Nasal temperature drops during stress, especially social speech stress

Thermal video of the nose tracked an objective “stress dip,” and it lined up with body-type anxiety symptoms more than self-rated stress.

In healthy adults, nasal skin temperature dropped during two lab stressors and rebounded during recovery, but did not fully return to baseline in five minutes. A speech-based social stressor produced a bigger temperature drop than a mental arithmetic stressor. Self-reported perceived stress did not track these temperature changes, but self-reported somatic anxiety symptoms did.

Quick summary

What the study found: Continuous infrared thermal imaging showed a significant “nasal dip” (lower nose temperature) during stress tasks versus baseline; the dip was larger during a speech task than during arithmetic; perceived somatic anxiety correlated with the size of the speech-task dip.
Why it matters: A non-contact, objective signal like nose temperature could help monitor stress responses when self-report is limited or biased, including in clinical and nonverbal populations.
What to be careful about: The sample was small; recovery was variable and not clearly linked to self-reports; nasal temperature can be influenced by factors beyond stress (including breathing and ambient conditions).

What was found

The journal article [The nose knows: Thermal responses to active psychological stressors] tested whether nose skin temperature changes track psychological stress. The researchers used continuous infrared thermal imaging (IRT), meaning a thermal camera measured skin temperature without touching the body.

Across 29 participants, nasal temperature significantly decreased during both stress tasks compared with a white noise baseline. The speech task (a social stressor) produced a larger drop than the arithmetic task (a cognitive stressor), and the minimum temperature during speech was significantly lower than the minimum during arithmetic.

After the tasks, temperature rose during a five-minute recovery period, ending higher than the task minima but still significantly lower than baseline. Recovery varied a lot: some people did not return to baseline within five minutes, while others went above baseline.

The study also included a “no-stress” manipulation: white noise. Nasal temperature increased significantly from initial temperature after room habituation (acclimatising to the room) to the maximum baseline temperature during the white noise period, supporting the idea that the nose shows bidirectional change across lower-stress versus higher-stress periods.

In a subset of 16 participants with extremely high frequency coding (more granular time sampling), a larger temperature drop was associated with a longer time to reach the minimum temperature. In plain terms: bigger “nose dips” tended to unfold more slowly, regardless of whether the person did speech or arithmetic first.

What it means

The simplest takeaway is practical: nasal skin temperature, measured continuously and without contact, behaved like a sensitive physiological marker of an acute stress response. The study links stress induction to a measurable cooling of the nose tip, which then partially reverses during recovery.

The authors interpret the temperature drop as consistent with sympathetic vasoconstriction, a stress-related narrowing of blood vessels driven by the sympathetic nervous system (the “fight-or-flight” branch), which can reduce blood flow near the skin and lower skin temperature. That is a plausible mechanism for why a stressed person’s nose tip might cool.

The difference between speech and arithmetic also matters. Even when participants sometimes rated arithmetic as more stressful, the speech task still produced a larger nasal temperature decrease. That pattern suggests nasal temperature may be more sensitive to certain classes of stress—especially socially evaluative stress—than to perceived stress ratings alone.

The study’s psychological pattern was sharp: perceived stress measured by the Perceived Stress Scale (PSS) did not relate to nasal temperature changes, but anxiety measured by the State-Trait Inventory for Cognitive and Somatic Anxiety (STICSA) did. Specifically, STICSA total score and the somatic subscore (physical symptoms) correlated with the temperature drop during the speech task, while cognitive anxiety (worry and thoughts) was not highlighted as related in the excerpts.

Where it fits

This work sits in the broader push for objective stress measurement that can complement (not replace) self-report. Self-report is valuable, but it can be limited by memory, social desirability, lack of insight into bodily cues, or inability to communicate (for example, in infants, some disabled populations, or high-stakes environments where people underreport stress).

Thermal imaging has an appealing promise here because it can be continuous and non-contact. “Non-contact” matters for ecological validity, meaning how well a measurement approach can work in more natural, real-world conditions without strapping sensors onto someone, which can be uncomfortable, stigmatizing, or simply impractical.

The study also indirectly underscores a common issue in stress research: “stress” is not one thing. Some measures reflect chronic load (ongoing strain), while others reflect acute reactivity (moment-to-moment response). The Perceived Stress Scale is framed here as more about chronic coping and may be less sensitive to short acute laboratory stressors, while STICSA is described as better at differentiating temporary and chronic anxiety.

Finally, the recovery results fit a broader idea used in physiology: recovery dynamics can be as important as peak reactivity. The authors considered thermal recovery rate as potentially similar in spirit to heart rate variability (variation between heartbeats, often used as an index of autonomic flexibility), but in this study thermal recovery was not associated with the self-report measures and was highly variable.

How to use it

If you work in mental health, coaching, education, or human performance, treat this as a proof-of-concept signal: a person’s nose temperature may drop during acute stress, and speech-based social evaluation may provoke a stronger response than cognitive load alone. The practical implication is not “use nose temperature to diagnose stress,” but “consider objective physiological channels when subjective ratings and observed behavior disagree.”

A grounded way to apply the idea is as a trend monitor within an individual. For example, if someone is practicing exposure to public speaking, you could track whether their physiological reactivity (the size of the nasal dip) decreases over sessions, while also tracking their subjective experience. That dual-track approach can reveal different targets: body reactivity versus cognitive appraisal.

Another use-case is settings where speech is limited or reporting is unreliable. The authors point to translation to clinical and nonverbal populations across the lifespan. The key is that thermal imaging does not require the person to answer questions, and it does not require physical contact that can itself raise stress.

To keep expectations realistic, any real-world use needs strict control or adjustment for basic confounds highlighted in the study. Ambient temperature and age were associated with nasal thermal measures, and breathing can affect facial temperature variability. In practice, that means you would need consistent environmental conditions or calibration methods before drawing conclusions from changes over time.

Limits & what we still don’t know

The participant group was relatively small (29 total, with 16 in the high-frequency timing analysis). Even with adequate statistical power for large effects, small samples can produce unstable estimates and may not generalize well to other ages, cultures, health statuses, or real-world stressors.

The method is also labor-intensive as described. Without automated region-of-interest tracking (software that follows the same facial location across frames), the approach relies on manual identification of facial landmarks frame-to-frame, which is time-consuming and can be vulnerable to motion artifacts (measurement errors when the face moves).

Temperature changes at the nose are not uniquely caused by stress. The authors note external and physiological factors that can influence infrared thermal imaging nasal changes, including breathing, and they also flag that thermal imaging may not yet distinguish affective valence—whether the arousal is experienced as positive or negative. In plain language: a “nose dip” may reflect arousal or threat response, but not necessarily whether the person interprets it as good stress (challenge) or bad stress (threat).

Recovery remains an open question. Although temperatures rose during recovery, they stayed below baseline on average after five minutes, and recovery rate did not associate with the stress and anxiety self-reports used here. That means we cannot yet treat “faster thermal recovery” as a validated resilience marker based on this study alone.

Closing takeaway

In this study, the nose behaved like a readable stress sensor: temperature fell during stress tasks, fell more during a social speech stressor than during arithmetic, and rose again during recovery without consistently returning to baseline in five minutes. The strongest psychological link was not “perceived stress,” but somatic anxiety—the physical sensations of anxiety—which tracked a larger nasal temperature drop during the speech task. Thermal imaging is promising as an objective, non-contact window into stress reactivity, but real-world deployment will require careful control for environment, movement, and other drivers of facial temperature.

Data in this article is provided by PLOS.

Tagged Anatomy, Anxiety, Arithmetic, Body temperature, Classical mechanics, Emotions, Face, Head, Mathematics, Mechanical stress, Mental health and psychiatry, Nose, Physics, Physiological parameters, Psychological stress, Signal processing, Skin temperature, Thermal stresses, White noise