Development and validation of wearable devices for real-time sweat biochemical sensing

Abstract

Wearable devices have made significant advances in the real-time, continuous monitoring of human physiology, offering higher sampling frequencies for rapidly changing physiological signals than currently achievable through conventional clinical methods. However, nearly all commercially available wearable devices are limited to measuring biophysical parameters, such as heart rate and activity, using electrical, mechanical, or optical sensors. In contrast, medical diagnostics for small-molecule biomarkers typically rely on invasive blood sampling and instrumentation that is costly, complex, and unsuitable for point-of-care or wearable use. Electrochemical biosensors have emerged as a powerful transduction modality for portable biochemical sensing, exemplified by the clinical success of continuous glucose monitors. Yet, there remains a critical need for noninvasive wearable biochemical sensing platforms. Sweat is a promising biofluid in this context, containing a wide array of physiologically relevant biomarkers including ions, amino acids, metabolites, hormones, peptides, and proteins. Sweat can also be accessed noninvasively and stimulated locally on demand via iontophoresis. This dissertation presents the design, development, and on-body validation of wearable sweat-sensing devices. Two distinct device architectures are demonstrated, reflecting different levels of complexity and design considerations. We report real-time, on-body lactate sensing during stationary cycling exercise, as well as successful detection of nicotine in sweat using a novel microbial redox enzyme-based biosensor. Sweat was locally induced for nicotine sensing, and study participants included heavy and light nicotine users as well as non-users. The wearable nicotine sensor exhibited superior analytical accuracy compared to the gold standard of mass spectrometry. Furthermore, we demonstrate multiplexed sensing targeting both nicotine and cortisol in sweat using a single wearable platform. Finally, we explore the use of electrochemical impedance spectroscopy as a system identification tool for the development and optimization of biosensors.