Progress and Strategies of Wearable Sensors in Human Health Monitoring Based on Physiological, Biochemical, and Environmental Signals

Abstract

Flexible wearable sensors, emerging as pivotal tools for health diagnosis and treatment, have garnered significant attention in the present. They facilitate real-time monitoring of human physiological, biochemical, and environmental parameters, demonstrating considerable potential for personalized health monitoring. However, existing literature reviews on wearable sensors predominantly focus on single-signal monitoring. In this paper, the progress of flexible wearable sensors in health monitoring is systematically reviewed, encompassing physiological, biochemical, and environmental signals relevant to human health. Simultaneously, this review provides a comprehensive overview of the design and sensing principles of wearable sensors as well as the strategies for improving their performance. Finally, the development prospects are also discussed, aiming to offer references for the advancement of wearable sensors in health monitoring.

1. Introduction

Wearable sensors are generally defined as miniaturized sensing devices that can be attached to the surface of human skin, or integrated into daily items, such as clothing and accessories. − Specifically, these devices are engineered for continuous collection and real-time monitoring of data pertaining to users’ physiological states, activity patterns, and ambient environmental information. − By supporting continuous, noninvasive health monitoring during daily activities, wearable sensors enable a transformative shift from reactive to proactive healthcare paradigms. In traditional healthcare, medical intervention usually occurs after symptoms appear or the condition worsens significantly. Conversely, wearable devices facilitate persistent data acquisition in health monitoring, thereby equipping healthcare professionals with contemporaneous insights into individual physiological states. The continuous data flow of wearable devices promotes early detection of health problems, enabling timely intervention and prevention of complications before they occur. Such a proactive strategy not only enhances clinical outcomes but also empowers individuals to manage their health proactively through data-driven decision. For disease diagnosis and treatment, wearable sensors can provide a comprehensive understanding of physiological and pathological indicators by monitoring common disease markers. Meanwhile, they can capture symptomatic alterations via dynamically tracking evolving physiological parameters throughout treatment courses, enabling effective disease state surveillance and treatment efficacy evaluation. ,

Currently, wearable sensors have achieved substantial progress in physiological signals monitoring including respiratory rate, blood oxygen saturation, , electrocardiogram (ECG), electroencephalogram (EEG), heart rate, blood pressure, , and other supplementary physiological indicators. There are also some research on human metabolic indicators such as glucose, lactate, electrolytes and other biomarkers information. The development of wearable sensors contributes to effective health monitoring and is expected to be a key component of healthcare delivery in the future. In this article, the development trends of flexible wearable sensors over the past decade are investigated by the aid of statistical and visualization analysis tools from the literature. A systematic bibliometric analysis was performed on publications retrieved from the Web of Science Core Collection (January 2016-June 2025) using the search terms “wearable sensors” and “health monitoring”, with a focus on publication metrics and keyword clustering. As depicted in Figure a, the search results show that the data records of wearable sensors exceed 8000, with annual publications increasing steadily and exceeding 1000 in the last five years. Obviously, wearable sensors have emerged as a prominent research focus, as evidenced by the surge in academic attentions. Their noninvasive monitoring capabilities and broad prospects have spurred substantial research engagement in wearable devices. Analysis of the research keywords in Figure b reveals that most studies focus on enhancing the functionality and performance of wearable sensors, indicating that performance research and functional development are key concerns for contemporary researchers. To further elucidate research trends in wearable sensors over the past five years, Figure c presents a key network distribution diagram (2020–2025). Evidently, during the initial research phase (prior to 2020), the focus primarily centered on acquiring single physical signals, with representative keywords including “gait”, “pressure”, and “strain sensors”. Over the past three years, driven by advances in two-dimensional (2D) materials (e.g., “graphene” and “MXene”) and biosensing technologies (e.g.,“biosensors”), the research focus has simultaneously shifted toward biochemical sensing studies. This shift is reflected in the increasing emphasis on biochemical markers (e.g.,“glucose”) and their correlation with health diagnostic indicators. Concurrently, the research intensity in biosensing detection has significantly increased, indicating that wearable sensor research is evolving from focusing on “physical signals” toward integrated development that fuses “physical, chemical, and biological signals”. In recent years, with the rapid advancement of artificial intelligence, terms such as “deep learning” and “machine learning” have frequently appeared in network diagrams. Simultaneously, they have increasingly been associated with performance metrics like “performance”, “sensitivity”, and “accuracy”. This trend clearly demonstrates that performance optimization and accuracy remain the main demands for wearable sensors. Research in wearable sensors continues to advance multidimensional health monitoring and precision medicine.

1.

Development and research trends in wearable sensors: (a) statistical analysis of published articles on wearable sensors (January 2016 to June 2025); (b) keyword distribution statistics from research articles on wearable sensors (January 2016 to June 2025); and (c) the keyword research trend of wearable sensors (January 2020 to June 2025).

At present, there have been many excellent reviews on wearable sensors, including the research progress of functional materials, sensing mechanisms, and applications. These reviews mostly cover the analysis of advanced wearable materials (graphene materials, gels, , nanomaterials, etc.), as well as the summary report of physical and chemical signals detection (such as electrophysiology, strain, sweat, ,, respiratory gas, etc.). However, most of the existing reviews are limited to single-mode signal analysis, lacking comprehensive exploration of physiological, biochemical, and environmental signals. Consequently, this review systematically addresses physiological, biochemical, and environmental signal sensing pertinent to human health (Scheme ), detailing recent advancements in wearable sensors for health monitoring. It also evaluates the sensing mechanisms underlying diverse sensors and corresponding strategies on materials and structures design. Ultimately, this paper aims to provide a comprehensive and in-depth reference for the development of wearable sensors in health monitoring.

1. Schematic Diagram of Wearable Sensors: Application in Physiological/Biochemical/Environmental Sensing and Corresponding Material-Structure Design Strategies.

2. Physiological Signals

Physiological signals encompass the characteristics of human physiological structures and functional activities. Continuous monitoring of physiological signals via wearable devices is crucial for promoting health management and preventive medicine. According to their generation mechanism, these signals can be categorized into three types: electrophysiological signals (such as action potentials from electrically active tissues, such as neurons and muscle cells), biomechanical signals (generated by musculoskeletal kinematics or external forces), and supplementary biophysical indicators. This section summarizes common physiological signals, such as electrophysiological signals, , mechanical motion signals, − and temperature − signals. The strategies and performance of wearable sensors for physiological signals monitoring were summarized in Table .

1. Construction Strategies and Performance of Wearable Physiological Sensors.

signals	substrates	sensitive materials	sensitivity	detection range	detection limits	refs
EMG	melamine resin sponge	PEDOT:PSS/PVA/EG	9.82 dB	/	/
EMG	HA-PBA/TA hydrogel	Ti3C2T x MXene/HA-PBA/TA	35.63 dB	/	/
ECG	gelatin/PCA-Na biogel	gelatin/PCA-Na	24.5 dB	/	/
strain	elastic textile	MXene/PDMS	GF = 18.0; TCR = 1.8%/°C	0–45%	/
strain	fibers	SBS/MXene	GF = 1.63 × 104	0.5–350%	0.5%
strain	hydrogel	MXene/P(AA-SMA)	GF = 2.2(<400%); GF = 1.78 (>400%)	1–1500%	1%
strain	TPU nanofiber films	MXene nanosheets/AgNWs	GF = 117 (0–30%), GF = 1117 (30–80%), GF = 4720 (80–120%)	0–120%	∼0.0645%
strain	gelatin matrix	MXene nanosheets	GF = 2.80 (0–100%), GF = 5.66 (100–300%)	0–300%	/
temperature	gelatin matrix	MXene nanosheets	TCR = 1.543%·°C1– (23–60 °C); 0.367%·°C1– (60–100 °C)	23–100 °C	/
temperature	TPU fiber	PEDOT	0.95%·°C1–	20–40 °C	0.2 °C
temperature	PDMS	PNIPAM/PEDOT:PSS/CNT	2.6%·°C1–	25–40 °C	0.5 °C
strain/temperature	polystyrene elastomer	poly(AMPS-co-AAm)/ rGO/thermochromic elastomer	GF = 15.1; 0–55 °C	0–300%; 0–55 °C	/
strain/temperature	PVA/PVP hydrogel	P(NIPAAm-co-VBIMBr)	GF = 2.6; TCR = −1.41%/°C	0–930%; /	/

2.1. Electrophysiological Signals

Common electrophysiological signals include electrocardiogram (ECG), electromyogram (EMG), electroencephalogram (EEG), and electrooculogram (EOG). They capture distinct physiological processes through wearable sensors. ECG records cardiac electrical activity during cardiac cycles, providing critical insights into cardiovascular health. Continuous long-term ECG monitoring enables the early diagnosis and management of cardiovascular pathologies. EMG detects the electrical signals generated by skeletal muscle contraction, facilitating neuromuscular assessment, and is particularly valuable for stroke and Parkinson’s disease patients. Surface-mounted electrophysiological sensors acquire these signals from facial, limb, and truncal muscle groups. EEG originates from postsynaptic potential fluctuations in the apical dendrites of cortical pyramidal cells. The signal amplitude is correlated with the neural activation intensity, permitting the detection of stimulus-response dynamics and pathological states (e.g., epileptiform activity). EOG measures the corneoretinal potential during ocular movement, which is generated by ion transfer across the retinal pigment epithelium. This pattern assists in diagnosing retinal disorders and sleep pathology, including narcolepsy.

In order to improve the sensitivity and stability of electrophysiological sensors, various design strategies have been explored in terms of materials and structures. For instance, highly conductive materials, including metal-based compounds, graphene and conductive polymers, have been engineered to improve detection sensitivity. As illustrated in Figure a, Ying et al.reported a biodegradable biogel with excellent ionic conductivity. The introduction of sodium pyrrolidone carboxylic acid into gelatin hydrogel significantly enhances the mechanical strength and conductivity, resulting in higher ionic conductivity and better water retention capacity. It can serve as an effective interface for a high-quality and long-term electrophysiological signal recording. Two-dimensional materials, including graphene and MXene, have emerged as preferred candidates for their ultrahigh specific surface area and exceptional electrical conductivity. As depicted in Figure b, Jeong et al. developed a fabric-based lamina-emergent MXene electrode for wearable biosensing. To further enhance the mechanical and oxidation resistance properties of MXene, cellulose nanofibers and poly­(carboxylate ether) were added. By introducing this composite material, MXene-based layered electrodes can be seamlessly integrated into daily clothing, enabling noninvasive and sensitive monitoring of ECG and EMG signals.

2.

Wearable electrophysiological sensors: (a) schematic illustration of a portable EEG acquisition head ring for wireless transmission, alongside EEG artifact signals induced by blinking as collected via electrodes coated with the biogels. Reproduced with permission from ref Copyright 2024 John Wiley and Sons. (b) Fabric-based, lamina-emergent MXene-integrated wearable bioelectrode designed for stable skin-electrode contact. Reproduced with permission from ref Copyright 2024 Springer Nature. (c) Smart, disposable, and flexible electronic tapes for long-term biopotential monitoring. Reproduced with permission from ref Copyright 2023 Springer Nature.

During long-term and continuous monitoring, electrophysiological electrodes may undergo degradation, resulting in poor signal quality. Therefore, some elastic or adhesive materials are often introduced to enhance the adhesion to the skin. For instance, Wang et al. developed a 10-μm-thick, polyurethane nanomesh-reinforced, gas-permeable hydrogel sensor that can self-adhere to human skin for continuous electrophysiological monitoring over 8 days under daily living conditions. Specifically, the electrospun nanofiber nets were immersed in a dilute gelatin hydrogel solution at an elevated temperature, and ultrathin hydrogel membranes were then fabricated after gelation. These nanofiber nets significantly enhanced the mechanical durability of the ultrathin hydrogel, while its ultrathin geometry and porous structure endow it with air permeability. Owing to the strong and reversible chemical and physical bonds formed at the hydrogel-skin interface, coupled with the ultrathin properties of the film, hydrogel sensors achieve robust adhesion and long-term stability. Besides, Huang et al. proposed an intelligent electronic tape for collecting bioelectrical potentials (Figure c). Due to the presence of dry single-sided tape containing carbon nanotubes, ECG and EMG can be collected with stability and durability.

Additionally, advanced manufacturing techniques can be employed to optimize electrode structures and personalized design. Ameri et al. proposed an ultrathin freestanding hydrogel electronic tattoo sensor with a parylene-hydrogel bilayer structure. By means of traditional tattoo paper, this sensor can be applied to the skin like a temporary tattoo. Its ultrathin structure and strong adhesion endow it with excellent performance, reliability, high fidelity, and long-term reusability. Qin et al. developed a flexible nanosilver electrode array using the multimaterial electric field-driven microjet three-dimensional (3D) printing technology. When this array is integrated with a self-developed biosensing and regulation platform, it can reliably collect high-quality electrophysiological signals. The nanosilver electrode array with flexible property significantly enhances cell-electrode coupling by facilitating a tight wrap of the cardiomyocyte membrane around it. This innovative approach results in superior signal quality when compared to traditional planar electrodes, highlighting the potential of nanosilver electrode array in improving the performance and sensitivity of bioelectronic interfaces. It is worth mentioning that some personalized wearable sensors can also be developed with 3D printing technology. By adjusting the size of the sugar grains to control porosity and pore size, different porous structures can be prepared to meet the manufacturing needs of personalized flexible sensors. Owing to the 3D microcellular network structure interconnected inside conductive materials, the biosensors exhibited low density and significant flexibility, thus being able to successfully detect human electrophysiological signals, including EMG and EEG.

2.2. Mechanical Strain Signals

Strain sensors can convert mechanical deformations into quantifiable electrical signals (e.g., resistance, capacitance, current). , These devices are capable of monitoring large-scale bodily movements, including finger flexion, wrist articulation, and knee motion, as well as subtle physiological activities, such as arterial pulsation, respiration, phonation, facial expressions, and pulse waves, through deformation-based sensing mechanisms.

Generally, flexible polymer substrates (including poly­(dimethylsiloxane), polyurethane, rubber, thermoplastic elastomers, and gels) are commonly employed to enhance wearing comfort. These materials can compounded with conductive fillers to achieve superior electrical conductivity and sensing sensitivity, , such as carbon-based materials, conductive polymers, liquid metals and ionic conductors. , To achieve conformal coverage on the human body, the aforementioned materials are usually fabricated into diverse configurations to prepare wearable strain sensors, including fibers, textiles, and epidermal patches (Figure ). As shown in Figure a, the flexible fiber-based strain sensors were produced via coaxial wet spinning of poly­(styrene-butadiene-styrene)/MXene composites. The fiber sensors can detect physiological strains ranging from subtle pulses (0.5% detection limit) to large-angle joint flexion. For smart textile sensors, conductive nanomaterials (e.g., carbon nanotubes, graphene, metallic networks) can be integrated into flexible substrates through dip-coating, spray deposition, or electrospinning techniques. MXene, as the two-dimensional transition metal carbides/nitrides, has excellent conductivity, specific surface area, and thermal stability, making it an ideal choice for wearable electronics. To construct a high-performance MXene-based sensor and simultaneously improve the durability, Luo et al. engineered a waterproof and breathable smart textile with multicore–shell structure (Figure b), which was fabricated by depositing MXene on polydopamine-modified elastic fabric and then encapsulated with poly­(dimethylsiloxane). The poly­(dimethylsiloxane) can not only protect MXene from oxidation but also endow the textiles with superhydrophobicity and corrosion resistance. The smart textiles exhibit excellent mechanical, photothermal, and electrothermal response properties for sensitive motion monitoring.

3.

Different types of flexible strain wearable sensors for mechanical motion sensing: (a) conductive fibers featuring an SBS/MXene composite shell surrounding an SBS core for strain sensing; Reproduced with permission from ref Copyright 2023 American Chemical Society. (b) The conductive smart textile fabricated by depositing MXene on polydopamine-modified elastic fabric; Reproduced with permission from ref Copyright 2020 Elsevier. (c) The hydrogel films with a double-cross-linked poly­(acrylic acid-stearyl methacrylate) composite MXene; Reproduced with permission from ref Copyright 2023 American Chemical Society. (d) Wearable strain sensors for human motion monitoring. Reproduced with permission from ref Copyright 2023 American Chemical Society.

In addition, elastic and flexible substrates are commonly used for patch strain sensors, such as hydrogels, ionic gels, and conductive elastomers. Hydrogels are widely used due to their high transparency and flexibility. Yang et al. fabricated a double-cross-linked poly­(acrylic acid-stearyl methacrylate)/MXene hydrogel with strong mechanical extensibility and self-healing properties, in which reversible coordination bonding between Fe³⁺ ions and −COOH groups and hydrophobic interactions among the neighboring C₁₈H₃₇ chains are introduced into the polymer network. Based on the synergistic effects, the hydrogel film exhibits high tensile strength (525 kPa), good tensile performance (∼2600%), and self-healing properties, and it can be used to monitor subtle human movements, different handwriting, and gestures (Figure c,d). To improve the stability of hydrogels and prevent the water evaporation (i.e., dehydration), our group also introduced the eutecto-/Hydrogel as excellent flexible conductor and substrates. The eutecto-/Hydrogel can maintain a dynamic balance between hydration and dehydration, exhibiting high mechanical strength (Young’s modulus: 216.6 kPa), favorable ionic conductivity (0.7405 mS/cm), transparency (96.3%), temperature tolerance (−20–60 °C), and stable performance in strain monitoring (0–100%). Recent advances have further introduced adhesive and high-wear-resistant formulations to ensure long-term stability. For instance, Shao et al. developed a conductive hydrogel sensor with exceptional mechanical properties (tensile strength: 179.7 kPa, elongation at break: 634%) and robust adhesion to human skin. The enhancement mechanism of the conductive hydrogel involves Al³⁺ mediated metal–ligand coordination with −COOH groups on both poly­(acrylic acid) and carboxylated carbon nanotubes, simultaneously improving self-healing capability.

Furthermore, engineered microstructures including network architectures, origami configurations, and microfeatures can enhance sensitivity and operational stability. Wu et al. engineered mesh-structured hydrogel sensors via direct ink writing of PVA/tannic acid/polyacrylamide conductive hydrogels, achieving high structural fidelity. The mesh structure can optimize stress distribution, endowing the ability to achieve high strength and ductility. As a result, these networked electrodes enable enhanced sensitivity for detecting physiological signals ranging from subtle muscle tremors to macroscopic joint movements. Complementarily, Deng et al. implemented laser direct writing to pattern flexible substrates with high-consistency microcones, fabricating 64 sensor units within a 7.5 mm × 7.5 mm area (array density: 113 units/cm²). The mechano-responsive microlattice tactile sensor array provides multipoint spatial resolution and enables precision object manipulation, demonstrating significant potential for electronic skin.

2.3. Temperature

Body temperature is physiologically regulated by the hypothalamic thermoregulatory center, and typically maintained within the range of 35.9 to 37.5 °C. Clinically, abnormal temperatures often signal underlying pathological conditions, such as infections, inflammations, or other systemic disorders. Long-term body temperature monitoring can further delineate individual-specific physiological patterns, facilitating the early assessment of health risks. Meanwhile, continuous monitoring can be used to identify heat stress states that require preventive interventions. Besides, the efficacy of pharmacotherapy and treatment responses can be evaluated by tracking body temperature parameters. Therefore, as a key biomarker of body physiological status, temperature monitoring plays a vital role in health management and diagnose.

Flexible wearable temperature sensors typically exploit the thermoresistive effect of functional materials for temperature sensing. This phenomenon describes the temperature-dependent modulation of electrical resistivity in conductive materials, arising from thermally induced variations in charge carrier density, mobility, and lattice vibration scattering. These collective mechanisms drive the corresponding alterations in electrical conductivity and resistance. To achieve high sensitivity of sensors, thermally responsive conductive composites, together with thermocouples, semiconductor devices, and field-effect transistors, form the main sensing elements. As illustrated in Figure a, a stretchable temperature sensor array can be fabricated with polyethylene terephthalate film, Ecoflex film and polyaniline nanofibers. The polyaniline nanofibers are composited with alloyed liquid metal, which was mainly composed of gallium (68.5%), indium (21.5%), and tin (10%). Due to the thin-film structure and the flexibility of Ecoflex, the temperature sensor arrays can be attached to the skin, exhibiting a high resistance sensitivity and a response time of 1.8 s in the temperature range from 15 to 45 °C. In addition, highly conductive nanomaterials (including graphene, carbon black, carbon fibers, carbon nanotubes, and multiwalled carbon nanotubes) are the preferred conductive fillers for thermally responsive conductive composites due to their excellent electrical conductivity and operational stability in temperature sensing. By compounding with the polymer matrix, the highly sensitive temperature-responsive conductive network can be constructed to improve the detection performance of the sensors.

4.

Flexible wearable temperature sensors with different sensing principles: (a) a thermoresistive temperature sensor array consisting of poly­(ethylene terephthalate) film and polyaniline nanofibers composited with alloyed liquid metal. Reproduced with permission from ref Copyright 2018 John Wiley and Sons. (b) The thermophase transition sensors composed of poly­(N-isopropylacrylamide-co-acrylamide) networks. Reproduced with permission from ref Copyright 2022 American Chemical Society. (c) Thermochromic fiber sensors for temperature sensing. Reproduced with permission from ref Copyright 2020 American Chemical Society.

Thermal phase transition materials can also be utilized in wearable temperature sensors. Ma et al. constructed a double-network hydrogel by introducing a poly­(vinylpyrrolidone)/tannic acid/Fe³⁺ cross-linked network into the N,N-methylene diacrylamide cross-linked poly­(N-isopropylacrylamide-co-acrylamide) networks in Figure b. Due to the incorporation of the poly­(N-isopropylacrylamide-co-acrylamide) network, the hydrogels demonstrate temperature responsiveness and undergo a bulk phase transition. When the temperature is lower than the bulk phase transition temperature, the migration of ions in the hydrogel is accelerated as the temperature increases, resulting in a decrease in the resistivity. Consequently, it can be used as a wearable temperature sensor for real-time monitoring of body temperature and fever infections. Fan et al. also reported an ionic conductive hydrogel by introducing covalent cross-linked networks of N-isopropylacrylamide and 1-vinyl-3-butylimidazolium bromide into the hydrogel skeleton composed of poly­(vinyl alcohol) and poly­(vinylpyrrolidone). Since the presence of 1-vinyl-3-butylimidazolium bromide and N-isopropylacrylamide, the hydrogel sensors demonstrate good thermal response. In addition, thermochromic materials are also suitable for the preparation of temperature sensors. For instance, a thermochromic fiber with core–shell structure by combining thermochromic microcapsules and elastomers was prepared by our group (Figure c). The fiber sensors can be used for temperature sensing (0–55 °C) through thermochromic microcapsules appearing in different colors. Evidently, the temperature sensors based on the thermoresistive effect typically feature high sensitivity and a wide detection range, making them suitable for high-precision measurements. The sensors based on thermal phase transition can achieve accurate responses through material phase changes, with distinct signal variations and high specificity to temperature signals, which renders them suitable for the development of multisensing signal systems. Although the measurement accuracy of thermochromic-based sensors is relatively low, they allow temperature reading via intuitive color changes and boast the advantage of convenient operation, making them suitable for daily wearable sensing scenarios.

In order to make temperature sensors more sensitive for human body temperature detection, temperature-sensitive materials can be deposited on flexible substrate, usually forming a thin-film structure. By strategically modulating film thickness, composition, and microarchitecture, temperature sensor performance can be precisely engineered. Furthermore, micronano structures such as biomimetic structures, microneedle arrays, and nanofiber networks can significantly improve detection sensitivity. Ha et al. reported a skin-attachable temperature sensor with a poly­(dimethylsiloxane) adhesive layer mimicking the microstructure of the octopus sucker, i.e., the rim and spin-coated poly­(N-isopropylacrylamide) corresponding to the infundibulum and muscle, respectively. The temperature-responsive poly­(N-isopropylacrylamide) gels can swell and shrink when in contact with the skin, resulting in a decrease in the suction cup pressure and allowing the sensor to adhere to the skin. This temperature sensor can accurately detect skin temperature changes as low as 0.5 °C within the range of 25–40 °C, and demonstrate high performance and mechanical stability after repetitive bending and stable temperature-sensing long-term attachment to the skin. Similarly, to further optimize the structures and enhance temperature sensitivity, Cuartero et al. developed a microneedle temperature sensor filled with a temperature-responsive conducting polymer­(poly­(3,4-ethylenedioxythiophene):polystyrenesulfonate) through conductive 3D printing. The rational design of robust microstructure with high resolution in the micrometer domain endows the temperature sensor with good sensitivity, repeatability, and medium-term stability.

3. Biochemical Signals

Compared with invasive sample collection and testing in hospitals or laboratories, wearable biochemical sensors enable convenient and noninvasive real-time monitoring. They occupy a unique position in the field of health monitoring due to their ability to directly detect biochemical information at the molecular level. In the aforementioned studies on physiological signals, most physiological sensors can convert physical quantities (force, temperature, light, and electrical activity) directly into electrical signals based on physical effects (such as piezoresistivity, piezocapacitance, thermoelectricity, and photoelectricity). While for biochemical signals analysis, special biochemical recognition reactions are usually required to convert the biochemical molecules (including electrolytes, metabolites, trace elements, and drug metabolites) into electrical signals or optical signals. Besides, the concentration of target signals in biochemical sensors is generally relatively low, and these signals easily interfere with other components in biological samples or irrelevant biochemical reactions. As a result, biochemical sensors impose higher demands and challenges on sensing materials and structures. Currently, the essential components of biochemical sensors include flexible substrates, biorecognition elements (e.g., enzymes, antibodies, and aptamers), signal transduction modules, and data processing units. These biorecognition materials play a core role in the process of biological information acquisition, enabling selective biomolecular detection and signal transduction.

For biochemical sensors, noninvasive biochemical samples mainly include biofluids (sweat, saliva, tears) and exhaled gas. Among them, sweat is currently one of the most widely studied biological fluids due to its unique advantages. First, sweat is naturally produced by the human body during daily activities or exercise, enabling continuous sampling through wearable devices such as patches or wristbands. Meanwhile, sweat contains a wealth of analytes, including electrolytes (Na⁺, K⁺, Cl^–), metabolites (glucose, lactate, urea), hormones (cortisol), and small-molecule drugs. These components can provide valuable information for assessing metabolic health and exercise physiology. Saliva, another fully noninvasive and easily accessible biological sample, is secreted by the salivary glands into the oral cavity. It is rich in various substances, such as hormones, immunoglobulins, and inflammatory markers, making it suitable for stress level evaluation, oral health diagnosis, and medication adherence monitoring. Tears represent a highly promising biological sample. The concentration of proteins and metabolites in tears is even higher than that in saliva, exhibiting a good correlation with the concentrations in the blood. This unique characteristic provides a distinct pathway for the accurate diagnosis of diabetes and ocular diseases, such as dry eye syndrome. Compared with biofluids, the collection of exhaled gas is more convenient. It can achieve noninvasive, comfortable, and continuous sampling during natural breathing with wearable devices such as masks or nasal clips. This section will comprehensively review the collection and analysis strategies of wearable biochemical sensors for monitoring biological fluids (sweat, saliva, and tears) and exhaled breath, as well as their research progress. The strategies applied in wearable biochemical sensors were analyzed in Table .

2. Construction Strategies and Performance of Wearable Biochemical Sensors.

signals	substrates	method	sensitive materials	sensitivity	detection range	detection Limits	refs
sweat glucose	cotton fabric	electrochemical/enzyme	GOx/PANI/Pt/CNT	39.86 nA·μM–1	0–250 μM	/
sweat glucose	paper	electrochemical/enzyme	GOx/MB/Ti3C2T x	2.4 nA·μM–1 (0.08–0.45 mM); 1.0 nA·μM–1 (0.45–1.25 mM)	0.08–1.25 mM	17.05 μM
sweat glucose	PI film/PDMS microfluidic patch	electrochemical	Pt/MXene/PVA/MXene	3.43 μA·mM–1·cm–2	0–1 mM	29.15 μM
sweat glucose	PVA/sucrose hydrogel	colorimetric/enzyme	glucose oxidase/ /peroxidase/chromogen	/	0–2 mM	∼0.028 mM
sweat pH	cross-linked starch	colorimetric/color reaction	methyl red	4	4.0–7.0	/
salivary glucose	poly(ethylene terephthalate glycol)	electrochemical/enzyme	GOx/CA/PMEHB/Pt/Ag/AgCl	/	1.75–10,000 μmol/L	/
tear cortisol	soft contact lens	electrochemical/immunoassay	graphene/C-Mab/AgNF/AgNW/NFC chip	1.84 ng/mL per 1% ΔR/R 0	1–40 ng/mL	10 pg/mL
tear cholesterol	silicone elastomer	electrochemical/enzyme	ChOx/PB/Nafion/Au/Cr/NFC chip	1% current change per 0.043 mM	0.01–1.2 mM	9.91 μM
salivary lactate	PET	electrochemical/enzyme	LOx/PPD/PB/Ag/AgCl	0.553 μA mM–1	0.1–1.0 mM	0.05 mM
salivary glucose	ABS/silicone nipple	electrochemical/enzyme	GOx/Chitosan/PB/Ag/AgCl/NFC module	0.69 ± 0.04 nA mM–1	0.1–1.4 mM	0.04 mM
exhaled acetone	acrylic	electrochemical/chemiresistive	CS-rGO composite	27.89% (10 ppm)	0.3–10 ppm	/
exhaled acetone	YSZ nanofiber membrane	chemiresistive/photoactivated	TiO2-A/R	27.4% ppm–1 (2–6 ppm)	0.15–75 ppm	0.15 ppm
exhaled NH3	PAN nanofiber membrane	colorimetric; chemiresistive	PAN/BCG; PAN/PANI/rGO	colorimetric: 0.66 ppm–1; chemiresistive: 1.183 ppm–1	0.3–50 ppm	0.3 ppm
exhaled NH3	porous eutectogel	colorimetric	eutectogel/BCG	/	0.5–10 ppm	0.5 ppm

3.1. Biofluids

Metabolites are abundant in human body fluids, including electrolytes (e.g., sodium, potassium, chloride, ammonium and calcium), small-molecule metabolites (e.g., glucose, lactate and alcohol), trace elements (e.g., iron, zinc and copper), cytokines, and other small molecules (e.g., cortisol, urea and tyrosine). The health information on the human body can be reflected by monitoring this biochemical information. Sweat, saliva, and tears are usually selected for noninvasively detecting metabolites in human body fluids.

In order to effectively collect sweat samples, two approaches are typically employed: passive and active methods. Passive methods involve individuals engaging in vigorous physical activities such as running, cycling, or other exercises to naturally induce sufficient sweat secretion. The composition of sweat collected by this method is closer to the natural physiological state, which can effectively avoid potential interference caused by the introduction of stimulants and adverse skin reactions. For instance, Lin et al. designed an agarose-based gel patch for natural sweat collection and in situ electrochemical analysis, which was used to analyze and detect target substances, such as caffeine and lactic acid in Figure a. However, this collection relies on the user’s exercise capacity and physical state, making it challenging to achieve controllable sweat acquisition during rest or daily activities. To enable more controllable sweat collection, some researchers have also adopted active collection methods. The active methods refer to obtaining sweat samples through external stimuli, such as iontophoresis, thereby enabling the collection of sweat samples under nonexercise conditions. As presented in Figure b, after applying a voltage between the iontophoretic electrodes of the sensor, an electric current can be generated in the skin layer, and then, the stimulants transport to the sweat glands, thereby stimulating the secretion of sweat to obtain sweat samples. This method eliminates the reliance on physical exercise, facilitating continuous sweat monitoring under resting conditions. Meanwhile, it is also important to note that the additional stimulation units (such as electrodes and stimulant reservoirs) will increase the complexity and power consumption of the devices.

5.

Different methods for sweat collection in sweat sensors: (a) an agarose-based gel patch for natural sweat collection. Reproduced with permission from ref Copyright 2017 American Chemical Society. (b) iontophoretic electrodes for sweat collection via stimulation. Reproduced with permission from ref Copyright 2018 John Wiley and Sons. (c) A multimodal microneedle sensing platform for efficient collection of biological fluids. Reproduced with permission from ref Copyright 2019 American Chemical Society. (d) Paper-based microfluidic sandwich-structured sensor patch for sweat collection. Reproduced with permission from ref Copyright 2023 American Chemical Society.

In addition, microneedle patches and microfluidic devices are also commonly designed for sweat collection due to their high throughput and efficiency. Microneedle patches utilize an array of micrometer-scale needles to penetrate the stratum corneum of the skin, enabling direct interaction with the interstitial fluid (ISF) in the dermis or sweat gland tissues. Notably, microneedles do not rely on the active secretion of sweat glands, thus eliminating the need for exercise induction or iontophoretic stimulation. Meanwhile, they also avoid errors caused by fluctuations in sweat rate, epidermal contamination, and evaporative concentration. Wang et al. developed a multimodal microneedle sensing platform integrated into the same sensor array patch, which can effectively collect biofluids and sensitively detect levodopa (Figure c). The manufacturing process for conventional microneedle arrays is relatively complex and costly. Combining paper-based materials with microfluidic technology represents a more economical and environmentally friendly approach. With microfluidic technology, sweat secreted at different time intervals can be sequentially channeled through microchannels into independent storage chambers for preservation and analysis, thereby ensuring the sample integrity and stability. Meanwhile, paper substrates, with their porous and hydrophilic structures as well as strong capillary action, enable rapid and spontaneous adsorption and filtration of sweat. This not only facilitates efficient sweat collection but also removes interfering substances such as sebum, further enhancing the reliability of subsequent biochemical analysis. Feng et al. developed a paper-based sandwich-structured pH sensor that can filter the sebum mixed in sweat. The overall structure resembles a sandwich, with the top and bottom layers serving as sebum absorption layers and the middle layer as a filter paper layer. This design is primarily intended to filter out sebum from sweat that may affect sensor performance before it enters the core detection component of the sensor. As shown in Figure d, the hydrophilicity, microstructure and microfluidic properties of paper-based materials enable the sensor to collect sweat while filtering out over 90% of sebum within it, greatly enhancing the sensor’s reliability and accuracy. Beyond mitigating matrix interferences like contaminants and sebum in wearable pH sensors, robust anti-interference performance under dynamic or extreme conditions is also crucial. To tackle this challenge, Wang et al. developed a wearable pH sensor based on tungsten oxide aerogel (TOA). Specifically, they constructed a TOA sensing material with a 3D interconnected porous network. Thanks to its ultralarge specific surface area and self-supporting structure, the sensor not only achieves high sensitivity (−63.70 mV/pH) but also exhibits excellent shock resistance. Even under an impact pressure of 118.38 kPa, the sensitivity variation of the sensing chip is only 7.17%. This integrated design of material and structure provides new insights for sweat pH detection under complex scenarios, such as motion monitoring and outdoor operations.

Since sweat tends to weaken the adhesion of the device to the skin, the substrate is required to possess hygroscopicity, flexibility, and stretchability. Typical substrates include flexible polymers, paper-based substrates, fibers, and their woven fabrics. Among these, flexible polymers (e.g., poly­(vinyl alcohol)) exhibit excellent flexibility and biocompatibility. They can be easily integrated with functional materials through various processing methods, making them suitable for sweat analysis. As illustrated in Figure a, Li et al.fabricated a conductive hydrogel using MXene and poly­(vinyl alcohol), which was then integrated with a microfluidic patch for sweat collection. This integrated device enables the monitoring of glucose in sweat. To improve comfort and breathability, paper-based materials are also used in the fabrication of sweat patches. Li et al.developed a paper-based sweat patch by combining a foldable paper substrate with two-dimensional Ti₃C₂T _x conductive materials, allowing for the detection of lactic acid and glucose (Figure b). In particular, by folding the printed paper into a multilayer structure and connecting the hydrophilic areas of each layer, a three-dimensional sweat diffusion path in the vertical direction was formed. This can improve the diffusion efficiency of sweat from the skin surface to the interior of the sensor. The paper-based substrates may suffer from poor mechanical strength. They tend to soften and break easily when they are in a wet state, which ultimately leads to insufficient durability. Consequently, they are generally suitable for disposable patches. One-dimensional fibers, which possess core–sheath or porous structures, can actively adsorb sweat and reduce sweat dilution, thus serving as substrates for sweat sensors. By compositing with conductive materials such as carbon nanotubes, these fibers enable the detection of various metabolites (e.g., lactate, glucose). Additionally, the fibers can also be fabricated into two-dimensional textile fabrics through mixing with other conductive materials, coating, or weaving processes. These textile fabrics exhibit extreme biocompatibility with human skin, along with excellent breathability and wearing comfort, making them suitable for long-term and large-area sweat monitoring. Wang et al. constructed a wearable electrochemical fabric sensor by embroidering a composite of cotton sheaths and carbon nanotube fibers onto a superhydrophobic fabric substrate. As shown in Figure c, the hydrophilic sensing yarn facilitates the absorption and enrichment of sweat within the fibers, while the superhydrophobic substrate can prevent sweat from spreading on its surface, thereby significantly improving the efficiency of sweat capture and collection.

6.

Different types of wearable sweat sensors for electrochemical and colorimetric analysis: (a) a conductive hydrogel sweat sensor composed of MXene and poly­(vinyl alcohol) for electrochemical analysis; Reproduced with permission from ref Copyright 2023 American Chemical Society. (b) A paper-based electrochemical sweat patch fabricated by combining a foldable paper substrate with 2D Ti₃C₂T _x . Reproduced with permission from ref Copyright 2022 Elsevier. (c) A textile-based electrochemical sweat sensor fabricated by embroidering a composite of cotton sheaths and carbon nanotube fibers. Reproduced with permission from ref Copyright 2022 John Wiley and Sons. (d) Colorimetric sweat sensor integrated with paper-based microfluidics. Reproduced with permission from ref Copyright 2016 The American Association for the Advancement of Science. (e) Microfluidic wearable wristband integrating a colorimetric timer and biochemical detection modules. Reproduced with permission from ref Copyright 2024 The American Association for the Advancement of Science.

Efficient detection methods represent another key aspect of sweat analysis. Typically, target analytes can be analyzed by electrical or optical signals generated during biochemical reactions. Electrochemical assays are widely employed for sweat analysis, owing to their high sensitivity. The biochemical reaction between specific molecular recognition elements and target analytes can be converted into quantifiable changes in the current, potential, or resistance. Common molecular recognition elements include enzymes, aptamers, antigen–antibody pairs, and molecularly imprinted polymers. For instance, oxidoreductases such as glucose oxidase, lactate oxidase, and cholesterol oxidase can be immobilized and modified on the electrode surfaces. Upon interaction with target analytes, these enzymes can catalyze redox reactions, generating measurable electrical signal changes that enable the detection of specific metabolites. For example, Yao et al. fabricated a glucose sensor by immobilizing glucose oxidase onto a PET-based thin-film gold electrode for sweat glucose sensing. Additionally, conductivity materials and nanomaterials are preferred for electrode modification to reduce signal-to-noise ratios and enhance the detection sensitivity. Shu et al. developed a carbon nanotubes/multiwalled carbon nanotubes/poly­(dimethylsiloxane) electrode and modified it with a Ni–Co metal–organic framework material with enzyme-mimetic activity. This modified electrode exhibited excellent electrocatalytic performance and achieved high sensitivity with a detection limit as low as 6.78 μM. Although electrochemical methods possess rapid response and high sensitivity, they inevitably rely on external power supplies, which limits miniaturization and long-term wearability. To address this issue, Chen et al. recently reported an integrated wearable biofuel cell (w-BFC). The system employs a Au-rGO dual hydrogel as the anode to catalyze the oxidation of ascorbic acid (AA) and PtCu bimetallic hydrogels as the cathode to catalyze O₂ reduction. It enables self-powered detection of AA without an external power supply, representing a promising battery-free electrochemical sensing solution for sweat analysis.

Colorimetric analysis achieves qualitative or semiquantitative detection based on color changes induced by the specific chemical reaction between target analytes and chromogenic reagents. The mechanisms of molecular recognition and signal conversion in colorimetric analysis can be categorized into two types. One is an enzyme-catalyzed chromogenic reaction. Similar to electrochemical analysis, oxidoreductases catalyze the generation of H₂O₂, which then reacts with chromogens and form colored products. The intensity of the color is positively correlated with the concentration of the analyte. The other type is an ion/molecule-specific chromogenic reaction. For example, the color reaction between sweat pH and methyl, resulting in the color changes. Rogers’ group has performed a series of research on colorimetric sensors. The embedded chemical analysis endows colorimetric change responses to markers such as chloride, glucose, and lactate (Figure d). Recently, they also developed a microfluidic wearable wristband with an integrated colorimetric timer and biochemical detection module. By incorporating a colorimetric silver nanoplasmonic assay and a dye-conjugated silica nanoparticle-agarose composite, the device achieved real-time measurement of pH and lactate concentration in sweat (Figure e).

Saliva and tears are other biological samples that can be used for noninvasive monitoring. Due to the extremely low concentration of target analytes in these fluids, electrochemical analysis is frequently employed for detection. Tear sensors can be attached to framed eyeglasses or contact lenses. Similar to the electrochemical analysis of sweat, electrochemical tear sensors rely on the chemical reaction between the target analyte and the active molecule immobilized on the sensor to produce changes in electrical signal changes. As illustrated in Figure a, a multifunctional contact lens sensor was fabricated for wireless tear diagnosis. This sensor employed a transparent flexible electrode made of graphene-silver nanowire composites and immobilized glucose oxidase to detect electrical signals during glucose oxidation reactions and glucose concentration in tears.

7.

Different types of wearable tear sensors and saliva sensors: (a) wearable contact lenses based on graphene-silver nanowire composites and immobilized glucose oxidase for tear glucose sensing; Reproduced with permission from ref Copyright 2017 Springer Nature. (b) Wearable contact lenses incorporating immobilized cholesterol oxidase for cholesterol sensing in tears. Reproduced with permission from ref Copyright 2022 John Wiley and Sons. (c) Smart contact lenses integrated with the cortisol sensor, capacitor, resistor, and integrated circuit chip for cortisol detection. Reproduced with permission from ref Copyright 2020 American Association for the Advancement of Science. (d) A dental guard integrated with a lactate oxidase-based salivary lactate sensor for lactic acid sensing. Reproduced with permission from ref Copyright 2014 Royal Society of Chemistry. (e) Wearable cellulose acetate-coated mouthguard biosensor for salivary glucose measurement. Reproduced with permission from ref Copyright 2020 American Chemical Society. (f) A saliva sensor composed of pump-driven saliva delivery and enzyme-encapsulated electrodes. Reproduced with permission from ref Copyright 2019 American Chemical Society.

Wearable tear sensors for cholesterol and cortisol detection have also been developed by immobilizing enzymes or immunoassays. As depicted in Figure b, a wearable contact lens sensor was prepared by integrating an electrochemical cholesterol biosensor with an integrated circuit chip into a soft wearable contact lens. Specifically, cholesterol oxidase was immobilized on the working electrode by Nafion (as a cation exchange polymer) and catalyzed by cholesterol oxidation to generate hydrogen peroxide. This intermediate subsequently underwent further oxidation–reduction reactions with prussian blue on the working electrode, producing an electrochemical signal. The resulting current changes can be recorded via ammeter method to quantify cholesterol concentration. Additionally, cortisol immunosensors have been developed by conjugating cortisol-specific monoclonal antibodies with graphene field-effect transistors. As illustrated in Figure c, the smart contact lenses capable of cortisol detection were integrated with the cortisol sensor, capacitor, resistor, and integrated circuit chip. This device can achieve a detection limit as low as 10 pg/mL and a linear range of 1–40 ng/mL, making it suitable for monitoring cortisol levels in human tears.

Similar to tear diagnostics, saliva testing can be realized by integrating microfabricated biosensing units with oral devices through enzyme-mediated sensing strategies. As illustrated in Figure d, Kim et al. fabricated a salivary lactate sensor by immobilized lactate oxidase on a flexible poly­(ethylene terephthalate) substrate. This sensor can be affixed to a mouthguard with the adhesives, enabling noninvasive detection of salivary lactate. To enhance the analytical convenience, they further improved the approach by modifying uricase onto screen-printed flexible electrodes and integrating wireless electronic modules to enable saliva uric acid monitoring. This integration allows for the real-time acquisition of sensing data and facilitates disease diagnosis. Generally, salivary components such as ascorbic acid and uric acid may hinder the accurate saliva analysis. To reduce the influence of interferents in saliva analysis, a mouthguard-type glucose sensor with a cellulose acetate membrane was developed for monitoring salivary glucose in Figure e. The cellulose acetate membrane served as an interference rejection membrane and inhibited the reactions of ascorbic acid and uric acid on the microneedle electrode, enabling the accurate detection of salivary glucose (1.75–10,000 μmol/L). For monitoring newborns, a pacifier-style wearable saliva sensor has also been developed. As depicted in Figure f, peristaltic movements of infant’s mouth facilitated pump-driven saliva delivery inside the pacifier, enabling unidirectional flow of saliva from the oral cavity to the electrochemical chamber. Subsequently, saliva glucose detection was achieved using enzyme-encapsulated electrodes within the chamber.

3.2. Exhaled Gas

Human exhaled breath primarily comprises nitrogen, oxygen, carbon dioxide, hydrogen, inert gases, and water vapor. It also contains inorganic volatile compounds (e.g., nitric oxide, ammonia, hydrogen sulfide) and organic volatile compounds (e.g., acetone, ethanol). When inhaled air enters the alveoli, metabolites diffuse into the air and are subsequently expelled in exhaled breath. Therefore, exhaled gas usually carries endogenous metabolic markers, which enable effective monitoring of metabolic status and disease treatment efficacy via breath analysis. Compared with traditional blood tests and endoscopic procedures, breath analysis provides advantages of rapidity, noninvasiveness, painlessness, low cost and simplicity. It can achieve early disease diagnosis and real-time physiological monitoring for the abundant metabolites and biomarkers in human respiratory gases. Currently, substantial research has been conducted on sensors for exhaled gases, including acetone, ammonia, hydrogen sulfide, and nitric oxide. For example, acetone concentration in exhaled breath enables the assessment of diabetes mellitus severity (≤0.9 ppm in healthy individuals, 0.9–1.8 ppm in moderate cases, and up to tens of ppm in severe cases), as well as the current physiological state of diabetic patients. , The concentration of ammonia (NH₃) in exhaled gas deviating from normal can provide early warnings for liver or kidney diseases including renal failure and liver dysfunction. Hydrogen sulfide correlates with the severity of respiratory diseases and airway inflammation, serving as a marker for airway inflammation, respiratory infections, and oral/dental health status. Additionally, abnormal concentration of exhaled nitric oxide may indicate pulmonary or respiratory diseases such as asthma, chronic obstructive pulmonary disease, and cystic fibrosis. ,

To facilitate highly sensitive gas detection, conductive active materials such as carbon-based materials, metals, and semiconductors are usually introduced. As shown in Figure a, Su et al. employed chitosan and reduced graphene oxide as sensitive materials to produce a wearable breath biosensor, which can detect the acetone in exhaled gas and achieve noninvasive diagnosis of early diabetes mellitus. The high surface area and excellent electrical conductivity of reduced graphene oxide contribute to improving the sensor’s response rate and sensitivity, with a response rate about 5 times that of pure chitosan materials. Figure b depicts a wearable oxide semiconductor gas sensor, which was fabricated by depositing anatase/rutile TiO₂ homojunctions onto a flexible yttria-stabilized zirconia nanofiber substrate through atomic layer deposition. Moreover, the TiO₂ homojunctions with nearly perfect lattice matching significantly promoted the separation efficiency of photogenerated carriers and strengthened oxygen adsorption capacity, thereby enhancing the gas sensing performance under light irradiation. As a result, this wearable gas sensor achieved a detection limit of 0.15 ppm for acetone at room temperature. Wang et al. designed a smart wearable mask sensor to monitor various respiratory infectious diseases (COVID-19, H5N1, and H1N1) on the respiratory valve of the mask via ion-gated transistors in Figure c. Based on the sensitive gating effect of ion gels, the aptamer-functionalized transistors can detect liquid samples at trace levels (0.3 μL) and gaseous media samples with ultralow viral protein concentrations (0.1 fg/mL). Besides, gas-sensitive polymers can also be utilized to achieve high adsorption and sensitive detection of gases. Figure d depicts a resistive ammonia sensor developed by our group, which utilizes polyacrylonitrile/polyaniline/reduced graphene oxide nanofiber membranes as sensing material. When exposed to an ammonia atmosphere, the protonation state of polyaniline changes, leading to the formation of ammonium ions and an increase in the resistance of polyaniline components. The integration of reduced graphene oxide further enhances the sensor’s conductivity and sensitivity, thereby achieving highly sensitive detection of ammonia. Recently, the eutectic gel was further introduced as ammonia adsorption materials to prepare wearable ammonia sensors. The eutectic gel possesses an abundant internal hydrogen bond acceptor and exhibits a strong affinity for NH₃. This enables efficient capture and enrichment of NH₃ upon exposure to exhaled breath (Figure e).

8.

Different types of wearable respiratory biochemical sensors: (a) a biosensor incorporating chitosan and reduced graphene oxide for acetone detection in exhaled breath. Reproduced with permission from ref Copyright 2020 Elsevier. (b) a wearable oxide semiconductor gas sensor fabricated by depositing TiO₂ homojunctions onto flexible yttria-stabilized zirconia nanofiber substrates. Reproduced with permission from ref Copyright 2019 American Chemical Society. (c) Smart wearable mask sensor integrated into the mask’s respiratory valve via ion-gated transistors for monitoring various respiratory infectious diseases (COVID-19, H5N1 and H1N1). Reproduced with permission from ref Copyright 2022 Elsevier. (d) An ammonia sensor based on polyacrylonitrile/polyaniline/reduced graphene oxide nanofiber membranes. Reproduced with permission from ref Copyright 2023 American Chemical Society. (e) Ammonia sensors fabricated via 3D printing using eutectic gels as ammonia adsorption materials. Reproduced with permission from ref Copyright 2025 Elsevier.

Notably, the micronanostructures or porous architectures can effectively enhance the contact efficiency between gases and the sensing layer, thereby improving sensing sensitivity. A variety of strategies such as nanopore fabrication, nanowire growth, and thin-film deposition have been used to adjust the microstructure of materials to increase the specific surface area of sensors, improving gas adsorption capacity and optimizing response kinetics. For instance, our group utilized 3D printing technology to fabricate the aforementioned eutectic gel into different structures (helical, cylindrical, diamond and planar), and systematically investigated their impact on NH₃ enrichment and colorimetric performance. The results showed that the helical structure exhibited optimal performance in NH₃ detection due to its maximum specific surface area and optimal NH₃ retention capacity, with a rapid response time (3 s) and low detection limit (0.5 ppm). Jang et al. developed a gas sensor based on Pd nanoparticles-decorated W₁₈O₄₉ nanowires, which exhibits sensitive and selective detection of hydrogen and ammonia. The introduction of Pd nanoparticles significantly enhanced the oxygen adsorption capacity on the surface of W₁₈O₄₉ nanowires, and improved the chemical affinity between gases and sensing materials (such as the spillover effect of H₂ and acid–base interactions of NH₃), thereby enhancing the response performance. This sensor was further integrated into a microelectromechanical system, resulting in a compact device with low power consumption and high response values of 35.3 and 133.8, respectively.

4. Environmental Signals

Human health is closely associated with the surrounding environment. We can perceive the changes of the surrounding environment and improve our life quality via wearable environmental sensors. At present, flexible wearable environmental sensors mainly include monitoring of UV light, humidity, temperature, and polluted gases. This section mainly focuses on UV and humidity sensors. The strategies and performance of sensors for environmental signal monitoring were summarized in Table .

3. Construction Strategies and Performance of Wearable Environmental Sensors.

signals	substrate	sensitive materials	sensitivity	detection limits	refs
UV	bacterial cellulose	photochromic dyes	discoloration <1 min; recovery <2 min	/
	chitosan	ZnO nanowires	55 A/W	0.1 μW cm–2
	office paper	porous ZnO nanostructures/CMC	432 ± 48 mA W1–	0.3 mW cm–2
humidity	PET film	CNF/CB/PVP/glycerol	9.6–10.0% RH–1	20–90% RH
	PVD membrane	PANI/CTAB	226.0% (R/R 0)	11–98% RH
	cotton textile	SWNT/PVA/LiCl	600%	40–100% RH
	silver interdigital electrodes	CNF/Ti3C2T x  MXene/PAAS	2.11 pF/% RH (10–41%); 60.97 pF/% RH (41–84%)	10–90% RH
	cellulose acetate	HPC/[Bmim]2[NiCl4]/ethylene glycol	163 (impedance ratio)	30–90% RH

4.1. Ultraviolet (UV) Irradiation

UV radiation in sunlight can stimulate the synthesis of vitamin D in the human body and enhance the absorption of calcium. However, excessive UV exposure may cause damage to the human retina and various skin diseases. Currently, UV detectors are primarily categorized into two types. One type is the photoelectric UV sensor, which converts ultraviolet radiation into electrical signals by using photoelectric response materials. Another type is the optical-signal UV sensor based on photochromic materials. These two types of UV sensors exhibit distinct characteristics in terms of sensing principle properties, material selection, and structural construction strategies. This section will discuss them in detail, along with the latest research developments on UV sensors.

Photosensitive materials are frequently employed in optoelectronic ultraviolet sensors. After these materials absorb photons, they trigger the separation of electron–hole pairs, and quantifiable electrical signals (changes in current, voltage, or resistance) are formed through the migration of charge carriers. Two categories of materials are usually the preferred photoelectric materials for fabricating UV sensors: wide-band gap semiconductors and perovskites. This is because they possess excellent UV photon capture capabilities and carrier separation abilities, which provide a solid foundation for the stable operation of UV sensors. Upon exposure to UV irradiation, conduction band electrons in photoresponsive semiconductors are excited to the valence band, leaving behind holes (electron vacancies) in the conduction band. These holes migrate toward the semiconductor surface, inducing unpaired free electrons to move from the negative electrode to the positive electrode, thereby generating an electric current. It is noteworthy that metal oxides, including zinc oxide, titanium oxide, nickel oxide, and zinc stannate, as well as other semiconductor materials, are widely employed in manufacturing photoresponsive sensors. For example, Fan et al. proposed a wearable UV wristband for real-time UV monitoring based on a flexible chalcogenide photodetector. Benefiting from the excellent UV response properties of the CH₃NH₃PbCl₃ chalcogenide film, the device demonstrated exceptional performance with high responsivity of 359.03 mA/W and fast rise/fall time of 3.91/4.55 ms in Figure a.

9.

Wearable sensors for UV sensing: (a) a wearable UV wristband based on a self-powered flexible perovskite photodetector for real-time UV monitoring. Reproduced with permission from ref Copyright 2022 Elsevier. (b) Degradable high-performance UV photodetectors based on ZnO nanowires on flexible chitosan substrate for UV sensing. Reproduced with permission from ref under a Creative Commons Attribution 4.0 License. (c) The screen-printed carboxymethyl cellulose/ZnO UV sensors with porous ZnO nanostructures; Reproduced from ref under the terms of the Creative Commons Attribution (CC BY) license. (d) A visual UV sensor prepared with photochromic bacterial cellulose (PBC). Reproduced with permission from ref Copyright 2024 John Wiley and Sons. (e) Self-healing UV detection patch with color change upon UV irradiation. Reproduced with permission from ref Copyright 2023 John Wiley and Sons.

To enhance the sensitivity of UV sensors, different strategies have been considered, such as incorporating other nanomaterials, increasing the specific surface area, and introducing porous micronano structures. For example, Figure b displays a degradable high-performance UV photodetector based on ZnO nanowires, which was located on the flexible chitosan substrates. The UV photodetectors demonstrated excellent and repeatable response under different bias voltages, accompanied by a high responsivity, superior detectivity, and higher gain value. In order to improve the comfort and practicality of wearable sensors, zinc oxide nanowires and carbon nanotubes were also compounded with highly elastic Spandex fibers to fabricate wearable UV fiber optic sensors. By coating the ZnO nanowires with hydrophobic phosphonic acids, the sensors can effectively avoid signal instability and maintain high sensitivity, even when exposed to water or sweat. Besides, ZnO nanomaterials with porous structure are useful for sensors since the large number of pores provides high specific surface areas for molecules to absorb. For instance, Martins et al. composited porous ZnO nanostructures with carboxymethyl cellulose to synthesize screen-printed inks and prepare the UV sensors on office paper, and the paper-based UV sensor exhibited consistent performance in multiple UV irradiation cycles under a working voltage of 2 V (Figure c).

For photochromic UV sensors, the methods of fabrication typically involve the combination of photochromic dyes and flexible substrates. Common photochromic dyes include azobenzenes, spiropyrans, spirooxazines, and their derivatives. These dyes display selectivity and reversibility under UV light, which is accompanied by color changes. For example, spiropyrans undergo a spirocyclic ring-opening reaction under UV irradiation, with their color changing from colorless to purple (and the absorption wavelength red-shifting from ∼300 to 580 nm); zobenzenes undergo cis–trans isomerization, changing color from yellow to orange; whereas spirooxazines present a color change from colorless to blue. The integration of flexible substrate materials not only enables effective immobilization and encapsulation of dye molecules but also enhances the sensor’s flexibility and durability. Commonly substrates include transparent polymers PDMS poly­(dimethylsiloxane), colorless hydrogels, and bacterial cellulose. Despite their lower sensitivity and limited accuracy, photochromic UV sensors offer advantages, such as visual detection, zero power consumption, and simplicity. They require no external electric field and can be detected either visually or with a smartphone. Xu et al. exploited commercial photochromic dyes (spiroxazine, spiropyran, and their mixtures) as UV photosensitizers and uniformly distributed them in a three-dimensional nanonetwork of bacterial cellulose to develop a UV sensor (Figure d). The uniform distribution of photochromic dyes and the three-dimensional network structure of bacterial cellulose provided the UV sensor with more surface area to receive and utilize the photon energy from UV rays, thereby enhancing the efficiency of the photochromic process. Abdon et al. developed a novel multifunctional photochromic elastomer by incorporating phosphomolybdic acid into a self-healing polyurethane polymer network cross-linked by dynamic disulfide bonds. They also prepared the UV sensing stickers and UV sensing wristbands by patterning a photochromic layer bonded to the adhesive layer, which allows for the assessment of UV intensity changes based on the reference and colors on the wristband (Figure e). The phosphomolybdic acid composite with polyurethane polymers displayed photochromic properties, providing an electron donor to enable color changes and UV sensing. Besides, self-healing property of the elastomer makes wearable wristbands more durable in daily life. The photochromic tattoo has also been reported as an intradermal UV radiometer, and prepared using the photchromic ink containing photochromic dyes and biocompatible materials. The tattoo demonstrated exceptional color responsiveness to UV light, enabling real-time monitoring of the UV exposure intensity through visual observation. As the intensity of UV irradiation increases, the color of the tattoo changes from light to blue. The sensor can be implanted through minimally invasive surgery that lasts only a few seconds but retains functionality for months to years, demonstrating its potential for long-term intradermal UV sensing.

4.2. Environmental Humidity

Humidity (typically expressed as relative humidity) is an important environmental parameter closely related to human life. High humidity can impair the body’s ability to sweat and conduct heat, leading to heat stress and increasing morbidity and mortality. On the other hand, low humidity will cause water loss and dry skin, which is also detrimental to human health.

Generally, humidity sensors can be classified into resistive and capacitive types based on their sensing mechanisms. Resistive humidity sensors are typically characterized by relatively simple structure, low cost, and ease of measurement. The humidity sensing mechanism of the sensor is the variation of the material conductivity caused by the adsorption of water molecules on the sensing surface. Consequently, the relative humidity can be assessed by the change in the resistance. For example, Chou et al. reported a resistive humidity sensor based on poly­(vinyl alcohol)/lithium chloride through the wet-spinning and solvent exchange process (Figure a). Interestingly, the sensor exhibited excellent stability during deformation and high sensitivity under a wide range of relative humidity conditions. This ultrafast response of the sensor is attributed to the rapid deliquescence of lithium chloride upon contact with water molecules, resulting in changes in conductivity caused by ion generation and migration. By the integration of high-performance humidity sensing fibers into textile substrates, it can be customized to monitor microclimate conditions in real time with high sensitivity.

10.

Flexible humidity sensor for environment monitoring: (a) Textile-based humidity sensors made of single-walled carbon nanotube/poly­(vinyl alcohol)/lithium chloride nanocomposite filaments through the wet-spinning and solvent exchange process. Reproduced with permission from ref Copyright 2022 Elsevier. (b) A flexible, environmentally friendly, and highly sensitive humidity sensor composed of cellulose, MXene, and sodium polyacrylate. Reproduced with permission from ref Copyright 2025 Elsevier. (c) Fully inkjet-printed humidity sensor based on hydroxypropyl cellulose and the ionic liquid. Reproduced with permission from ref Copyright 2025 American Chemical Society. (d) Single-sided integrated flexible humidity sensor with polyaniline and poly­(vinylidene fluoride) microporous membrane. Reproduced with permission from ref Copyright 2021 Elsevier.

For capacitive humidity sensors, the capacitance is mainly related to the dielectric constant of the sensing material and air as well as the area and distance between the capacitor plates. When humidity increases, the sensing material will adsorb polar water molecules, subsequently enhancing polarization and causing a change in the dielectric constant, which in turn leads to a change in capacitance. For example, Liu et al. developed a humidity sensor composed of cellulose, MXene and sodium polyacrylate (Figure b). Benefiting from the rich polar functional groups in sodium polyacrylate, the humidity sensor possesses high sensitivity and a wide detection range. Moreover, the hygroscopic sodium polyacrylate can absorb a large number of water molecules under heavy humidity conditions and form multiple water layers, which provide more ion diffusion pathways for the release of sodium ions from the sodium polyacrylate, thereby resulting in excellent sensitivity.

Moisture sensitive materials are crucial components of humidity sensors. Based on the excellent hydrophilicity of cellulose and ionic liquids, Gerardo et al. also proposed an inkjet printable ink based on hydroxypropyl cellulose and ionic liquid ([Bmim]₂[NiCl₄]) to prepare the humidity sensor in Figure c. Apart from the aforementioned ionic conductors, other humidity-sensitive materials, such as carbon-based materials (e.g., graphene, graphene oxide, carbon nanotubes, carbon nitride, reduced graphene oxide, etc.), metal oxides, and conductive polymer materials, are also widely used in humidity sensors. For example, Huang et al. adopted hydrophobic poly­(tetrafluoroethylene) as the substrate and graphene oxide as the sensitive material to fabricate silver fork electrodes and capacitive humidity sensors by screen printing. Wang et al. fabricated a flexible humidity sensor using vertically aligned carbon nanotubes as electrodes and graphene oxide as a sensing material. This sensor exhibited an ultrafast response speed of ∼20 ms, as well as high sensitivity, low hysteresis, long-term stability and excellent flexibility.

Furthermore, in order to make humidity sensors more prone to absorbing moisture, micronano structures such as porous, wrinkled, and uniformly stacked structures can be applied to humidity sensors. Zhang et al. constructed a humidity sensor by printing a grid-structured electrode with cellulose acetate butyrate. Due to the large front and back surface areas of the capacitor plates, the capacitance of this sensor responds quickly when contacted with water vapor. Combined with electrospinning technology, a humidity-sensitive membrane based on tin dioxide/reduced graphene oxide can also be manufactured on a polyimide film, showing a fast response and recovery time. Compared with pristine graphene oxide, the tin dioxide/reduced graphene oxide nanocomposites exhibited a significantly larger specific surface area, which provides more available active sites for water molecules and accelerates the adsorption and desorption processes. Therefore, by a combination of reduced graphene oxide with tin dioxide, the humidity sensing performance has been greatly enhanced. In addition, the porous structure can improve the permeability and provide a high surface area for the adsorption of water molecules, resulting in a high sensitivity. It should be noted that the specific characteristics of the porous structure (e.g., pore size, tortuosity, and connectivity) directly regulate the sensor’s response/recovery time and sensitivity by impacting the diffusion rate of water molecules and the number of adsorption/desorption sites. The designed pore size should be moderate to ensure a large specific surface area while avoiding excessive diffusion resistance. Simultaneously, it is also essential to optimize tortuosity to reduce the complexity of water molecule transport pathways as well as maintain pore connectivity to facilitate rapid water molecule diffusion. Yang et al. developed an integrated flexible humidity sensor via in situ depositing polyaniline to form a uniform layer of polyaniline nanoparticles on the surface and within the pores of a poly­(vinylidene fluoride) microporous membrane in Figure d. This delicate design not only increases the specific surface area inside the sensor but also maintains good air permeability of the substrates. As a consequence, the micronano structures of the humidity sensor promote the adsorption/desorption and diffusion of water molecules on the integrated membrane, with rapid and reversible humidity response and small hysteresis (5% RH). Additionally, even when the flexible sensor was deformed, the micronano structure mitigated the damage to polyaniline conductive network, demonstrating stable humidity response.

5. Conclusions and Prospects

5.1. Conclusions

In this review, we comprehensively summarize the current research progress of wearable sensors in physiological, biochemical, and environmental signals monitoring and discuss different strategies to improve their performance from the perspectives of materials and structures. Wearable sensors are currently primarily utilized for physiological monitoring, including electrophysiological, mechanical strain, and body temperature signals. Regarding the optimization of sensor performance, researchers have focused predominantly on refining the properties of chemical materials, composite formulations, and structural designs. Specifically, the incorporation of materials with high flexibility, elasticity, conformability, and conductivity can significantly enhance their sensing sensitivity. Furthermore, the integration of micronano structures such as grids, serpentine patterns, and cracks further improves sensor performance. Additionally, additional functionalities including self-healing, degradability, and antibacterial properties have also been designed, so as to broaden the application of wearable sensors. For biochemical signal monitoring, the integration of flexible substrates and microsensors enables the development of wearable sensors, which can be seamlessly incorporated into wristbands, patches, fabrics, mouthguards, glasses, and masks for real-time detection of analytes in sweat, saliva, tears, and exhaled gases. Analogous to physiological signal sensors, the strategic design of substrate materials with high electrical conductivity and flexibility continues to serve as a cornerstone for enhancing the sensing performance in biochemical sensors. However, due to the low concentrations of biochemical markers in biological fluids and matrix interference, designing recognition elements such as specific enzymes and antibodies is crucial for improving sensor specificity. Furthermore, the deployment of functional materials featuring properties such as high permeability and adsorption, in tandem with advanced structural designs (e.g., microfluidics), facilitates optimized sample collection and detection, thereby bolstering the sensitivity and long-term stability of biochemical sensing platforms. In the context of environmental signals, this work primarily focuses on the monitoring of UV and humidity signals. For the UV sensor, it typically involves the incorporation of photoelectric materials, photosensitive materials, and photochromic materials to fabricate wearable UV sensors. Additionally, humidity-sensitive materials can be utilized to develop humidity sensors such as resistive and capacitive sensors. Furthermore, the design of porous micronano structures and the adoption of advanced fabrication techniques (including 3D printing and electrospinning) significantly enhance the sensing performance.

5.2. Challenges and Prospects

Although flexible wearable sensors have achieved progress in the fields of physiological, biochemical, and environmental signal monitoring, several key issues still need to be addressed for their practical application.

5.2.1. Multifunctional Sensing

In the research of wearable sensors, most devices remain focused on single functions, including physical signals (e.g., pressure, temperature) or chemical markers (e.g., glucose, pH). While they satisfy basic monitoring needs for specific scenarios, they struggle to enable multidimensional characterization of health indicators. With advances in materials science, chemistry, and bioengineering, there has been growing research on multifunctional sensors, encompassing studies of multiple physical, chemical, or environmental signal monitoring. These multifunctional wearable sensors can be classified into two types based on the construction strategy. One type is a multifunctional sensor that achieves multiple sensing capabilities by exploiting the material’s response characteristics to various stimuli, eliminating the need for complex module assembly. The other is the multimodal wearable sensor that combines multiple material modules with independent functions to achieve multisignal detection. For multifunctional sensors, they rely on the multiphysical properties of a single material such as its response to heat, electricity, light, and chemical stimuli to achieve multifunctionality. Commonly, these materials include carbon-based conductive materials (e.g., carbon black, composite carbon nanotubes, graphene, MXene), piezoelectric materials, and light-responsive polymers. For instance, graphene can be utilized for multifunctional sensor capable of simultaneously monitoring UV light, humidity, and temperature. Bi et al. also fabricated carbon nanotube composite flax fabric that enable strain sensing via stress-induced conductive network reconstruction, temperature sensing via temperature-induced carrier mobility changes, and humidity sensing through humidity-driven alterations in ionic conductivity, thereby achieving simultaneous detection of pressure, temperature, and humidity. For multifunction sensors with multimodule integration, these sensors rely on independent functional units. Different modules are usually integrated using micronano structures or flexible substrates. Through the cross-module coupling of mechanical, electrical, optical, and chemical sensing mechanisms, the synchronous monitoring of multiple indicators can be achieved. For example, the MXene/prussian blue composite hydrogel can serve as the biochemical sensing module (detecting glucose, lactic acid, and pH), while laser-induced graphene acts as the pressure sensing module, enabling simultaneous monitoring of biochemical markers and mechanical pressure. Similarly, carbon nanotube-modified cotton fabric was employed as the pressure sensing module, and the PVA/glycerol/LiCl composites can be selected for the temperature module. By exploiting the temperature-dependent variation in ionic conduction properties, a stable temperature response was achieved within the range of 20–50 °C. For the humidity module, TiO₂/P­(VDF-TPFE) nanofibers were loaded. The synergy between TiO₂’s hydrophilicity and P­(VDF-TPFE)’s piezoelectric effect enables efficient humidity sensing. Generally, the construction of such multimodal sensors involves multiple design and manufacturing strategies, including spinning techniques, layered module assembly, and integrated 3D printing. For instance, in our group’s previous work, the wet spinning was used to synchronously integrate conductive hydrogel fiber segments (as the pressure module) with a thermochromic elastomer-based temperature sensing module, facilitating collaborative detection of pressure and temperature.

However, the development of multifunctional wearable sensors still faces several challenges. For multifunctional sensors based on multiresponsive materials, signal overlap frequently occurs when they are exposed to multiple stimuli. For example, the resistance change in carbon nanotube sensors is simultaneously affected by the pressure and temperature, leading to significant crosstalk errors. Although preparing multifunctional sensors through multimodular integration can reduce interference, it requires additional structural design. Furthermore, current multifunctional sensors rely on high-precision manufacturing processes and complex assembly steps, resulting in high costs. In future development, novel materials, such as stimulus-specific materials or the design of biomimetic layered sensing structures, could be developed to enhance signal specificity and reduce interference. Additionally, integrating mass production techniques (e.g., textile manufacturing and 3D printing) with low-cost processes could effectively reduce material and production costs.

5.2.2. Sensors Integration

Wearable devices typically consist of sensing units, data processing units, data transmission components, and power supply systems. Integrating these elements onto a single platform requires complex assembly procedures, involving cumbersome processes and significant manufacturing challenges. Additionally, since wearable sensors are worn on the body or implanted internally, these sensors are generally compact in size and designed for high comfort. Beyond the signal interference issues discussed in Section , miniaturization also impacts the sensing performance, data transmission capabilities, and energy supply. For instance, data processing units may suffer reduced computational power due to microscale constraints. Data transmission requires consideration of microplatform spacing and external interference. Furthermore, traditional power sources may become inadequate on microplatforms, necessitating more compact and convenient energy solutions.

In future development, the integration of sensing units and data modules can be achieved through novel functional materials and advanced manufacturing methods such as 3D printing and microelectromechanical systems (MEMS), thereby reducing multistep assembly and enabling miniaturization. Concurrently, high-computing-power chips can be incorporated to achieve data storage and computation functions, minimizing power consumption. For data transmission, wireless communication, such as Bluetooth modules and NFC can be introduced. The different transmission methods can be coordinated based on the required transmission distances. Regarding energy supply, novel self-powered devices can be developed, such as those harnessing human motion (e.g., triboelectric nanogenerators and piezoelectric materials) or biofluids like sweat for power generation, thereby reducing reliance on traditional batteries.

5.2.3. Integration of Artificial Intelligence (AI) and Wearable Sensors

The application of artificial intelligence (AI) in wearable sensors requires further exploration. Current research primarily focuses on optimizing algorithms for data processing and noise reduction in single signals, making it challenging to analyze multimodal data. Future efforts should promote the deep integration of AI with wearable sensors to achieve precise and personalized health management. By employing multimodal data fusion algorithms, we can construct feature spectra of multiple biomarkers and establish correlations between physical and chemical signals, enabling accurate identification of health conditions. This enables early disease warning and intervention, assisting individuals in transitioning from traditional passive healthcare management to proactive health management via cloud platforms. Additionally, personalized model training can be developed to establish diverse databases for different age groups, genders, and occupations, providing tailored guidance based on individual needs. For instance, through the analysis of user requirements such as integrating dietary and exercise data, personalized dietary guidance can be offered to diabetic patients and competitive athletes.

5.2.4. Commercialization

Although numerous wearable sensors have been studied in the literature, few have reached commercialization. Many materials for wearable sensors currently involve relatively high preparation costs, preventing large-scale production. Besides, clinical validation and regulatory compliance present significant challenges. The application of wearable devices in healthcare requires rigorous clinical validation and regulatory agency approval, which is typically time-consuming and costly. Future approaches may involve adopting mature technologies such as screen printing and roll-to-roll printing or developing low-cost alternative materials to reduce production expenses. In terms of clinical validation and regulatory compliance, it is important to strengthen collaboration between research institutions and healthcare providers, conducting large-scale clinical trials to accumulate reliable clinical evidence and meet medical regulatory requirements.

Overall, wearable sensors have achieved rapid development in physiological, biochemical, and environmental signal sensing. Although multiple challenges remain, we believe that with the synergistic advancement of interdisciplinary technologies including materials science, electronics, artificial intelligence, and medicine, wearable sensors will ultimately facilitate a shift from passive treatment in traditional hospital settings to proactive health management (Figure ). This will enable convenient cloud-based healthcare services and drive further progress in precision and personalized medical health management.

11.

Schematic illustration of wearable sensors in facilitating the transition from traditional passive healthcare to active home health management and their future outlook.

The authors declare no competing financial interest.