"to [try and] advance the understanding of potential mechanisms underlying insomnia in adults with ASD"

The paper by Vanessa Hohn and colleagues [1] (open-access available here) provides the blogging fodder today and their findings in relation to the presence of insomnia as a "common source of distress in adults with autism spectrum disorder (ASD)."

Insomnia is essentially a regular problem in getting to sleep. It has a myriad of 'causes' ranging from uncomfortable sleeping conditions (bed, noise, temperature) to the influence of things like stress, depression and anxiety. Insomnia is no joke. Outside of the potential wide-ranging effects on physical and mental health [2] it puts a person at some heightened risk of all-manner of adverse outcomes [3].

As part of a suite of sleeping-related problems, insomnia is also no stranger to diagnostic labels like autism. With autism in mind, sleep problems that begin in childhood are pretty likely to carry over to adulthood. Nobody knows the precise reason(s) why those on the autism spectrum are more likely to experience sleep problems including insomnia. Some have speculated that the presence of autistic traits may be a risk factor for such sleeping issues (see here) but there is currently little detail regarding the bridge from core autism features to sleep problems. Others have observed an *association* between the presence of other somatic issues occurring alongside autism and sleep problems (see here and see here) but again, further clarity is needed on the hows-and-whys of any relationship and, importantly, what can be done to alleviate such issues.

Hohn et al started with the premise that: "Two characteristics of ASD could be relevant to insomnia complaints by hampering the entrainment of a circadian sleep-wake rhythm." They specifically talk about examining the link between sensory responsiveness and social skills in the context of insomnia in adults with ASD.

"Data were obtained from the Netherlands Autism Register (NAR), which is a longitudinal register including approximately 2000 individuals with ASD." Some 630 participants with autism were included for study, where "individual responses given to three measures of interest during an online survey in 2015 were analyzed." Those 'measures' were the Insomnia Severity Index (ISI), the Sensory Perception Quotient and the Autism Spectrum Quotient-28. The combined data were analysed.

Results: bearing in mind that this was a study exclusively including self-report from participants with autism - with no non-autistic control group - and one that relied on a single temporal snapshot of insomnia symptoms, one has to be a little careful with the findings. So: "The mean ISI score in the present sample was 9.50 (SD = 6.01), which is indicative of subthreshold insomnia and higher than means reported for the general population ranging between 1 and 7... but lower than values obtained from insomnia patients ranging between 17 and 20." Further, about half of participants scored in the "absence of insomnia" category with only around 20% of the group recording either moderate or severe insomnia. This tells us that, again according to self-report (and not using actigraphy for example), insomnia was not necessarily a widespread issue in this cohort but present in a not-significant number: about 1 in 5 people.

Next: "Primary analyses revealed that each of the covariates had a significant impact on the ISI total score." What this means is that biological sex, gender and medication use potentially played a role in insomnia. Women participating in the study tended to score higher on the ISI than men, and those aged 45-65 also seemingly showed more of a tendency towards insomnia. I don't think anyone would be really surprised with the idea that age and insomnia might be connected [4].

Some further statistical analysis of their results led researchers to opine on: "positive associations of insomnia severity with general and visual sensory hyper-reactivity and with impairment of social skills." I'm not too au fait with the specifics of the statistical modelling technique used by the authors - hierarchical multiple linear regression analyses (HMLR) - but can see what they did and how they did it, including "controlling for confounding effects of covariates" such as "medication, biological sex, intelligence, and age" (as previously mentioned). That being said, I'll draw your attention to some other conclusions reached by the authors: "The rather small effect sizes reported in the present study suggest that other factors not assessed in this study contribute to the emergence and persistence of sleep problems in adults with ASD."

The Hohn study does add something to the literature on sleep difficulties being experienced by quite a few people (children and adults) on the autism spectrum. I'm not overly convinced that they've stumbled upon evidence that core autistic features such as social skills problems and sensory issues are necessarily central to insomnia in their cohort but am willing to concede that they might play a role. Personally, I'd be inclined to go back to that data on 'comorbid' issues like gastrointestinal (GI) problems (see here) or breathing issues like apnoea (see here) as being something to explore further; also with the assumption that such issues are potentially 'treatable' and could have an important knock-on effect for something like insomnia and other sleep issues with autism in mind...

Oh, and don't forget the molecular handyperson that is melatonin and it's potential role in all of this (see here and see here). That and other potential avenues of research/clinical interest (see here)...

----------

[1] Hohn VD. et al. Insomnia Severity in Adults with Autism Spectrum Disorder is Associated with sensory Hyper-Reactivity and Social Skill Impairment. J Autism Dev Disord. 2019 Feb 9.

[2] Fernandez-Mendoza J. & Vgontzas AN. Insomnia and its impact on physical and mental health. Curr Psychiatry Rep. 2013;15(12):418.

[3] Garbarino S. et al. Insomnia is associated with road accidents. Further evidence from a study on truck drivers. PLoS One. 2017;12(10):e0187256.

[4] Foley DJ. et al. Sleep complaints among elderly persons: an epidemiologic study of three communities. Sleep. 1995 Jul;18(6):425-32.

----------

Evidence of blurring on the autism spectrum edges

A quote to begin: "multiple types of genetic risk for ASDs [autism spectrum disorders] influence a continuum of behavioral and developmental traits, the severe tail of which can result in diagnosis with an ASD or other neuropsychiatric disorder."

That was the bottom line of the findings reported by Elise Robinson and colleagues [1] looking at data from "several large ASD consortium and population-based resources" (N~38,000) including the fabulous resource that is ALSPAC. Specifically analysing for various types of genetic 'risk' related to autism including inherited genetic issues and those so-called de novo variants (see here), researchers found that the genes thought to be influential in cases of autism also crossed over in to those who don't have autism in relation to skills such as social communication. To zoom in on some specific media interpretation of the findings: "The same genes involved in predisposing people to autism appear to influence social skills in the wider population, suggesting that the autism spectrum has no clear cut-off point, scientists have discovered."

Personally, I'm not so surprised that these findings have turned out the way they have. As I was told many, many years ago when it comes to the presentation of [early] autism, the [core] behaviour(s) associated with autism in many respects, mirror that seen during various stages of child development. It is the intensity, duration and effect on daily life that merit the diagnosis. If one assumes that genes play a significant role in the presentation of certain facets of autism (or even certain types of autism [2]), the Robinson findings fit rather well into the whole autism as a "continuum model." I say this bearing in mind that common genetic variations have been previously discussed with autism in mind (see here) although seemingly no one-size-fits-all genetic pattern related to all autism has yet been reported (see here).

Questions and caveats still remain however. First is the continued focus on structural genetics over and above other influences on gene expression such as that suggested by the science of epigenetics. I know there is a lot of hype in this area but there is also some good science emerging on aspects like the methylome and autism (see here) and some important links between the epigenome and autism and immune function for example (see here). I might also direct you to some interesting research suggesting that those fossil viruses that we all carry around in our genome might also exert some effect (see here). In short, variations in the structure of the genome are probably not the whole story when it comes to the presentation of autism or any other label (or anything sub-threshold-wise, around the diagnostic edges).

Second, and added to the first point, is the idea that 'genes for autism' are probably not exclusive 'genes for autism' insofar as their potential links to other behaviours/labels. We know from other research (see here) that genes perform lots of functions either working as individuals or more likely working as groups. Combined with the idea that some of the behaviours/symptoms noted in cases of autism might have other 'associations' (see here) and the idea of pleiotropy comes to the forefront [3]. This also combines with the idea that autism rarely exists in some sort of diagnostic vacuum (see here). I might add that I'll be coming to the paper by Wen and colleagues soon enough...

Finally, although it is tempting to use the Robinson findings as 'proof' that blurred genetic edges might mean 'we're all a little bit autistic', I'm happy to see that some commentators have countered such an assertion. As I hinted at earlier, the behaviours noted in autism are not somehow totally unconnected to the human experience insofar as all can be seen at some time or other during typical development. But... and it is an important point, autism is a diagnosis based on how said behaviours go well beyond that 'typical' categorisation in terms of their intensity, duration and importantly, effect on a person's life. Much like the pop quizzes asking 'are you autistic?' to say 'we're all a little autistic' potentially belittles what autism means to many, many people day-after-day-after day. And with recent reports on where autism in the extreme can sometimes lead (see here), caution is very much advised in this context.

----------

[1] Robinson EB. et al. Genetic risk for autism spectrum disorders and neuropsychiatric variation in the general population. Nat Genet. 2016 Mar 21.

[2] Robinson EB. et al. Autism spectrum disorder severity reflects the average contribution of de novo and familial influences. Proc Natl Acad Sci U S A. 2014 Oct 21;111(42):15161-5.

[3] Hagenaars SP. et al. Shared genetic aetiology between cognitive functions and physical and mental health in UK Biobank (N=112 151) and 24 GWAS consortia. Mol Psychiatry. 2016 Jan 26.

----------

Robinson EB, St Pourcain B, Anttila V, Kosmicki JA, Bulik-Sullivan B, Grove J, Maller J, Samocha KE, Sanders SJ, Ripke S, Martin J, Hollegaard MV, Werge T, Hougaard DM, iPSYCH-SSI-Broad Autism Group, Neale BM, Evans DM, Skuse D, Mortensen PB, Børglum AD, Ronald A, Smith GD, & Daly MJ (2016). Genetic risk for autism spectrum disorders and neuropsychiatric variation in the general population. Nature genetics PMID: 26998691

Kids need other kids (for social communication skills)

A quick post today bringing the findings from Angela Barber and colleagues [1] to your attention and the idea that both young children diagnosed with autism and young children not diagnosed with autism might similarly benefit from peer mediated intervention (PMI). PMI basically means that peers are active agents in the instruction of one or more skills (see here for further information).

For the Barber study, 3 pairs of pre-school children (one diagnosed with an autism spectrum disorder and one not) were brought together. The child not diagnosed with autism was taught a specific PMI - "the Stay, Play, Talk PMI" - which as the name suggests involves being instructed to "Stay with their friend, Play with their friend, and Talk to their friend." For two 20 minute sessions over 6-8 weeks features of social communication were monitored.

Researchers reported that all children - those with autism and this without - seemed to benefit from this type of approach in terms of social responsiveness. That being said, social initiations were described as "variable across dyads" and when examined 2-months post-interventions were are told that gains "were not maintained."

You might have read that last paragraph and thought 'what's the point' given that gains were not maintained and important things like social initiations did not seem to be uniformly affected by peer involvement. I don't share that view given a few important points: (1) in a real-world setting, peers (generally also known as friends!) don't just interact in two 20 minute sessions for 6-8 weeks, it's more of an on-going thing, and (2) other studies of PMI have come up with slightly more favourable and generalisable results [2] with the need for quite a bit more study. Indeed, in these days of searching for ever more cost-effective strategies for screening and intervention with autism in mind (see here), I'd be minded to suggest that peers (including siblings) represent valuable partners who might also benefit from the whole PMI process too...

---------

[1] Barber AB. et al. Peers as clinicians: Examining the impact of Stay Play Talk on social communication in young preschoolers with autism. J Commun Disord. 2015 Aug 7;59:1-15.

[2] Katz E. & Girolametto L. Peer-mediated intervention for pre-schoolers with ASD: Effects on responses and initiations. Int J Speech Lang Pathol. 2015 Apr 17:1-12.

---------

Barber AB, Saffo RW, Gilpin AT, Craft LD, & Goldstein H (2015). Peers as clinicians: Examining the impact of Stay Play Talk on social communication in young preschoolers with autism. Journal of communication disorders, 59, 1-15 PMID: 26606507