For professionals
When autism masks the actual diagnosis
An autistic client has been reporting exhaustion and anxiety for months. In the team, one sentence comes quickly: that is just part of the autism. Sometimes that is true. Sometimes an independent depression or anxiety disorder lies underneath and remains untreated as a result.
Research calls this pattern diagnostic overshadowing: new or additional difficulties are attributed to a diagnosis that is already known. The following answers show which mechanisms can be confused in the process.
The following answers were generated in German in the app on 29 September 2026, unedited and unabridged. The English text is a faithful translation of those answers.
1. Course over time, not a label
The first answer separates exhaustion from load from exhaustion with its own course.
User question
As a therapist, how do I know whether an autistic client’s exhaustion or anxiety is an independent issue or simply part of the autism?
Autistic Mirror (outside view/environment, therapist)
The question touches one of the most consequential decisions in work with autistic clients, because the direction of treatment differs fundamentally.
A helpful first arbiter is the course over time. Exhaustion that depends on increased demands, sensory load or social masking phases, and that also recovers with relief, follows a different pattern from a depression with its own trajectory of severity, which rises and persists independently of the sensory context. If the client can tell you "after the holiday I was myself again" or "in the quiet summer week without office/school/obligations it eased", that points more strongly to a budget-exhaustion pattern that belongs to the autistic profile. If the exhaustion does not ease even in quiet weeks and is accompanied by a persistent sense of worthlessness or hopelessness that is not tied to contexts, it more likely has an independent affective aetiology.
For anxiety, the distinction between intolerance of uncertainty and an independent anxiety disorder matters. Intolerance of uncertainty is an autistic mechanism: the predictive coding system, that is, the way the brain compares expectations against incoming input, weights sensory input more strongly relative to the internal prediction in autistic people. Unclear situations therefore produce real physiological alarm via the stress axis, even when there is no objective danger. This is not a disorder phenomenon but the direct consequence of this processing pattern. An independent anxiety disorder, by contrast, has its own course, often with physical anxiety symptoms that also appear when the environment is currently manageable and predictable, and that do not behave proportionally to uncertainty cues.
In practice, a biographical longitudinal inquiry that pays attention to masking helps here. If the client describes anxiety or exhaustion appearing massively whenever social demands rise, whenever she had to "function", whenever she spent long periods in sensorily demanding environments, that points to the masking-load pattern. Masking means that automatic autistic reactions are controlled and suppressed prefrontally while social content is supposed to be processed at the same time. That permanently costs capacity, produces a cortisol curve over the day and accumulates. If the client describes this in retrospect as "I was always exhausted, but I thought that was normal", that is a clear indication of the autistic connection, not of an independent disorder.
A second diagnostic axis is alexithymia. Many autistic clients cannot reliably distinguish between exhaustion, anxiety and diffuse physical stress, because interoceptive resolution is lower, that is, the ability to perceive body signals and assign a feeling word to them works less reliably. This means: what is reported as "anxiety" may be overload tension, and what is described as "exhaustion" may contain a depressive dip. So do not primarily ask about the feeling word, but about the situation: When? What happened before? How long? What made it better?
Important background: autistic burnout and depression can coexist and reinforce each other. Burnout describes a loss of abilities after accumulated masking load over weeks to years, which differs from the exhaustion of a single day and is not fully reversible even after recovery. Depression is an independent affective diagnosis with a lowered expected value, that is, the feeling that what comes afterwards is not worth it. Both can run in the same person. If that is the case, treatment needs both strands, but with autism-sensitive adaptations for both.
The treatment consequence is why the decision matters so much: exhaustion that arises from masking load does not improve through activation, but through relief. Anxiety that arises from intolerance of uncertainty does not need exposure logic, but predictability and structure. Anyone who treats the autism-specific part with standard CBT may increase the masking load instead of reducing it.
2. Four states behind one word
The second answer separates four states that all look like withdrawal from the outside.
User question
An autistic client has been describing lack of drive for months. The team says it is just autistic withdrawal. Which mechanisms could be confused here?
Autistic Mirror (outside view/environment, therapist)
Lack of drive in an autistic client has several possible origins, each with very different diagnostic consequences. "Autistic withdrawal" is not an explanation but a description that can cover four mechanistically different states.
Autistic burnout is the first candidate that should be ruled out first when loss of drive has lasted for months. Autistic burnout arises from accumulated load from masking, constant sensory strain and lack of recovery, and it shows as a loss of abilities: things that were still possible months ago no longer work now. That is the diagnostically important difference from weak motivation. If the client describes being able to do certain everyday things before and not being able to now, even though he wants to, burnout is the more likely mechanism. The physiological basis is a chronically raised cortisol load from ongoing masking, which exhausts prefrontal control capacity over the long term and weakens the sensory filter permanently.
Autistic inertia explains another aspect. Autistic inertia means that switching between states, that is starting, stopping, changing over, is neurologically disproportionately expensive. This is due to differently wired basal ganglia and resting-state circuits. In experience: the client knows what he wants to do and still cannot get going. From the outside that looks like lack of drive, but it is not a lowering of wanting, it is a block at the state switch. Inertia can accompany burnout and amplify it, but it should be named as a mechanism of its own: as an amplifier it increases the perceived action block on top of what burnout has already taken away in capacity.
Depression is the third diagnosis to separate and has an independent affective aetiology. In depression, the expected value of the action itself drops: it is not that switching is expensive or capacity is missing, but that nothing seems worthwhile. Autistic people have an empirically clearly increased risk of depression, and there is a known causal chain: persistent camouflaging, that is the strategic concealment of autistic traits, leads via ongoing cortisol load to chronic exhaustion, social loneliness and self-alienation, from which depressive episodes develop. The difference from inertia is clinically accessible: in inertia the wanting remains, in depression the expected value is missing. In burnout the two overlap, because exhaustion affects both axes.
Shutdown as a fourth mechanism can also be described as withdrawal, but it is acute and reactive. Shutdown is a parasympathetic collapse response via the dorsal vagus, triggered by sensory overload, with withdrawal inwards, reduced responsiveness and often limited speech. It lasts hours, not months. If the team observes "withdrawal" over months and explains it with shutdown, that confuses time scales.
The practical differentiating question for work with the client is: was there a point in the last months up to which certain abilities were still present, and afterwards not any more? If so, that points to burnout. Is the wanting present in principle, but starting is systematically blocked? Then inertia is involved. Is drive missing even for things that used to be enjoyable, regardless of sensory load or switching costs? Then depression needs its own diagnostic consideration, not as a side effect but as its own aetiology with its own treatment axis.
3. Why anxiety disappears into autism
The third answer explains why independent anxiety is so easily overlooked in autistic adults.
User question
Why is anxiety in autistic adults so often interpreted as part of the autism rather than as its own co-occurring condition?
Autistic Mirror (outside view/environment, therapist)
Because the visible symptoms of both states overlap strongly, and because diagnostic criteria for anxiety disorders were developed on neurotypical presentations.
Predictive coding explains part of the confusion. The autistic brain weights incoming sensory information more strongly than its own predictions, while neurotypical processing relies more on the predictive model and dampens details in the background. Every uncertainty therefore produces a larger matching effort: the unannounced hits the system with full force, because there is no dampening upstream. From the outside that looks like anxiety, and it also feels physically similar, because the same stress axis is activated. The difference lies in the trigger and the mechanism: in autistic processing it is a computational mismatch between what is and what was predicted. In an anxiety disorder there is additionally a clinically independent dysregulation of threat appraisal that goes beyond mere uncertainty processing.
The second problem is the HPA axis. Autistic people often show higher cortisol peaks and slower normalisation afterwards. Chronic masking, sensory load and social demands keep the stress axis running at a raised level permanently. Clinical observations capture this constant-cortisol state without knowing its cause and categorise it as generalised anxiety. That is then not a misdiagnosis in the narrowest sense, but an incomplete one, because the generating context is missing.
The third problem is structural: diagnostic interviews ask about social anxiety, exam anxiety, avoidance behaviour, physical anxiety symptoms. All of this is present in autistic adults too, but partly for different reasons. Social situations are exhausting because of the decoding effort in the double empathy problem and because of masking load, not primarily because of anticipated threat. Avoidance can be exhaustion management rather than anxiety avoidance. Alexithymia makes it harder still, because autistic people often cannot reliably distinguish whether what they perceive physically is exhaustion, overstimulation or anxiety: the insula, which translates interoceptive signals into recognisable body states, works differently, and assigning a feeling word stays blurred. In the assessment what arrives is "I don't know, I'm just churned up", and that gets coded as anxiety.
The distinction is clinically relevant because the interventions differ. Standardised anxiety treatment, for example exposure with response prevention for social anxiety, assumes that avoidance arises from anxious thoughts and decreases through habituation. But if the underlying load is sensory exhaustion and decoding effort, exposure without environmental adaptation produces no habituation, but retraumatisation or burnout. An independent anxiety disorder can still be present in addition, and it then needs its own consideration too. But without the autistic starting point as context, treatment stays on the surface.
What is supported in this framing
Diagnostic overshadowing has been described in autistic adults, as has the increased risk of anxiety disorders and depression. The answers apply this to therapeutic differentiation. The distinguishing features mentioned are orientation, not a validated assessment tool.
Statements about cortisol, the HPA axis, prefrontal capacity, the insula, the basal ganglia and the dorsal vagus are models, not measurements in the individual person. Polyvagal theory is widely used but not undisputed. Sensory filter exhaustion is a descriptive term without its own first publication and is therefore not in the source list.
The first German answer contains a typo that comes from the unedited output. Its sentence on standard CBT is not an assessment of a form of therapy, but a note on masking load when autistic processing is not taken into account. Diagnostic and therapeutic decisions need the full context of the person and the responsible professionals.
Autistic Mirror explains the mechanisms behind exhaustion and anxiety. It does not diagnose and does not replace a clinical history.
Sources
- Mason et al. (2019). DOI: 10.1007/s10803-019-04049-2
- van Steensel, Bögels & Perrin (2011). DOI: 10.1007/s10567-011-0097-0
- Hudson, Hall & Harkness (2019). DOI: 10.1007/s10802-018-0402-1
- Raymaker et al. (2020). DOI: 10.1089/aut.2019.0079
- Buckle, Leadbitter, Poliakoff & Gowen (2021). DOI: 10.3389/fpsyg.2021.631596
- Vasa et al. (2018). DOI: 10.1002/aur.1916
- Bird & Cook (2013). DOI: 10.1038/tp.2013.61
- Pellicano & Burr (2012). DOI: 10.1016/j.tics.2012.08.009
- Hull, Petrides, Allison, Smith, Baron-Cohen, Lai & Mandy (2017). DOI: 10.1007/s10803-017-3166-5