Attachment style

Bowlby’s (1982) proposal that infants form attachment bonds with caregivers, and that the pattern of those bonds durably shapes cognition, emotion regulation, and physiology — “from the cradle to the grave” (Bowlby 1979). The three styles relevant to interoception:

• Secure — develops when a caregiver consistently provides sensitive, attentive care. The child uses the caregiver as a “secure base,” can manage anxiety with trust, and recruits others to resolve threatening or novel events.
• Avoidant — develops when a caregiver avoids responding to immediate needs, delays comfort, or responds frighteningly or inconsistently. The individual becomes compulsively self-reliant, minimizes negative affect, and directs attention away from threat cues.
• Anxious — develops from the belief that the caregiver is available but only conditionally, and will withdraw support if standards aren’t met. The individual becomes hypervigilant to relational threat, exaggerates helplessness, and seeks excessive reassurance.

The interoceptive claim

Oldroyd et al. (2019) propose that responsivity to bodily cues mirrors responsivity to interpersonal cues — the same strategy the person applies to relationships, they apply to their own body. This predicts qualitatively distinct interoceptive profiles rather than a single good/bad axis:

• Anxious → hypervigilance. Excessive monitoring of bodily sensation for threat, and catastrophic interpretation of benign sensation. Empirically: higher MAIA Noticing and Emotional awareness, much lower Not-worrying (r = -0.43). The authors link this to the hypochondriacal/health-anxiety pattern (Salkovskis & Kobori 2015).
• Avoidant → hypoarousal and distrust. Attention diverted from bodily cues, action tendencies suppressed. Empirically: lower MAIA Attention regulation and Trusting. This is the classic psychophysiological dissociation profile — the avoidant person presents as calm while showing elevated heart rate and cortisol (Spangler & Grossmann 1993; Diamond et al. 2006), a “preemptive” disengagement from distress before negative affect is encoded (Fraley et al. 2000).
• Secure → congruence. Comfort acknowledging distress, so self-report and physiology track each other. Operationalized in Study 2 via self-report-physiology-congruence.

Note the asymmetry this creates for measurement: the anxious profile scores higher on several self-report interoception measures while arguably having worse interoceptive function. This is a concrete case of why Farb et al. insist on separating sensibility from accuracy — see maia.

Proposed mechanisms

Three, per oldroyd-2019-attachment-interoception — see that page for detail:

1. hpa-axis dysregulation altering interoceptive signal strength (via stroke volume) and signal processing (via cortisol).
2. Neural architecture — anxious/avoidant children show lower insular volume and surface area (Kühn & Gallinat 2013; Sheffield et al. 2013; Lim et al. 2014); avoidant adults show decreased insular activation (DeWall et al. 2011). Early maltreatment also alters ACC (van der Werff et al. 2013; Teicher et al. 2014) and OFC (Schore 2005). See insular-cortex, anterior-cingulate-cortex.
3. Construction of the bodily self through caregiver mirroring — see social-origins-of-interoception and embodied-selfhood.

Caveats

The evidence is thinner than the theory. The Oldroyd et al. correlations are small (|r| = 0.18-0.26, excepting Not-worrying), cross-sectional, and in Study 1 wholly self-report; the causal-developmental story is argued rather than tested. In particular, the prediction that avoidant individuals would score higher on attention regulation (a deliberate-control reading of avoidance) was contradicted by the data, which found them lower — the theory accommodates either result, which is a sign it is not yet doing much predictive work.