Stretchable bioelectronics have advanced beyond conforming to soft tissues to sustaining stable signal acquisition during continuous tissue deformation^[1,2]. Extending this reliability to molecular sensing is more demanding because electrochemical readouts depend on stable charge transport, controlled electrode–biofluid interactions, and interfacial redox reactions^[3,4]. Previous stretchable electrochemical sensors have mitigated strain-induced failure using serpentine or mesh interconnects^[5,6], hydrogel electrodes^[7,8], composite conductors^[5,9,10], liquid-metal conductors^[11], soft adhesive interfaces^[9,12], and implantable electrodes^[13] [Table 1]. However, these strategies stabilize continuity in only one segment of the sensing pathway, rather than the electrochemical interface as a whole, while strain mismatch in multilayer stacks can drive delamination^[14-16]. Because deformation alters resistance, electrochemically active area, charge-transfer kinetics, and sensing-layer permeability^[17], the central challenge is not stretchability itself but preserving quantitative molecular transduction during deformation^[14,18].

To address this coupled problem, Xu and colleagues introduced an intrinsically stretchable interface for resilient electrochemical sensing (SIRES)^[15]. Unlike earlier geometric designs that route strain around rigid sensing regions, SIRES makes the entire interface deformable, shifting the design strategy from avoiding strain to compensating for its electrochemical effects. As shown in Figure 1A, SIRES is a fully elastomeric trilayer comprising a liquid-metal strain-resilient conductor (SRC) for charge transport, a carbon nanotube (CNT)–polyurethane electrically tunable interface (ETI) for electromechanical coupling, and a stretchable functional coating (SFC) hosting the sensing chemistry. These layers are covalently integrated within one polyurethane matrix, enabling them to deform together without delamination^[15]. Under strain, the SFC exposes a larger electrochemically active area, increasing faradaic current and lowering charge-transfer resistance (R_ct)^[15,19], while elongation of the conductive pathway increases the series resistance that would otherwise shift or attenuate redox peaks. Because the ETI resistance is tunable through CNT loading^[10,15], its strain-dependent increase is matched to the area-driven decrease in R_ct. These opposing contributions therefore cancel, and the total resistance remains nearly constant, as captured by a modified Randles model [Figure 1B and C]. Quantitatively, R_ct decreases from 253.6 to 173.2 Ω between 0% and 200% strain, while the double-layer capacitance increases by approximately 23%; their product therefore remains within 3% of its initial value at 100% strain. Despite these changes, the voltammetric response is largely retained, with the redox peak current varying by no more than 14% over the same strain range. However, different layers contribute to this stability depending on the analyte. ETI resistance matching operates where faradaic current and electrode area determine the signal, as in voltammetric detection. Amperometric and potentiometric readouts instead depend more strongly on the functional coating and stretchable reference electrode, which preserve the sensing chemistry and potential baseline. Wearable and organ-mounted demonstrations [Figure 1D and E] confirm that this principle extends to complete devices.

Yet signal stability under strain should not be interpreted as calibration stability during chronic use. The reported robustness over 1,000 cycles at 100% strain was established on a clean interface under repeatable deformation^[15], rather than under the fouling-rich, continuously perturbed conditions of on-body or implanted use. More fundamentally, electrochemical compensation assumes that the electrochemically active area varies only with strain. In practice, biofouling, nonspecific adsorption, and biofluid residues progressively block reaction sites, and this area loss distorts the calibrated signal through a pathway that the strain-tuned circuit cannot sense^[19,20]. Moreover, a preserved voltammetric readout does not necessarily indicate calibration stability: glucose sensitivity still shifts by approximately 4.7% at 300% strain, and such residual deviations can matter when quantifying subtle biomarker changes, requiring device-specific and potentially strain-state-specific recalibration. Likewise, because glucose, lactate, pH, H₂O₂, and uric acid rely on distinct enzymes, mediators, and membranes, the optimal electrochemical compensation and calibration cannot be assumed to transfer unchanged among them. Validation is also largely uniaxial, whereas organs deform biaxially and nonuniformly over curved surfaces, often with concurrent shear and torsion^[21-23]. Applying biaxial, shear, and torsional loading at organ-relevant amplitudes and frequencies, and reporting calibration stability across directions and repeated cycles rather than signal stability alone, would bring the evaluation closer to actual operating conditions. Reference-electrode drift, mediator instability, and liquid-metal isolation further compound these effects, making calibration stability in complex biofluids during prolonged deformation and implantation challenging^[24,25].

Looking forward, the next step is to move from strain-resilient interfaces to organ-specific molecular sensing systems [Figure 1F]. Because each organ deforms differently, calibration should become deformation-aware. Data-driven or artificial intelligence (AI)-assisted models could enable this approach^[26-29]. Electrochemical compensation addresses the immediate effect of deformation, whereas the model would address slowly accumulating drift. Using ETI resistance and periodic impedance measurements, it could estimate the mechanical state of the interface, predict drift for a given organ and measurement schedule, and indicate when recalibration is needed, with predictions verified against concurrent reference measurements. Organ-specific design is equally important: acidic gastric distension, intestinal peristalsis, and cyclic bladder filling impose different chemical and mechanical demands, so adhesives, encapsulation, and sensing chemistry cannot be interchangeable^[23,30]. Chronic stability must also be built into the sensing layer itself, which should remain selective for the target analyte while resisting fouling that degrades performance. This requires sterilizable encapsulation and long-term tissue compatibility^[24,25]. By reframing stretchability as preservation of electrochemical accuracy rather than mechanical survival alone, SIRES marks a conceptual turning point for the field. Building on this foundation, next-generation biointerfaces could enable molecular monitoring that remains reliable across the full range of tissue motion, bringing organ-level biochemical monitoring closer to clinical reality.