I want to start with a number.
75%.
That’s the percentage of autistic adults who meet the criteria for at least one mental health condition. Anxiety, depression, PTSD, OCD — the rates in the autistic community are not slightly elevated. They are significantly, substantially, consistently higher than in the general population.
Now here’s the other number.
Most of those people are being offered therapy that was not designed for their brains. Delivered by clinicians who had, on average, minimal training in autism. Running protocols built for neurotypical minds. And when it doesn’t work — when the autistic person sits there doing the homework and thinking the thoughts and still feeling exactly the same — the conclusion that gets reached, too often, is that the person is resistant to treatment.
Not that the treatment was wrong. That the person was.
I have things to say about this.
The CBT problem nobody wants to say plainly
Cognitive Behavioral Therapy is the gold standard. Everybody says this. Insurance covers it. Clinicians train in it. It’s what gets recommended when you tell your doctor you’re struggling and they refer you somewhere.
Here is what the research actually says: most clinical tools were designed for allistic (non-autistic) brains. When we apply those tools to autistic distress, it’s like trying to run Mac software on a PC — the code just doesn’t match.
Standard CBT has limited efficacy in reducing anxiety for autistic youth, and outcomes are mixed for adults. Without proper adaptations, CBT often fails to address the sensory, social, and processing differences that come with autism.
Here’s one of the specific reasons this matters:
CBT focuses on insight and logic, helping clients become more aware of how their thoughts impact their emotions and behaviors. But many late-diagnosed autistic individuals already possess a high level of self-insight. Years of masking, internal rumination, and self-analysis mean that the “insight work” CBT emphasizes might not only feel redundant — it can feel invalidating.
If your brain has been running a 24/7 analysis of everything you’ve said and done and gotten wrong since approximately age seven, being asked to “reflect on your thinking patterns” is not illuminating. It’s just the thing your brain already does, formatted as homework.
And there’s another layer. CBT is built around the premise that anxious thoughts are often irrational and can be restructured. The fears that CBT therapists train their patients to view as irrational — “if I say the wrong thing, I’ll lose my job” — are completely rational for autistic people, and rooted in genuine experience.
The autistic person who is afraid of social situations is afraid because social situations have consistently been punishing for them. The anxiety is not a cognitive distortion. It’s an accurate read of a pattern that is actually there. Telling someone to challenge that belief isn’t therapy. It’s gaslighting.
Why autistic people go to therapy for years and still feel the same
I’ve had this conversation more times than I can count. Someone tells me they’ve been in therapy for years — real therapy, with a real therapist who is a decent person and is trying — and they still feel like something fundamental isn’t being reached. Like they’re doing all the right things and the core of the thing is still sitting there, untouched.
CBT doesn’t take into account alexithymia or interoception differences, sensory trauma, social communication differences, masking, or environmental triggers that many autistic people experience.
Let me unpack what some of that means in practice
Alexithymia is difficulty identifying and describing your own emotional states — it affects a significant portion of autistic people. If you can’t clearly identify what you’re feeling, CBT’s approach of tracking thoughts and emotions assumes a level of emotional access that isn’t always there. It’s not that you’re not trying. The access point the therapy is designed to use just isn’t available in the same way.
Interoception differences — difficulties reading your body’s internal signals — mean that the physical cues therapists use to help clients identify when they’re getting dysregulated may not register the same way. You may not feel anxious in the ways you’re being asked to notice. And then by the time you notice something is wrong, you’re already in it.
And masking — the years-long performance of neurotypicality — creates a specific problem in the therapy room. A lot of autistic people are very, very good at presenting as fine. At giving the therapist the responses that seem right. At functioning within the session in ways that obscure what’s actually happening. Existing CBT packages designed for specific anxiety diagnoses in the general population may not be appropriate if the interaction between anxiety and autism is not addressed, and the treatment plan does not take sufficient account of that interaction.
The therapist sees a functional person who is working the program. They don’t see the cost of the performance. They don’t see what the drive home looks like.
What the mental health system does instead of fixing this
Here’s what I want to say about the system, because the individual therapist who is doing their best with the training they have is not entirely the problem. The system is the problem.
Few mental health professionals have adequate training or experience in treating autistic people. The concept that a person can be both autistic and struggle with mental health disorders is relatively new in the social sciences.
Relatively new. In 2026. The idea that an autistic person might also have depression or PTSD or anxiety is still being integrated into clinical training.
So what happens is: autistic adults seek mental health support. They are placed in the standard pathways — CBT, medication, the usual protocols. These things help some people. They help autistic people less, and less reliably. The autistic person is offered more of what isn’t working, or told they’re treatment-resistant, or quietly concludes that therapy just doesn’t work for them and stops trying.
Meanwhile the mental health crisis in the autistic community continues, and the research on better approaches exists and is not making it into clinical practice fast enough, and the people who need the most tailored support are receiving the most generic.
What actually helps
I want to be clear that this is not a “therapy bad” post. Therapy can be transformative. I have experienced this personally. The question is whether you’re getting the version that can reach you.
Traditional CBT isn’t “bad.” It’s just not designed for autistic neurology. What is designed for autistic neurology looks different.
Somatic approaches — body-centered therapy that works with physical sensation and nervous system regulation rather than primarily through cognitive restructuring — tend to be more accessible for autistic people, particularly those with interoception differences. Getting out of the head and into the body in a supported way can reach things that talk therapy doesn’t.
Trauma-informed approaches that understand autistic experience as often involving significant trauma — the years of being wrong in rooms that didn’t know how to hold you — rather than treating the trauma as a separate complication.
Therapists who actually know what autism looks like in adults. Who aren’t going to spend the first ten sessions asking you to make eye contact or interpreting your communication style as avoidance. Who understand masking and don’t require you to perform wellness at them.
And community. This one isn’t therapy, technically, but the research on what reduces mental health struggles in autistic people consistently points toward autistic community — spaces where the translation labor disappears, where you don’t have to explain yourself, where someone else in the room already knows.
The thing I most want clinicians to hear
If an autistic person has been in therapy and it hasn’t helped, the problem is probably not the autistic person.
The problem is very likely that they were given a tool designed for a different kind of brain and told it was the best available thing. That’s not their failure. That’s a systemic gap.
The autistic people in your caseload who seem “resistant” to treatment deserve a second look — not at whether they’re trying hard enough, but at whether what you’re offering is actually built to reach them.
Most of the time, it isn’t. Not yet. But it could be.
And for autistic people reading this who have tried therapy and walked away feeling like something was wrong with them: something might have been wrong with the approach. You deserved better than what the system had available. A lot of us did.
That’s not an excuse to stop trying. It’s an invitation to try something different — with a clinician who actually knows your brain, in a form that’s actually built to reach it.
You are not treatment-resistant. You are under-served.
Those are not the same thing.
