Last week, we published the first of three installments about a paper claiming a urine panel could detect whether or not a child is autistic. The first post addressed some of the issues related to the foundation (or what there was of a foundation) for the work described in the published study. This week, I’m digging into the many study-related questions I’d have wanted answered had I been a reviewer on this work.

First and foremost: The authors claim that their study is predicated on the idea that autistic people have different gut bacteria compared to non-autistic people, and that these different microbes produce different chemicals that then get discarded in the urine. The differences, you see, in turn identify who’s autistic and who isn’t.

When someone mentions gut and autism to me, the very first thing I think of – from very long experience – is stress and anxiety. The gut is an anxiety and stress meter like no other for the autistic people I know, seizing up and freaking out at the slightest change in routine, time zone, or social context. So, you might think that a study to address what autistic people’s guts do might do a little discerning of the factors that affect what guts do. Gut microbes are definitely among those factors, as are treatments for the gut distress autistic people experience.

Yet this study does not mention consideration of gastrointestinal symptoms or conditions in either the autistic or the non-autistic children they compared to each other. But if we can assume, as the authors state, that 40% of autistic children have some sort of GI issues, that means about 20 or so of the 52 autistic children included in this study may have had such issues. Indeed, for the key metabolites the study assessed in urine, the average rate of autistic people with high levels of these chemicals was 42% (how high were those levels? No idea. More on that later). If the study really did accurately detect something to some modest extent, was it really detecting the effects of anxiety and treatment on the gut and its microbial residents?

Obviously, anxiety and gut issues affect what we eat, so diet will play a role, too. The study does not suggest any consideration of diet, and most especially, diet around the time the study samples were taken. That matters, too, because as far as I can tell from the study, a single urine sample from each participant was analyzed. It’s a snapshot in time of what was in the urine, a bodily output that changes constantly with what we put in our bodies.

The authors claim that “8–9 out of 10 young children with [autism] in this cohort had high concentrations of one or more microbially-related metabolites (78%).” What we don’t know is what the ranges were or what constitutes “high” versus “normal range” vs “low” metabolites, as only this “high” rate gets reported in the main paper. In the supplemental material, we learn that the 50th percentile values for many of the measure metabolites were quite similar, and the minimum and maximum values are not given, only the interquartile ranges (25th and 75th percentiles). In other words, we have no idea what the values were for these groups of children beyond the barest bones tossed to the reader in this paper.

Among these bones are “area under the curve” or AUC values. These are supposed to give a look at the cumulative effect of, in this case, a combination of urine values in identifying autistic children or not. The closer they get to 1, the more meaningful the measure is. The “maximum” AUCs in the study largely fall below 0.8, in the range of “fair to good” accuracy. The authors omit confidence intervals from their reported and repeatedly highlighted AUC values, which is super weird, as these intervals usually are considered important for understanding the real import of AUC results. A 95% confidence interval gives the range in which a true value exists with 95% confidence, given the method applied. Providing only point values for AUC – and only the maximum values, at that – omits the range, leaving us with no confidence whatsoever.

Then there’s the question of sex: The authors report that the autistic group of children had 41 boys and 11 girls, compared to the “typically developing” group of 47 children, which had 20 boys and 27 girls. They don’t produce any analysis of whether this difference is significant, so I did a lil one for them. My chi-square analysis indicates that yes, this is a very significant difference between these groups (two-tailed P value, just for strictness: 0.0002). And that matters because boys and girls can diverge in terms of the microbiome, especially with the approach of puberty. The children included in this study were as old as 10 and probably as old as 12, based on the data provided (which does not define the values around the average given, but which I infer are standard deviation). There was no analysis by sex in this study, either, which could have unveiled whether there was an effect of these sex differences.

There also is a question of exactly how many children even made it into the final sensitivity and specificity analyses, as 7 and 10 for the autistic and typically developing groups, respectively, seem to be missing from this analysis, for reasons that are not clear. That is a good chunk!

There are other confounders that go unaddressed: what was the history of antibiotic use? The study recruited children from four centers (more on that next time), but does not show an analysis by location/geography. And again, the study does not show an analysis based on sex.

The very last sentence of the limitations of this study (which does not bring up any of these potential and obvious issues) offers the first and only notice that this is a “pilot” study. Yet the authors then go on (and on, and on) proposing a “new phenotype” that really is trademarked, and yes, they are definitely trying to make MDMs “microbially-derived metabolites” a thing:

Based on our findings that a large subset of children with ASD in our multi-site cohort have elevated concentrations of one or more MDMs, we propose a new ASD phenotype, ‘ASD associated with Microbially-Derived Metabolites’ (ASD-MDM), which is defined by quantitative laboratory measurements of MDMs in urine samples in young children. The development of the MDM System(TM) as a diagnostic screening tool for ASD is promising. Further validation of the MDM System(TM) in another cohort is underway by our research team, but the 40-plus studies of elevated concentrations of individual MDMs in ASD provide strong literature support for the MDM System(TM).

I’d ask them about these data that have raised questions for me, but here’s what they say about data sharing:

The data supporting this research are not publicly available due to patent-pending restrictions 96001970.d. Data may be made available upon reasonable request to the corresponding author, subject to a signed NDA.

Next time: Speaking of patents, how about conflicts of interest, baybee?

News you can use

• An “influencer” husband and wife went online to tell the world that the wife had undergone an abortion after they learned that their fetus bore an extra chromosome 21 and would be born with Down syndrome (NYT gift link). Although the timing is unclear, the abortion likely took place in the second trimester of the pregnancy, as it reportedly occurred after amniocentesis results confirmed the condition.

  In news coverage and their social media responses to the outrage that ensued from many quarters, including anti-abortion groups and self-advocates and allies in the Down syndrome community, the couple seems to have overstated statistics related to medical complications associated with the syndrome and other factors. The connection being made in both their words and actions is that Down syndrome is such a dire prospect that the only reasonable, thoughtful thing to do is to not bring someone into the world who bears that extra 21st chromosome in some or all of their cells.

  This argument is very similar in nature to the “monster,” “stolen child,” “vaccine-damaged child,” and other language that pervades especially the antivaccine crowd who wants to prevent autistic people. Indeed, autism and Down syndrome often co-exist, with about 39% of people with Down syndrome also being autistic, and share these stigmatizing, dangerous characterizations.

  Creating the specter of a ruined life or a child so unbearably damaged that preventing their birth is viewed as practically a kindness is one of the darkest aspects of prenatal testing. The list of “undesirable outcomes” to test for began with conditions like Down syndrome and other conditions continued to be added, with people riding on RFK The Lesser’s bandwagon talking openly about preventing them all, autistic people included.

  I am not here to finger-wag at anyone about the decisions they make about their reproductive lives and their families. But I will say that the language hammering on burdens, disease, monstrousness, unbearbable costs, and unworthiness of life is not even a shade different from the language of Nazis and other committed eugenicists in arguing for the extermination and prevention of disabled people. It is extreme language designed to scare people into extreme measures. Deciding to terminate a pregnancy based on manufactured threat narratives instead of the realities and humanity, the giving and receiving hearts and lives of people with Down syndrome or any other developmental condition, is to fall in line with those who spin these dark, twisted lies. Their tactics will have worked, bringing them one step closer to what they perceive as the ultimate desideratum: the complete erasure of all disabled or otherwise ‘atypical’ people from existence.

  Decisions people make about abortion should be evidence-based and accurately informed, not the result of terror tactics from anyone, on any side, whether it’s alleged psychological damage or cancer risk from having an abortion or the alleged shitshow that having a child with Down syndrome or other developmental condition is claimed to be. There are many, many false narratives around abortion that lead to decisions people might later regret and wish they had made differently. Meanwhile, people with Down syndrome, autistic people, and lots of other people with disabilities can hear, read, and understand what those narratives say, and know that there are people out there who not only want them gone – they want them never to have existed in the first place. - From the TPGA editors: Do not ask how much lower RFK the Lesser can go and you won’t be disappointed. He is -

  pursuing federal government access to most Americans’ medical records, in a quest to research a link between vaccines and autism — a connection the medical establishment studied for decades and flatly rejects.

• Autistic and/or ADHD patients with eating disorders (EDs) may experience heightened psychological distress and functioning difficulties, especially if they have AuDHD. In a new study, ADHD was particularly linked to more intense ED behaviors. This authors say that “screening for neurodivergence in ED services may support person-centered care and improve outcomes.”

• ASAN’s Colin Killick has a letter published (gift link) in the NY Times regarding the outlet’s recent article on how the ABA industry prioritizes profits over children:

  Reducing spending on Medicaid disability services almost always leads to substantial harm. In this case, however, covering alternatives instead of spending billions on A.B.A. would actually benefit autistic people.

• “It’s never really been a great system, so I think it’s hard to capture how bad it is now, given that it’s always been difficult.” The Disability Rights Education and Defense Fund found that the Social Security Administration policy and process changes implemented by the second Trump administration created significant barriers for those seeking to apply for and maintain disability benefits, as well as the attorneys and benefit specialists who represent them. Here’s their detailed report.
• TPGA is trying out Threads, tentatively. Check us out, give us a follow!

New at TPGA

Autistic But Misdiagnosed Until Adulthood: Talking With Author Laura Richmond — THINKING PERSON'S GUIDE TO AUTISM

In which we talk with comedy writer Laura Richmond about her adult autism diagnosis, and also the parts of her life that were—& are not—funny.

People you should know

• Darrell Britt-Gibson, who played the character Bishop on Euphoria (which I admit I’ve never watched, but I’ve … heard and read things), says that in his view, Bishop was autistic, so Britt-Gibson sought to play him that way.[ALERT: Spoilers at the link!]. Britt-Gibson says that Bishop was “like an onion — every time you see him, there’s another layer being peeled back.” For playing the autistic contributions to that onion, Britt-Gibson says:

  I don’t think Black people get to play that enough. I don’t like the way that people who are on the spectrum are spoken about. There are a lot of people in my life and my family who are on the spectrum, and they are the most beautiful, deep, soulful people. I thought, ‘If I get to do this and play him like that, I want to be able to do it justice and make it thoughtful and layered.’ That was something I inherently thought about. I wanted to give it that care.

Thanks for reading, and here’s to the people in our lives giving it that care.

Got something autism-related to share with us? Send it along to editorial@thinkingautism.com.

Got a comment? We’d love to hear from you, so drop us a line below. Please note that comments are moderated per TPGA guidelines.