I had hoped that the press would have learned from their experience with Andrew Wakefield and the Great MMR Hoax. Perhaps now they would be more circumspect in reporting “breakthroughs” in the science of autism. But on the evidence of recent reports in the Mail , the Express and the Telegraph it seems my hopes were misplaced. All three papers carried reports on recently published research into Urinary Metabolic Phenotyping and Autism .
The Express story was the shortest and its headline was suitably cautious.
NEW TEST MAY SPOT AUTISM IN CHILDREN
It has all the essentials.
Autism diagnosis is a long process.
A simple test that could be administered before the signs emerge would enable therapy to start earlier.
This would improve outcomes.
New research has found a chemical fingerprint in the urine of autistic children that could form the basis of such a test in the future.
The Mail also maintains a note of caution in its headline
Revolutionary urine test for autism could soon diagnose children with the condition
Its story contains the same elements as the Express but adds that
The diagnostic process is not only lengthy but also traumatic for young children.
It is proven that autistic children have different microbes in their gut.
The test should be available in 5 years time.
Both papers seem to have taken their story direct from the press release issued by Imperial College London . Both make much of the lengthy diagnostic procedures and suggest that a simpler biomedical test could replace these procedures. In this they misunderstand the nature of autism and the problems we face in getting a diagnosis.
At present the problem is not so much the length of the assessment itself but the time spent waiting between appointments. First you need a referral, usually from a GP. When you finally get to see the diagnostic team there is not a uniform procedure. Then you wait for the report. Then, armed with your diagnosis the real work begins of identifying and accessing whatever services are available in your area. It would be better if everyone got to see a consultant paediatrician or child psychologist along with assessments for speech and language therapy, occupational therapy and a full medical work-up that led seamlessly from diagnosis to appropriate services tailored to the child’s individual needs.
If a urine test can be developed it will be more in the nature of a screening tool. It might conceivably expedite the initial referral but it will not replace the need for a detailed follow up.
The Telegraph is rather more forthcoming in its headline.
Autism test could make the condition ‘preventable’
This goes way beyond what was claimed in the press release and is based on this quote from one of the researchers, Professor Nicholson.
“Children with autism have very unusual gut microbes which we can test for before the full blown symptoms of the disease come through.
“If that is the case then it might become a preventable disease.”
Professor Nicholson is an eminent and well respected scientist in the field of biological chemistry. But nobody on the research team has a primary qualification or research interest in autism. Dr Yap is a member of the National Autistic Society. Dr Angley is a pharmacist who has turned to autism research and published papers on biomedical interventions for autism. But they appear to have been ill-advised on the present state of academic research into autism, particularly in regard to the gastro-intestinal tract. This may be down to the fact that they offer
thanks also to Dr. D. Granpeesheh (C.A.R.D, Los
Angeles, Ca.) for helpful discussions on the manuscript and
Doreen Granpeesheh has been a long time colleague and supporter of Andrew Wakefield and her C.A.R.D. organization continues to provide educational services to parents who attend Thoughtful House, the Texas clinic that Wakefield established after his departure from the UK. This may explain the uncritical acceptance of papers by Wakefield and others on the fringe of autism research that informs this paper’s position on a unique disturbance of microbes in the gut of autistic people. LBRB and Countering Age of Autism both discuss the questionable nature of these sources.
Even so, Professor Nicholson is going far beyond the study’s findings with these speculations. He is suggesting that the alleged abnormal microbial environment in the gut is what causes autism. The journalist, Telegraph Science Correspondent, Richard Alleyne extrapolates from this to suggest that
Eventually the link between the learning difficulties and the gut microbes could be established and that could lead to “probiotic” treatments or cures.
It is not always clear to this reader whether Alleyne is expressing his views or paraphrasing Professor Nicholson. Either way, the entire piece displays intense ignorance about autism.
We are told that early intervention can prevent permanent psychological damage. Diagnosis presently occurs after the damage has been done. Early intervention is delayed because it is currently difficult to establish a firm diagnosis until children begin speaking.
But even if we had a reliable urine test at 6 months of age what sort of intervention is envisaged at that age? Glenn Doman’s quackery?
I do not understand why Imperial College issued a press release about this. It is a pilot study. The results are interesting but will not be significant until they are replicated on a larger scale. The study itself is very honest about its limitations and the need for more work. If there is to be a simple test for autism it is still years away.
It is easy to criticize the media for misleading coverage of scientific affairs.But in this case the media seem to have been misled by ICL. The press release begins thus:
Children with autism have a different chemical fingerprint in their urine than non-autistic children, according to new research published tomorrow in the print edition of the Journal of Proteome Research.
No matter how much they qualify this opening statement, that is the headline for any news editor. But is it true? To answer that we have to look at the study itself.
I have read the paper in question. It is not an easy read and I do not pretend to understand all of it, especially the technical discussion of methodology and the statistical analysis of results. But, having read a lot of autism research in recent years, a number of questions occur to me.
Normally studies like this go into some detail about how the subjects were recruited. This is because so many details can confound a study. Are subjects matched for age, gender and socio-economic status? As we are looking at biomedical markers in urine at what time of day was it collected? Were all subjects controlled for diet, supplements and medication? We are not told. Bizarrely, part of the control group was recruited in Switzerland. all the rest were Australian. Why? We are not told.
The “unique fingerprint” claim suggests to a lay audience (and the claim was made in a press release aimed at a lay audience) that they tested the urine and the results showed three distinct results - autistics, siblings of autistics and controls. However, the way they tested showed so many variations between individuals that autism specific variations were not that obvious. They had to run the results through a sophisticated statistical analysis to identify any significant differences. Different methods of analysis revealed different results.
Visually, allowing for the interindividual variability, the urinary spectra were very similar, but the autistic individuals showed subtle differences in urinary succinate, N-methyl nicotinic acid (NMNA) and N-methyl nicotinamide (NMND) compared to the controls, as evidenced from the median spectra shown in Figure 1. Multivariate Statistical Analysis of the NMR Spectral
To further explore the metabolic differences between the three groups of participants, multivariate statistical analyses were employed on the full resolution NMR data set consisting of 34 controls, 28 siblings and 39 autistic urine samples to extract useful metabolic information. PCA was carried out on UV-scaled data to identify any inherent differences within the data set. The resulting scores plot of PC1 versus PC2 (Figure 2A) showed no clear differences between the three groups, all pairwise combinations of PCs down to PC3 were examined, which showed no discrimination indicating that the major source of variation in the data was not related to autism, but was rather dominated by interperson variability.
However by utilizing group information in PLS-DA analysis, systematic differences could be observed between the three groups (Q2 ) 15%; R2 (goodness of fit) ) 65.7%). The corresponding crossvalidated PLS-DA scores plot (Figure 2B) showed clear separation between autistic individuals and the controls and partial separation between siblings and the controls.
I take that to mean that when looking at the data for each individual there was not a single feature, a chemical fingerprint, that identified all the autistic individuals. But when they looked at each chemical in turn and checked its prevalence in each of the three groups they did find a statistically significant difference between groups. They may have identified potential fingerprints but they have no reliable way, as yet, of using them to identify autistic subjects.
There is a lot of discussion of how the results may be indicative of metabolic dysregulation that fits with certain hypotheses about possible metabolic pathways in autism, always with the caveat that further studies are required. I feel that this discussion is premature. All the weaknesses above - the over hyping of results and the attempts to present a preliminary study as supporting a hypothesis about the gut and autism that is much disputed and lacking in hard evidence - all this finds its basis in similar overreaching assumptions in the paper itself. Something is going on here in addition to the science that threatens to subtract from any merit the study might have.
And the study is not without merit. It would indeed be a step forward if we could establish biological markers for autism. But any follow up will have to address the waeakneses identified in this excellent commentary from NHS Choices
The research has several limitations:
The researchers point out that, as it is not possible to tell whether these differences indicate a cause or consequence of the disease, further research is needed in a larger group of children over time.
Different statistical analyses had different results, some showing differences in certain chemical levels in autistic children, while others did not.
The researchers did not assess the medications the children with autism were taking for their condition or the diet they were following. Both would affect the chemicals they found in the children’s urine samples.
Finally, these children had already been diagnosed with autism, and the study design was cross-sectional, looking at their urine samples from only one point in time. It is not possible to say whether there would be any differences in the chemicals found in the urine in younger children prior to standard diagnosis, and whether it could be used as a diagnostic tool.
This is encouraging research, but it is too early to say whether this research would be of benefit in terms of providing an additional diagnostic tool for autism in children.