autism. This is analogous to Justice Stewart Potter’s opinion in a Supreme Court case of pornography. Potter wrote “I can’t define pornography, but I know it when I see it”.

    Nowhere is this most represented by the presence of Angelman Syndrome cases in the AGRE dataset. Hakonarnon claims the AGRE data set replicated the finding of genetic variants in the region of 15q11–13 and significantly, mutations in the single gene UBE3A.

    These are the same genetic variants reported in Angelman’s Syndrome. Researchers specializing in Angelman’s Syndrome note that a diagnosis is based on mutations in these genetic regions. The features of Angelman’s Syndrome are severe mental retardation, seizures and ataxia. Angelman’s Syndrome has commonly been suggested as an Autism Genetic Syndrome by those specializing in autism research, yet those who specialize in Angelman’s Syndrome have noted that Angelman’s Synrome is often misdiagnosed as cerebral palsy or autism.

    http://www.angelman.org/stay-informed/

    Autism is so ill defined that healthy persons with acquired neurological damage have been found to have acquired autistic personality impairments based on Baron-Cohen’s autism-spectrum quotient test

    To date, there is no genetic mutation that has been found to be specific to autism. The associations with ill-defined autism are based on interpretations always based on the bias of the beliefs of the interpretor.

    Tom, or Sullivan, name one specific autism gene, just one.

    • Sullivan July 31, 2009 at 14:44 #

      RAJ, next time we do a post on genetcs, feel free to cut and paste this comment there.

      ETA: Tom, Prometheus, and anyone else: you are welcome to continue the discussion should you like.

  8. Tom July 31, 2009 at 19:41 #

    RAJ,

    You can lead a horse to water…

    http://gene.sfari.org/

  9. Joseph July 31, 2009 at 21:33 #

    It as everything to do with your post. You cherrypick two studies you agree with as disproving all other studies with different results eg:

    http://www.ncbi.nlm.nih.gov/pubmed/10547242?

    @RAJ: Seriously, you don’t know what in the heck you’re talking about. How in the world does the non-controlled study above contradict anything in the post?

    Besides, I’m sure Sullivan understands that the entire body of science on this issue is what matters, not one study here or there. If you look at the science as a whole, there’s no convincing evidence thus far that autism is associated with GI disease.

  10. MJ August 4, 2009 at 01:08 #

    “The Bristol group’s study came out last week.”

    A minor nitpick, this study came out much earlier this year. I wrote about it in April although I believe it was published in March.

    This is really a weak study and there are a number of problems with it:

    The results were based on a parental report that was done at wide spaced intervals (4 weeks, 6 months, 18 months, 30 months, 42 month). I would have to question how much a parent would remember about incidents of diarrhoea, consistency of stools over a year period unless they were keeping detailed notes along the way (unlikely). And of course, tracking of other problems like abdominal pain is tricky at best when the children can’t talk (no data is included about the verbal status of the children).

    The number of children with autism (76) is much, much smaller than the control population (12,805). For this to be relevant to autism I would think there would have to more children with autism included.

    The data included in this study is from children born in 1991 and 1992, so it very dated. I think using a more recent group of children would have made the results much more useful and relevant.

    There were other problems as well.

    When you consider all of the problems this study does not tell us anything significant.

Trackbacks/Pingbacks

  1. Twitted by Izziland - July 29, 2009

    […] This post was Twitted by Izziland […]

  2. Twitted by justthevax - July 29, 2009

    […] This post was Twitted by justthevax […]

  3. The Autism Gut Question - People of NI - Norn Iron - July 29, 2009

    […] studies have been published that cast doubt on the validity of the autism gut connection. These are ably discussed on LB/RB.The first study compared the stool patterns of autistic children and non-autistic children and […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: