Author(s): CHAGNEAUD, R., BOUGARD, C., DIAZ, E., MARSAC, L., DESCHODT-ARSAC, V., Institution: IMS LABORATORY, CNRS UMR 5218, Country: FRANCE, Abstract-ID: 1337

INTRODUCTION:
Performance emerges from the interaction between cognitive and physical demands, with training and competition often imposing repeated cognitive challenges. Within the brain–heart interaction framework, Heart Rate Variability (HRV) emerges as a relevant marker of autonomic adaptations to cognitive demands, particularly through parasympathetic activity and complexity-related indices [1,2]. Time-varying HRV analyses have shown high sensitivity to cognitive load detection [3] and allow tracking autonomic dynamics over time.

This study aimed to determine whether time-varying HRV markers provide additional insights compared to phase-averaged markers when assessing autonomic adaptations during repeated cognitive challenges.


METHODS:
Twenty-nine healthy young adults (19.6 ± 2.0 years; 8 females) performed two visual cognitive tasks (COG1, COG2) interleaved with two neutral baseline periods (BASELINE1, BASELINE2: viewing a neutral documentary). HRV was continuously recorded using an H10 Polar chest belt. Time-varying analyses were conducted during COG1 and COG2 to characterize cardiac dynamics through LF and HF trends (LF_TV, HF_TV). In addition, time- (RMSSD), frequency- (LF, HF, total power, HFnu), and nonlinear-domain (multiscale entropy) HRV indices were computed as phase-averaged values for each condition.

All indices were statistically compared across phases (COGs vs. BASELINEs and COG1 vs. COG2).

To assess the sensibility of dynamic trends (LF_TV and HF_TV) to discriminate COG1 from COG2, receiver operating characteristic (ROC) analyses were performed.


RESULTS:
As the main result of the present study, LF_TV was the most sensitive index to discriminate the two successive cognitive tasks (ROC curve, AUC = 0.71; Youden index = 0.38) among all indicators. This demonstrates that analyzing temporal trends of autonomic dynamics provides added value beyond classical HRV markers for capturing adaptations during repeated cognitive engagement.
Additionally, markers of parasympathetic modulation (HFnu: χ² (3) = 18.074, p < 0.001, Kendall’s W = 0.312) and cardiac complexity (multiscale entropy: χ² (3) = 30.414, p < 0.001, Kendall’s W = 0.524) primarily discriminated COGs from BASELINEs, reflecting a global change in cognitive state.

Together, these findings highlight the complementary roles of time-varying and phase-averaged HRV indices in characterizing both dynamic adaptations and overall cognitive load.


CONCLUSION:
Time-varying HRV markers, particularly LF_TV, improve discrimination between successive cognitive engagements, highlighting their relevance for monitoring autonomic adaptations under repeated cognitive demands.

REFERENCES:
[1] Forte G et al., Appl. Cogn. Psychol., 2022
[2] Blons E et al., Entropy, 2019
[3] Boffet A et al., Sensors, 2025