Fluid-based wearable sensors: a turning point in personalized healthcare

**Background:** Wearable sensor technology has evolved from large benchtop instruments to miniaturized, flexible devices capable of noninvasively monitoring biomarkers in body fluids. This review focuses on sensors for saliva, sweat, tears, and interstitial fluid (ISF), highlighting their potential for personalized healthcare, early disease detection, and remote patient monitoring. The paper analyzes nearly 50 articles published between 2017 and 2023.

**Methods:** The authors conducted a narrative review of the literature, categorizing wearable sensors by biofluid type: saliva, sweat, tears, and ISF. For each category, they summarized key studies, sensor platforms (e.g., mouth guards, wristbands, contact lenses, microneedles), transduction methods (primarily electrochemical and optical), and target analytes. Performance metrics such as sensitivity, detection limits, and linear ranges were extracted from the reviewed studies.

**Key Results:** - **Saliva-based sensors:** Platforms include mouth guards, pacifiers, wristbands, and tooth tattoos. Wang et al. developed the first wearable real-time mouth guard sensor for lactate using screen-printed enzymatic electrodes. Kim et al. created a mouth guard for uric acid detection with a modified screen-printed electrode system. Mitsubayashi et al. designed a mouth guard for salivary glucose (range 1.75–10,000 μmol/L). Mishra et al. developed a ring sensor for simultaneous THC and alcohol detection (alcohol linear range 0.1–0.6 mM). Laura et al. created a pacifier sensor for glucose monitoring in newborns. A 2022 wristband sensor tracked phenylalanine in serum and saliva (dynamic range 20–1000 μM). Eom et al. reported a cholesterol sensor with a limit of detection (LOD) of 2 μM. - **Sweat-based sensors:** Platforms include wristbands, headbands, gloves, skin patches, and tattoos. Jagannath et al. developed a wristband for cytokines (interleukin-6, -8, -10, tumor necrosis factor) with a reproducible response range of 2–200 pg/mL. Ghoorchian et al. created a headband sensor for Na⁺ using Na₀.₄₄MnO₂ (linear range 0.21–24.54 mmol/L). Bolat et al. designed a soft epidermal microfluidic device for sweat glucose combining iontophoretic pilocarpine delivery. Liu et al. presented a flexible photoelectrochemical skin patch for glucose with a LOD of 22.2 pM. Cui et al. developed an optical glucose skin pad using fluorescent nanohybrids. Wang et al. reported an ultrasmall sweat biosensor for glucose, lactate, Na⁺, and K⁺ after 10 min of cycling. A 2023 study used a flexible LSPR aptasensor for cortisol on PDMS. - **Tear-based sensors:** Contact lens platforms dominate. Moreddu et al. developed microfluidic contact lenses for pH (sensitivity 12.23 nm/pH, LOD 0.25 pH units), glucose, protein, and nitrite. Kajisa et al. created a hydrogel FET glucose sensor. Zhou et al. used FeₓCoᵧO₄-rGO on flexible carbon electrodes for glucose (LOD 0.07 M). Huang et al. developed a flexible GFET nanosensor for L-cysteine (LOD 0.043×10⁻⁶ M in artificial tears). Park et al. designed colorimetric contact lenses with cerium oxide nanoparticles for glucose, tested in diabetic rabbits. Deng et al. created smart contact lenses with a glucose fluorescence probe (range 23 μM–1.0 mM). Sempionatto et al. integrated a tear alcohol biosensor into eyeglass nose bridge pads. Kalasin et al. developed glasses for tear creatinine using Cu-BDC MOF/GO-Cu(II)/Cu₂O NP (selectivity 95.1%, range 1.6–2400 μM). - **ISF-based sensors:** Platforms use reverse iontophoresis (RI) or microneedles (MNs). Yang et al. combined RI and MNs for Epstein-Barr virus cell-free DNA extraction (95.4% efficiency in 10 min). Cheng et al. developed a touch biosensor for glucose using solid MNs and RI, increasing extraction flow by ~1.6 times. Bolella et al. created a highly porous gold MN biosensor for glucose (linear range 0.1–10 mM). Zheng et al. used silk fibroin-based MN patches for glucose. Kemp et al. introduced magnetohydrodynamic extraction for glucose (sensitivity 0.8 mA/mol·M·cm²). Teymuryan et al. developed a continuous ketone body monitoring MN platform (LOD 50 μM). Sempionatto et al. integrated sensors for blood pressure, heart rate, glucose, lactate, caffeine, and alcohol. Kim et al. developed a dual device for sweat-alcohol and ISF-glucose. Goud et al. created an MN patch for apomorphine (sensitivity 45 nA/μM, R²=0.998).

**Clinical Implications:** Wearable sensors enable noninvasive, real-time monitoring of biomarkers, facilitating early disease detection and personalized treatment. Saliva sensors offer painless sampling for glucose, uric acid, and drugs; sweat sensors track electrolytes, metabolites, and cytokines; tear sensors provide ocular and systemic biomarker data; ISF sensors allow minimally invasive glucose and drug monitoring. Challenges include low biomarker concentrations, contamination, calibration, biocompatibility, and standardization. Future directions include multiplexed sensing, improved materials, and integration with wireless data transmission for remote healthcare.