If you’re reading this, you’re probably not surprised by the idea that autistic and neurodivergent people deal with comorbid or co-occurring conditions (i.e., other health conditions in addition to autism or the “primary condition of interest”) at higher rates than the general public.
“Examples of autistic comorbid conditions range from physical conditions, like sleep & GI problems, to conditions like anxiety and depression, and other brain differences under the neurodivergent umbrella, like OCD or ADHD.”
But how do we classify these co-occurring conditions? Are they co-occurring at birth or do they develop during the lifetime? Do they develop in part due to autism, either because of shared genetic factors or environmental impacts (e.g., access to fair & supportive healthcare)? Is autism even part of the co-occurring condition’s development or do these conditions occur in higher proportions among autistic people for some other, yet discovered reason? Do these distinctions even matter?
In a paper by Eric Rubenstein and Lauren Bishop‐Fitzpatrick, published in late 2018, they discuss the importance of specific classification of co-occurring conditions to Autism Spectrum Disorder (ASD), how improved classification may also improve research and treatment, and proposed three categories for classifying co-occurring conditions.
Let’s begin by introducing the categories proposed as the authors state them, then we’ll go into a deeper description of each.
(1) “truly co‐occurring”: share an etiologic origin with ASD in utero and are a defining characteristic of a subphenotype
(2) “resulting”: caused by ASD related disparity or the health effect of behaviors developed to cope with ASD symptoms
(3) “associated”: conditions more common in individuals with ASD with etiology not yet known or hypothesized, or an artifact of diagnostic process or trends”
The authors include this visualization of each category. It may give you a better idea of what each category means before we dive in.
Autism – Truly Co-occuring Conditions
-
“share an etiologic origin with ASD in utero and are a defining characteristic of a subphenotype”
Authors note on this section of the figure: As we know very little about health conditions in older adults with ASD, we highlight a potential condition (Parkinson’s disease) that has been hypothesized to share an etiologic link with ASD but would not present until later in life (Starkstein et al., 2015). We convey this uncertainty with a dashed line.
-
“Share an etiologic origin with ASD in utero”: a common cause (“etiologic”), such as genetic influences or something that occurs during the pregnancy (“in utero”)
The primary considerations for this category are development of the co-occurring condition during pregnancy, risk factors present in the weeks leading up to and immediately following birth, shared genetic origins, and the condition’s onset and effect across the course of one’s life.
-
“Are a defining characteristic of a subphenotype”: the common origin between a co-occurring health condition and ASD may help to better define the ASD group and subgroups within ASD, which tend to be very different and varied (heterogenous) in terms of both genetics and symptom expression
Conditions labeled as truly co-occurring share a root cause with autism, originating at the same time as autism. These conditions may be “late on-set”, meaning they are triggered by something in the environment or by age, but the tendency to develop the co-occurring condition exists from birth.
Shared causes with co-occuring conditions may also lead to better identification of the autistic subgroup (versus the general population), as well as possible subgroups within autism. Identifying these groups can lead to better tailored and more effective treatments for reducing the health impact of co-occurring conditions.
The authors note that the knowledge gained from subsetting groups by shared origins between conditions and ASD should “guide efforts to reduce exposure to risk factors with the goal of preventing the [frequency] of health conditions…and [reduce] the burden of health conditions and their interaction with ASD [symptoms]” [1]. I feel like this statement is supposed to be getting at one of the concerns cropping up for me:
“Will determining these origins result in trying to prevent autism, either with ‘cures’ or greater terminations of pregnancies with potential ASD genetic markers? I believe they’re trying to say that this information would be used to prevent and better treat the co-occurring health conditions, knowing the context of ASD and the common origins, not prevent or cure ASD.”
Though this concern remains otherwise unaddressed in the article.
Autism – Resulting Conditions
“caused by ASD related disparity or the health effect of behaviors developed to cope with ASD symptoms”
-
“Related disparity”: a systematic difference (disparity) that is related to being autistic, such as access to healthcare or social support
“There are differences between the lived experiences of autistic people and the general population. These differences, such as levels of participation and acceptance in a community, sensory processing challenges, or learning to mask one’s autistic symptoms, contribute to disparities in autistic health, access to healthcare and health outcomes.”
Autistics may receive subpar healthcare compared to the general population due to a host of ASD-related reasons: executive dysfunction making setting an appointment or remembering what to talk about difficult, sensory processing problems (smells, lights, change in routine) in doctor office settings, meltdowns & selective mutism during appointments, unconscious masking of symptoms during appointments, and I believe there is a tendency to take autistics less seriously than your average neurotypical individual. These are just a few possibilities.
In addition to healthcare inequalities, all people may experience social and economic injustice that may impact one’s health. The authors point to research from the general population suggests health conditions can result from this social or economic injustice experienced throughout life, and it would make sense that this “experienced injustice” also affects the autistic population. Autistic people tend to experience higher rates of experienced injustice as well, such as lack of social support or social isolation, financial and employment problems, and generally fewer opportunities. Experiencing both ASD-specific access and health problems, compounded by economic, social support, ethnic or other disparities may lead to even poorer health outcomes.
-
“Health effect of behaviors developed to cope with ASD symptoms”: health effects stemming from symptoms of ASD, such as trouble processing social cues, or stemming from behaviors developed to cope with ASD. So, a health effect like increased stress hormone levels (cortisol) could arise from trouble processing social cues, a common characteristic in ASD, or from masking, a behavior developed to cope with ASD.
Autism has many different symptom expressions and many autistic people develop coping strategies that might impact us physically. The physical impact of symptoms of coping mechanisms may be a contributing factor for some co-occurring conditions. However, we don’t really know how autistics are impacted on the physical level by autism and its expression. I haven’t heard of research about the impact of stimming on muscles and joints, or the connection between stimming and brain activity. We also don’t know how meltdowns, shutdowns, and episodes of mutism impact autistics physically and mentally over the long haul in terms of hormone levels, brain activity, or other impacts (and please tell me if you have seen research in this space!). We can draw some connections, like we know that increased cortisol levels (stress hormones) are associated with increased risk of heart disease, and there is evidence that autistic people often have higher levels of cortisol in different situations than the general public (at least, it’s been well studied for autistic children. Here are a couple articles related to cortisol and stress hormone response in autistic children one [2],two [3],three [4]). These connections still need further exploration, especially in adults.
This lack of research among autistic adults, especially older adults, limits our understanding of how later on-set co-occurring conditions develop. We’re missing data collected across the lifetime for many research areas. The lack of research may make the “resulting” label harder to use, but also could provide context for many new research studies!
Autism – Associated Conditions
-
“conditions more common in individuals with ASD with etiology not yet known or hypothesized, or an artifact of diagnostic process or trends”
Example: Autistic people are often diagnosed with anxiety too, but it’s unclear if the anxiety:
(1) shares a common origin with autism;
(2) stems from social or societal challenges that come with being autistic in a neurotypical world;
(3) results from behaviors developed to cope with autism and (2); or
(4) if there’s a cause independent of autism entirely.
In some cases, a doctor or therapist might be able to make an educated guess for the category and use that as context for treatment
This category is a bit of a catch all, encompassing many unknowns. The authors point out that there is much research to be done within this category.
-
“An artifact of diagnostic process or trends”: The diagnostic criteria has changed over the years for ASD, most recently in 2013 with the DSM-V, the manual for diagnosing mental health conditions and includes neurodivergent conditions as well. Trends in research may also impact what diagnoses are given to an individual. Previous diagnoses are not consistently handled when a new diagnosis of ASD is given.
When someone receives a late diagnosis of ASD, they have likely received one or more diagnoses previously. Two things can happen to these previous diagnoses: (1) diagnostic accretion: keep the original diagnosis or (2) diagnostic substitution: replace the original diagnosis with ASD [1].
I feel the authors make a really important point here: “In population level data, [autistic] individuals may carry previous diagnoses as an artifact, but not truly have these conditions that alter…treatment approaches.” Meaning that some of the link found between autism and other diagnoses might be due to old diagnoses that no longer apply, but stick around in the data and charts (keeping a diagnosis, diagnostic accretion, when it should have been replaced, diagnostic substitution! I’m not making you learn jargon for nothing, I swear.). The impact of these artifact diagnoses reaches from research done with this data to the individuals carrying these potentially outdated diagnoses – at the very least altering their treatments.
I have experienced a version of this personally – I was initially diagnosed with anxiety and depression, and took medication for several years. In hindsight, I feel we were treating symptoms caused by a “resulting co-occurrence”. Once I received my autism diagnosis as well (and I maintained my anxiety and depression diagnoses),
“I wanted to see what my baseline was without medication and within my new context of autism. My anxiety and depression have lessened in a way I was never able to achieve with therapy or medication before, I believe primarily due to working within this new autistic context.”
Yet, almost every time I visit the doctor, they want to put me on a this antianxiety or that antidepressant – but personally I feel this is unnecessary and perhaps even counterproductive at this time. My anxiety doesn’t seem to be caused by an innate chemical imbalance, but by having to cope with the sensory and social challenges I face in the world as an autistic person and the behaviors I developed in response.*
* I admit I don’t know for sure that my anxiety and depressive episodes aren’t caused by a chemical imbalance. But for the years I was medicated, my GAD-8 and PHQ-10 (anxiety and depression scales, respectively) were high and varied only a handful of points over the course of those two years. Those scores are many points higher compared to where I rate now, post-diagnosis. And there’s a subjective change as well: the anxiety feels different now. It’s still a part of my daily life, but its impact is less. And this all isn’t to say I wouldn’t try medication again in the future – I would.
“This inaccuracy in determining co-occurrence can go in the other direction too – autistics, especially girls and women, are often misdiagnosed at first and some may never receive a professional diagnosis of autism.”
The authors don’t mention self-diagnosis. However, self-diagnosis is important particularly because of the other inequalities for autistics in access to healthcare. An autism diagnosis might be barred for financial/insurance reasons, or even because your primary care physician refuses to refer you to a specialist for a diagnosis. An individual’s treatment is impacted immensely if they never receive a diagnosis, and the difference in access to the effective treatments likely impacts their co-occurring health conditions, perhaps in a more severe way than other autistic people. As I mentioned above, my anxiety and depression were much worse, and very little seemed to help, before I had the context of autism.
As data science and big data is on the rise, sorting out instances of inconsistent prior diagnosis change or retention, finding and collecting self-diagnosed data, and collecting data collected over the course of weeks, months or years (longitudinal data) from the autistic community will be very important, however challenging, if we hope to gain an accurate picture!
Closing Thoughts
The language we use is important. In this case, it seems important not only for autism but anything you can define as the “condition of primary concern”. The origins of co-occurring conditions and when and how they arise is likely an important step to developing different and more effective treatments; allowing us to focus on treating the cause, instead of just the symptoms.
The authors hope this will spur discussion in the research community as well as among autistics and advocates, and they mentioned several times the impact of the autistic and advocate community on their research. Personally, I would love to see more work impacted by our voices – so let’s get to discussing!
What are your thoughts?
Have you experienced loss or retention of other diagnoses and any consequences due to this loss or retention? Do you feel your co-occuring conditions impact your treatment effectively? Are all of your diagnoses considered when deciding treatment or is each treated more individually? What concerns or comments do you have on the proposed categories for co-occurring conditions?
Sullivan is an autistic yoga teacher striving to share more coping tools (such as yoga & meditation) with the neurodivergent community and beyond. Blending her background in psychology and mental health with yoga, Sullivan strives to share the peace, self-acceptance, and physical awareness yoga and mindfulness has brought her. Sullivan has a BS in Psychology from the University of Washington, and completed her 200-hour yoga teacher training in 2019.
References
-
Rubenstein, E., & Bishop-Fitzpatrick, L. (2018). A matter of time: The necessity of temporal language in research on health conditions that present with autism spectrum disorder. Autism Research, 12(1), 20–25. doi: 10.1002/aur.2010
-
Corbett, B., Mendoza, S., Abdullah, M., Wegelin, J., & Levine, S. (2006). Cortisol circadian rhythms and response to stress in children with autism. Psychoneuroendocrinology, 31(1), 59–68. doi: 10.1016/j.psyneuen.2005.05.011
-
Spratt, E. G., Nicholas, J. S., Brady, K. T., Carpenter, L. A., Hatcher, C. R., Meekins, K. A., … Charles, J. M. (2011). Enhanced Cortisol Response to Stress in Children in Autism. Journal of Autism and Developmental Disorders, 42(1), 75–81. doi: 10.1007/s10803-011-1214-0
-
Schupp, C. W., Simon, D., & Corbett, B. A. (2013). Cortisol Responsivity Differences in Children with Autism Spectrum Disorders During Free and Cooperative Play. Journal of Autism and Developmental Disorders, 43(10), 2405–2417. doi: 10.1007/s10803-013-1790-2
Discussion about this post