Autism and Anxiety: Why So Many Autistic Adults Are Misdiagnosed with Anxiety Disorders

*Anxiety is the most common presenting complaint in late-diagnosed autistic adults — and one of the most commonly misunderstood. This article examines why so many autistic people spend years being treated for anxiety disorders when the primary condition driving their distress was never identified.*

She had been in therapy for anxiety since her mid-twenties. She had tried CBT, mindfulness, medication, and more than one approach that had helped for a time before the anxiety returned, as it always did, in a new configuration. By the time she sat in our consulting room at the age of forty-six, she had accumulated a clinical history that read as a textbook case of generalised anxiety disorder with some social anxiety features. What it did not read as — because no one had ever thought to look — was autism.

Her diagnosis, when it eventually came, was not a surprise to her. It was a recognition. The anxiety that had defined her clinical presentation for two decades was real — it remains real, and it continues to require attention. But its origins were not where the treatment had been directed. The anxiety was not, at its root, a disorder of cognition or a dysregulation of the threat-detection system in isolation. It was the predictable, understandable response of an undiagnosed autistic woman navigating a neurotypical world without the tools, the accommodations, or the self-understanding to do so sustainably.

This story is not unusual. It is, in the clinical experience of those who specialise in late autism diagnosis, remarkably common — particularly for women, and for anyone who learned early and effectively to mask. This article examines why anxiety and autism are so frequently confused, what the clinical consequences of that confusion are, how to think about the relationship between the two, and what an autism assessment at our Harley Street clinic can offer when the existing clinical picture is not providing adequate answers.

The Scale of the Problem

The research on anxiety in autistic populations is unambiguous on one point: anxiety is extraordinarily prevalent. Studies consistently find that between 40 and 60 per cent of autistic adults meet diagnostic criteria for one or more anxiety disorders, with some estimates considerably higher depending on methodology and population. Generalised anxiety disorder, social anxiety disorder, specific phobias, panic disorder, and OCD are all significantly over-represented in autistic adults compared to the non-autistic population.

What the research is less consistent about — and what the clinical literature has only recently begun to address seriously — is the direction of causality, the nature of the relationship between autism and anxiety, and the clinical implications of that relationship for assessment and treatment. Anxiety in autistic adults is not a coincidental co-occurrence. In most cases, it is a direct product of living as an unrecognised or unsupported autistic person in a world that was not designed for the autistic nervous system. Understanding this changes both how the anxiety should be assessed and how it should be treated.

Why Autistic People Develop Anxiety: The Structural Causes

Anxiety in autistic adults rarely emerges from nowhere. It develops, in most cases, from a comprehensible set of structural causes — the specific stresses and challenges that the autistic nervous system faces in a predominantly neurotypical environment — that remain active and unaddressed unless the autism itself is recognised and accommodated.

Social unpredictability is among the most significant. The social world, from an autistic perspective, is characterised by a high degree of ambiguity: of implicit communication, unstated expectations, shifting norms, and rapid conversational dynamics that are not accessible through explicit instruction and do not resolve into reliable patterns. For a nervous system that is oriented towards pattern recognition and consistency, and that has difficulty processing the implicit signals through which most social communication travels, the social world is a sustained source of uncertainty — and uncertainty, for the threat-detection system, tends to register as threat.

The experience of repeated social failure or misattunement — of saying the wrong thing, missing the unspoken cue, responding too literally to what was said rather than to what was meant — compounds this anxiety over time. Many autistic adults describe having learned, very early, that social situations are dangerous: that they are likely to make a mistake, to be laughed at, excluded, or judged. This learning becomes encoded as social anxiety that persists long after the specific situations that generated it have passed.

Sensory overload is another significant source. Environments that are loud, crowded, visually complex, or temperature-variable — the ordinary environments of workplaces, public transport, social gatherings, and urban life — place a continuous high demand on the autistic sensory system. Managing that demand consumes cognitive resources that are then unavailable for other processing. It also generates physiological arousal that is, neurologically, indistinguishable from anxiety. The autistic person who feels chronically on edge in public spaces may have been told they have generalised anxiety disorder. What they have is sensory overload that has never been named as such.

The effort of masking — the sustained performance of neurotypical social behaviour — generates anxiety both directly and indirectly. Directly, because the performance of a self that is not authentic is inherently stressful: there is always the implicit threat of being found out, of the mask slipping, of the underlying difference being exposed and judged. Indirectly, because masking consumes the resources that would otherwise be available for managing the ordinary stresses of daily life, leaving the autistic person operating with depleted capacity and a much lower threshold for anxiety activation.

How Autism Presents as Anxiety: The Diagnostic Overlap

The clinical presentations of autism and anxiety share enough surface features that misidentification is understandable, even inevitable, when the clinician’s framework does not include autism as a possibility. This is particularly true for presentations in women and in adults who mask effectively.

Social anxiety disorder is perhaps the most common misdiagnosis in late-diagnosed autistic adults. Both conditions involve significant distress in social situations, avoidance of social encounters, and anticipatory anxiety about social performance. The critical distinction — and one that requires careful clinical assessment to identify — is in the underlying mechanism. Social anxiety disorder, as classically understood, involves fear of negative evaluation: the person fears that they will be judged, embarrassed, or humiliated by others. Autistic social anxiety tends to involve a different set of concerns: the fear of not understanding the social rules, of making a mistake that reveals difference, of not being able to maintain the performance of normality, or of the sensory and cognitive overwhelm that sustained social interaction produces. These distinctions are clinically significant because they predict different therapeutic needs.

Generalised anxiety disorder — characterised by persistent, excessive worry across multiple domains — is another frequent misidentification. Autistic adults often describe themselves as chronic worriers, and this is frequently accurate. But the content and structure of autistic worry tends to differ from classical GAD. Much of the worry in autistic presentations is focused on social situations, changes to plans or routines, uncertainty about what is expected or will happen next, and the management of overwhelming sensory or cognitive environments. It is worry that is, in important respects, a rational response to genuine difficulty — and treating it as an irrational cognitive distortion without addressing its underlying causes is unlikely to produce lasting change.

Panic disorder may develop in autistic adults as a consequence of repeated autistic meltdowns that are misunderstood as panic attacks — either by the clinician assessing them or by the autistic person themselves. Meltdowns and panic attacks share physiological features but differ significantly in their triggers and their management. Treating an autistic meltdown with the interventions designed for panic disorder may produce minimal benefit, and may actively increase the autistic person’s sense that they are fundamentally broken in a way that therapy cannot reach.

OCD is a particularly complex area of overlap. The repetitive behaviours and routines of autism — which serve functions of self-regulation, predictability, and sensory management — can be superficially similar to OCD compulsions, and this similarity has led to significant misidentification in both directions. Genuine OCD does occur in autistic people at elevated rates, and distinguishing autistic routine from OCD compulsion — which requires attention to the function of the behaviour, the presence or absence of ego-dystonic intrusive thoughts, and the degree of distress associated with interruption — is essential clinical work.

The Consequences of Missed Autism Diagnosis

The clinical consequences of spending years or decades being treated for anxiety when the primary unaddressed condition is autism are significant and deserve to be stated directly.

The first consequence is treatment that does not work, or does not work well enough. Many autistic adults who are eventually diagnosed describe having had meaningful but limited responses to anxiety treatment: some improvement, some periods of better function, but a persistent sense that the treatment is managing something rather than addressing its origin. Anxiety that is structural in nature — produced and maintained by the ongoing demands of living as an unidentified autistic person — is not fully addressable through interventions that target the anxiety in isolation. The structural causes remain active, and the anxiety continues to be generated.

The second consequence is the accumulation of self-blame. The narrative that is constructed, over years of partial treatment, is often one in which the person is fundamentally defective in some way that clinical intervention cannot quite reach — too anxious, too fragile, too sensitive, too treatment-resistant. This narrative is not only inaccurate. It is actively harmful. It attributes to character or constitution a difficulty that is, in significant measure, a product of environmental mismatch and insufficient understanding.

The third consequence is the opportunity cost: the years spent without the self-understanding that an autism diagnosis provides, without the accommodations it enables, without the framework that allows a person to build a life that is genuinely sustainable for their neurology rather than one that perpetually demands more than the system can sustain.

What Distinguishes Autistic Anxiety from Primary Anxiety Disorders

For clinicians and for autistic adults who are trying to make sense of their own presentation, there are a number of clinical features that tend to distinguish anxiety that is primarily autistic in origin from anxiety that constitutes a primary anxiety disorder independent of autism. These are not absolute rules — the picture is always individual and always complex — but they are clinically useful patterns.

• The anxiety is most pronounced in social situations involving ambiguity, implicit communication, or unpredictability — rather than in situations involving specific feared objects or generalised across all domains.
• The anxiety is significantly relieved when the person is alone or in the company of a small number of trusted people with whom masking is not required — suggesting that the social demand itself, rather than a generalised threat response, is the primary activating factor.
• There is a longitudinal history of social difficulty, sensory sensitivity, and the need for routine or predictability that predates the anxiety presentation and cannot be fully explained by it.
• Cognitive interventions targeting the anxiety — particularly those that focus on identifying and challenging catastrophic thoughts about social evaluation — have produced limited or partial benefit, despite genuine engagement with the therapy.
• The person describes a sense of not understanding the social rules as they are being played, or of watching social interactions from the outside and attempting to decode what is happening and what is expected, rather than participating spontaneously.
• There is a history of masking — of effortful performance of social normality — that the person can identify in retrospect even if they did not have a name for it at the time.

None of these features is, on its own, diagnostic. Together, and in the context of a comprehensive assessment, they point towards an autistic presentation that deserves to be properly evaluated.

Treating Anxiety in Autistic Adults: What Actually Helps

When anxiety in an autistic adult has been properly understood as arising, at least in part, from autistic experience, the treatment approach changes — not completely, but significantly.

Standard anxiety management approaches — including some CBT techniques, mindfulness, and psychoeducation about the anxiety response — retain utility for many autistic adults. The modification is not in abandoning these tools but in contextualising them correctly: as tools for managing the anxiety response rather than as treatments for its root cause, and as tools that may need to be adapted to autistic cognitive and communicative styles.

The more fundamental therapeutic work involves addressing the structural conditions that generate the anxiety. This means working with the autistic adult to understand and reduce masking where possible; to identify the environmental conditions that are most depleting and to advocate for accommodations or changes; to develop a genuinely autism-informed self-understanding that replaces the narrative of personal deficiency with one of neurodevelopmental difference; and to build sustainable life structures that are designed for the actual nervous system involved.

It also means, in many cases, working with the grief and anger that accompany the recognition of how much the absence of autism identification has cost. These are not merely emotional responses. They are clinically significant material that, if left unaddressed, can themselves become sources of sustained distress and can undermine the integration of the new understanding.

At London Trusted Therapy, our autism assessment and post-diagnosis therapeutic services are designed precisely for adults who arrive with a long history of anxiety treatment that has not told the full story. We offer a comprehensive assessment process that takes the possibility of autism seriously, conducted by clinicians with specific expertise in late-diagnosed adult presentations and in the particular way that autism presents in women and in highly masking individuals. And we offer therapeutic support, following assessment, that is genuinely informed by neurodevelopmental understanding rather than adapted from a neurotypical framework.

If something in this article has resonated — if the description of an anxiety that has always felt structural, that has responded partially to treatment but never fully resolved, that is most pronounced in social situations and most relieved in solitude — we would encourage you to contact us. The anxiety you have been living with may be entirely real. The question of what is generating it, and what would actually address its origins, may not yet have been properly asked. We are here to ask it with you.

To enquire about autism assessment or autism-informed therapy in London, please contact us via the website or telephone to arrange a confidential initial consultation. 115A Harley Street, London W1G 6AL

Originally published at https://londontrustedtherapy.com on June 4, 2026.