Summary
Alexithymia — from the Greek a (without), lexis (words), thymos (emotion) — is the difficulty identifying and describing one’s own emotions. It is not a diagnosis or a disorder in its own right; it is typically understood as a dimensional personality trait that varies in severity across the population. What makes it essential for understanding autism is its remarkably high co-occurrence rate: approximately 50% of autistic people meet clinical thresholds for alexithymia, compared to roughly 10% of the general population. This is not just a statistical curiosity. An influential and growing body of research argues that many of the emotional difficulties traditionally attributed to autism — reduced empathy, poor emotion recognition, limited emotional expression — may actually be driven by co-occurring alexithymia, not by autism itself.
Definition and measurement
Alexithymia was originally described in psychosomatic medicine by Peter Sifneos in the 1970s. Its core features include difficulty identifying feelings and distinguishing them from bodily sensations, difficulty describing feelings to others, a constricted imaginal capacity, and an externally oriented thinking style (focused on external events rather than inner experience).
The most widely used measure is the Toronto Alexithymia Scale, 20-item version (TAS-20) (Bagby, Parker & Taylor, 1994). It is a self-report questionnaire with three subscales: Difficulty Identifying Feelings (DIF), Difficulty Describing Feelings (DDF), and Externally Oriented Thinking (EOT). A total score of 61 or above is typically used as the clinical cut-off for alexithymia, with scores of 52–60 indicating possible alexithymia.
The TAS-20 has been validated across many languages and cultural contexts, with a 25-year review confirming good reliability and factorial validity (Bagby et al., 2020). However, its reliance on self-report is inherently paradoxical for a construct defined by poor self-awareness. Alternative measures include the Bermond-Vorst Alexithymia Questionnaire (BVAQ), which additionally captures the fantasy/imagination dimension.
Prevalence in autism
A systematic review and meta-analysis by Kinnaird, Stewart and Tchanturia (2019, European Psychiatry) examined 15 studies comparing TAS-20 scores between autistic and non-autistic groups (total N = 714). They found that autistic people scored significantly higher on all TAS-20 subscales, with medium to large effect sizes. The prevalence of alexithymia in the autistic samples was approximately 50% (range: 33–63%), compared to approximately 5% in the non-autistic comparison groups.
Critically, this also means that roughly half of autistic people do not have alexithymia. Alexithymia is common in autism but not universal, and it is not specific to autism — it occurs at elevated rates in many other conditions including depression, eating disorders, substance use disorders, and PTSD.
The alexithymia hypothesis
In a landmark 2013 paper, Geoffrey Bird and Richard Cook proposed what has become known as the alexithymia hypothesis: the idea that many emotional processing difficulties assumed to be intrinsic to autism are actually attributable to co-occurring alexithymia. They marshalled evidence showing that difficulties with empathy, emotion recognition, and attention to the eyes — all considered hallmark features of autism — were predicted by alexithymia scores rather than autism diagnosis when the two were statistically separated.
Subsequent research has supported and extended this argument. Cook, Brewer, Shah and Bird (2013) demonstrated that it was alexithymia, not autism, that predicted poor recognition of emotional facial expressions. Shah, Hall, Catmur and Bird (2016) showed that alexithymia, not autism, was associated with impaired interoceptive accuracy. Brewer, Cook and Bird (2016) proposed that alexithymia represents a “general deficit of interoception” — a broad difficulty in reading the body’s internal signals, of which emotional difficulty is one consequence.
The alexithymia hypothesis has significant implications for how we understand autism. If the “emotional coldness” stereotype is driven by alexithymia rather than autism per se, then many autistic people without alexithymia may have typical or even heightened emotional responsiveness — they simply experience and express emotions differently. This aligns with the growing recognition that autistic people often experience intense emotions but may have difficulty labelling or communicating them.
Connection to interoception
The link between alexithymia and interoception is theoretically coherent and empirically supported. If emotions are partially constituted by bodily sensations — the racing heart of anxiety, the gut-clench of fear, the warmth of affection — then difficulty perceiving those bodily sensations (poor interoception) would naturally produce difficulty identifying those emotions (alexithymia).
Brewer, Cook and Bird (2016) found that alexithymia was associated with reduced interoceptive accuracy across both autistic and non-autistic participants, and proposed that alexithymia should be understood as a consequence of impaired interoception. Mul et al. (2018) found that autistic adults with alexithymia showed lower interoceptive sensibility, reduced interoceptive awareness, and lower empathy than autistic adults without alexithymia — suggesting an alexithymic subgroup within autism with distinct interoceptive processing patterns.
This interoception-alexithymia connection is one of the strongest arguments for including interoception in sensory processing assessments and in the prikkelprofiel (see: building-an-individual-prikkelprofiel, interoception-in-autism).
Implications for practice
- Screen for alexithymia alongside autism. The TAS-20 is brief and freely available. Knowing whether an autistic person has co-occurring alexithymia changes the picture significantly: the person’s emotional difficulties may be amenable to interoceptive awareness training rather than (or alongside) social skills instruction.
- Don’t assume emotional absence. An autistic person with alexithymia may experience strong emotions but be unable to identify or label them. This is very different from not having emotions. Treating an autistic person as if they lack empathy because they have difficulty recognising their own feelings is both inaccurate and harmful.
- Support interoceptive awareness. If alexithymia is at least partly an interoceptive difficulty, then interventions that improve body awareness — mindfulness-based approaches, body-scanning, interoception curricula like Mahler’s (2017) — may help.
- Build emotional vocabulary collaboratively. Help the person develop their own language for internal states, which may look different from conventional emotion labels. “My chest feels tight and my stomach is churning” is a more useful starting point than “I feel anxious.”
- Account for alexithymia in sensory assessment. A person who cannot easily identify their own emotions may also have difficulty reporting on their own sensory experiences. Self-report instruments like the MAIA or the Adolescent/Adult Sensory Profile may underestimate difficulties in this group.
Open questions
- Why do autism and alexithymia co-occur at such high rates? Is there a shared neurodevelopmental pathway, or is the co-occurrence driven by environmental factors (e.g., does growing up autistic in a non-accommodating world produce alexithymia through chronic emotional suppression)?
- Can alexithymia be reduced through intervention, or is it a stable trait? Early evidence suggests that interoceptive training and some psychotherapeutic approaches can improve emotional awareness, but the evidence base is still small.
- How does alexithymia present in autistic people with intellectual disability, who are rarely included in research and for whom self-report measures are inapplicable?
- Should the diagnostic criteria for autism be revised to account for the alexithymia hypothesis? If many “autistic” emotional symptoms are actually alexithymic, this has implications for how autism is defined and assessed.
Key sources
- Bird, G. & Cook, R. (2013). Mixed emotions: the contribution of alexithymia to the emotional symptoms of autism. Translational Psychiatry, 3, e285. doi: 10.1038/tp.2013.61
- Bagby, R.M., Parker, J.D.A. & Taylor, G.J. (1994). The twenty-item Toronto Alexithymia Scale — I: Item selection and cross-validation of the factor structure. Journal of Psychosomatic Research, 38(1), 23–32.
- Kinnaird, E., Stewart, C. & Tchanturia, K. (2019). Investigating alexithymia in autism: a systematic review and meta-analysis. European Psychiatry, 55, 80–89. doi: 10.1016/j.eurpsy.2018.09.004
- Shah, P., Hall, R., Catmur, C. & Bird, G. (2016). Alexithymia, not autism, is associated with impaired interoception. Cortex, 81, 215–220. doi: 10.1016/j.cortex.2016.03.021
- Brewer, R., Cook, R. & Bird, G. (2016). Alexithymia: a general deficit of interoception. Royal Society Open Science, 3(10), 150664. doi: 10.1098/rsos.150664
- Cook, R., Brewer, R., Shah, P. & Bird, G. (2013). Alexithymia, not autism, predicts poor recognition of emotional facial expressions. Psychological Science, 24(5), 723–732. doi: 10.1177/0956797612463582
- Mul, C.-L., Stagg, S.D., Herbelin, B. & Aspell, J.E. (2018). The feeling of me feeling for you: interoception, alexithymia and empathy in autism. Journal of Autism and Developmental Disorders, 48, 2953–2967. doi: 10.1007/s10803-018-3564-3
- Mahler, K. (2017). Interoception: The Eighth Sensory System. Shawnee, KS: AAPC Publishing.
- Livingston, L.A. & Livingston, L.M. (2016). Commentary: Alexithymia, not autism, is associated with impaired interoception. Frontiers in Psychology, 7, 1103. doi: 10.3389/fpsyg.2016.01103