Wearable sensors for monitoring the physiological and biochemical profile of the athlete

Dhruv R. Seshadri, Ryan T. Li, James E. Voos, James Rowbottom, Celeste M. Alfes, Christian A. Zorman, Colin K. DrummondNpj Digit. Medicine 2019 · DOI

TL;DR

This review explores the burgeoning field of wearable sensors for athletes, focusing on their potential to non-invasively monitor physiological and biochemical markers from sweat and saliva. The paper highlights the current state of technology, challenges, and future opportunities for these devices to revolutionize athlete performance monitoring, recovery, and injury prevention.

What Problem Does This Paper Address?

The paper addresses the critical need for more comprehensive and real-time monitoring of athlete health and performance beyond traditional metrics like movement and heart rate. While wearable technologies have advanced significantly in tracking external workloads, there's a gap in understanding the athlete's internal physiological and biochemical state. This gap hinders personalized recovery protocols, hydration strategies, and the early detection of fatigue and stress, ultimately impacting an athlete's competitive edge and long-term health. The authors emphasize that current methods often rely on invasive blood draws or are limited in their ability to provide continuous, actionable data.

Methods

The paper reviews a range of wearable sensor technologies and approaches, categorized by the biofluid they analyze (saliva or sweat) and the parameters they measure. Key methods discussed include:

  • Electrochemical Sensors: These are central to many wearable devices, leveraging techniques like amperometry and potentiometry to detect specific analytes. Examples include glucose oxidase for glucose detection and lactate oxidase for lactate measurement.
  • Microelectromechanical Systems (MEMS): MEMS technology is employed for miniaturizing sensor components and creating microfluidic channels for sweat collection and analysis.
  • Flexible and Stretchable Electronics: The development of these materials is crucial for creating comfortable, conformable sensors that can be integrated into clothing or worn directly on the skin without impeding movement.
  • Biosensors: These devices utilize biological recognition elements (e.g., enzymes, antibodies) coupled with transducers to detect specific biomarkers.
  • Physiological Signal Monitoring: This includes established methods like electrocardiography (ECG) for heart rate variability (HRV) and electrodermal activity (EDA) or galvanic skin response (GSR) for skin conductivity.
  • Iontophoresis: This technique is discussed as a method to artificially induce sweat, overcoming limitations of low sweat rates or skin contamination.
  • Machine Learning and Data Analytics: The paper highlights the increasing importance of these tools for processing the vast amounts of data generated by wearable sensors, translating raw signals into actionable insights for coaches and medical staff.

Datasets

Not discussed in the paper. The paper focuses on a review of existing technologies and literature rather than presenting new experimental data or introducing novel datasets.

Key Insights and Ideas

  • The "Internal Workload" Gap: The most significant insight is the unmet need to quantify an athlete's internal physiological and biochemical state, moving beyond external metrics. This internal view is crucial for personalized recovery and injury prevention.
  • Sweat as a Rich Biofluid: Eccrine sweat is identified as a highly promising source for non-invasive, continuous monitoring of a wide array of biomarkers, including electrolytes, metabolites, and even neuropeptides.
  • Saliva's Potential and Limitations: Saliva offers an attractive, non-invasive alternative, but its utility is currently limited to sports requiring mouthguards, and analyte concentrations are generally lower than in sweat.
  • The Rise of "Sports Scientists": The increasing employment of sports scientists by teams underscores the need for expertise in interpreting complex wearable sensor data and translating it into practical applications.
  • Multimodal Sensing is Key: The future lies in integrating various sensor types (biochemical, physiological, motion) into a single wearable platform to provide a holistic view of an athlete's status.
  • The Challenge of Translational Utility: A recurring theme is the gap between technological development and the practical application of data. "What do we do with the data?" remains a critical question.

Key Results

The paper does not present specific quantitative results from new experiments. Instead, it summarizes findings from existing literature, highlighting:

  • Successful Detection of Biomarkers: Numerous studies have demonstrated the feasibility of detecting biomarkers like lactate, glucose, sodium, potassium, cortisol, and orexin A from sweat and saliva using wearable sensors.
  • Correlation with Traditional Methods: Many of these wearable sensors show good correlation with established laboratory methods for biomarker analysis.
  • Feasibility of Continuous Monitoring: Advances in flexible electronics and microfluidics are enabling continuous, real-time monitoring of certain parameters.
  • Potential for Injury Prevention: The ability to monitor hydration status (e.g., sodium levels) and fatigue markers (e.g., lactate) is seen as a direct pathway to reducing soft-tissue injuries.
  • Stress Monitoring Capabilities: Wearable sensors can monitor physiological indicators of stress like HRV and skin conductivity, offering insights into an athlete's mental and physiological state.

Limitations and Future Work

The authors acknowledge several limitations and point towards future research directions:

  • Clinical Validation: A significant unmet need is the clinical validation of many developed wearable devices in real-world, non-stationary athletic conditions.
  • Sensitivity and Selectivity: Improving the sensitivity and selectivity of sensors, especially for low-concentration biomarkers, is crucial.
  • Power Consumption and Miniaturization: Reducing device power consumption and further miniaturizing components are essential for practical, long-term wearability.
  • Data Analytics and Actionable Insights: Developing robust data analytics platforms and machine learning models is critical to translate raw sensor data into actionable protocols for athletes and medical staff.
  • Differentiation of Stressors: There is a need to differentiate between physical and mental stress to provide a more nuanced understanding of an athlete's well-being.
  • Cost and Scalability: Scaling up sensor production through methods like roll-to-roll manufacturing is necessary to reduce costs and increase accessibility.
  • Long-Term Stability and Durability: Ensuring the long-term stability and durability of sensors, especially under the demanding conditions of athletic activity, remains a challenge.
  • Integration into Garments: Further development is needed for seamless integration of sensors into athletic apparel.

Summary updated

BibTeX
@article{seshadri2019wearable,
  author = {Dhruv R. Seshadri and Ryan T. Li and James E. Voos and James Rowbottom and Celeste M. Alfes and Christian A. Zorman and Colin K. Drummond},
  title = {{Wearable sensors for monitoring the physiological and biochemical profile of the athlete}},
  journal = {Npj Digit. Medicine},
  year = {2019},
  doi = {10.1038/s41746-019-0150-9},
}
Figure