Sweat Collection and Sensing for Non-Invasive Health Monitoring

Sweat is increasingly recognized as a promising biofluid for health monitoring. Its main advantage lies in the fact that it can be accessed non-invasively and, in principle, continuously or semi-continuously. This unique accessibility sets it apart from established diagnostic fluids such as blood or urine and creates new opportunities for monitoring outside traditional clinical settings. While sweat has often been associated with sports and hydration, its potential value for clinical and medical applications is far greater. Non-invasive, personalized, and even home-based health monitoring could become feasible if sweat can be reliably sampled and measured. At present, correlations between sweat and systemic biomarkers are not yet complete, and sensing platforms remain under development. Nevertheless, the central message of this thesis is that sweat should not be overlooked: understanding its composition, variability, and measurement challenges is an important step toward future diagnostic use. The thesis begins with an overview of electrochemical sensing strategies for sweat analysis. Electrochemical methods are highly relevant for wearable applications because of their sensitivity, low power requirements, and compatibility with miniaturized and flexible platforms. Different transduction principles, including potentiometry, amperometry, voltammetry, impedimetry, and transistor-based approaches, offer diverse opportunities depending on the target analyte. Despite these advantages, challenges such as sensor stability, selectivity, and reproducibility in the complex sweat environment remain significant. Addressing these challenges is crucial if sweat is to move from an experimental medium to one that contributes meaningfully to clinical practice. As a model system, a solid-contact sodium-selective electrode is developed and validated. Sodium is one of the most abundant sweat ions and provides clinically relevant information on hydration and electrolyte balance. The fabricated sensors demonstrate a wide linear range, high selectivity against interfering ions, and stable performance across physiologically relevant concentrations. Validation against atomic absorption spectroscopy confirms their reliability for small-volume measurements. These findings demonstrate the feasibility of electrochemical sensing in sweat and highlight the potential of translating this principle to other analytes of medical relevance. The thesis then explores the feasibility of immune-related protein analysis in sweat. Using a multiplex immunoassay, a panel of 22 inflammatory markers is studied in sweat from healthy volunteers. Several proteins, including IL-1α, IL-1β, IL-1RA, IL8, IL-12p70, IL-18, MCP-1, MIF, EGF, VEGF, ACE, and CRP, are detected. These results indicate that sweat can provide information about local inflammation and, potentially, systemic immune activity. Establishing sweat as a source of inflammatory markers is a first step toward non-invasive monitoring of immune-related diseases, which would represent a significant advance over current invasive methods. Recognizing that interpretation of sweat data strongly depends on sampling, this thesis also examines the influence of sweat gland variability. A modeling framework is introduced to describe the statistical distribution of active sweat glands and its implications for sampling strategies. Because accurate interpretation is not possible until sampling strategies are properly established, the ability to measure sweat rate locally or even at the level of individual glands is emphasized as an important step toward reliable analysis. Such models provide a foundation for understanding how gland activity affects sweat composition and help guide experimental design and sensor integration. They are particularly important for bridging controlled laboratory analysis with future on-body applications. Finally, a proof-of-concept platform for controlled sweat collection and transport using electrowetting-on-dielectric (EWOD) technology is presented. The device manipulates droplets on patterned electrodes, enabling compartmentalization and minimizing mixing. A compartmentalized design with an array of electrodes ensures robust operation during stimulated sweating, while integrated capacitive sensing provides feedback on droplet motion and volume, which can be used to estimate sweat rate. This demonstrates the feasibility of precisely collecting and analyzing sweat in a way that can be scaled toward wearable formats. Together, these studies underline why investigating sweat is important. Reliable sensing platforms, biological validation, and improved collection methods are all required to move sweat monitoring beyond sports applications and toward meaningful clinical use. While sweat cannot yet replace established diagnostic fluids, its accessibility and non-invasive nature make it a compelling target for future diagnostics. By advancing our understanding of sweat and demonstrating practical sensing and collection strategies, this thesis contributes to laying the foundation for sweat to become a clinically relevant biofluid in personalized, preventive, and home-based healthcare.