Study at a glance
- Authors: Denise DuBois, Stephanie H. Ameis, Meng-Chuan Lai, Manuel F. Casanova, Pushpal Desarkar
- Year: 2016
- Journal: International Journal of Developmental Neuroscience
- DOI: 10.1016/j.ijdevneu.2016.05.001
- N: Systematic review; PRISMA search (99 records → 16 full-text → 5 original studies reporting interoception in ASD)
- Design: Systematic literature review (PubMed, CINAHL, EMBASE; Jan 1980–Dec 2015)
Key findings
- The wiki’s first dedicated review of interoception in autism — earlier (2016) and more focused than the autism paragraph the wiki carried through Quadt et al. (2018). Its verdict: between-group differences are repeatedly demonstrated but heterogeneous in degree and direction, with a slight tendency toward hyporeactivity in interoceptive awareness in ASD — a hedge, not a result, and the honest one.
- Only five original studies existed measuring interoception in ASD as of end-2015 (fMRI ×1, self-report ×1, heartbeat paradigm ×2, qualitative ×1) — the review’s most consequential datum is how thin the evidence base was. All samples were under 50, mostly male, normal-IQ, verbal; every study excluded comorbid developmental disability, though >50% of ASD individuals have one (Baio 2012), so findings do not generalize to the wider spectrum.
- Accuracy findings split by age/sample. Heartbeat-counting accuracy was intact in ASD children (Schauder et al. 2014, n=21 vs 24 — no group difference, but ASD sustained attention to heartbeat more constantly across durations) yet impaired in ASD adults (Garfinkel et al. 2016, n=18) — the same heterogeneity Quadt et al. later report.
- Sensibility findings openly conflict — the review’s most wiki-relevant tension. Garfinkel et al. (2016) found elevated interoceptive sensibility (Porges BPQ) in ASD adults; Fiene & Brownlow (2015, n=74 vs 228) found sharply decreased body/thirst awareness (BAQ d=−1.26, TAQ d=−1.02, both p<0.001); Elwin et al. (2012, qualitative autobiographies) found lifelong hyposensitivity to interoceptive cues and pain. So the ITPE’s ‘high sensibility + low accuracy’ signature is one of at least two directions the sensibility literature runs.
- Garfinkel et al. (2016), reported first-hand here: impaired interoceptive accuracy, no group difference in metacognitive awareness (heartbeat error score), elevated sensibility (BPQ) → an enlarged interoceptive trait prediction error (ITPE = objective accuracy minus subjective sensibility) that differed between groups and predicted state and trait anxiety. This is the source study behind the wiki’s ITPE material.
- Neural arm (Barttfeld et al. 2012): during an interoceptive state (focus on breathing/heart) connectivity increased in ASD vs controls (the reverse held for an exteroceptive oddball task); ASD networks were more compact within cingulo-opercular, fronto-parietal and DMN systems, and the degree of interoceptive-state hyperconnectivity correlated positively with ADOS symptom severity. Resting-state work is read as insula hypo-/hyperconnectivity (posterior→anterior insula disconnection; salience and DMN alterations), with directionality unresolved.
- Frames ASD interoception through predictive coding (Seth 2013; Seth, Suzuki & Critchley 2012): prediction errors in interoceptive systems as a candidate substrate, converging with the ‘autism as a disorder of prediction’ reading the wiki also meets on Sensory over-responsivity (SOR). Notes clinical promise (biofeedback, TMS/DBS, wearables) but no interventions yet run through interoception.
Limitations
- A review of five small studies, not new data. Its conclusions inherit every weakness of the primary literature: n<50 throughout, male-skewed, normal-IQ/verbal only, comorbid developmental disability excluded, and cardiac-dominant (both physiological studies used heartbeat paradigms) — the single-channel problem, in a population where the review itself argues evolutionarily salient cues (hunger, thirst, bladder urgency) may matter more.
- The Alexithymia confound is under-addressed here relative to Quadt et al. — the possibility that atypical interoception tracks co-occurring alexithymia rather than autism per se is not centrally examined, though it is the standard caution on this literature.
- Schauder et al. excluded children who could not detect their heartbeat (potentially the lowest-sensibility cases), and introduced a control task partway through — so the ‘intact accuracy in children’ finding is selection-limited.
- Fiene & Brownlow’s ASD sample was self-disclosed diagnosis via online survey with unclear recruitment, and the Thirst Awareness Questionnaire was still undergoing item refinement.
- Narrative synthesis of a five-study literature: no meta-analysis, no effect-size pooling; the ‘slight tendency toward hyporeactivity’ is an authorial read across incommensurable designs, not a quantitative estimate.
Connections
- Methods: Heartbeat detection task, Resting-state functional connectivity, Multidimensional Assessment of Interoceptive Awareness (MAIA)
- Concepts: Interoception in autism (ASD), Interoceptive trait prediction error (ITPE), Interoceptive psychopathology, Interoceptive taxonomy (Farb et al. Box 1), Insular cortex, Interoceptive inference, Sensory over-responsivity (SOR), Alexithymia, Salience network, Default mode network
- Debates: Is interoception domain-general or organ-specific?, Is more interoceptive awareness better?
- Related studies: Quadt et al. (2018), Seth & Friston (2016), Schwarzlose et al. (2023), Greenwood & Garfinkel (2025)
The wiki had met autism interoception only in passing — a paragraph in Quadt et al. (2018), the ASD row on Interoceptive psychopathology, and the source study behind the ITPE (Garfinkel et al. 2016) named but not read. This is the dedicated review, published two years earlier, and its main service is to show how little there was: a systematic PRISMA search of three databases across 35 years returned exactly five original studies that measured interoception in ASD. Everything the wiki says about autism and interoception rests, ultimately, on this handful.
The five studies
| study | design | n | interoceptive finding |
|---|---|---|---|
| Barttfeld et al. (2012) | fMRI (rest / oddball / interoceptive focus) | 12 hfASD, 12 TD (adults) | Interoceptive-state connectivity increased in ASD (exteroceptive task the reverse); hyperconnectivity correlated with ADOS severity |
| Garfinkel et al. (2016) | 3-dimension model; heartbeat tracking + BPQ | 18 ASD, 18 TD (male adults) | Impaired accuracy, elevated sensibility, awareness intact → enlarged ITPE predicting anxiety |
| Schauder et al. (2014) | Heartbeat counting (children) | 21 ASD, 24 TD (8–17y) | No accuracy difference; ASD sustained attention to heartbeat more constantly |
| Fiene & Brownlow (2015) | Online BAQ + Thirst Awareness Q | 74 ASD, 228 TD (adults) | Decreased body/thirst awareness (d=−1.26 / −1.02) |
| Elwin et al. (2012) | Qualitative (10 autobiographies) | 10 | Lifelong hyposensitivity to interoceptive cues and pain |
Why it matters to the wiki: the sensibility direction is genuinely contested
The wiki’s ITPE and Interoceptive psychopathology pages carry the “low accuracy + high sensibility” signature (Garfinkel’s BPQ result) as the autism interoceptive profile, with a one-line caveat that “at least one study finds decreased subjective body awareness.” This review names and quantifies that dissenting study — Fiene & Brownlow (2015), a large-effect decrease in body and thirst awareness — and adds a second (Elwin’s qualitative hyposensitivity). So the ASD sensibility literature does not point one way. Two of the three sensibility-relevant studies find reduced sensibility; the ITPE’s elevated-sensibility premise comes from one. The review’s own overall read — a “slight tendency toward hyporeactivity in interoceptive awareness” — sides with the majority, against the ITPE story it also reports. This is the tension worth carrying, and it is filed on Interoception in autism (ASD) (created with this ingest). It is not a hard contradiction with existing wiki content: the ITPE page already flags the inconsistency; this review supplies the underlying studies.
The instrument problem, sharpened by the population
Both physiological studies used heartbeat paradigms, so the wiki’s single-channel worry applies here with unusual force: the review itself argues that in ASD the evolutionarily salient interoceptive cues — hunger, thirst, urgency to void, gastrointestinal discomfort (all elevated in autism) — are where the action likely is, yet the measured literature is cardiac. Fiene & Brownlow’s own result dramatizes the point: ASD participants’ thirst awareness dissociated entirely from their fluid intake. The review’s methodological prescription is the one the whole wiki keeps arriving at — multidimensional designs combining neuroimaging, physiology and self-report, guided by an explicit model — plus, specific to autism, adapted non-verbal measures for the majority-of-the-spectrum the five studies excluded (the non-verbal measurement gap that also thins the developmental and dementia ends of this wiki).
Where it lands
- Interoception in autism (ASD) (created here) — the construct’s home page: mixed accuracy, contested sensibility, the ITPE signature and its dissent, insula connectivity, and the predictive-coding framing.
- Interoceptive trait prediction error (ITPE) — supplies the Garfinkel (2016) source study first-hand, and names the countervailing Fiene & Brownlow finding.
- Interoceptive psychopathology — the dedicated ASD review behind that page’s autism row; earlier and more focused than Quadt et al.
- Sensory over-responsivity (SOR) / Exteroception Boundary — the interoceptive half of the “atypical sensory processing in autism” story whose exteroceptive half Schwarzlose et al. carry; both invoke “autism as a disorder of prediction.”
- Insular cortex — ASD insula hypo-/hyperconnectivity (posterior→anterior disconnection; salience/DMN alterations), added as a lesion-free connectivity dissociation.
Brakes
A narrative review of five small, selected, cardiac-heavy, male, normal-IQ samples reaches a hedged conclusion by authorial judgement, not meta-analysis. Its value is not a verdict but a census: it dates and bounds the autism-interoception literature, exposes that its headline sensibility signature is one of two directions, and states plainly that the spectrum’s majority — non-verbal, developmentally disabled — has never been measured at all.