Regressive vs. non-regressive autism: limited chemical differences noted

The paper published by Antonio Gomez-Fernandez and colleagues [1] examining whether or not there may be some potentially important biological differences as a function of reported regression vs. no regression in autism provides the blogging fodder today. Not for the first time has the immune system and 'regressive autism' been mentioned in the peer-reviewed science literature (see here and see here), but the current work focuses on the examination of various immune system and other related compounds: in a seemingly well-defined cohort: "Analyses of plasma molecules, such as cathepsin, IL1β, IL6, IL8, MPO, RANTES, MCP, BDNF, PAI NCAM, sICAM, sVCAM and NGF."

"Fifty-four children (45 males and nine females) aged 2-6, who were diagnosed with ASD [autism spectrum disorder], and a control group of 54 typically-developing children of similar ages were selected." Authors relied on quite an extensive battery of assessments looking at behaviour, alongside their use of the DSM-5 diagnostic criteria for autism (see here). Also accompanying physical examination "with a special emphasis on neurological and nutritional status", authors garnered blood samples from participants (overnight fasting) for their immune system and related functioning evaluations.

"The group of ASD children was further divided into two subgroups based on the presence or absence of neurodevelopmental regression during the first two years of life, which was assessed using a five-item questionnaire following the guidelines used by the Autism Diagnostic Interview-Revised (ADI-R) for the evaluation of this process." The ADI-R has been previously discussed on this blog in relation to regression in autism (see here). And just in case you might not be totally convinced that regression can be part of a pathway to autism, here's some more evidence for you (see here)...

Results: "there were 20 children included in the AMR [neurodevelopmental regression] subgroup and 32 in the ANMR [without neurodevelopmental regression] subgroup; two children could not be classified in these subgroups because they were adoptees, allocated by a national adoption agency." Bearing in mind that we cannot rule out any recruitment bias that might have leaned towards including those with regressive autism on the Gomez-Fernandez study, the figure of approaching 40% of their cohort showing such a regressive profile is notable. I'd also draw your attention to the finding that the behavioural profile for the regressive group (AMR) was also significantly different from the non-regressive group (ANMR) insofar as perhaps painting a picture of greater [group] autism severity...

Interestingly, the study did not show too many immune system and other compound differences between those diagnosed with autism and the asymptomatic (for autism) control group. So: "No differences were found between the two groups in terms of the cytokine and adhesion molecule levels studied, except for NGF [nerve growth factor], in which the group of ASD children was found to have twice the plasma levels compared to the control group." NGF is no stranger to autism research, and other studies have come to a similar conclusion [2].

When it came to examining results based on comparing the regressive (AMR) and non-regressive (ANMR) groupings, things got slightly more interesting but again no complicated pattern of difference was noted. So, for the ANMR (non regression) grouping: "lower plasma levels of the NCAM adhesion molecule were detected compared to the levels in the AMR subgroup and the control group. This ANMR group also exhibited higher NGF levels than the typically-developing children, which could indicate an alteration in neuronal development." Again, adhesion molecules have been mentioned in other autism research (see here).

"In conclusion, the results of this study show that there is not a typical profile for the expression of relevant plasma cytokines, adhesion molecules or growth factors in children with ASD compared with that in typically-developing children." OK, there are caveats to the phrasing used by the authors; not least that the participant numbers were quite small in the Gomez-Fernandez study and the idea that within the very heterogeneous autism spectrum, there may be smaller groupings (phenotypes) that perhaps show a tendency to greater immune system and related 'issues' (see here). But there are also some strengths attached to the Gomez-Fernandez study; not least the study "benefits from a careful selection of children of similar ages, as well as the complete diagnosis of ASD with multiple tests, clinical follow-up and associated complementary tests."

Questions still remain. Perhaps an important one is the question around why some children show a regressive pattern of behaviour as part of their path to a diagnosis of autism? Yes, issues such as infection do seem to be part-and-parcel of the clinical profile for some (see here and see here for examples) and perhaps more detailed focus is required in such areas. But much like a group showing the opposite of regressive autism - those who seemed to 'grow out' of autism - currently thought to include as many as one in ten (see here), a wider range of biological as well as psychometric measures are required to help pick out potentially important mechanisms pertinent to the idea that autism is not necessarily 'hard-wired' for all...

----------

[1] Gomez-Fernandez A. et al. Children With Autism Spectrum Disorder With Regression Exhibit a Different Profile in Plasma Cytokines and Adhesion Molecules Compared to Children Without Such Regression. Front. Pediatr. 2018. September 26.

[2] Dinçel N. et al. Serum nerve growth factor levels in autistic children in Turkish population: a preliminary study. Indian J Med Res. 2013 Dec;138(6):900-3.

----------

Autistic traits in adult schizophrenia

"Results of this study indicate the existence, in a sample of patients with a diagnosis of schizophrenia, of a distinct group of subjects with ASD [autism spectrum disorder] features, characterized by specific symptomatological and cognitive profile."

So said the findings reported by Stefano Barlati and colleagues [1] continuing a research theme from this group [2] looking at the potential overlap between autism spectrum disorder and schizophrenia.

Reiterating my interest in how the autism and schizophrenia spectrums can and do collide (see here) both at a condition and trait level, the Barlati findings provide some pretty in-depth analysis of what autism *might* look like in the context of schizophrenia. They report evaluation of their cohort - "Seventy-five schizophrenia patients (20 females, mean age 42 ± 12)" - with two of the gold-standard autism assessment instruments: the Autism Diagnostic Observation Schedule (ADOS) and the Autism Diagnostic Interview-Revised (ADI-R) alongside other "clinical, neuropsychological, and psychosocial functioning measures."

It's important to say that, in these days of pluralisation of behavioural and/or psychiatric labels (see here and see here), quite a few participants (47/75) assessed as part of the Barlati study turned up "negative to all the autism scales administered." This tells us that it's not necessarily a straight-forward nor universal relationship when it comes to autism and schizophrenia (and vice-versa). More likely is the possibility that there either may be subgroups within the diagnosis of schizophrenia that present with significant autistic traits or possibly even that the timing or severity or grading of schizophrenia and its symptoms may predispose to autistic traits being more or less likely to be presented. That last point relies on the idea that various traits or characteristics of labels like schizophrenia and autism might not be as immutable as many people believe...

For however the participants diagnosed with schizophrenia who turned up clinically significant autistic traits in one or other or total domains/scores using the ADOS and ADI, further research is indicated. Further research on what this phenotype might look like longitudinally, how frequent it might manifest, and whether there may be unique challenges associated with it. It also might have some implications for intervention too (see here).

And yet again, such findings provide more fodder for the idea that autistic traits are not exclusively just part and parcel of a diagnosis of autism (see here), and the pressing need for formal, professional assessment when autism is suspected...

Oh, and then there's more...

----------

[1] Barlati S. et al. Autistic traits in a sample of adult patients with schizophrenia: prevalence and correlates. Psychol Med. 2018 Mar 20:1-9.

[2] Barlati S. et al. Autism Spectrum Disorder and Schizophrenia: Do They Overlap? International Journal of Emergency Mental Health and Human Resilience. 2016; 18: 760-763.

----------

Low grade intestinal inflammation and autism

The suggestion that low grade intestinal inflammation might be related to some autism comes from the findings reported by Katarina Babinská and colleagues [1] (open-access available here).

Researchers set out to "assess the concentrations of fecal calprotectin in a sample of children with ASD [autism spectrum disorder] and to investigate the correlations of this inflammatory marker with the core behavioral symptoms of ASD."

Faecal calprotectin (FC) is a measure of the amount of calprotectin in a stool (poo) sample. It's typically released in response to the presence of inflammation and, here in Blighty at least, is indicated as "an option to support clinicians with the differential diagnosis of inflammatory bowel disease (IBD) or irritable bowel syndrome (IBS) in adults with recent onset lower gastrointestinal symptoms for whom specialist assessment is being considered." 

In terms of research history looking at autism and FC, there is some peer-reviewed science on the topic; also having been included as a parameter in the important paper by Laura de Magistris and colleagues [2] talking about 'leaky gut' in the context of some autism (see here) and how "FC was elevated in 24.4% of patients with autism and in 11.6% of their relatives." Such research is set in the more general context that bowel or gastrointestinal (GI) issues are absolutely no stranger to a diagnosis of autism (see here).

This time around Babinská et al measured FC (via ELISA) in some 87 children diagnosed with an autism spectrum disorder (ASD) aged between 2 and 17 years of age. The authors use the term 'low functioning' to describe this portion of their participant group but I'm rather less enamoured with such labels (see here) despite the well-deserved focus on a group very much under-represented in autism research and other areas. Alongside, over 50 age-matched controls (not-autism) and 29 siblings of children with ASD also provided samples for analysis and comparisons.

Results were not exactly as cut-and-dried as one might have expected. So: "In non-relatives significantly lower values of fecal calprotectin were observed than in both subjects with ASD and their siblings." What this means is that based on group results, those with autism and the siblings of those with autism seemed to manifest higher levels of FC than non-related controls. Based on individual results, where elevated levels of fecal calprotectin was set at 50 µg/g of feces or higher according to test producers guidance as being a level of concern, the frequency of such a finding was greater in those with autism (22%) and their siblings (20%) than in non-related controls (9%) but this difference was reported as 'non-significant'.

Authors also did a little work on another important area in relation to bowel symptoms/pathology and autism: how *might* something like intestinal inflammation 'interact' with the behavioural signs and symptoms of autism? Well, we are told that those diagnosed with autism "had to meet criteria for ASD" on two gold-standard diagnostic tools: the Autism Diagnostic Observation Schedule – second edition and the Autism Diagnostic Interview-Revised (ADI-R). Data from the ADI was examined in the context of the FC findings and lo and behold: "In the group with ASD significant correlations of fecal calprotectin with all domains of the ADI-R diagnostic tool were found: qualitative abnormalities in reciprocal social interaction and communication, restrictive and repetitive patterns of behavior." I say this bearing in mind that similar analyses between FC values and ADOS ratings do not seem to have been either done or reported on for some reason.

When the authors talk about low grade intestinal inflammation as potentially being relevant to some autism, they seem to be accurate insofar as the measured levels of FC in some participants and the *correlation* with autism scores on one of the gold-standard assessment instruments. That being said, there is quite a bit more to do in this area before anyone gets too carried away with the results as they stand. So for example, all that chatter about inflammatory bowel disease (IBD) being related to some autism (see here and see here) did not seem to register in this particular study insofar as the guidance on FC being a marker for possible IBD, albeit based on higher levels of FC being detected: "Active, symptomatic inflammatory bowel disease 200 – 40,000 mg/kg."

I also note that the authors report an important limitation when it came to their research: "Additional factors that might have been a cause of elevated FC levels, such as nutritional or gastrointestinal factors were not analysed." Nutritional factors eh? Y'mean like milk type for example [3] or other dietary and/or environmental factors such as the implementation of a gluten-free diet [4] positively affecting FC levels? Indeed, there are lots of potential factors that could cause a 'false-positive' when it comes to elevated FC such as infections like C. diff or gastrointestinal conditions such as coeliac disease, many of which have shown some important connections to autism (see here for example).

It looks like there is still much more research to do in this area but investigations should definitely continue.

To close, my brood have just discovered the brilliant film 'The Great Escape'. As well as setting up many, many discussions about war, bravery and captivity, they've also commented on the theme tune...

----------

[1] Babinská K. et al. Fecal calprotectin levels correlate with main domains of the autism diagnostic interview-revised (ADI-R) in a sample of individuals with autism spectrum disorders from Slovakia. Physiol Res. 2017 Dec 30;66(Supplementum 4):S517-S522.

[2] de Magistris L. et al. Alterations of the intestinal barrier in patients with autism spectrum disorders and in their first-degree relatives. J Pediatr Gastroenterol Nutr. 2010 Oct;51(4):418-24.

[3] Ho S. et al. Comparative effects of A1 versus A2 beta-casein on gastrointestinal measures: a blinded randomised cross-over pilot study. Eur J Clin Nutr. 2014 Sep;68(9):994-1000.

[4] Balamtekın N. et al. Fecal calprotectin concentration is increased in children with celiac disease: relation with histopathological findings. Turk J Gastroenterol. 2012;23(5):503-8.

----------

Subgroups in autism (without intellectual disability)

"Children with ASD [autism spectrum disorder] without ID [intellectual disability] could be differentiated into Moderate and Severe Social Impairment subgroups when core ASD symptoms were more closely examined."

So said the findings reported by Felicity Klopper and colleagues [1] looking at an important part of the autism research scene related to the 'plurality' of the term autism and the seemingly vast range of presentations included under the label. Reliant on data obtained from "the ‘gold standard’ ASD diagnostic instruments" (including the ADOS and ADI), researchers looked at the "presence of phenotypic subgroups" in their cohort.

As per the opening sentence to this post, there were some differences to be seen in the cohort, and in particular, how social interaction issues might be a key part of any differentiation. The authors talk about how social interaction issue differences seemed to tie into other core behavioural features such as communication and the presence of restricted/repetitive behaviours. They concluded: "both categorical and dimensional approaches may be useful in classifying ASD, with neither alone being adequate."

It is not necessarily new news that the label of autism is good for diagnosis but seemingly says little about the range of presentation included under the heading (see here for example). Indeed, in these days of ESSENCE I might forward the view that even the label autism might be part of a wider heterogeneous presentation (see here) and one should further expand those subgroup notions at the label as well as symptom level. The focus on overt behaviour (as assessed by those gold-standard instruments) in the Klopper study is but one part of looking at such 'heterogeneity' (see here for example) as the authors argue that: "The dissociated profiles of ASD features could represent different underlying neurobiological mechanisms for each subgroup." At least one of the authors on the Klopper paper probably, more than most, realises that fact (see here).

There are other key areas to this focus on the presentation of autism that also need to be factored in: sex differences and comorbidity profiles. Specifically, the growing realisation that girls and boys on the autism spectrum probably show subtle differences in presentation (see here) and, minus any sweeping generalisations, should be considered in future studies in this area. Oh, and keep in mind that those diagnosed with autism with an intellectual disability (ID) could also be 'sub-grouped' according to symptom presentation too with similar caveats. The question is: how many sub-groups of autism will we eventually end up with?

Music, and because Spring has Sprung... In Bloom.

----------

[1] Klopper F. et al. A cluster analysis exploration of autism spectrum disorder subgroups in children without intellectual disability. Research in Autism Spectrum Disorders. 2017; 36: 66-78.

----------

Felicity Klopper, Renee Testa, Christos Pantelis, & Efstratios Skafidas (2017). A cluster analysis exploration of autism spectrum disorder subgroups in children without intellectual disability Research in Autism Spectrum Disorders : 10.1016/j.rasd.2017.01.006

Machine learning applied to autism screening going big time?

Machine learning, when machines, er.. learn, is of growing interest to the autism research field. The names Wall and Duda have filled quite a few posts on this blog (see here and see here for example) on this topic and their suggesting that applying machine learning algorithms to something like autism screening and detection could cut down on time taken and resources used.

As per the publication of the paper by Daniel Bone and colleagues [1] it appears that others working in autism research are also waking up to the idea that this might be a useful area to investigate. So: "In this work, we fastidiously utilize ML [machine learning] to derive autism spectrum disorder (ASD) instrument algorithms in an attempt to improve upon widely used ASD screening and diagnostic tools." Fastidiously is such a lovely word (particularly in the context of science).

The tools in question were the Autism Diagnostic Interview-Revised (ADI-R) and Social Responsiveness Scale (SRS) (both of which have already been machine learning 'applied') and their scores "for 1,264 verbal individuals with ASD [autism spectrum disorder] and 462 verbal individuals with non-ASD developmental or psychiatric disorders, split at age 10." And the results... well, let's just say that the authors were not disappointed - or at least less disappointed than on previous research occasions [2] - as they reported on created algorithms that "were more effective (higher performing) than the current algorithms, were tunable (sensitivity and specificity can be differentially weighted), and were more efficient (achieving near-peak performance with five or fewer codes)." Indeed: "We present a screener algorithm for below (above) age 10 that reached 89.2% (86.7%) sensitivity and 59.0% (53.4%) specificity with only five behavioral codes." Sensitivity and specificity are important concepts when it comes to something like screening instruments in terms of identifying 'all' those with a specific condition and making sure that no 'not-cases' aren't mistakenly identified as 'cases'. The nearly 90% sensitivity rate presented by Bone et al on the basis of 5 behavioural codes is not to be sniffed at.

The addition of one Cathy Lord to the authorship of the Bone paper also adds an air of inevitability that applying machine learning to autism research (and practice) is going to continue and increase. Not only because of her historical connection to the ADI-R [3] (which is a hefty document in anyone's book) but also given her very prominent role in autism research history. Who knows, I might one day be blogging about more big autism research names talking about Wall/Duda things including autism screening triage by YouTube? The final question is: outside of just behavioural variables, who would be brave enough to talk genetics/epigenetics/biology machine learning as the next step in autism screening and/or assessment?

----------

[1] Bone D. et al. Use of machine learning to improve autism screening and diagnostic instruments: effectiveness, efficiency, and multi-instrument fusion. J Child Psychol Psychiatry. 2016 Apr 19.

[2] Bone D. et al. Applying machine learning to facilitate autism diagnostics: pitfalls and promises. J Autism Dev Disord. 2015 May;45(5):1121-36.

[3] Lord C. et al. Autism Diagnostic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord. 1994 Oct;24(5):659-85.

----------

Bone D, Bishop S, Black MP, Goodwin MS, Lord C, & Narayanan SS (2016). Use of machine learning to improve autism screening and diagnostic instruments: effectiveness, efficiency, and multi-instrument fusion. Journal of child psychology and psychiatry, and allied disciplines PMID: 27090613

Boiling down ADOS for autism detection (again)

Today I want to direct your attention to the paper by Kosmicki and colleagues [1] (open-access) reporting that the use of "machine learning algorithms" could help "streamline ASD [autism spectrum disorder] risk detection and screening."

Regular readers of this blog might have already cottoned on to the fact that any talk about applying "computational and statistical methods" to autism screening and/or diagnosis can really mean only one person and research group: Dennis Wall from Stanford University. To quote from his institutional website on this area of research, the aim is to "evaluate the degree of redundancy of the ADOS and ADIR and if so determine whether a reduced set of uncorrelated features could correctly classify individuals with the same accuracy as the gold-standard diagnostic tests." ADOS and ADI-R by the way, are some, if not the, gold-standard schedules when it comes to the assessment of autism. The idea is that boiling down these respective schedules might save both time and resources when it comes to identifying those where a diagnosis of ASD is indicated. In case you'd like some history about this line of work, look no further than here...

The latest paper from the Wall group continues the research journey looking this time at modules 2 and 3 of the ADOS where previous work looked at module 1 (see here). In case you're not familiar with the concept of modules in ADOS, it's all about selecting the correct module according to verbal fluency (see here) where module 1 is for those who have very little or inconsistent phrase speech and modules 2 and 3 represent increasing phrase speech with also a little more focus on the use of age-appropriate props.

The results? Based on the development of 'classifiers' for each module, several machine learning algorithms were developed and tested (see here). One of the algorithms, ADTree, is by the way, the same classifier used in the previous module 1 ADOS work [2]. But ADTree did not perform best on this occasion: "The logistic regression classifier based on analysis of archival records from ADOS module 2 consisted of nine items, 67.86% fewer than the complete ADOS module 2, and performed with 98.81% sensitivity and 89.39% specificity in independent testing." Further: "The SVM module 3 classifier based on analysis of archived ADOS module 3 records consisted of 12 items, 57.14% fewer than the complete ADOS module 3, and performed with more than 97% sensitivity and specificity in testing."

The authors conclude: "These results support the notion that fewer behaviors when measured using machine learning tools can achieve high levels of accuracy in autism risk prediction."

Anyone who has either professional or personal experience of undertaking an ADOS will know that this is a highly specialised assessment schedule which often requires some time to complete. It's nothing like as time-consuming as the ADI but still, significant efforts and resources are needed to carry out the assessment and do so with skill and reliability (and maintain those all-important reliability stats). Wall et al have really started to shake the establishment when it comes to ADOS (and ADI) when asking just how much of the schedule is really needed to assess for autism/ASD. This on top of their other work talking about assessment 'triage' via YouTube videos using, horror of horrors, non-clinical raters (see here). I'm not saying that these approaches are ready for clinical practice; quite a bit more replicative work is required [3] including crossing geographical boundaries. But something like the idea that "mobile health approaches that ultimately enable individuals to receive more expedient care than is possible under the current paradigms" is a tantalising prospect.

So: Start by The Jam.

----------

[1] Kosmicki JA. et al. Searching for a minimal set of behaviors for autism detection through feature selection-based machine learning. Transl Psychiatry. 2015 Feb 24;5:e514.

[2] Wall DP. et al. Use of machine learning to shorten observation-based screening and diagnosis of autism. Transl Psychiatry. 2012 Apr 10;2:e100.

[3] Bone D. et al. Applying Machine Learning to Facilitate Autism Diagnostics: Pitfalls and Promises. J Autism Dev Disord. 2014 Oct 8.

----------

Kosmicki JA, Sochat V, Duda M, & Wall DP (2015). Searching for a minimal set of behaviors for autism detection through feature selection-based machine learning. Translational psychiatry, 5 PMID: 25710120

Efficacy of foetal stem cell transplantation in autism...

The recent news that researchers might be one step closer to 'curing' type 1 diabetes following the publication of the paper by Pagliuca and colleagues [1] brought back into focus how stem cell therapy might hold some promise for all manner of conditions. The idea that researchers could generate "hundreds of millions of glucose-responsive β cells from hPSC [human pluripotent stem cells] in vitro" still faces a few challenges, including overcoming the immune assault central to the autoimmune condition that is type 1 diabetes. I have but one comment to make about the immune system and autoimmunity in this context: worm pills (see here)...

The question of whether an advance has been similarly made following the publication of the paper by Jeff Bradstreet and colleagues [2] (open-access available here) is perhaps open to some discussion with their observations that: "Statistically significant differences (p<0.05) were shown on ATEC/ABC scores for the domains of speech, sociability, sensory and overall health, as well as reductions in the total scores when compared to pre-treatment values" based on the use of foetal stem cells (FSCs) "in treating children diagnosed with ASDs [autism spectrum disorders]". Further details about the study can also be found in the latter slides of the presentation shown here.

Stem cell therapy in the context of autism is still a scientific hot potato. I've covered previous, very preliminary, forays into this research area before on this blog (see here). It is with the same cautions and caveats that I discuss the latest paper from Bradstreet et al.

So:

• This was a study of some 45 children diagnosed with an autism spectrum disorder (ASD) (mean age = 6-7 years). Diagnosis was confirmed by some of the gold-standard assessment instruments including ADOS and ADI. There were quite a few exclusion criteria applied to study entrants such that those with epilepsy, or "a neurological or co-morbid psychiatric disorder" were not examined. Learning disability without autism was also "considered exclusion criteria" as was a diagnosis of Asperger syndrome.
• The study was based in Kiev in the Ukraine where "stem cells harvested from 5-9 weeks old human fetuses following voluntarily – elective pregnancy terminations (legally available in the Ukraine)" were used. I don't doubt that there may be some who have strong views about this practice as per commentary from other authors (see here). Hematopoietic stem cells (HSCs) after harvesting were tested for various bacterial, fungal and viral infections as were the women who previously carried.
• Long quote coming up... "Stem cell transplantation of suspensions containing cryopreserved fetal stem cells were preceded by pre-medication of the subject via intravenous slow infusion of diphenylhydramine (Darnitsa, Ukraine) 10 mg and prednisone (Darnitsa, Ukraine) 15 mg on Day 1 and diphenhydramine (Darnitsa, Ukraine) 10 mg on Day 2". At this point, I'll draw your attention to some other work previously discussed on this blog on a possible role for corticosteroid therapy for some types of autism (see here) which included the use of prednisolone, the active metabolite of prednisone. After which the stem cells were administered...
• Results: "Early post-transplantation effects were reported in 78% of children: 26% of these children became calmer; eye contact was improved in 9%, while 29% had better appetite and 23% had an improved affect". Importantly, the authors report that no adverse effects were initially noted and "No transmittable diseases were noted during the 12 month follow-up". They also make mention of how initial effects may well have been [partly] as a consequence of the corticosteroid and other medication initially administered.
• Scores on the ATEC and ABC bore out the positive group changes noted between baseline (before stem cell therapy) and at 6 and 12 month follow-up which were also accompanied by various immunological changes "indicative of improved cell-mediated immunity in children".

OK. Despite these results the authors themselves are still cautious about their findings and stress: "future research studies are urgently needed and larger randomized -placebo controlled trials are needed to further characterize potential FSC-associated improvements in ASDs". This was a straight forward observational trial (before and after) which lacked control groups and in particular a placebo-controlled element so one has to be slightly hesitant about the strength of any findings. For those however who might be pulling on this study because of the use of something like the ATEC to measure autistic presentation, I'll draw your attention to some work suggesting that this instrument might be rather useful for monitoring intervention options for autism (see here).

As previously described, feelings run deep about the use or not of stem cells when it comes to autism not least because of the lack of data on long-term safety (and efficacy) in this context, the source 'material' for stem cells and the lack of information on just what might be going on in biological terms consequent to the behavioural results described. Examining this research from a cold, dispassionate, scientific point of view, I have to say that I'm becoming rather interested in what might be potentially going on during this and other studies [3] if not just as a function of other work by the late Paul Patterson and colleagues overlapping with this area [4] (discussed in a previous post). 

That being said, I'd like to see a lot more research done in this area before this kind of intervention enters anything like mainstream autism practice...

----------

[1] Pagliuca FW. et al. Generation of Functional Human Pancreatic β Cells In Vitro. Cell. 2014 Oct 9;159(2):428-439.

[2] Bradstreet JJ. et al. Efficacy of fetal stem cell transplantation in autism spectrum disorders: an open-labeled pilot study. Cell Transplant. 2014 Oct 9.

[3] Lv YT. et al. Transplantation of human cord blood mononuclear cells and umbilical cord-derived mesenchymal stem cells in autism. J Transl Med. 2013 Aug 27;11:196.

[4] Hsiao EY. et al. Modeling an autism risk factor in mice leads to permanent immune dysregulation. Proc Natl Acad Sci U S A. 2012 Jul 31;109(31):12776-81.

----------

Bradstreet JJ, Sych N, Antonucci N, Klunnik M, Ivankova O, Matyashchuk I, Demchuk M, & Siniscalco D (2014). Efficacy of fetal stem cell transplantation in autism spectrum disorders: an open-labeled pilot study. Cell transplantation PMID: 25302490

An observation-based classifier for rapid detection of autism risk

"Keep clear of the moors"

Among the many researchers and research groups admired on this blog for their contribution to the world of autism research, the name Dennis Wall is fast becoming a real favourite. Aside from mention of the words 'systems biology' in his profile at Stanford University, I'm particularly interested in the way the Wall research group are looking at trying to apply machine-learning approaches to things like autism assessment.

I've covered a few of their past research reports with regards to instruments like the Autism Diagnostic Interview (ADI) and the Autism Diagnostic Observation Schedule (ADOS) previously (see here and see here respectively). More recently was the work suggesting that YouTube videos and non-expert raters might be a useful resource for autism triage (see here). That last report certainly set the cat among [some] pigeons...

Today I'm talking about another paper from the Wall laboratory by Marlena Duda and colleagues [1] (open-access) and the suggestion that: "reductions in the process of detecting and monitoring autism are possible". The ADOS was once again the focus of the study following on from their previous 'preliminary' foray [2].

The paper is open-access but here are a few choice details:

• If I'm reading the paper correctly, this was a follow-up study to the previous Wall paper [2] testing the accuracy of the "observation-based classifier (OBC)" which I think was previously called/included the ADTree algorithm. This time around "a cohort of archival score sheets of over 2600 subjects, including more than 280 assessments of non-spectrum controls" were included in the study derived from ADOS and ADOS-2 algorithms. ADOS-2 by the way, represents the revised algorithms used to score ADOS reported by Gotham and colleagues [3]. I've talked about the Gotham paper before on this blog and how it seemed to 'predict' diagnosis of autism in DSM-5 (see here).
• The aim was to test whether a boiled down version of the ADOS / ADOS-2, the OBC, that: "presently contains eight behaviors... that are often impacted in children with autism, including eye contact, imaginative play and reciprocal communication" might be able to distinguish autism from not-autism and "shorten screening and diagnostic processes overall and potentially enabling more families to receive care far earlier and during timeframes when interventions have the most positive benefits".
• Results: "The OBC was significantly correlated with the ADOS-G (r=−0.814) and ADOS-2 (r=−0.779) and exhibited >97% sensitivity and >77% specificity in comparison to both ADOS algorithm scores". These figures aren't bad at all, if a little down on the previous Wall data [2]. The authors add: "Less than 5% of all tested cases were misclassified by the OBC and 78% of the misclassified individuals were given a low OBC score".

Obviously there is much more investigation needed in this area of autism research before one might start shortening ADOS assessments (or indeed doing away with trained ADOS raters altogether). The issue of comorbidity is, for example, something that needs to be included in any further study and whether that might interfere with any results obtained [4]. I might also add that ADOS is only part of the diagnostic assessment for autism and does not replace reasoned clinical opinion.

I am however drawn to the authors suggestion that: "use of the OBC as a web-based assessment in advance of a clinical visit may enable clinicians to quickly prioritize patients according to symptom severity, scheduling shorter, more immediate diagnostic appointments for individuals that can be clearly identified as on or off the autism spectrum, and allowing longer time periods for deeper evaluation of children that exhibit clinically challenging symptoms". Certainly with the numbers of children/adults seemingly coming through the various referral systems, this kind of triage might yet hold some usefulness. And it seems other groups are getting in on the computer-assisted act [5]...

Music to close and The Marcels with Blue Moon. "Keep clear of the moors" as we were once told...

----------

[1] Duda A. et al. Testing the accuracy of an observation-based classifier for rapid detection of autism risk. Translational Psychiatry. 2014; 4: e424.

[2] Wall DP. et al. Use of machine learning to shorten observation-based screening and diagnosis of autism. Transl Psychiatry. 2012 Apr 10;2:e100.

[3] Gotham K. et al. The Autism Diagnostic Observation Schedule: revised algorithms for improved diagnostic validity. J Autism Dev Disord. 2007 Apr;37(4):613-27.

[4] Leyfer OT. et al. Overlap between autism and specific language impairment: comparison of Autism Diagnostic Interview and Autism Diagnostic Observation Schedule scores. Autism Res. 2008 Oct;1(5):284-96.

[5] Hashemi J. et al. Computer Vision Tools for Low-Cost and Noninvasive Measurement of Autism-Related Behaviors in Infants. Autism Research and Treatment. 2014. 935686.

----------

M Duda, J A Kosmicki, & D P Wall (2014). Testing the accuracy of an observation-based classifier for rapid detection of autism risk Translational Psychiatry, 4 : 10.1038/tp.2014.65

The developmental regression rate and autism

I have already made mention of the paper by Robin Goin-Kochel and colleagues [1] in a previous post on the topic of developmental regression and autism (see here). On that occasion it was to substantiate that approximately 40% of children diagnosed as being on the autism spectrum were reported to have shown some kind of developmental regression as part and parcel of their presentation. Given however the assertion in the Goin-Kochel paper about using "the largest, most comprehensively phenotyped sample to date" reflecting their participant numbers (N=2105) derived from the Simons Simpex Collection (SSC), I thought it worthwhile to include a separate blog entry for this important paper.

I'm pretty sure that the topic of developmental regression and autism needs no further introduction. Suffice to say that as the name suggests, it's all about a loss of previously acquired skills and how this might map on to some of those 'autisms' that I seem to talk about quite a lot. After a bit of a laboured start (see here) there is now a general acceptance that regression can occur in cases of autism and sometimes pretty rapidly (see here) although I hasten to add, not seemingly present for everyone with autism.

I have to thank Natasa for passing me a full-text copy of the Goin-Kochel paper, and without further ado, here are a few of the main points:

• The SSC was the source for participants, which meant that diagnoses of autism or autism spectrum disorder (ASD) were about as reliable as one could hope for. Interestingly, the authors do make mention of the whole DSM-IV to DSM-5 transition and what that might mean for databases such as the SSC but that's perhaps fodder for another day.
• "Regression was operationalized using the skill-loss items from both the ADI-R and a supplemental interview that captured information about additional and more subtle skill losses". I've talked before about how ADI-R was a really important contributor to raising the profile of regression and autism as a function of it including items about 'loss of language/other skills'. In this case, researchers defined various types of 'loss' to include codings such as 'full losses' and 'subthreshold losses'. They also analysed data according to "skill loss at/before 36 months and those who experienced skill loss after 36 months".
• Results: "Overall, 36.9% of children had some type of regression". Such regression included a regression in language and/or other skills and included those with full losses and those with subthreshold losses. Another quote: "When combined, 585 (27.8%) children experienced some degree of language loss and 568 (27.0%) experienced some degree of other loss".
• The authors focused quite a lot on the pre-36 months loss side of things as a function of quite a lot of discussion on age of first parental concern being rooted in that critical period. So: "those with any degree of language loss (full or subthreshold) at/before 36 months scored significantly lower than those with no language loss" when looking at how regression might affect "cognitive and adaptive-functioning outcomes". 
• Finally, a couple of other important details to mention: (a) "subthreshold losses of other skills occurred later than any type of language regression and later than full losses of other skills", and (b) duration of loss of skills is also discussed: "most children regained their skills by 3.5-5 years of age". They conclude that their study: "lends support to the argument that children whose parents report regression represent a distinct ASD subtype that may be associated with lower cognitive functioning".

The participant group size and the use of the SSC are important features which make these results something to shout about. Alongside the meta-analysis by Barger and colleagues [2] autism research is beginning to build up quite a good picture on the rate of regression being reported in the autism research literature and some details on presentation and even outcome.

That being said, there is still an awful lot more to do in this area. The Goin-Kochel did not set out to look at "etiological mechanisms" but the next stage in this research should surely be further study which looks at possible correlates to such regressive accounts. I wouldn't necessarily agree with the sentiments of the authors when they say: "An acknowledged limitation of this study is its reliance on parent-report data" especially when it comes to parent report and other issues such as gastrointestinal (GI) problems (see here) and the conclusions reached by Phillip Gorrindo and colleagues [3] (open-access here). Recall may well be affected by things like telescoping effects (see here) but in these days of home movies [4] and the rise of social media, many infants see the end of camera lens very, very early following their entry into the big wide world. Where regression occurs, I'm sure many parents would have their own ideas about the hows and whys which could be investigated including some slightly more biological / genetic based investigations [5].

I've two more points to make on this study and research area and that's all. First is an issue which the authors touch upon in their discussion of their results: "... the SSC is largely comprised of Caucasian families". The recent abstract from Adiaha Spinks-Franklin and colleagues (which, at the time of writing, I don't yet think is published in a peer-reviewed form) hints at some racial differences in the occurrence of regression in autism. The question is whether this might also translate into a different pattern of regressive symptoms and indeed, any different effects on longer term outcome?

Second is that issue of long-term outcome. As we're starting to see with the optimal outcome work (see here), longitudinal studies looking at outcome from all those developmental trajectories which make up the autisms is starting to yield some interesting results. The next question would be whether similar studies looking at regressive vs. non-regressive presentation (even different types of regression [6]) would similarly provide important data on the extent to which regression might affect future development and presentation and ways these could be positively affected. I'd also add that looking at any effects from regression on comorbidity outside of the core dyad of symptoms might also be included in any future work knowing what we (think we) know about this important issue.

To close, Katy Perry and Last Friday Night (a song my brood are really enjoying at the moment).

----------

[1] Goin-Kochel RP. et al. Developmental regression among children with autism spectrum disorder: Onset, duration, and effects on functional outcomes. Res Autism Spectr Disord. 2014; 8: 890-898.

[2] Barger BD. et al. Prevalence and onset of regression within autism spectrum disorders: a meta-analytic review. J Autism Dev Disord. 2013 Apr;43(4):817-28.

[3] Gorrindo P. et al. Gastrointestinal dysfunction in autism: parental report, clinical evaluation, and associated factors. Autism Res. 2012 Apr;5(2):101-8.

[4] Palomo R. et al. Autism and family home movies: a comprehensive review. J Dev Behav Pediatr. 2006 Apr;27(2 Suppl):S59-68.

[5] Shoffner J. et al. Fever plus mitochondrial disease could be risk factors for autistic regression. J Child Neurol. 2010 Apr;25(4):429-34.

[6] Ozonoff S. et al. Parental report of the early development of children with regressive autism: the delays-plus-regression phenotype. Autism. 2005 Dec;9(5):461-86.

----------

Goin-Kochel, R., Esler, A., Kanne, S., & Hus, V. (2014). Developmental regression among children with autism spectrum disorder: Onset, duration, and effects on functional outcomes Research in Autism Spectrum Disorders, 8 (7), 890-898 DOI: 10.1016/j.rasd.2014.04.002

sPECAM pie and school-aged autism

I've talked about adhesion molecules with autism in mind before on this blog (see here). In that entry it was some interesting data out of the MIND Institute which caught my attention; specifically the selectins and their sticky siblings being 'generally' suggested to be lower in case of autism than control samples. Without repeating my previous post, it's all about the binding of leukocytes to the walls of blood vessels to begin their rolling journey towards the site of an injury and then inflammation, yadda, yadda...

Rolling stone & moss @ Wikipedia  

Anyhow, a new addition joins the voices suggesting issues with adhesion in cases of autism in the form of the paper by Yosuke Kameno and colleagues* (open-access paper available here).

The Kameno paper fills a bit of a gap in the literature in this area by looking at levels of platelet-endothelial adhesion molecule-1 (PECAM-1), platelet selectin (P-selectin), endothelial selectin (E-selectin), intracellular adhesion molecule-1 (ICAM-1), and vascular cell adhesion molecule-1 (VCAM-1) in serum samples from school-aged children (5-17 years old) diagnosed with autism compared with asymptomatic controls. The reasoning being that very young infants and young adults have been examined with these adhesion molecules in mind but not the intervening age group.

The results: well probably unsurprisingly, levels of at least some of the adhesion molecules were lower in cases of autism compared with the control group. So: "The serum levels of sPECAM-1 in subjects with high-functioning ASD were significantly lower than those of controls (U = 91.0, P<0.0001) (Table 1). Subjects with high-functioning ASD also had significantly decreased levels of sVCAM-1 compared with those in controls (U= 168.0, P = 0.0042)". The U by the way refers to the statistical test used (Mann-Whitney U test) to analyse results. That and the fact that attempts to correlate the biological findings with things like scores on the Autism Diagnostic Interview-Revised (ADI-R) didn't reveal any significant correlations.

There are also a few hidden gems in this paper not readily discussed too much. So for example: "To exclude inflammatory disease, serum C-reactive protein (CRP) levels were determined". CRP is another interesting compound which I've talked about before with regards to inflammation and autism or risk of autism (see here and here). Kameno didn't seem to find anything specific in their autism cohort aside from: "The CRP measurement of one subject with ASD was 2.30 mg/dl (this individual did not have subjective symptoms or a history of inflammatory disease)".

They also looked at a number of cytokines in their participant group and concluded: "We determined that plasma concentrations of IL-1β, IL-1RA, IL-5, IL-8, IL-12(p70), IL-13, IL-17 and GRO-α were 
significantly higher in subjects with ASD compared with the corresponding values of the matched controls, after correcting for multiple comparisons". I'm particularly interested in their observations on IL-17 given some previous work in this area (see here) and its [proposed] link to various autoimmune conditions.

So Kameno and colleagues have filled the age group gap in the work looking at adhesion molecules with autism in mind. Given the increasing strength of the evidence coming out of this area of autism research, one could make a good argument for quite a bit more detailed investigation?

To finish, there are potentially lots of rolling linked songs I could offer. But instead I'll go for burning....

----------

* Kameno Y. et al. Serum levels of soluble platelet endothelial cell adhesion molecule-1 and vascular cell adhesion molecule-1 are decreased in subjects with autism spectrum disorder. Mol Autism. 2013 Jun 17;4(1):19.

----------

Kameno Y, Iwata K, Matsuzaki H, Miyachi T, Tsuchiya KJ, Matsumoto K, Iwata Y, Suzuki K, Nakamura K, Maekawa M, Tsujii M, Sugiyama T, & Mori N (2013). Serum levels of soluble platelet endothelial cell adhesion molecule-1 and vascular cell adhesion molecule-1 are decreased in subjects with autism spectrum disorder. Molecular autism, 4 (1) PMID: 23773279

Pets win (prosocial) prizes?

Miss Ellie Dog @Wikipedia

Pets have never really been a great part of my childhood it has to be said. Aside from a cat allergy which sort of ruled out any would-be Top Cat staying at ours, the family home just wasn't graced with enough space to accommodate other animal companions.

I did have a goldfish called George for a short period of time; that is until he/she(?) passed away and went to the great WC in the sky. Sad memories indeed.

This lack of animal contact during my childhood is probably why I am a little ambivalent towards children keeping pets at home (that and a very unfounded phobia of T.gondii) whilst, at the same time, being thankful for school pets who undoubtedly 'earn their keep' in the petting stakes.

Where autism is in mind however some recent research by Grandgeorge and colleagues* (full-text) suggests that pet arrival might very well have prosocial prizes.

The study is open-access so only a brief summary needed:

• From quite a large bank of participants (N=260), two studies were carried out on two very much smaller groups: study 1: arrival of a pet at age 5 years (n=12) vs. never owned a pet (n=12); study 2: owned a pet since birth (n=8) vs. never owned a pet (n=8).
• Alongside a questionnaire on human-pet relationships, parents of participants undertook questioning based on the ADI-R (see recent post) across two time periods (T0 and T1) primarily unaware of the reason for study participation.
• Results: based on study 1, pet arrival between the ages of 4-5 years was associated with significant changes to 2 algorithm items on the ADI-R (53) offering to share and (55) offers comfort which "reflect prosocial behaviors". Having a pet from birth (study 2) did not seem to bestow the same changes.

There are a few obvious caveats to these findings based on the sample size and sole reliance on ADI-R to assess change at the same time of pet arrival. As with all studies of association, people don't generally live in a vacuum outside of real life, so one has to be slightly cautious about linking just pet arrival to the reported changes in behaviour particularly over quite a long period of time.

All that being said, I am really quite interested in these findings. I know some have talked about the whole theory of mind (ToM) issue as accounting for the results (see here). Whilst this remains a possibility, I have to say that I still remain unconvinced of a major link between animal associated prosocial behaviours and perspective-taking or vice-versa in this particular instance. Such cold hard psychology fails to take into account the concept of 'enjoyment' in having a pet and also the responsibility that comes with ownership outside of trying to understand the mental state or what their pet might be thinking.

On the other hand, I do rather like the idea that stress, and importantly a reduction/moderation in stress and anxiety responses following the introduction of a pet might be part and parcel of the results seen as per suggestions like this one from Virués-Ortega & Buela-Casal**. If there's one thing we know about autism, it's that stress and anxiety are very often in the background.

This study also reminds me of an earlier blog post concerning some research on animal magic and the amygdala which suggested that the amygdala might the place to be when it comes to animal identification and recognition. Exactly what role the amygdala might play is still unclear in autism but one can speculate that neuronal functioning may have potentially been affected by pet companionship in a sort of 'pet brain training scenario'. Let's wait for more evidence of this first though.

Whilst perhaps being a more outlandish link, my post on the appeal behind Thomas the Tank Engine to some cases of autism might also be relevant. Think about it: cats, dogs, hamsters, guinea pigs, etc. are fairly uncomplicated creatures by human standards. They don't talk (aside from the odd 'sausages' here and there), they don't use a wide variety of facial or other gestures, and in most cases, they pretty much like, and are responsive to, interaction without other complications. Just a thought.

I would like to see more investigation on this topic with autism in mind; a call echoed by an even more recent review of the use of assistance and therapy dogs for autism***. I hold back from suggesting that every child with autism should be automatically handed a pet at aged 4 or 5 years given that not every child probably wants a pet, but for some it might be a useful aid to their development. Dare I even suggest that the introduction of a pet to the family home might also have some knock-on effects to immune functioning as per articles like this one by Tse and Horner**** (full-text) in light of autism and the immune system research or am I just being a little bit silly?

----------

* Grandgeorge M. et al. Does pet arrival trigger prosocial behaviors in individuals with autism? PLoS ONE. 2012; 7: e41739.

** Virués-Ortega J & Buela-Casal G. Psychophysiological effects of human-animal interaction: theoretical issues and long-term interaction effects. Journal of Nervous & Mental Disease. 2006; 194: 52-57.

*** Berry A. et al. Use of assistance and therapy dogs for children with autism spectrum disorders: a critical review of the current evidence. Journal of Alternative & Complementary Medicine. September 2012.

**** Tse H. & Horner AA. Allergen tolerance versus the allergic march: the hygiene hypothesis revisited. Current Allery & Asthma Reports. 2008; 8: 475-463.

----------

Grandgeorge M, Tordjman S, Lazartigues A, Lemonnier E, Deleau M, & Hausberger M (2012). Does pet arrival trigger prosocial behaviors in individuals with autism? PloS one, 7 (8) PMID: 22870246

ADTree reloaded: classifying autism based on 7 ADI-R items?

A lengthy quote to begin this post:

"Deploying a variety of machine learning algorithms, we found one, the Alternating Decision Tree (ADTree), to have high sensitivity and specificity in the classification of individuals with autism from controls. The ADTree classifier consisted of only 7 questions, 93% fewer than the full ADI-R, and performed with greater than 99% accuracy when applied to independent populations of individuals with autism, misclassifying only one out of the 1962 cases used for validation".

Interested? The quote comes from this recent paper by Dennis Wall and colleagues* (full-text) following up some related research from this group based on another gold standard autism assessment schedule, the ADOS (see here).

I'm not going to dwell too much on the study details aside from saying:

• The ADI-R (Autism Diagnostic Interview - Revised) is probably one of the more time and resource-intensive interview questionnaires used for the assessment of autism, coming in at 93 questions long and asking about current behaviour and 'most abnormal' (aged 4-5 years) if appropriate.
• An abbreviated ADI-R would therefore be quite a useful measure in these austere times; particularly one which could maintain the same level of accuracy as completion of the full schedule.
• Similar to their previous study, Wall and colleagues applied various machine learning algorithms (n=15) to ascertain whether any might be able to determine which ADI-R items are most relevant to diagnosis based on the AGRE dataset.
• Once again, the Alternating Decision Tree (ADTree) model performed best: perfect sensitivity (1.0), a low false-positive rate (0.013) and "overall accuracy of 99.9%".
• Seven items of the ADI-R comprised the ADTree model: (i) comprehension of simple language at most abnormal (4-5 years), (ii) reciprocal conversation (regarding the ability to facilitate the flow of conversation), (iii) use of imaginative / pretend play at most abnormal (I have post about pretend play coming up fairly soon), (iv) social imaginative play with peers at most abnormal, (v) direct gaze at most abnormal, (vi) group play with peers at most abnormal (spontaneous games or activities) and (vii) age when abnormality was first evident.
• Testing accuracy was again carried out on participant data from the Boston Autism Consortium (AC) and the Simons Simplex Collection (SSC) from where the quite compelling data for the success of the ADTree model was derived.

So the ADI-R has been boiled down to 7 pertinent items. The ADOS boiled down to 8 distinguishing module 1 items. I think most people would stand up and take note of these findings even if further replication is still required (based on different geographical groups for example). Don't get me wrong, there is still a large degree of skill required to deliver the ADI-R and ADOS and maintain your reproducibility and diagnostic prowess so I don't think this combined data will be putting people out of work just yet; certainly not with the number of people estimated to be coming through the diagnostic process. That and the fact that these are assessment instruments and so are subservient to a final clinical opinion for an autism diagnosis or not.

Aside from the grand findings I am interested in the types of behaviours which are noted to be important for diagnosis. All very 'social-communicative' (sounds familiar) and not at all heavy on the 'restricted and repetitive behaviour' side of things. Indeed, looking back at the ADOS ADTree paper, I might be wrong but only one element, 'functional play with objects' seems to have any strong relation to issues with repetitive behaviours. This could be that we aren't asking the right questions about this area of behaviour, but I would hedge my bets that more likely is the stress on the social-communicative side of presentation as being key to diagnosis. I think I might have to look at this further in future posts.

Finally, I have previously talked about the lack of instruments to appropriately and accurately assess 'change' in autism (as a function of maturation or intervention or anything else). Y'know with these combined data, I think we might have the outline of something really quite useful...

----------

* Wall DP. et al. Use of artificial intelligence to shorten the behavioral diagnosis of autism. PLoS ONE. 2012; 7: e43855.

----------

Dennis P. Wall, Rebecca Dally, Rhiannon Luyster, Jae-Yoon Jung, & Todd F. DeLuca (2012). Use of artificial intelligence to shorten the behavioral diagnosis of autism PLoS ONE : 10.1371/journal.pone.0043855