Page 116 - AIH-2-4

Artificial Intelligence in Health Autonomic nervous system patterns in men

Cluster 2 (intermediate autonomic profile; n = 28) represents However, several limitations should be acknowledged
a moderate autonomic state, with HRV values significantly in this study. A key limitation is the homogeneous nature
lower than those of cluster 1 but higher than those of cluster of the sample, which consisted of 80 healthy young men
3. Cluster 1 (low-vagal-tone or sympathetic predominant from Macapá, Brazil. Consequently, the findings should
profile; n = 19) exhibits the lowest HRV values of be interpreted as proof of concept within this specific
parasympathetic activity and the highest relative LF power. demographic. The study emphasizes that the identified
While these individuals are clinically healthy, this profile cluster patterns may not be directly generalizable to women,
may indicate a subclinical state of autonomic imbalance or other age groups, ethnic backgrounds, or populations with
reduced adaptive capacity, potentially indicating elevated different health and lifestyle characteristics. Therefore,
long-term cardiovascular risk. validating the method is a priority, and future studies are
The identification of these three distinct profiles in a recommended to include larger, more diverse, and multi-
young, healthy population represents a central finding of center cohorts to determine the broader applicability of
this study. While the literature establishes that reduced these autonomic profiles. Furthermore, the cross-sectional
HRV is associated with disease, the present findings suggest design of the study limits the ability to draw conclusions
that a spectrum of autonomic function exists even in the about the temporal stability of these profiles or their
absence of clinical pathology. This observation aligns with predictive value for future health outcomes.
previous research, including that by Pasquini et al., who
54
applied similar techniques to identify autonomic states 6. Conclusion
during emotional reactivity. The present study further This study highlights that the application of combined
highlights interindividual variability at rest, suggesting PCA and K-means clustering to HRV data is a valid
that these autonomic profiles may represent individual and effective approach for identifying distinct patterns
physiological traits. of autonomic regulation in healthy men. Three distinct
The use of machine learning, particularly K-means physiological profiles, high vagal tone, intermediate vagal
clustering, underscores the potential of AI to enhance HRV tone, and low vagal tone, were identified, even within a
analysis and deepen understanding of ANS regulation. demographically homogeneous cohort. These findings
This data-driven approach enables the identification of reinforce the importance of HRV as a sensitive biomarker
distinct autonomic patterns that may be obscured by of cardiovascular health and underscore the potential of
traditional statistical methods. The combined PCA and machine learning techniques to advance personalized
K-means clustering method serves as an effective tool preventive strategies. However, future longitudinal studies
for exploring complex physiological data and identifying are needed to examine the stability of these autonomic
meaningful subgroups within a population. Furthermore, profiles and their association with long-term health
the application of machine learning techniques to predict outcomes across more diverse populations.
individual risk based on HRV profiles holds significant
promise for enabling more personalized and effective Acknowledgments
preventive strategies. 4,5,11-13
The author would like to thank the Physical Education
It is crucial to interpret these findings within the Department at the Federal University of Amapá for
context of the study’s design. Unlike studies that aim to allowing them to use their laboratory facilities and for
distinguish between healthy individuals and those with helping them to recruit participants.
cardiac disease, a supervised classification problem this
study addresses an unsupervised discovery question: Funding
what distinct autonomic profiles exist within a healthy
population? This approach is based on the hypothesis This research was funded by the Amapá Research
Support Foundation through its public call 003/2018,
that physiologically meaningful patterns may reflect
distinct long-term cardiovascular health trajectories, specifically within the “Research Program for the Unified
even among asymptomatic individuals, before the onset Health System (SUS): management in Health-PPSUS.”
of clinical disease. Therefore, the identification of three The funder had no role in study design, data collection
distinct clusters should not be viewed as a limitation due and analysis, decision to publish, or preparation of the
to the absence of a patient control group, but rather as a manuscript.
validation of the study’s primary objective, which explores Conflict of interest
an under-investigated area of autonomic profiling in
healthy populations. The author declares no conflicts of interest.

Volume 2 Issue 4 (2025) 110 doi: 10.36922/AIH025050006