When “Access” Isn’t Support: Rethinking Talking Therapy Benchmarks for the Autistic Community
More Than a Tick-Box: Why CBT Must Be Adapted for Autistic Needs
A Critique of LeDeR’s Use of Talking Therapy Metrics
Recent reporting from the LeDeR (Learning from Lives and Deaths) report highlights that 56% of autistic adults had access to talking therapies, including CBT, group therapy, or psychotherapy/counselling. At face value, this statistic appears to signal progress. However, emerging critical assessments in 2025–2026 suggest that this figure rests on a neurotypical benchmark assumption: that standard talking therapies are inherently effective and appropriate for autistic people.
This assumption is deeply problematic.
Access to a service does not equate to effective, appropriate, or safe support, particularly when the service model itself is not designed for the population being measured. By treating “access” as a proxy for adequacy, LeDeR risks overstating provision while masking systemic failures in mental health care for autistic adults.
As an autistic adult that has had alot of therapy over the years (CBT, ACT, Schema Therapy and EMDR), I personally found that CBT was the least appropriate or effective for my needs.
The Core Methodological Issue: Neurotypical Bias in Benchmarking
The LeDeR metric assumes that CBT and other talking therapies are neutral, universally applicable interventions. In reality, standard CBT and traditional talking therapies were developed for, trialled on, and normed against neurotypical populations. When these models are applied to autistic people without adaptation, the evidence consistently shows poorer outcomes.
Crucially, LeDeR does not capture whether therapies were adapted, whether therapists had autism-specific training, or whether autistic adults found the intervention helpful or harmful. The statistic therefore measures contact, not benefit.
In methodological terms, this is a category error: counting exposure to an intervention without measuring its suitability or efficacy for the population concerned.
Why Standard CBT and Talking Therapies Often Fall Short for Autistic People
It is important to be clear: CBT and talking therapies can be beneficial for autistic people. The problem is not the therapies themselves, but the assumption that standard, unadapted versions work equally well.
Current research and expert consensus identify several reasons why this assumption fails.
1. Neurotypical Bias in Therapy Models
Most talking therapies rely on neurotypical norms of communication, emotional insight, and social reasoning. These expectations can clash with autistic cognitive and emotional processing styles, leading to misunderstanding or disengagement.
2. Alexithymia
Around 50% of autistic adults experience alexithymia—difficulty identifying and describing emotions. Standard therapies often assume emotional labelling as a prerequisite skill, placing autistic clients at an immediate disadvantage.
3. Misinterpretation of Autistic Behaviours
Therapists without autism training may misinterpret stimming, avoidance of social situations, or shutdowns as symptoms of OCD, anxiety disorders, or “resistance,” rather than as autistic regulatory strategies. This can lead to inappropriate formulations and ineffective or harmful interventions—a phenomenon closely linked to diagnostic overshadowing.
4. Autistic Burnout
Autistic burnout—a state of profound physical and cognitive exhaustion caused by prolonged stress and masking—is frequently mislabelled as depression. Standard therapy models rarely account for this distinct experience, resulting in treatment plans that miss the root cause.
5. Cognitive Flexibility Assumptions
CBT often relies on cognitive reframing and flexible perspective-taking. For some autistic individuals, rigid or literal thinking styles make these techniques difficult without careful adaptation.
6. Communication Barriers
Abstract language, metaphors, and open-ended questioning—common in therapy—can be confusing or overwhelming for autistic clients, particularly in moments of distress.
What the Evidence Actually Shows (2023–2026)
Recent data underscore why LeDeR’s reliance on access metrics is insufficient:
Poorer Outcomes: A 2023 University College London study found autistic adults were 25% less likely to show improvement and 34% more likely to experience deterioration following standard talking therapy compared to non-autistic peers.
Lack of Adaptation Data: In September 2025, the National Autistic Society noted that LeDeR reporting frequently fails to capture whether mental health support was adapted for autism, directly contributing to diagnostic overshadowing.
Negative Experiences: Qualitative reviews consistently show that many autistic adults describe mental health care as predominantly negative, often citing a lack of practitioner understanding and autism awareness.
None of these issues are visible in a headline “56% access” figure.
When CBT Does Work: The Role of Adaptation
The research is equally clear that adapted CBT and talking therapies can be effective for autistic people, particularly for anxiety, depression, and OCD—when delivered appropriately.
Key adaptations include:
FeatureStandard CBT (Neurotypical Model)Adapted CBT for Autistic AdultsPacingFixed session length and tempoFlexible pacing and processing timeCommunicationAbstract language and metaphorsClear, concrete, literal languageStructureManual-driven protocolsIndividualised, formulation-led approachEnvironmentSensory-neutral assumptionsExplicit consideration of sensory needs
Additional evidence-based adaptations include therapist autism training, use of visual supports, incorporating special interests, and (where appropriate) involving trusted caregivers.
The National Autistic Society and Mind have both published free professional guidance outlining these adjustments—yet LeDeR does not measure whether they are in place.
Why the 56% Figure Is a Flawed Benchmark
By using access to talking therapies as a benchmark, LeDeR implicitly assumes:
the therapy was appropriate,
the therapy was adapted,
the therapist was autism-informed,
and the outcome was beneficial.
None of these assumptions are supported by the data collected.
This is particularly concerning given that reasonable adjustments are a legal requirement under the Autism Act and NICE guidelines. A metric that fails to capture whether reasonable adjustments were made risks normalising inadequate care and obscuring preventable harm.
In effect, the statistic may mask systemic failures rather than illuminate them.
What LeDeR Needs to Change
If LeDeR is to genuinely “learn from lives and deaths,” its methodology must evolve beyond neurotypical service metrics. At a minimum, future reporting should include:
Whether therapies were adapted for autism
Whether practitioners had autism-specific training
Autistic people’s own reports of usefulness or harm
Outcome measures meaningful to autistic individuals
Access without suitability is not support. Counting attendance without understanding experience is not learning.
CBT and talking therapies are not inherently ineffective for autistic people. But standard, unadapted approaches often fail, and LeDeR’s current reporting framework does not account for this reality.
By equating access with efficacy, LeDeR risks reinforcing neurotypical bias at the level of national oversight. A programme designed to reduce premature mortality among autistic people cannot afford metrics that mistake contact for care.
True learning will require asking a harder question—not who was seen, but who was actually helped.