Method at a glance
- Modality: autonomic stress induction (not an interoceptive measure)
- Measures: Nothing about interoception directly — it is a standardized acute-stress induction. Used here as an independent variable to perturb autonomic and (in the socially-evaluated variant) HPA state, so its downstream effect on interoceptive tasks can be measured.
- Procedure: The participant immerses one hand (to the wrist) in near-freezing water (0–3 °C) for a fixed interval, typically ~3 minutes; a warm-water bath (~32–35 °C) is the matched control. The socially-evaluated variant (SECPT) adds a same-sex experimenter and a (fake) video camera during immersion, which reliably recruits the HPA axis on top of the autonomic response. Cardiovascular measures and, where HPA activation is of interest, salivary cortisol are sampled before and after.
- Also called: cold pressor stress test, socially-evaluated cold pressor, CP test, SECPT
Strengths
- Reliably raises heart rate and blood pressure (sympathetic activation) and, in the SECPT form, salivary cortisol (HPA activation) — a genuine two-system stress response, not just arousal.
- Highly standardized and time-locked: the stress window is short and its timing is fixed across participants, which the TSST and public-speaking stressors cannot match (Schulz et al. 2013).
- Simple, cheap, and does not require deception beyond the fake camera.
Limitations
- Cold-specific side effects confound any downstream cardiac measure. The cold induces vascular constriction, venous pooling (which itself may amplify cardiac sensations) and pain — so an effect on a heartbeat task after a CP test may be driven by cold or pain physiology rather than by stress per se (see Schulz et al. (2013)).
- It raises the cardiac signal’s amplitude (contractility, stroke volume), which is precisely the cardiodynamic confound that makes heartbeat-perception scores rise for non-perceptual reasons — so it is a treacherous stressor to pair with a cardiac interoception task.
- Pain is time-limited to the immersion and declines quickly afterward, so tasks run after the CP are less pain-contaminated than tasks run during it — but the boundary is not clean.
- Involves a pain/aversion component that is itself interoceptive (Nociception), so ‘stress’ and ‘nociceptive interoception’ are entangled in the manipulation.
Connections
- Key studies: Schulz et al. (2013)
- Related methods: Affective picture viewing, Acoustic startle response
An acute-stress induction, not an interoceptive measurement — catalogued here because the wiki’s stress-and-interoception studies use it as their independent variable, and because its side effects bear directly on how their results should be read.
What it does
Immersing a hand in near-freezing water for a few minutes produces a rapid, reproducible autonomic stress response: heart rate and blood pressure climb within the immersion window and recover over the following minutes. The socially-evaluated variant (SECPT — an observer plus a camera during immersion) adds a psychosocial-evaluative component that recruits the HPA axis, so cortisol rises too. This makes it a fuller model of stress than pure arousal manipulations (caffeine, exercise) while being far more standardized in timing than the Trier Social Stress Test.
Why it is dangerous to pair with cardiac interoception
The very thing that makes it a good stressor — sympathetic drive raising cardiac contractility — also makes the heartbeat a louder mechanical signal, and a louder heart is easier to detect for reasons that have nothing to do with perceptual ability. In Schulz et al. (2013) the CP raised heart rate +8.2 bpm and systolic pressure +20.6 mmHg, and heartbeat-counting accuracy rose in lockstep — a result the authors read (and flag as unproven, because stroke volume went unmeasured) as the cardiodynamic confound rather than sharpened interoception. Add the cold-specific vascular effects and the nociceptive load, and any post-CP change on a cardiac task has several candidate causes the manipulation cannot separate. The lesson generalizes: a stressor that moves the signal is a poor way to test whether stress moves the perceiver.