Development of a wearable multimodal bio-microfluidic chip for tracking exercise metabolism

In situ monitoring of endogenous amino acids in sweat can provide physiological insights into health and metabolism. However, existing amino acid biosensors are unable to quantitatively assess metabolic status during exercise and are rarely used to establish blood-sweat correlations because they only detect a single concentration indicator and ignore sweat rate.

To solve the above problems, Wang Lili's team from the Institute of Semiconductors, Chinese Academy of Sciences, introduced a wearable multimodal bio-microfluidic chip that integrates advanced electrochemical electrodes and multifunctional microfluidic channels, capable of simultaneously quantifying multiple sweat indicators, including phenylalanine (Phe) and chloride ions, as well as sweat rate. This comprehensive measurement method revealed a negative correlation between sweat phenylalanine levels and sweat rate between individuals, further making it possible to identify individuals at high metabolic risk.

By tracking phenylalanine fluctuations caused by protein intake during exercise, and by normalizing sweat rate to concentration indicators to reduce inter-individual variation, a reliable method was demonstrated to correlate and analyze sweat-blood phenylalanine levels, which can be used for personal health monitoring. The study was published in the journal Nature Communications under the title "Interindividual- and blood-correlated sweat phenylalanine multimodal analytical biochips for tracking exercise metabolism".

This study introduces a wearable multimodal bio-microfluidic chip for sensing multiple metrics, including phenylalanine and chloride concentrations in sweat, as well as sweat rate, which together enable quantitative assessment of metabolic status during exercise. This bio-microfluidic chip integrates three functional modules for advanced in situ sweat detection (Figure 1a): (i) an electrochemical electrode (E-MIP) modified with an electrocatalytically active molecularly imprinted polymer (MIP) for direct and selective determination of phenylalanine in sweat; (ii) a well-designed multifunctional microfluidic module that allows rapid sweat sampling, concentration refreshing, and pH buffering to maintain a stable detection environment, as well as flow rate visualization for sweat loss measurement; and (iii) integration with matching wireless flexible circuits and mobile software. Using this bio-microfluidic chip approach, the changes in phenylalanine levels between two groups of human test subjects with different body mass index (BMI) values ​​were studied, which can be attributed to differences in sweat rate. Using this negative correlation, the researchers analyzed the possible mechanism of phenylalanine entry into sweat.

In addition, the researchers evaluated the exercise metabolic status and risk of volunteers with different physiological characteristics by using a composite indicator, phenylalanine secretion rate (SP), which is derived from sweat phenylalanine concentration (CP) and sweat rate (RW) (Figure 1b). Finally, the similarity and strong correlation between sweat and serum phenylalanine levels in different volunteers were demonstrated after normalization by sweat rate to reduce inter-individual variability. All these demonstrations reveal the potential practicality of wearable multimodal systems in sweat-based personalized exercise and diet management.

Figure 1 Schematic diagram of a wearable multimodal biochip assessing exercise metabolic risk and serum correlation through sweat analysis

Electrochemical characterization of the E-MIP sensor

Figure 2 shows the properties of the E-MIP sensor for phenylalanine detection, including the selective recognition of phenylalanine and the effective discrimination of other amino acids at physiologically relevant high concentrations. In addition, the sensor also demonstrated enantioselectivity for L-phenylalanine versus D-phenylalanine, which was due to the use of L-phenylalanine as an imprinting template. The electrochemical properties of the sensor were evaluated by cyclic voltammetry and electrochemical impedance spectroscopy, showing the stability of the response to phenylalanine under different pH conditions. By differential pulse voltammetry scanning, a clear increase in peak current readings was detected in the phenylalanine concentration range of 10 to 1500 μM, and two linear relationships were determined. These results indicate that the E-MIP electrode is superior to other detection electrodes in the direct, selective, and sensitive electrocatalytic oxidation of phenylalanine.

Figure 2 Characterization of the E-MIP sensor for phenylalanine detection

Design and performance characterization of multi-purpose microfluidics

The design and performance characteristics of the multifunctional microfluidic module designed for the wearable sweat phenylalanine sensor are shown in Figure 3. The results show that the microfluidic module can effectively improve the temporal resolution of sweat sampling and provide real-time readout of sweat loss through visual flow channels while maintaining a solution environment with neutral pH and high ionic strength to ensure accurate sweat phenylalanine detection.

Figure 3 Design and characterization of a versatile microfluidic system for sweat sampling

Evaluation of sensors for assessing metabolic risk of exercise

Figure 4 shows the in vivo evaluation of the wearable multimodal bio-microfluidic chip for dynamic exercise sweat analysis and metabolic assessment. Photos of subjects wearing the bio-microfluidic chip and screenshots of the smartphone application interface show the actual application and user interface of the bio-microfluidic chip. The ultrathin and flexible demonstration of the sensor emphasizes its design suitable for long-term wear. The hardware block diagram reveals the flexible circuit of the sweat sensor, including modules for exciting differential pulse voltammetry (DPV) potential and differential open circuit potential (OCP) measurements, as well as the corresponding multimodal biosignal retrieval, processing and transmission functions. The real-time continuous monitoring chart and DPV data obtained from the subject's forehead provide real-time information on the changes in sweat composition during exercise. Dynamic sweat phenylalanine measurement and corresponding box plots compare sweat phenylalanine levels of subjects with different BMI values, revealing metabolic differences between people with different physiques during exercise. The positive correlation between sweat phenylalanine secretion rate and sweat amino acid levels verifies its potential as an indicator for exercise metabolic risk assessment. By normalizing the sweat rate, the inter-individual variation can be reduced, allowing a more accurate assessment of the correlation between serum and sweat phenylalanine levels.

Figure 4 In vivo evaluation of a wearable multimodal bio-microfluidic chip for dynamic exercise sweat analysis and assessment

Sweat phenylalanine sensor for dietary management and evaluation of serum correlation

Figure 5 shows real-time sweat phenylalanine analysis for evaluating the effects of serum levels and protein dietary intake during exercise. The metabolic pathways of phenylalanine in sweat and the fluctuations in serum and sweat phenylalanine levels due to protein intake during exercise, as well as the dynamic changes in sweat phenylalanine levels in subjects with different BMI during three stages of exercise (including before and after protein dietary intake and after rest), are helpful in studying the metabolic differences between people of different physiques during exercise. The comparison of sweat phenylalanine concentrations measured by the sensor with enzyme-linked immunosorbent assay (ELISA) readings verified the accuracy of the sensor measurement. In addition, the study also demonstrated the relative cell viability of HaCaT cells after biocompatibility testing, evaluating the safety and biocompatibility of the bio-microfluidic chip. This indicates that the bio-microfluidic chip is suitable for long-term wear and multimodal detection of sweat. These results indicate that sweat phenylalanine secretion rate is a potential suitable indicator for studying the correlation between serum and sweat phenylalanine by introducing sweat rate to reduce inter-individual variability.

Figure 5 In situ sweat phenylalanine analysis for assessment of serum levels and the effect of dietary protein intake

In summary, this study used a wearable multimodal bio-microfluidic chip to monitor the concentration of phenylalanine and chloride ions in human sweat and sweat rate in real time during human exercise, thereby achieving a quantitative assessment of metabolic status. This bio-microfluidic chip integrates three functional modules, including electrochemical electrodes modified with electrocatalytically active molecularly imprinted polymers (MIPs), multifunctional microfluidic modules, and integration with wireless flexible circuits and mobile software. Using this bio-microfluidic chip approach, the research team explored the changes in phenylalanine levels between groups of human subjects with different BMI values ​​and analyzed the possible mechanisms by which phenylalanine enters sweat.

In addition, by using a composite indicator, phenylalanine secretion rate (SP), the researchers evaluated the exercise metabolic status and risk of volunteers with different physiological characteristics. Finally, the study demonstrated similar and strong correlations between sweat and serum phenylalanine levels, especially after reducing inter-individual variability by normalization of sweat rate. These findings reveal the potential practicality of wearable multimodal microfluidic systems in sweat-based personalized exercise and diet management.