Separating Acute Psychological Stress from Physical Exertion in Biometric Signals Journal: International Journal of Psychophysiology
Esther Bosch · 2026
Read Separating Acute Psychological Stress from Physical Exertion in Biometric Signals Journal: International Journal of Psychophysiology by Esther Bosch (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-15756.txt.
Introduction
Esther Bosch splits acute psychological stress from exertion in signals. Five signals were tracked in 19 people under cognitive stress and physical activity, alone and mixed. Stress came from an n-back task with pressure and cash reward. Tasks were sitting, walking, and cycling. Tonic electrodermal activity rose with both stress and exertion, with no interaction, making it the top pick for stress spotting. Heart rate and neck muscle signals tracked exertion almost alone, with no stress sign. RMSSD and breathing rate tracked activity, with little sign of load or stress.
Catalog record
| CREATOR | Esther Bosch |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45764 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.15756
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Separating Acute Psychological Stress from Physical Exertion in Biometric Signals Journal: International Journal of Psychophysiology Esther Bosch Abstract Acute psychological stress occurs in a wide range of everyday contexts, including transportation, occupational settings, and physical activity, where its reliable detection could enable adaptive system responses and support human well-being. Stress produces characteristic changes in autonomic and cardiovascular physiology, yet a persistent challenge in its automated recognition is disentangling the biometric characteristics of acute psychological stress from those of concurrent physical exertion. This study examined how five physiological signals (tonic electrodermal activity, left trapezius electromyography, heart rate, heart rate variability, and respiration rate) respond to cognitive stress and physical activity, both independently and in combination. Nineteen participants completed a 2 × 3 within-subjects design in which acute psychological stress was induced via an n-back arithmetic task combined with social pressure and financial reward across three activity conditions: idle sitting, walking, and stationary cycling. Multilevel linear mixed models and repeated-measures analysis of variance were used to decompose main effects and interactions for each sensor. Tonic electrodermal activity showed a robust, additive response to both cognitive stress (r = 0.48) and physical exertion (r = 0.67), with no interaction between the two, making it the most promising candidate for stress detection during physical activity. By contrast, heart rate and trapezius electromyography were driven almost exclusively by physical exertion, with no reliable sensitivity to the n-back task. Root Mean Square of Successive Differences of heart beats was strongly suppressed by activity and showed only marginal sensitivity to cognitive load. Respiration rate was dominated by physical activity; a marginal stress effect found in a subsidiary complete-cases analysis did not replicate in the primary multilevel model on the full sample. These findings provide a sensor-specific hierarchy for real-world stress detection and highlight tonic electrodermal activity as the most informative channel when cognitive stress must be identified in physically active populations. † † highlights: Tonic electrodermal activity (EDA) showed robust sensitivity to acute psychological stress across all physical activity conditions, making it the most suitable sensor for cognitive stress detection during movement. Heart rate and trapezius muscle activity (RMSA) were driven almost exclusively by physical exertion, with no reliable response to the stress task. Respiration rate was dominated by physical activity; a marginal stress effect found in a subsidiary complete-cases analysis did not replicate in the primary multilevel model, and respiration rate is therefore not considered a reliable stress indicator in the present data. Heart rate variability (RMSSD) was strongly suppressed by physical activity, largely masking any cognitive stress signal and limiting its utility in ambulatory stress detection. In real-world settings, decomposing the EDA signal into its stress- and movement-related components, for instance via concurrent accelerometry, is an important methodological challenge for field deployment. 1 Introduction As modern life increasingly exposes individuals to overlapping cognitive and physical demands, the reliable detection of psychological stress in naturalistic settings has become a central challenge in affective computing ( 39 ; 14 ) . Stress detection is inherently interdisciplinary, drawing on signal processing, machine learning, psychology, and physiology, yet the field lacks consensus on how to disentangle acute psychological stress (APS) from the physiological noise introduced by concurrent physical exertion. The difficulty is compounded in applied settings where cognitive stress and physical exertion co-occur, as is routinely the case in mobility, manual labour or emergency response work. In the field of human factors in mobility, an unresolved issue remains the difficulty of classifying affective state from biometric signals alone ( 3 ) . Given a transport participant and their elevated biometric signals, separating psychological arousal from physical exertion proves to be difficult. Real time stress classification systems consider this, but findings are limited ( 27 ) . Understanding the interplay between physical exertion, acute psychological stress, and sensor configuration would advance real-time stress classification and detection algorithms. This paper aims to evaluate the utility of varying sensors in ambulatory stress experiments. Affective state will be better identifiable due to characteristics produced via physical exertion. At an interdisciplinary level, this study will aim to decompose the interplay between physical exertion and acute psychological stress in biometric signals. 2 Background 2.1 Physiology and Sensor Configuration for Affective Stress Recognition The heart is simultaneously regulated by the sympathetic nervous system (SNS) and the parasympathetic nervous system (PNS). The heart rate (HR) at any given time is identified by the overall effect of vagus nerves (PNS) which slow the rate and sympathetic nerves which accelerate it ( 8 ; 40 ) . Increased physical exertion induces sympathetic nervous system arousal to meet oxygen demand to allow for microvascular vasodilation, thereby raising HR. Similarly, psychological stress activates the sympathetic-adrenal-medullary axis, distributing catecholamines, resulting in increased activity upon HR ( 5 ) . In both instances, increased activity in autonomic neural pathways, blood pressure and respiratory control systems create notable shifts in heart rate variability (HRV) that are of interest when distinguishing APS from physical exertion ( 25 ) . The Root Mean Square of Successive interbeat interval Differences (RMSSD) is a feature that indicates PNS and SNS modulation. RMSSD is the primary time-domain measure used to estimate parasympathetic modulation of heart rate variability ( 40 ) and is rescaled, but otherwise identical to the non-linear metric SD1, which is typically used when reflecting variability in more precise time intervals ( 40 ) . Due to the increased oxygen demand in the stress response, there is a subsequent demand upon the respiratory system. The combination of respiration rate with heart activity strengthens classification of stress detection as shown by 45 . Electromyography (EMG) is a electrodiagnostic sensor that detects changes in the current of musculoskeletal functioning. The trapezius muscle has heightened myoelectrical activity in both cognitive activity and physical stress ( 1 ) . Thus, utilising an EMG on the trapezius muscle is a common technique in stress-based experiments ( 1 ) . It has further been discovered that the trapezius muscle activation is of particular responsiveness in individuals sensitive to anticipatory stressors; typically consisting of those with exposure to early life trauma ( 28 ) . The Root Mean Square Amplitude (RMSA) is a common signal characteristic that measures musculoskeletal stress ( 44 ) . Although intended to detect myoelectric activity of the trapezius muscle, RMSA is extraneously impacted by peripheral properties of surrounding muscles and privy to disruption via the sensor configuration ( 22 ) . The Electrodermal Activity (EDA) is a measure of arousal formed by conductivity of the skin. Under psychophysiological arousal, skin conductance increases. Electrodermal activity is categorised into slow changing features that are unaffected by stimuli (known as tonic features), and fast event related changes (known as phasic features). Increased psychological arousal is associated with elevated skin conductance level, suggesting that conditions inducing APS would elevate skin conductance responses ( 4 ) . The tonic and phasic components together are important features that have a proven ability to differentiate cognitive and physical exertion ( 12 ) . 19 found that especially the tonic features of EDA correlate with acute psychological stress. Consistent with this, 2 showed that tonic EDA increases stepwise with exercise intensity. Palmar eccrine sweat glands are primarily activated by sympathetic arousal in response to psychological stress (