Aging bodies, aging emotions (MacCormack, Henry, Davis, Oosterwijk & Lindquist 2021)
The wiki’s first source on aging and the lifespan, and the primary study behind the age-related-interoceptive-decline page that had been carrying the finding as a stub. Two online studies from Lindquist’s lab, led by jennifer-maccormack, asking a question the body-emotion literature had almost entirely left to young-adult samples: does the association between emotion and the felt body change across adulthood? Their answer, from two methods and two samples, is yes — and specifically for high-arousal interoceptive sensations, which weaken from mid-life onward.
What it actually measures — and does not
This matters enough to state before the findings, because the word “interoception” in the title is doing lighter work than it looks.
- Study 1 is a property-association-task: participants rate how strongly a property (e.g. “heart racing,” “clenched fists,” “insulted”) comes to mind when they think of an emotion category. The interoceptive items measure conceptual knowledge about which bodily sensations go with which emotions — what MacCormack et al. call “interoceptive knowledge,” an idiographic/cultural store, not a perceptual ability.
- Study 2 is the Day Reconstruction Method (a low-burden experience-sampling variant): participants reconstruct the prior day into episodes and rate the self-reported intensity of sensations, behaviors, situations, and emotional/physical/cognitive states.
Neither study measures interoceptive accuracy (no heartbeat-detection-task) or physiology (no autonomic recording). So the paper is about the represented and felt body in emotion across age, and the inference to biological decline — peripheral demyelination, reduced autonomic reactivity, poorer objective interoception — is imported by citation (Khalsa, Rudrauf & Tranel 2009; Murphy et al. 2018; Verdú et al. 2000), not demonstrated here. This is the same limit the age-related-interoceptive-decline page already flags for Volynets: what changes with age is the felt intensity, and the mechanism is proposed, not measured. The two papers are methodologically complementary — Volynets colours a felt-body map, MacCormack rates emotion-property associations and daily-life sensation intensity — and reach the same shape of conclusion.
The core result: interoceptive-specific, high-arousal, and curved
Study 1’s design isolates interoception by comparison. Age moved the ratings of interoceptive properties (up to ~45, then down) and left behavioral and situational properties flat. That specificity is the paper’s strongest card against the obvious alternative — that older adults simply think about emotion differently in general, or remember worse. If it were general cognitive or semantic aging, situations and behaviors should have moved too; they did not. And situations, which even young children understand, are exactly what a “first in, last out” cognitive-aging account predicts would be preserved.
Within interoception, the effect was carried by high-arousal sensations (“blood pumping,” “racing heart”) — the ones tied to autonomic reactivity — and absent for low-arousal ones (“drained”). This arousal-specificity is the thread that ties the aging finding to core-affect: it is the high-arousal, bodily pole of affect that fades, not the low-arousal pole.
The Study 2 twist: linear, not curved — and the decoupling
Study 2 did not reproduce the inverted-U. Instead age linearly predicted less intense high-arousal interoceptive experience in daily life. The authors blame the missing older adults (only ~12% over 50) for flattening the late-life turn, and note the aging literature is itself split on linear vs curvilinear age effects. The wiki should hold the curvilinear claim loosely for that reason: it appears in the sample with better old-age coverage (Study 1) and vanishes in the one without (Study 2), which is as consistent with a sampling artefact as with a real difference between representation and experience.
The novel Study 2 result is structural. A network analysis of how modalities co-occur found that with age, interoceptive sensations grew less coupled to emotional situations — but stayed coupled to behaviors. Read through the authors’ lens (situations ≈ appraisals of the event), older adults increasingly experience emotion via situational meaning rather than felt body, but still feel the body when they are overtly acting. That is a specific, testable shape for “maturational-dualism,” and it converges on the wiki’s appraisal thread: the older adult leans on the situational/appraised side of emotion as the interoceptive side quiets.
Valence, and the positivity effect
Study 2 also replicated the well-known positivity effect — with age, more intense positive and less intense negative emotion — and added the arousal cut: less high-arousal and more low-arousal emotion. So the aging shift is jointly along both dimensions of core-affect: down on arousal, up on positive valence. MacCormack et al.’s proposal is that these are not independent facts — if fewer, weaker high-arousal bodily sensations accompany emotion, then unpleasant, highly activated emotions become both less frequent and easier to regulate, which is one mechanistic route to the positivity effect that motivational accounts (socioemotional selectivity) and expertise accounts do not require. This is offered as complementary to those accounts, not a replacement.
Why it is not a hard contradiction, and what it adds
Nothing here overturns existing wiki content. It is consonant with the constructionist thread (theory-of-constructed-emotion, core-affect, allostasis) that senior author Lindquist helped build, and it corroborates the incidental age effect the wiki first met in Volynets et al. (2020) with a study designed to find it, across two methods. What it adds beyond Volynets:
- A purpose-built continuous-age design rather than an effect that fell out of a universality paper.
- Arousal-specificity — the decline is in the high-arousal body, which localizes it on core-affect’s arousal dimension.
- A structural claim — interoception decouples from situations but not behaviors — that gives maturational dualism a concrete prediction.
- A bridge to the decision literature. The discussion notes the paradox that older adults regulate emotion better yet make worse affect-based decisions (trusting scammers, poor financial choices; Damasio 1994; Clithero & Rangel 2014). If the same fading interoceptive input eases regulation and starves somatic-marker guidance, one mechanism covers both — a genuine, if speculative, link to does-somatic-feedback-guide-decisions and the accuracy↔decision-quality thread.
Where it sits in the queue
First file of the topical Aging/ folder, and the first of the wiki’s aging sources. Its own citation of Khalsa, Rudrauf & Tranel (2009), “Interoceptive awareness declines with age” was the next file in Aging/ and is now ingested — the objective heartbeat-task counterpart to this paper’s self-report/conceptual evidence, and the source MacCormack et al. lean on for the claim that interoceptive ability really does fall with age. It supplies exactly what this paper’s design cannot: an objective accuracy decline (age = 30% of variance on a heartbeat-discrimination task), so the felt/represented decline measured here now sits beside first-hand evidence that the ability declines too. What Khalsa still cannot supply — and neither can this paper — is the afferent-physiology mechanism both sets of authors invoke: his objective decline is confoundable with reduced cardiac signal strength, so the biological chain stays proposed rather than instrumented. The two studies are the accuracy and self-report legs of one convergent claim, and share one gap.