Alexithymia Trait

Alexithymia is a personality trait characterized by difficulty identifying and describing one’s own emotions, difficulty distinguishing between emotional feelings and physical sensations of emotional arousal, and an externally oriented thinking style focused on concrete details rather than internal experience. The term was introduced by Peter Sifneos in 1972 and developed primarily through the empirical work of Graeme Taylor, Michael Bagby, and James Parker. It is a dimensional trait, not a categorical diagnosis: people score along a continuum rather than having or not having alexithymia.

The three-component structure operationalized in the Toronto Alexithymia Scale (TAS-20) is the dominant contemporary framework: difficulty identifying feelings (DIF), difficulty describing feelings (DDF), and externally oriented thinking (EOT). Population prevalence of clinically significant alexithymia (typically defined as TAS-20 scores at or above 61) is approximately 10% in general adult samples, with higher rates in clinical populations including autism spectrum disorder (approximately 50%), depression (approximately 30%), and eating disorders.

The contemporary picture treats alexithymia as a well-supported construct with substantial empirical literature, while recognizing that it overlaps with several adjacent phenomena and that its causal status (predisposing factor, consequence of trauma, comorbid feature, or independent trait) varies by context. Kinnaird, Stewart and Tchanturia (2019) meta-analyzed alexithymia in autism across 49 studies (N = 4,346) and confirmed roughly 50% prevalence in autistic populations compared to 10% in non-autistic comparison groups. Luminet, Bagby and Taylor (2018) reviewed the broader construct and its measurement issues. The trait is robust; its specific etiology in any individual case is more uncertain than popular framings suggest.

Alexithymia matters at three distinct levels.

For self-understanding. Many adults with high alexithymia scores describe a lifetime of confusion about emotional experience — not knowing whether they are anxious or hungry, struggling to answer questions about how a situation made them feel, finding therapy frustrating because the questions assume access to inner emotional information they cannot easily produce. The construct gives a name to this experience and connects it to a research literature that explains why standard emotional-introspection approaches do not work the same way for everyone.

For autism and ADHD diagnosis. The overlap between alexithymia and autism is substantial enough that the “alexithymia hypothesis” (Bird & Cook 2013) proposes that several of the emotional-processing differences attributed to autism may actually reflect co-occurring alexithymia rather than autism itself. The hypothesis remains debated but has shaped how researchers and clinicians think about emotional differences in autism. The construct is similarly relevant for ADHD, where high-trait alexithymia is more common than in the general population and may contribute to emotional dysregulation patterns.

For therapy and intervention. Alexithymia predicts poorer response to insight-oriented psychotherapies that depend on identifying and articulating emotional experience. People with high alexithymia scores tend to benefit more from skills-based approaches, body-based and interoception-focused interventions, and structured emotion-recognition training. Recognizing alexithymia at the outset of therapy can shape treatment selection and reduce the experience of being asked questions a person genuinely cannot answer.

The term “alexithymia” was coined by Boston-based psychiatrist Peter Sifneos in 1972, from the Greek roots a- (without), lexis (word), and thymos (emotion) — literally “no words for feelings.” Sifneos was working with patients who had psychosomatic conditions and noticed a recurring pattern: difficulty identifying emotional states, externally oriented thinking, and a particular style of communication that focused on the concrete details of situations rather than emotional or interpersonal content. The pattern did not fit cleanly into existing diagnostic categories and seemed clinically meaningful enough to warrant a name.

The construct was developed empirically over the following decades, primarily through the work of a group of researchers at the University of Toronto. Graeme Taylor, Michael Bagby, and James Parker developed the Toronto Alexithymia Scale through successive iterations: the TAS (1985), TAS-Revised (1992), and TAS-20 (Bagby, Parker & Taylor 1994). The TAS-20 became the dominant measurement instrument and has been translated into more than 30 languages with consistent factor structure across most translations.

Two parallel developments have shaped contemporary research. The Bermond-Vorst Alexithymia Questionnaire (BVAQ, Vorst & Bermond 2001) added two dimensions to the three-component TAS-20 framework: fantasizing (the capacity for imaginative thought about emotional content) and emotionalizing (the magnitude of emotional response to events). The BVAQ five-component model has theoretical appeal but has not displaced the TAS-20 in most empirical work.

The neuroimaging and biological work has been substantial. Neural correlates of alexithymia have been identified in the anterior cingulate cortex, insula, and amygdala, with reduced activation during emotional processing tasks. Bird, Silani et al. (2010) showed that participants with high alexithymia (regardless of autism status) had reduced empathy-related activation in the anterior insula, supporting the hypothesis that some emotional-processing differences attributed to autism may actually reflect alexithymia. The same group's broader work (Bird & Cook 2013) formalized the alexithymia hypothesis as an alternative or complementary account of autistic emotional processing.

The contemporary state: alexithymia is a well-supported dimensional construct, the TAS-20 is the dominant measurement tool with adequate psychometric properties, and the construct is empirically distinguishable from depression, anxiety, and autism while also showing substantial comorbidity with all three. The causal pathways — whether alexithymia is trait-like and stable, state-related and context-dependent, or both depending on the individual — remain an active research question.

The construct is best understood through its three (or five) component dimensions and the cognitive processes they reflect.

1. Difficulty identifying feelings (DIF). The first TAS-20 dimension. A person high on DIF has trouble distinguishing one emotional state from another and from physical sensations. Asked “how are you feeling?” they may respond with physical descriptions (“tired”, “tense”) rather than emotional labels, or with broad uncertainty (“I don’t know”). This is the most-studied dimension and the one most consistently associated with the broader clinical picture.
2. Difficulty describing feelings (DDF). The second TAS-20 dimension. Even when emotional states are identified, finding words for them is difficult. The dimension is conceptually distinct from DIF: a person can have access to an emotional experience without having the vocabulary to articulate it precisely. DDF is partly cultural and developmental as well as trait-like.
3. Externally oriented thinking (EOT). The third TAS-20 dimension. A cognitive style focused on concrete external details rather than internal emotional or interpersonal content. The dimension has been the most controversial of the three: factor-analytic studies sometimes find EOT items load less consistently with the other two dimensions, and some researchers argue EOT measures a cognitive-style trait that is partly independent of alexithymia proper.
4. Fantasizing (BVAQ only). The capacity for imaginative thought about emotional and interpersonal content. Low fantasizing in the BVAQ framework is part of the alexithymic profile but is treated as an independent dimension rather than a feature of EOT.
5. Emotionalizing (BVAQ only). The magnitude of emotional response to events. The dimension is conceptually distinct: a person can have intense emotional responses they cannot identify or describe (high emotionalizing, high DIF and DDF) or muted emotional responses they can identify and describe (low emotionalizing, low DIF and DDF). The BVAQ framework distinguishes these patterns; the TAS-20 framework does not.

Two broader cognitive processes are proposed to underlie the components. The first is interoception: the perception of internal bodily states, including emotional arousal. Reduced interoceptive accuracy has been documented in high-alexithymia samples and is one of the candidate mechanisms linking alexithymia to physical-symptom focus and emotional-recognition difficulty. The second is emotional granularity: the cognitive resolution at which a person distinguishes between emotional states (e.g., distinguishing anxiety from frustration from disappointment rather than experiencing all as “bad”). Low granularity is associated with high alexithymia and has independent associations with clinical outcomes.

The current best understanding is that alexithymia is not a single cognitive deficit but a clustering of related differences in interoception, emotional vocabulary, granularity, and cognitive style that tend to co-occur and that produce the characteristic clinical picture.

Alexithymia is measured primarily through self-report instruments, with adequate psychometric support.

Toronto Alexithymia Scale (TAS-20). The dominant instrument since the early 1990s. Twenty self-report items rated on a 5-point Likert scale, producing a total score and three subscale scores (DIF, DDF, EOT). The conventional cutoffs are: ≤51 non-alexithymic, 52–60 possible alexithymia, ≥61 alexithymic. Internal consistency (Cronbach’s alpha) is typically 0.75–0.85 in adult samples. Test-retest reliability is moderate. The factor structure replicates well across most translations, though the EOT subscale shows weaker psychometric properties than DIF and DDF.

Bermond-Vorst Alexithymia Questionnaire (BVAQ). A 40-item alternative with five subscales (the three TAS-20 dimensions plus fantasizing and emotionalizing). More comprehensive but less widely used. Better suited for research designs that need to distinguish the cognitive components of alexithymia from the affective response component.

Toronto Structured Interview for Alexithymia (TSIA). A clinician-administered interview developed to address concerns that self-report alexithymia measures may underestimate the trait in people who do not recognize their own difficulty. The interview format produces moderately higher scores than self-report in some samples. Less common in research but used in clinical settings where self-report may not capture the trait reliably.

Behavioral and physiological measures. Various tasks attempt to measure alexithymia through performance rather than self-report: emotion-recognition tasks (Levels of Emotional Awareness Scale, LEAS), interoception tasks (heartbeat detection paradigms), and physiological response measures. These avoid the self-report paradox (people with poor introspective access may underreport) but have their own limitations and modest correlations with self-report instruments.

What the LBL Adult Autism Self-Inventory accounts for. The LBL-AAS does not include a dedicated alexithymia subscale but its emotion-recognition and inner-experience items capture closely related dimensions. Because alexithymia is approximately 50% prevalent in autistic adults compared to 10% in non-autistic comparison groups, the inventory’s treatment of emotional self-recognition reflects this base-rate difference. For users specifically interested in alexithymia measurement, the published TAS-20 remains the standard instrument; the inventory captures the broader pattern relevant to autism screening without claiming to be a dedicated alexithymia measure.

The construct sits in a neighborhood of related phenomena that are frequently conflated.

• vs. autism spectrum disorder. Alexithymia is a dimensional personality trait; autism is a developmental neurotype with characteristic features across social, communication, sensory, and repetitive-behavior domains. The two are empirically distinguishable and have substantial but incomplete overlap. The alexithymia hypothesis (Bird & Cook 2013) proposes that several emotional-processing differences attributed to autism may actually reflect co-occurring alexithymia. Important practical implication: not all autistic adults have alexithymia, and not all people with alexithymia are autistic.
• vs. AuDHD and ADHD. High alexithymia scores are more common in adults with ADHD than in general-population comparisons, and the highest prevalence is in adults with co-occurring autism and ADHD (AuDHD). The construct may contribute to the emotional-regulation difficulties commonly reported in adult ADHD presentations, though the directionality is contested (does ADHD cause alexithymia features, or are they parallel manifestations of a broader neurodevelopmental pattern?).
• vs. emotional avoidance. Emotional avoidance is the active or learned tendency to suppress, distract from, or escape emotional experience. Alexithymia is the underlying difficulty in identifying and describing emotion even when not actively avoiding it. The two can co-occur (a person with alexithymia may also develop avoidance strategies) but are distinct: avoidance can be reduced through exposure-based therapy in ways that alexithymia trait scores typically cannot.
• vs. interoception. Interoception is the perception of internal bodily states; alexithymia includes interoceptive accuracy as one component but is broader. Reduced interoceptive accuracy contributes to alexithymia but does not exhaust it. A person can have adequate interoception and still struggle with the emotional-labeling and externally oriented thinking dimensions.
• vs. depression. Approximately 30% of people with major depression score above clinical alexithymia cutoffs, more than three times the general-population rate. The relationship is bidirectional and contested: alexithymia may be a state feature that increases during depressive episodes, a trait that increases vulnerability to depression, or both depending on the individual. The TAS-20 has some sensitivity to depressive states; this is a known measurement limitation.
• vs. trauma response and dissociation. Both can produce alexithymia-like presentations through different mechanisms. Trauma-related emotional numbing and dissociative experiences can present as difficulty identifying and describing feelings, particularly in PTSD. Distinguishing trait alexithymia from trauma-related dissociation matters clinically because trauma-focused interventions can reduce trauma-related presentations in ways they typically cannot reduce stable trait alexithymia.
• vs. mentalizing deficits. Mentalizing is the capacity to understand others’ mental states. Alexithymia involves difficulty with one’s own emotional states. The two are related but separable: a person can have intact mentalizing for others while struggling with self-recognition (high alexithymia, adequate mentalizing) or vice versa.

Example 1 — The therapy session

A 33-year-old in his fourth therapy session is asked how a recent work conflict made him feel. He pauses for several seconds, then says he is not sure. His therapist asks him to try a single word. He says “tired.” The therapist asks whether he means physically tired or emotionally tired. He says he cannot tell the difference. The session continues; he is able to describe the work conflict in detail, including specific things his colleague said and the order of events, but cannot produce an emotional account of it that satisfies the therapist’s questions.

This is a recognisable pattern in alexithymic presentations. The person is not being evasive or resistant to therapy; he genuinely does not have the introspective access to emotional content that the therapeutic frame assumes. Standard insight-oriented therapy can be slow and frustrating for both parties under these conditions. Approaches that work better include body-based interventions, emotion-recognition skill training, and frameworks that do not require the client to identify their emotions before discussing them.

Example 2 — The medical appointment

A 47-year-old visits her doctor for fatigue and stomach pain that have persisted for several weeks. The doctor asks about stress in her life and whether she has been feeling anxious or depressed. She says her life is fine and she is not stressed. The doctor orders tests. The tests come back normal. The fatigue and stomach pain continue. Six months later her sister mentions that the patient has been at home almost continuously since a difficult work transition the previous summer.

This is the classic somatic presentation associated with alexithymia, and the reason Sifneos developed the construct from psychosomatic-medicine work. The patient is not in denial about her circumstances; she does not have ready introspective access to the emotional dimension of what is happening. Her body is registering the impact in physical symptoms her conscious self-report cannot connect to emotional cause. The medical evaluation, focused on physical pathology, finds nothing and recommends no further investigation. The pattern is common enough that some primary-care training now includes screening for alexithymia and externally oriented thinking patterns as a complement to standard depression and anxiety screening.

The construct is well-supported, but several caveats are routinely missed in popular accounts.

• The self-report paradox. Self-report measures of alexithymia depend on the respondent recognising their own difficulty with emotional identification and description. People with severe alexithymia may underreport precisely because they have limited introspective access to their own emotional difficulty. The Toronto Structured Interview for Alexithymia (TSIA) was developed in part to address this; the gap between self-report and clinician-rated alexithymia is small but real.
• State versus trait. Alexithymia scores are partly stable across time and partly responsive to current mood, particularly depression. The construct is treated as a trait in most research, but state-related variation is substantial enough that single-time-point measurement may overestimate trait alexithymia in people currently depressed.
• The EOT subscale is the weakest. Factor-analytic studies consistently find the externally oriented thinking subscale of the TAS-20 has weaker psychometric properties than DIF and DDF. Some translations have particular problems with EOT item loading. The dimension is theoretically meaningful but measured less reliably than the other two.
• Causal direction is often unclear. In any individual case, alexithymia features may be: stable from early development (trait alexithymia), produced or amplified by depression or trauma (secondary alexithymia), associated with a neurodevelopmental difference like autism (co-occurring alexithymia), or partly reflective of cultural and developmental differences in emotional vocabulary. The category labels are useful but the individual mechanism often requires careful clinical assessment.
• Cultural variation. Emotional vocabulary, the cultural acceptability of discussing emotions, and the expected level of emotional articulation vary substantially across cultures. The TAS-20 has been translated into more than 30 languages with reasonable factor structure replication, but cross-cultural comparison of absolute scores is more complex than within-culture comparison. What looks like alexithymia in one cultural context may partly reflect cultural norms about emotional expression.
• The construct is not a deficit framing. Alexithymia is a dimensional trait; high scores indicate difficulty with a particular kind of introspective and communicative task, not a fundamental impairment in being a person. The clinical relevance comes from the consequences of the difficulty in specific contexts (therapy, medical communication, intimate relationships), not from the trait itself being pathological. This distinction matters for how clinicians and the people described by the construct relate to it.

If you suspect alexithymia may be part of a broader pattern that includes autism features, the LBL Adult Autism Self-Inventory screens for the broader autism profile and includes emotion-recognition and inner-experience items that reflect the elevated alexithymia base rate in autistic adults. The inventory does not claim to be a dedicated alexithymia measure; for that purpose the validated Toronto Alexithymia Scale (TAS-20) remains the standard instrument. The two assessments complement each other for adults exploring whether alexithymia features are part of a broader neurodevelopmental picture.

Alexithymia is a personality trait characterized by difficulty identifying and describing one’s own emotions, difficulty distinguishing emotional feelings from physical sensations, and an externally oriented thinking style. Introduced by Peter Sifneos in 1972 and developed empirically through the Toronto Alexithymia Scale (TAS-20; Bagby, Parker & Taylor 1994), it is a dimensional trait rather than a categorical diagnosis. Population prevalence of clinically significant alexithymia is approximately 10% in general adult samples, with roughly 50% prevalence in autism (Kinnaird, Stewart & Tchanturia 2019), 30% in major depression, and elevated rates in eating disorders and ADHD. The alexithymia hypothesis (Bird & Cook 2013) proposes that several emotional-processing differences attributed to autism may reflect co-occurring alexithymia rather than autism itself. The construct is well-supported and clinically meaningful; the causal pathways in any individual case (trait, state, comorbid feature, or trauma-related) require careful assessment. Recognition has practical implications for therapy selection: insight-oriented approaches that assume access to emotional introspection are slower and more frustrating for people with high alexithymia scores, who tend to benefit more from skills-based, body-based, and structured emotion-recognition interventions. The LBL Adult Autism Self-Inventory includes emotion-recognition items reflecting the elevated base rate of alexithymia in autistic adults without claiming to be a dedicated alexithymia measure; the published TAS-20 remains the standard instrument for direct alexithymia assessment.

This entry is intended as a citable scholarly reference. Choose the format that matches your context. The retrieval date should reflect when you accessed the page, which may differ from the entry's last-reviewed date shown above.

LifeByLogic. (2026). Alexithymia: Sifneos, TAS-20, and Autism Overlap. https://lifebylogic.com/glossary/alexithymia/

LifeByLogic. "Alexithymia: Sifneos, TAS-20, and Autism Overlap." LifeByLogic, 13 May 2026, https://lifebylogic.com/glossary/alexithymia/.

LifeByLogic. 2026. "Alexithymia: Sifneos, TAS-20, and Autism Overlap." May 13. https://lifebylogic.com/glossary/alexithymia/.

@misc{lblalexithymia2026,
  author = {{LifeByLogic}},
  title = {Alexithymia: Sifneos, TAS-20, and Autism Overlap},
  year = {2026},
  month = {may},
  publisher = {LifeByLogic},
  url = {https://lifebylogic.com/glossary/alexithymia/},
  note = {Accessed: 2026-05-13}
}