From Sensitivity to Temporal Fidelity in Wearable Sweat Sensing

Wearable sweat sensors are judged mainly by analytical sensitivity, yet a detection limit cannot tell when a measured concentration was secreted. At low or intermittent sweat rates, fluid retained in the device delays and mixes successive samples, so each reading reflects a distribution of sample ages rather than a single moment. In this Perspective, we argue that temporal fidelity should stand alongside sensitivity as a design target. We treat it as a property of the whole measurement chain, from biological transfer through sampling topology to sensor kinetics. Minimal transport and sensor models relate the tolerable transport time to the timescale of the physiological feature, and published sweat glucose and chloride traces test the resulting envelopes. Under the illustrative fidelity criteria used here, no transport threshold is universal. Hour-scale trends tolerate mean transport times on the order of 10 min, whereas features of 30 min or shorter generally require renewal within a few minutes or less. For a 30 min feature read by a 1 min sensor, the lowest reported resting sweat rate limits the coupled sensing volume over a 3 mm collection area to about 1 nL. Transport and sensor lags can also be indistinguishable in the device output, so each stage must be characterized separately. We therefore map five sampling architectures to the temporal claims each can support and propose a decision-first validation sequence with minimum reporting items. Sensitivity remains necessary, but interpretable sweat data also require sample age and renewal to be measured rather than assumed.