A Wearable Electrochemical Biosensor For The Monitoring of Metabolites and Nutrients

Articles
https://doi.org/10.1038/s41551-022-00916-z
A wearable electrochemical biosensor for the

monitoring of metabolites and nutrients
Minqiang Wang1,7, Yiran Yang1,7, Jihong Min1,7, Yu Song1, Jiaobing Tu1, Daniel Mukasa2, Cui Ye1,
Changhao Xu1, Nicole Heflin3, Jeannine S. McCune4, Tzung K. Hsiai5, Zhaoping Li6 and Wei Gao 1 ✉
Wearable non-invasive biosensors for the continuous monitoring of metabolites in sweat can detect a few analytes at sufficiently
high concentrations, typically during vigorous exercise so as to generate sufficient quantity of the biofluid. Here we report the
design and performance of a wearable electrochemical biosensor for the continuous analysis, in sweat during physical exercise
and at rest, of trace levels of multiple metabolites and nutrients, including all essential amino acids and vitamins. The biosensor
consists of graphene electrodes that can be repeatedly regenerated in situ, functionalized with metabolite-specific antibody-like
molecularly imprinted polymers and redox-active reporter nanoparticles, and integrated with modules for iontophoresis-based
sweat induction, microfluidic sweat sampling, signal processing and calibration, and wireless communication. In volunteers,
the biosensor enabled the real-time monitoring of the intake of amino acids and their levels during physical exercise, as well
as the assessment of the risk of metabolic syndrome (by correlating amino acid levels in serum and sweat). The monitoring of
metabolites for the early identification of abnormal health conditions could facilitate applications in precision nutrition.
C
irculating nutrients are essential indicators for overall health sensors to monitor an individual’s health state and to enable timely
and body function1. Amino acids (AAs), sourced from intervention under home- and community-based settings15–23; it is
dietary intake and gut microbiota synthesis, and influenced also increasingly important to monitor a person’s long-term cardio-
by personal lifestyles, are important biomarkers for a number of metabolic and nutritional health status after recovery from severe
health conditions (Fig. 1a)2. Elevated branched-chain amino acids COVID-19 infection using wearables to capture early signs of
(BCAAs), including leucine (Leu), isoleucine (Ile) and valine (Val), potential endocrinological complications such as T2DM12.
are associated with obesity, insulin resistance and future risk of Sweat is an important body fluid containing a wealth of chemicals
type 2 diabetes mellitus (T2DM), cardiovascular diseases (CVDs) reflective of nutritional and metabolic conditions24–27. The progres-
and pancreatic cancer3–5. Deficiencies in AAs (for example, argi- sion from blood analyses to wearable sweat analyses could provide
nine and cysteine) could hamper the immune system by reduc- great potential for non-invasive, continuous monitoring of physio-
ing immune-cell activation6. Tryptophan (Trp), tyrosine (Tyr) and logical biomarkers critical to human health28–38. However, currently
phenylalanine (Phe) are precursors of serotonin and catecholamine reported wearable electrochemical sensors focus primarily on a
neurotransmitters (dopamine, norepinephrine and epinephrine), limited number of analytes including electrolytes, glucose and lac-
respectively, and play an important role in the function of complex tate, owing to the lack of a suitable continuous monitoring strategy
neural systems and mental health7,8. A number of metabolic finger- beyond ion-selective and enzymatic electrodes or direct oxidation
prints (including Leu, Phe and vitamin D) are linked to coronavirus of electroactive molecules25–27,34–40. Thus, most clinically relevant
disease 2019 (COVID-19) severity9,10. Health disparities in nutri- nutrients and metabolites in sweat are rarely explored and unde-
tion also correlate well with the alarming racial and ethnic dispari- tectable by existing wearable sensing technologies. Moreover, cur-
ties that are worsened by COVID-19 vulnerability and mortality11. rent wearable biosensors usually require vigorous exercise to access
Moreover, organ and tissue dysfunction induced by severe acute sweat; although a few recent reports use pilocarpine gel-based ion-
respiratory syndrome coronavirus 2 could result in an increased tophoresis for sedentary sweat sampling22,30,36, this approach suffers
incidence of cardiometabolic diseases12. from short sweat periods and low sensing accuracy due to the mix-
Metabolic profiling and monitoring are a key approach to ing of sweat and gel fluid and the lack of dynamic sweat sampling.
enabling precision nutrition and precision medicine13. Current gold In this Article, we present a universal wearable biosensing
standards in medical evaluation and metabolic testing heavily rely strategy based on a judicious combination of the mass-producible
on blood analyses that are invasive and episodic, often requiring laser-engraved graphene (LEG), electrochemically synthesized
physical visits to medical facilities, labour-intensive sample pro- redox-active nanoreporters (RARs) and molecularly imprinted
cessing and storage, and delicate instrumentation (for example, polymer (MIP)-based ‘artificial antibodies’, as well as unique
gas chromatography–mass spectrometry (GC–MS))14. As the cur- in situ regeneration and calibration technologies (Fig. 1b). Unlike
rent COVID-19 pandemic remains uncontrolled around the world, bio-affinity sensors based on antibodies or classic MIPs, which
there is a pressing need for developing wearable and telemedicine are generally for one-time use and require multiple washing steps
1
Andrew and Peggy Cherng Department of Medical Engineering, Division of Engineering and Applied Science, California Institute of Technology, Pasadena,
CA, USA. 2Department of Applied Physics and Materials Science, Division of Engineering and Applied Science, California Institute of Technology, Pasadena,
CA, USA. 3Department of Electrical Engineering, Division of Engineering and Applied Science, California Institute of Technology, Pasadena, CA, USA.
4
Department of Hematologic Malignancy Translational Sciences, Beckman Research Institute at City of Hope, Duarte, CA, USA. 5Division of Cardiology,
David Geffen School of Medicine, University of California, Los Angeles, CA, USA. 6Division of Clinical Nutrition, David Geffen School of Medicine, University
of California, Los Angeles, CA, USA. 7These authors contributed equally: Minqiang Wang, Yiran Yang, Jihong Min. ✉e-mail: weigao@caltech.edu
Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng 1225

Articles NaTure BIomedIcal EngIneerIng
a b
+/– Target
Current
Voltage
Diet Lifestyle
regeneration
Recognition
In situ
Gut microbiota
R Artificial antibody
H
OH Oxidation Reduction
H2 N
O
Amino acids
COVID-19 Central fatigue Cardiac hypertrophy Hepatic lipid storage T2DM

System on body
c d e f
Sweat biosensors T sensor PI film
Sensors
Carbachol gel
Channel layers
Sweat Inlet layers

stimulation
electrodes
Skin
g h i
Protection circuit Iontophoresis
Boost converter
MCU
electrodes
GPIO Current source IP +
VREF IP –
SPI 1 Dual DAC
Charging/load sharing circuit
Sweat
Potentiostat
CPU biosensors
interface
CE
Battery
ADC 1 TIA 1 WEDPV1

WEDPV2
ADC 2 TIA 2 RE
ADC 3 In-amp WEPOT
Voltage regulator
ADC 4 Voltage divider T sensor
UART Bluetooth User interface
SPI 2 Display
Fig. 1 | Schematics and images of the wearable biosensor ‘NutriTrek’. a, Circulating nutrients such as AAs are associated with various physiological
and metabolic conditions. b, Schematic of the wearable ‘NutriTrek’ that enables metabolic monitoring through a synergistic fusion of LEG, RARs and
artificial antibodies. c,d, Schematic (c) and layer assembly (d) of the microfluidic ‘NutriTrek’ patch for sweat induction, sampling and biosensing. T,
temperature. e,f, Images of a flexible sensor patch (e) and a skin-interfaced wearable system (f). Scale bars, 5 mm (e) and 2 cm (f). g, Block diagram
of electronic system of ‘NutriTrek’. The modules outlined in red dashes are included in the smartwatch version. CPU, central processing unit; POT,
potentiometry; In-Amp, instrumentation amplifier; MCU, microcontroller; TIA, trans-impedance amplifier; IP, iontophoresis; CE, counter-electrode; RE,
reference electrode; WE, working electrode. h, Custom mobile application for real-time metabolic and nutritional tracking. i, ‘NutriTrek’ smartwatch with a
disposable sensor patch and an electrophoretic display. Scale bars, 1 cm (top) and 5 cm (bottom).
to transduce the bio-affinity interactions in standard ionic solu- biomarkers in biofluids including all nine essential AAs as well
tions41,42, this approach enables the demonstration of sensitive, as vitamins, metabolites and lipids commonly found in human
selective and continuous monitoring of a wide range of trace-level sweat (Supplementary Table 1). Seamless integration of this unique
1226 Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng

NaTure BIomedIcal EngIneerIng Articles
approach with in situ signal processing and wireless communication (Fig. 2b,c and Supplementary Fig. 14). Linear relationships between
leads to a powerful wearable sweat sensing technology ‘NutriTrek’ peak height current densities and target concentrations with sensi-
that is able to perform personalized and non-invasive metabolic and tivities of 0.63 µA µM−1 cm−2 and 0.71 µA µM−1 cm−2 respectively for
nutritional monitoring towards timely intervention (Fig. 1b). The the LEG–MIP Tyr and Trp sensors were observed (Supplementary
incorporation of the carbachol iontophoresis-based sweat induc- Fig. 15). It is worth noting that choices of monomer/crosslinker/
tion and efficient microfluidic-based surrounding sweat sampling template ratios and incubation periods have substantial influences
enables prolonged autonomous and continuous molecular analysis on sensor response while sample volume does not (Supplementary
with high temporal resolution and accuracy across activities, dur- Fig. 10). The Tyr and Trp sensors can be readily and repeatably
ing physical exercise and at rest. Using five essential or condition- regenerated in situ without any washing step with a high-voltage
ally essential AAs (that is, Trp, Try and three BCAAs (Leu, Ile and amperometry current–time (IT) that oxidizes the bound targets at
Val)) as exemplar nutrients, we corroborated the system in several their redox potentials (Fig. 2d).
human trials by enroling both healthy subjects and patients towards As the majority of metabolites and nutrients (for example,
personalized monitoring of central fatigue, standard dietary intakes, BCAAs) are non-electroactive and cannot easily be oxidized
nutrition status, metabolic syndrome risks and COVID-19 severity. under operational conditions, we herein utilize an indirect detec-
tion approach involving an RAR layer sandwiched between the
Results LEG and MIP layers to enable rapid quantitation (Fig. 2e). The
Design and overview of the autonomous wearable biosensor selective adsorption of the target molecules onto the imprinted
technology. The flexible and disposable sensor patch consists of polymeric layer decreases the exposure of the RAR to the sample
two carbachol-loaded iontophoresis electrodes, a multi-inlet micro- matrix. Controlled-potential voltammetric techniques such as DPV
fluidic module, a multiplexed MIP nutrient sensor array, a tempera- or linear sweeping voltammetry (LSV) can be applied to measure
ture sensor and an electrolyte sensor (Fig. 1c–f and Supplementary the RAR’s oxidation or reduction peak, where the decrease in peak
Fig. 1). All flexible electrode and sensor designs are based on the height current density corresponds to an increase in analyte levels.
LEG, which has large surface area, has excellent electrochemi- For example, using Prussian Blue nanoparticles (PBNPs) as the
cal properties and can be produced at a large scale directly on a RAR (Supplementary Fig. 11), we developed an MIP–LEG Leu sen-
polyimide (PI) substrate via CO2 laser engraving (Supplementary sor with a log-linear relationship between the peak height decrease
Fig. 2). The sensor patch can be easily attached to skin with confor- and Leu concentration and a sensitivity of 702 nA mm−2 per decade
mal contact and interfaces with a miniaturized electronic module of concentration (Fig. 2f). We established this approach to quan-
for on-demand iontophoresis control, in situ signal processing and tify the physiologically relevant range of all nine essential AAs (that
wireless communication with the user interfaces through Bluetooth is, Leu, Ile, Val, Trp, Phe, histidine (His), lysine (Lys), methionine
(Fig. 1g and Supplementary Figs. 3 and 4). A custom mobile app (Met) and threonine (Thr)) (Fig. 2g and Supplementary Fig. 16) as
‘NutriTrek’ was developed to process, display and store the dynamic well as a number of vitamins, metabolites and lipids (vitamins B6,
metabolic information monitored by the wearable sensors (Fig. 1h C, D3 and E, glucose, uric acid, creatine, creatinine and cholesterol)
and Supplementary Video 1). The wearable system was also inte- (Fig. 2h and Supplementary Fig. 17). In addition to these nutrients
grated into a smartwatch with an electronic paper display (Fig. 1i and metabolites, this approach can be easily reconfigured to enable
and Supplementary Fig. 5). the monitoring of a broad spectrum of biomarkers ranging from
hormones (for example, cortisol) to drugs (for example, immuno-
Biosensor design and evaluation for universal metabolic and suppressive drug mycophenolic acid) (Supplementary Fig. 18 and
nutritional analysis. Universal detection of AAs and other metab- Supplementary Tables 2 and 3). Most of these targets are undetect-
olites/nutrients with high sensitivity and selectivity was achieved able continuously by any existing wearable technology. Considering
through careful design of the selective binding MIP layer on the that a total level of multiple nutrients (for example, total BCAAs) is
LEG. MIPs are chemically synthesized receptors formed by polym- often an important health indicator, a multi-template MIP approach
erizing functional monomer(s) with template molecules. Although can be used to enable accurate and sensitive detection of the total
MIP technology has been proposed for sensing, separation and concentration of multiple targets with a single sensor (Fig. 2i,j).
diagnosis42,43, it has not yet been demonstrated for continuous These indirect LEG–RAR–MIP sensors can be regenerated in situ
wearable sensing as classic MIP sensors require washing steps by applying constant potential to the working electrode, which
for sensor regeneration and the detection is generally performed repels the bound target molecules from the MIP layer, achieving
in standard buffer or redox solutions. In our case, the functional prolonged re-usability (Fig. 2k).
monomer (for example, pyrrole) and crosslinker (for example, The LEG–MIP sensors show stable responses during repeatable
APBA) initially form a complex with the target molecule; follow- use: the PBNP-based RAR showed stable redox signals through-
ing polymerization, their functional groups are embedded in the out 60 repetitive cyclic voltammetry (CV) scans (Fig. 2l and
polymeric structure on the LEG; subsequent extraction of the tar- Supplementary Fig. 11); minimal output changes were observed
get molecules reveals binding sites on the LEG-MIP electrode that throughout a 42-day storage period (Supplementary Fig. 19a,b);
are complementary in size, shape, and charge to the target analyte the sensors also showed no substantial relative signal shift when
(Supplementary Fig. 6). Two detection strategies—direct and indi- used continuously over 5 days (Supplementary Fig. 19c). Compared
rect—are designed on the basis of the electrochemical properties of with traditional MIP preparation processes, the electrodeposited
the target molecules (Fig. 2). Optimizations and characterizations MIP layer on the mass-producible LEG leads to high reproduc-
of the LEG–MIP sensors are detailed in Supplementary Note 1 and ibility in selectivity, sensitivity and device-to-device consistency
Supplementary Figs. 7–13. (Supplementary Figs. 20 and 21). The choice of LEG as the MIP
For electroactive molecules in sweat, the oxidation of bound tar- deposition substrate also showed advantages in sensor sensitivity
get molecules in the MIP template can be directly measured by dif- compared with classic electrodes such as glassy carbon electrode,
ferential pulse voltammetry (DPV) in which the peak current height printed carbon electrode and Au electrode (Supplementary Fig. 22).
correlates with analyte concentration (Fig. 2a). Considering that Other RARs such as anthraquinone-2-carboxylic acid (AQCA) can
multiple electroactive molecules can be oxidized at similar poten- also be used for indirect AA sensing with stable performance (nega-
tials, this LEG–MIP approach addresses both sensitivity and selectively scanned DPV was used here to monitor AQCA reduction)
tivity issues. For example, Tyr and Trp, two AAs with close redox (Fig. 2m and Supplementary Fig. 23). As illustrated in Fig. 2n, the
potentials (~0.7 V), could be detected selectively with this strategy LEG–AQCA–MIP sensors could be directly regenerated in a raw

a b c d
8 10
ΔJ (µA mm–2) Potential (V)

Current density (µA mm–2)

Direct detection 3 Tyr 4 Trp 1.0 DPV
ΔJ (µA mm )
ΔJ (µA mm )
DPV
–2
–2
NH2 N 2 0.8 IT IT
7 2
HO
1 0.6
DPV O +Target 8 0.4
V I 0 0
6 Time
0
0
0
0
0
0
20
40
20
40
0.8
T Oxidized V (Tyr) (µM) (Trp) (µM)
6 0.6
5
n
NH
HN 0.4
NH
HO O H 0.2
HN N
B
N
4 4 0
O O
n
LEG 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0 6 12 18

Potential (V) Potential (V) Number of regeneration
e f g h
0 3.6 3.6
Indirect detection Lys Phe VC Creatine

Val Thr VB6 Creatinine
NH2 CH3 Leu His Cholesterol
VD3
ΔJ (µA mm–2)
ΔJ (µA mm–2)
HO
2.4 Ile 2.4
LSV CH3 +Target VE
V O I 2 Met
Uric acid
ΔJ (µA mm )
–2
–5 Trp
Glucose
Block 1
T V 1.2 1.2
CH3
n
NH CH3 NH 0
HO O 1 2 3
HN HN N log10((Leu) (µM)) Leu
B O O –10 0 0
n
RAR 0 0.1 0.2 0.3 1 1.5 2 2.5 1 1.5 2 2.5

LEG
Potential (V) log10((amino acid) (µM)) log10((analyte) (µM))
i j k l
0 ΔJ (µA mm–2) Potential (V) 24

Multi-template 0.4 LSV LSV Initial Fe4(Fe(CN)6)3
0.2 30 cycles
CH3 CH3 CH3 –2 12 60 cycles
0
R= IT IT
3.2
–0.2
CH3 CH3 CH3
ΔJ (µA mm )
–2
–4 Time 0
1.6 1.2
N O O HN R2 0.8
NH –6 0 –12
O O
1.5 2.5 3.5 0.4
n
HN N
R1 NH B O O BCAAs
log10((BCAAs) (µM))
–8 0 –24
n
RAR
LEG 0 0.1 0.2 0.3 0 4 8 12 –0.1 0.1 0.3 0.5
Potential (V) Number of regeneration Potential (V)
m n o p
Concentration by sensor (µM)

0 0.4 150
Raw sweat Leu sensor Sweat Tyr

1 Sweat Trp
Sweat Leu
ΔJ (µA mm–2)
–4 100
ΔJ (µA mm–2)
1.8
ΔJ (µA mm )
0.2
–2
0.5
0.9
–8 50
0
1 2 3 S
SIleVal
Leu log10((Leu) (µM)) 0 0 S
–12 s r S Leu 0
AA Ty Trp u STyrTrp
–0.8 –0.6 –0.4 –0.2 0 0 4 8 12 BC Le IIe al SBCAAs 0 50 100 150
V
Potential (V) Number of regeneration Conc. by GC-MS (µM)
Fig. 2 | Schematics and characterizations of the LEG–MIP sensors. a, Direct detection of electroactive molecules using LEG–MIP sensors. b,c, DPV
voltammograms of the LEG–MIP sensors for direct Tyr (b) and Trp (c) detection. Insets, calibration plots with a linear fit. ∆J, peak height current density.
d, In situ continuous sensing and regeneration of an LEG–MIP Trp sensor in 50 µM Trp. e, Indirect molecular detection using LEG–RAR–MIP sensors.
f, LSV voltammograms of indirect Leu detection with LEG–PBNP–MIP sensors. Inset, calibration plot with a linear fit. g,h, Indirect detection of all essential
AAs (g) and multiple vitamins, lipids and metabolites (h) using LEG–PBNP–MIP sensors. Dashed lines represent linear-fit trendlines. VC, vitamin C;
VB6, vitamin B6; VD3, vitamin D3; VE, vitamin E. i, Schematic of multi-MIP AA sensors. j, LSV voltammograms of an LEG multi-MIP sensor for BCAA
quantification. Inset, calibration plot with a linear fit. k, In situ continuous sensing and regeneration of an LEG–PBNP–MIP Leu sensor in 50 µM Leu.
l, Repetitive CV scans of an LEG–PBNP electrode in 0.1 M KCl. m, DPV voltammograms of indirect Leu detection with LEG–AQCA–MIP sensors. Inset, the
calibration plot. n, In situ regeneration of an LEG–AQCA–MIP Leu sensor in a raw sweat sample. o, Selectivity of the Trp, Tyr, Leu, Ile, Val and BCAA sensors
against other AAs. p, Validation of Tyr, Trp and Leu sensors for analysing raw exercise sweat samples (n = 20) against GC–MS. All error bars represent the
standard deviation (s.d.) from three sensors.
human sweat sample, resolving a main bottleneck of wearable bio- measurements in raw human sweat samples have been validated
sensing. The MIP–LEG AA sensors have excellent selectivity for against GC–MS (Fig. 2p and Supplementary Figs. 26 and 27).
other analytes in sweat (including AAs with similar structures) at
physiologically relevant concentrations (Fig. 2o, Supplementary Wearable system design for autonomous sweat induction, sam-
Fig. 24 and Supplementary Table 3). The LEG–MIP technology pling, analysis and calibration. To enable on-body continuous
showed a comparable sensitivity with the current gold-standard metabolic and nutritional monitoring, the flexible sensor patch
laboratory-based GC–MS44 (Supplementary Fig. 25); the sensor was designed to comprise an iontophoresis module for localized

a b c d
Na
+ Temperature 6 6
sensor O With recognition 75 µM Trp
DPV Without recognition 50 µM Trp
HO OH H 2N NH2 25 µM Trp
Outlet
ΔI ' of STrp (µA)

ΔI of STrp (µA)
Inlets O O 4 4
Ref. Tyr
Trp OH ΔIs
HO NH2 ΔIb
IP anode H 2N O H 2N O 2 2
IP cathode OH OH
Gels
NH NH 0 0
LEG ISE MIP 1 MIP 2 Gel 0 100 200 300 0 100 200 300
(Tyr) (µM) (Tyr) (µM)
e f g h
160 Pilocarpine Carbachol 5 4
Test data Carbagel, S1 Carbagel, S1
+ Inlets
Sweat rate (µl min–1)
Sweat rate (µl min–1)

Na compensation Carbagel, S2 Carbagel, S2
4
Trp concentration
120 Predicted Carbagel, S3 3 Carbagel, S3

by sensor (µM)
Sweat Pilogel, S1 Pilogel, S1

3 Pilogel, S2 Pilogel, S2
80 Sweat Pilogel, S3 2 Pilogel, S3
glands 2
40 1
1
Sudomotor 0 0
0 neuron
0 40 80 120 160 Human skin 0 60 120 180 0 30 60 90
Trp concentration by GC/MS (µM) Time (min) Time (min)
i
(Trp) (mM)
×10–2
8
j
IP cathode 3:00 3:20 3:29 3:30 4:30 5:45
Inlets
IP anode
Outlet
Fig. 3 | Wearable system design for autonomous sweat induction, sampling, analysis and calibration. a, Illustration of a multi-functional wearable sensor
patch. ISE, ion-selective electrode. b–d, The two-scan sensor calibration strategy enabling selective Trp sensing in situ in the presence of Tyr. ∆I, peak
height current; ∆I′, peak height difference caused by target recognition. Solid and dashed curves in c and d represent linear-fit trendlines. e, Electrolyte
calibration of the AA sensor reading, with a linear fit. f, Schematic of localized sweat sampling based on iontophoretic sweat extraction with muscarinic
agents: pilocarpine and carbachol. g,h, Localized sweat rates measured from the stimulated (g) and surrounding (h) skin areas after a 5-min iontophoresis
with pilocarpine and carbachol. Solid and dashed curves represent quadratic-fit trendlines. S, subject. i, Numerically simulated Trp concentration ([Trp])
distributions in the microfluidic reservoir at 120 s after the inlet fluid changed from 20 to 80 µM Trp (flow rate 1.5 µl min−1) (with varied designs in inlet
number, angle span, and inlet and outlet orientation). j, On-body evaluation of the optimized flexible microfluidic patch for efficient carbachol-based
iontophoretic sweat induction and surrounding sampling at rest. Timestamps represent the period (min) after a 5-min iontophoresis session. Black dye
was used in the reservoir to facilitate the direct visualization of sweat flow in the microfluidics. Scale bar, 3 mm.
on-demand sweat induction, a multi-inlet microfluidic module for and ionic strength on the AA sensors can be calibrated in real time
efficient sweat sampling, a multiplex LEG–MIP sweat nutrient sen- on the basis of the readings from an LEG-based strain-resistive
sor array for continuous AA analysis, and LEG-based temperature temperature sensor and an ion-selective Na+ sensor (Fig. 3e and
and electrolyte sensors for real-time AA sensor calibration (Fig. 3a). Supplementary Fig. 28). Considering that sweat rate during exer-
Unlike classic bio-affinity sensor’s detection in buffer or redox solu- cise was reported to influence certain biomarker levels, we could
tions, in situ sweat analysis poses more challenges due to complex use sweat Na+ level (which showed a linear correlation with sweat
and interpersonally varied sweat composition and demands tech- rate) to further calibrate the nutrient levels for personalized analy-
nological innovations for accurate on-body sensing. For example, sis. This unique transduction strategy involving both the two-step
for direct LEG–MIP Trp sensing, a DPV scan in sweat even before DPV scans and the temperature/electrolyte calibrations allows us to
target/MIP recognition could lead to an oxidation peak as a small obtain accurate reading continuously in sweat during on-body use
amount of electroactive molecules (for example, Trp and Tyr) can (Supplementary Fig. 29).
be oxidized on the surface of MIP layer; after recognition and To make this wearable technology broadly applicable, particu-
binding of Trp into the MIP cavities, a substantially higher current larly for sedentary individuals, we utilize here a custom-designed
peak height can be obtained; measuring difference of the two peak iontophoresis module consisting of the LEG anode and cathode
heights allows more accurate measurement of bound Trp directly in coupled with hydrogels containing muscarinic agent carbachol
sweat with high selectivity (Fig. 3b–d). The influence of temperature (carbagel) for sustainable sweat extraction. Carbachol was selected

a b c d e
38 6 90 800
concentration Temperature
Subject 1 Trp Raw data Baseline
Tyr, Trp level
Concentration (µM)
(°C)
35 3 70 600
Potential (mV)
Potential (mV)
32 0 50 400
50 8 Tyr 100 Automated peak
extraction
30
(mM)
Time 50
Na+
40 4 Tyr
Trp
30 0 10 0
0 20 40 60 0.5 0.6 0.7 0.8 16 32 48 64 0.2 0.4 0.6 0.8
Time (min) Potential (V) Time (min) Potential (V)
f g h i j
80 520 0.33
Trp 10 min
BCAA concentration (µM)

2 Subject 1 Subject 1 Subject 1
Trp, BCAA level
Trp concentration (µM)

Trp/BCAA ratio
Subject 2 Subject 2 Subject 2

60 Subject 3 390 Subject 3 Subject 3
Subject 4 Subject 4 Subject 4
Trp/BCAAs
100 min 0.24
0.5 0.6 0.7 0.8 0.9 40 260
Central
fatigue 10 min
5 0.15
Time 20 130
100 min
BCAAs Background
0 0 0.06
–0.6 –0.5 –0.4 –0.3 –0.2 0 25 50 75 100 0 25 50 75 100 0 25 50 75 100
Potential (V) Time (min) Time (min) Time (min)
k l m n o
6 100 4 80
Supplement level
Subject 1 Intake Tyr No intake Tyr

Intake Trp Trp
Concentration (µM)
Concentration (µM)
3 75 2 60
Potential (mV)
Potential (mV)
Trp Trp
0 50 0 40
No intake 8 4
Time 25 20
4 Tyr 2 Tyr
(Trp, Tyr) 0 0 0 0
Supplements 0.45 0.5 0.55 0.6 0.65 0.7 10 30 50 70 90 0.45 0.5 0.55 0.6 0.65 0.7 10 30 50 70 90
Potential (V) Time (min) Potential (V) Time (min)
Fig. 4 | Wearable system evaluation across activities towards prolonged physiological and nutritional monitoring. a–d, Continuous on-body Trp and Tyr
analysis using a wearable sensor array with real-time sensor calibrations during cycling exercise. e, Custom voltammogram analysis with an automatic
peak extraction strategy based on a polynomial fitting and cut-off procedure. f–j, Dynamic sweat Trp and BCAA analysis during physical exercise towards
central fatigue monitoring. Dashed lines in h–j represent quadratic fit trendlines. k–o, Dynamic analysis of sweat AA levels with and without Trp and Tyr
supplement intake at rest towards personalized nutritional monitoring.
from various muscarinic agents as it allows the most efficient, patch could provide reliable and accurate analysis of the dynamic
repeatable and long-lasting sweat secretion from the surrounding changes of the AA levels (Supplementary Figs. 34 and 35).
sweat gland owing to its additional nicotinic effects45 (Fig. 3f–h,
Supplementary Fig. 30 and Supplementary Note 2). In contrast, the Evaluation of the wearable system for dynamic physiological
classic sweat-inducing agent—pilocarpine—used by the standard and nutritional monitoring. Evaluation of the wearable system
sweat test and previously reported wearable systems22,30,36 offers was conducted first via sensing of sweat Trp and Tyr in human sub-
only a short period of sweat and very limited sweat rate from the jects during a constant-load cycling exercise trial (Fig. 4a–d and
neighbouring sweat glands (Fig. 3f–h). Furthermore, sampling the Supplementary Fig. 36). The DPV data from the sensors were wire-
mixture of the leaked sweat underneath the pilocarpine gel and lessly transmitted along with temperature and Na+ sensor readings
the gel fluid could result in substantial wearable sensor errors and to the mobile app that automatically extracted the oxidation peaks
fail to provide real-time information owing to the absence of effi- using a custom-developed iterative baseline correction algorithm
cient sweat refreshing. A very small current (50–100 µA) is used (Fig. 4e and Supplementary Fig. 37) and performed calibration
for our iontophoresis module, compared with commonly used for the accurate quantification of sweat Tyr and Trp. Considering
1–1.5 mA (refs. 22,30,36), greatly reducing the risk of skin irritation. that AAs (for example, Try and BCAAs) play a crucial role in cen-
To maximize the efficiency of low-volume sweat sampling and tral fatigue during physical exercise46, a flexible Trp and BCAA
improve the temporal resolution of wearable sensing, a compact sensor array was used to monitor the AA dynamics during vigor-
and flexible microfluidic module was carefully designed to isolate ous exercise (Fig. 4f–j and Supplementary Fig. 38). Both Trp and
sweat sampling areas from iontophoresis gels. Numerical simu- BCAA levels decreased during the exercise owing to the serotonin
lations were performed to optimize the geometric design of the synthesis and BCAA ingestion, respectively. The increased sweat
microfluidic module, including inlet number, angle span, orien- Trp-to-BCAA ratio was observed, which could potentially serve as
tation and flow direction with respect to the reservoir geometry an indicator of central fatigue, in agreement with a previous report
(Fig. 3i, Supplementary Note 3, Supplementary Figs. 31 and 32, on its plasma counterpart46.
Supplementary Video 2 and Supplementary Table 4). With the opti- The wearable iontophoresis-integrated patch enables daily
mized design for sweat induction and sampling, sweat can be con- continuous AA monitoring at rest beyond the physical exercise.
veniently induced locally and readily sampled with the multi-inlet As illustrated in Fig. 4k–o and Supplementary Figs. 39–42, rising
microfluidics over a prolonged period (Fig. 3g,j, Supplementary Trp and Tyr levels in sweat were observed from all four subjects
Fig. 33 and Supplementary Video 3). At the physiological sweat after Trp and Tyr supplement intake while the readings from the
rates ranging from 0.15 µl min−1 to 3 µl min−1, our wearable sensor sensors remained stable during the studies without intake. Such

a c Sweat Leu concentration (µM)
d
Healthy Obesity and T2DM 0 35 70 105 140
Serum BCAA concentration (mM)

1.6 200 210
Serum Leu concentration (µM)

Healthy
Obesity
Leu concentration (µM)

1.2 150 Obesity and T2DM
140
0.8 100
Healthy
Obesity and T2DM 70
Leu level
0.4 50
0 0 0
0 0.2 0.4 0.6 0.8 Sweat Serum
Serum Sweat
Sweat BCAA concentration (mM)
b e f
BCAA concentration BG concentration

300 840 300 690
concentration (µM)
concentration (µM)
concentration (µM)
Sweat Leu
Sweat Leu
200 560 200 460
BCAA
(µM)
100 280 100 230
BCAA intake Insulin response
Subject 1 Subject 2 0 Subject 1 Subject 2 0
0 0
BG concentration
Ins concentration
Ins concentration
Muscle Fat Other 120 130 400 130
(mg dL–1)
cell cell substrates
(mg dl–1)
(pM)
(pM)
60 105 200 105
BCAA intake Protein intake

0 80 0 80
0 30 60 90 120 0 30 60 90 120
Production Time (min) Time (min)
obesity and T2DM g
Healthy
225
h 800
Gluconeogenesis Obesity and T2DM COVID-19 negative
Serum Leu concentration (µM)

Sweat Leu concentration (µM)
Obesity COVID-19 positive

Healthy
600
150
Normal Mitochondrial BCAA metabolic
BCAA clearance dysfunction enzyme level 400
75
concentration
250
Obesity and T2DM 200
ΔLeu
(%)
125
Leu level
0
Healthy Decreased I II III
0 0
BCAA intake BCAA clearance 0 20 40 60 80 Subject
T
Time (min)
Fig. 5 | Personalized monitoring of metabolic syndrome risk factors using LEG–MIP BCAA sensors. a, Elevated BCAA levels identified in individuals with
obesity and/or T2DM. b, The close associations between BCAA metabolism and insulin response in healthy and obesity/T2DM groups. c, Correlation of
serum and sweat total BCAA and Leu levels obtained with the LEG–MIP sensors (n = 65). Dashed lines represent linear-fit trendlines. d, Box-and-whisker
plot of measured Leu levels in iontophoresis-extracted sweat and serum in three groups of participants: normal weight (group I, n = 10), overweight or
obesity (group II, n = 7) and obesity with T2DM (group III, n = 3), The bottom whisker represents the minimum, the top whisker represents the maximum
and the square in the box represents the mean. e,f, Dynamic changes of sweat Leu and total BCAAs, serum insulin (Ins) and blood glucose (BG) levels from
two healthy subjects with 5 g BCAAs (e) and standard protein diet (f) intakes. g, Sweat Leu dynamics collected from groups I–III after 5 g BCAA intake.
Inset, ratio of the Leu level at 50 min after BCAA intake and the level before intake. h, Evaluation of Leu as a metabolic fingerprint for COVID-19 severity in
serum samples from COVID-19-negative subjects (n = 8) and COVID-19-positive patients (n = 8). Error bars represent the s.d. from three measurements.
capability opens the door for personalized nutritional monitor- Note 4)3,4, which could lead to potential complications of severe
ing and management through personalized sensor-guided dietary COVID-19 (ref. 12). Recent studies have shown the potential use
intervention. It should be noted that our pilot study showed that of BCAA supplementation as a dietary intervention to ameliorate
sweat nutrient and electrolyte levels were independent of sweat rate insulin resistance48. Monitoring changes in essential nutrient levels
changes during the carbachol-based iontophoresis-induced sweat provides highly sensitive early detection of metabolic syndrome
(Supplementary Fig. 43). risks, enabling effective personalized dietary intervention (Fig. 5b).
To explore the use of sweat BCAAs as a non-invasive risk factor
Personalized monitoring of metabolic syndrome risk factors of metabolic syndrome, we performed a pilot study to investigate
using wireless biosensors. Metabolic syndrome, characterized by the correlations between serum and sweat BCAAs involving three
abdominal obesity and insulin resistance, is now on the rise as the groups of subjects: normal weight (I, n = 10), overweight/obesity
leading cause of morbidity and mortality, affecting more than a third (II, n = 7) and obesity with T2DM (III, n = 3) (Fig. 5c,d). Positive
of all adults in the United States47. Elevated circulating BCAAs levels Pearson correlation coefficients of 0.66 (n = 65) and 0.69 (n = 65)
are predictive of insulin-resistant obesity and metabolic syndrome, were observed between sweat and serum levels (all analysed by the
and are linked to CVDs and T2DM (Fig. 5a and Supplementary sensors) of Leu and total BCAA, respectively (Fig. 5c). Compared

with healthy participants in group I, substantially elevated sweat and metabolic syndrome risk monitoring. The substantial difference in
serum Leu levels (analysed by the sensors) were observed in groups Leu between COVID-19-positive and COVID-19-negative blood
II and III (Fig. 5d), consistent with previous reports that higher cir- samples indicates the potential of using this technology for at-home
culating BCAA levels were identified in individuals with obesity and COVID-19 management. We envision that this wearable technol-
T2DM3. Considering the well-established role of BCAAs on insulin ogy could play a crucial role in the realization of precision nutri-
production and inhibition of glycogenolysis, we also investigated tion through continuous monitoring of circulating biomarkers and
the post-prandial response of sweat Leu/BCAAs and blood glu- enabling personalized nutritional intervention. This technology
cose/insulin after BCAA supplement and dietary intake in healthy could also be reconfigured to continuously monitor a variety of
subjects (Fig. 5e,f). All biomarkers remained stable during the fast- other biomarkers towards a wide range of personalized preventive,
ing period; protein diet intake resulted in increases in both blood diagnostic and therapeutic applications.
glucose and insulin, while BCAA intake only led to a rapid insu-
lin increase. In both studies, sweat Leu and BCAAs first increased Methods
in the 30–60 min and then decreased. For subjects with different Materials and reagents. Uric acid, l-tyrosine, silver nitrate, iron(III) chloride,
metabolic conditions, Leu levels in iontophoretic sweat after BCAA dopamine hydrochloride, choline chloride, creatinine, pantothenic acid calcium
salt, citrulline, pyridoxine and lactic acid were purchased from Alfa Aesar.
vary differently: although a substantial increase in sweat Leu levels Sodium thiosulfate pentahydrate, sodium bisulfite, tryptophan, leucine, alanine,
was observed in all cases, healthy subjects showed a drastic percent- isoleucine, methionine, valine, lysine, thiamine hydrochloride, serine, sulfuric
age fluctuation and individuals with obesity/T2DM showed blunted acid, hydrochloric acid, AQCA, 3-aminophenylboronic acid (APBA), aniline,
fluctuation that may indicate the different metabolic stage of BCAA o-phenylenediamine, methylene blue, thionine, 2-(N-morpholino)ethanesulfonic
acid hydrate (MES), ethanolamine, N-(3-dimethyl-aminopropyl)-N′-ethy
in those individuals (Fig. 5g).
lcarbodiimide (EDC), N-hydroxysulfosuccinimide sodium salt (sulfo-NHS),
Considering that circulating elevated Leu has been reported as bovine serum albumin (BSA), tris(hydroxymethyl)aminomethane hydrochloride
a key metabolic fingerprint for COVID-19 severity, we also evalu- (Tris–HCl), streptavidin–peroxidase conjugate (Roche) and hydroquinone were
ated our biosensors for analysing the samples from patients with purchased from Sigma-Aldrich. Carboxylic-acid-modified magnetic beads (MBs;
COVID-19 and healthy individuals; substantially elevated Leu lev- Dynabeads, M-270) were obtained from Invitrogen. Potassium ferricyanide
and potassium ferrocyanide were purchased from Acros Organics. Acetic acid,
els were identified in COVID-19-positive samples compared with methanol, sodium acetate, sodium chloride, sodium dihydrogen phosphate,
the negative ones (415.6 ± 133.7 versus 151.5 ± 36.0 µM), indicating potassium chloride, potassium hydrogen phosphate, urea, l-ascorbic acid and
the great potential of our biosensors for at-home COVID-19 moni- dextrose (d-glucose) anhydrous, glycine, arginine, inositol, ornithine, aspartic acid,
toring and management (Fig. 5h). threonine, histidine, riboflavin, creatine, phenylalanine, nicotinic acid, folic acid,
glutamic acid and hydrogen peroxide (30% (w/v)) were purchased from Thermo
Fisher Scientific. Insulin capture antibody and biotinylated detector antibody were
Discussion purchased from R&D systems (Human/Canine/Porcine Insulin DuoSet ELISA).
Circulating metabolic biomarkers, such as AAs and vitamins, have Screen printed carbon electrodes and magnetic holder were purchased from
been associated with various health conditions, including diabetes Metrohm DropSens. Medical adhesives were purchased from 3 M and Adhesives
and CVDs. Metabolic profiling using wearable sensors has become Research. PI films (75 μm thick) were purchased from DuPont. PET films (12 μm
increasingly crucial in precision nutrition and precision medicine, thick) were purchased from McMaster-Carr.
especially in the era of the COVID-19 pandemic, as it provides not
Fabrication and preparation of the LEG sensors. The LEG electrodes were
only insights into COVID-19 severity but also guidance to stay fabricated on a PI film with a thickness of 75 μm (DuPont) with a 50 W CO2 laser
metabolically healthy to minimize the risk of potential COVID-19 cutter (Universal Laser System). When engraving the PI with a CO2 laser cutter,
infection. As the pandemic remains rampant throughout the world the absorbed laser energy is converted to local heat and thus leads to a high
and regular medical services are at risk of shortage, there is an localized temperature (>2,500 °C), chemical bonds in the PI network are broken
and thermal re-organization of the carbon atoms occurs, resulting in sheets of
urgent need to develop and apply wearable sensors that can moni-
graphene structures. The optimized parameters for the graphene electrodes and
tor health conditions via metabolic profiling to achieve at-home electronic connections were power 8%, speed 15%, and points per inch (PPI)
diagnosis and timely intervention via telemedicine. However, cur- 1,000 in raster mode with three-time scan. For the active sensing area of the
rent wearable electrochemical sensors are limited to a narrow range temperature sensor, the optimized parameters were power 3%, speed 18%, and
of detection targets owing to lack of continuous sensing strategies PPI 1,000 in vector mode with one-time scan. To prepare the reference electrode,
Ag was first modified on the corresponding graphene electrode by multi-current
beyond ion-selective and enzymatic electrodes. Though various electrodeposition with electrochemical workstation (CHI 832D) at −0.01 mA
bio-affinity-based sensors have been developed to detect a broader for 150 s, −0.02 mA for 50 s, −0.05 mA for 50 s, −0.08 mA for 50 s and −0.1 mA
spectrum of targets using antibodies or MIPs, they generally require for 350 s using a plating solution containing 0.25 M silver nitrate, 0.75 M sodium
multiple washing steps or provide only one-time use; these limita- thiosulfate and 0.5 M sodium bisulfite. To obtain the Ag/AgCl electrode, 0.1 M
tions have hampered their useability in wearable devices. Moreover, FeCl3 solution was further dropped on the Ag surface for 30 s, and then 3 µl
polyvinyl butyral (PVB) reference cocktail prepared by dissolving 79.1 mg of PVB
the majority of wearable biosensors rely on vigorous exercise to and 50 mg of NaCl in 1 ml of methanol was dropped on the Ag/AgCl electrode
access sweat and are not suitable for daily continuous use. and dried overnight. The Na+-selective electrode was prepared as follows: 0.6 µl of
By integrating mass-producible LEG, electrochemically syn- Na+-selective membrane cocktail prepared by dissolving 1 mg of Na ionophore X,
thesized RARs and ‘artificial antibodies’, we have demonstrated a 0.55 mg sodium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate, 33 mg polyvinyl
powerful universal wearable biosensing strategy that can achieve chloride and 65.45 mg bis(2-ethylhexyl) sebacate into 660 µl of tetrahydrofuran was
drop-casted onto the graphene electrode and dried overnight. To obtain the desired
selective detection of a broad range of biomarkers (including all stable Na+-sensing performance for long-term continuous measurements, the
essential AAs, vitamins, metabolites, lipids, hormones and drugs) obtained Na+ sensor was conditioned overnight in 100 mM NaCl.
and reliable in situ regeneration. Furthermore, to enable continu- The fabrication process of the LEG–MIP sensor array is illustrated
ous and on-demand metabolic and nutritional monitoring across in Supplementary Fig. 6. All the MIP layers are synthesized by
the activities, we have integrated the LEG–MIP sensor array and electro-polymerization. The polymerization solution was prepared by dissolving
5 mM template (for example, target AA), 12.5 mM APBA and 37.5 mM pyrrole
iontophoresis-based sweat induction into a wireless wearable tech- into 0.01 M phosphate-buffered saline (PBS) (pH 6.5). For multi-MIP BCAA
nology, with optimized multi-inlet microfluidic sudomotor axon sensor, 5 mM of each target (that is, Leu, Ile and Val) was used. Before MIP
reflex sweat sampling, in situ signal processing, calibration and deposition, the LEG was activated in 0.5 M H2SO4 with CV scans for 60 segments
wireless communication. Using this telemedicine technology, we (−1.2 to 1 V with a scan rate of 500 mV s−1). For the direct-detection LEG–MIP
have demonstrated the wearable and continuous monitoring of sensors, the target imprinted polymer was electrochemically synthesized on the
LEG electrode with CV deposition (0–1 V for ten cycles, 50 mV s−1) using the
post-prandial AA responses to identify risks for metabolic syn- prepared polymerization solution. The target molecules were extracted by soaking
drome. The high correlation between sweat and serum BCAAs the electrode into an acetic acid/methanol mixture (7:3 v/v) for 1 h. Subsequently,
suggests that this technology holds great promise for use in the resulting electrode was immersed into 0.01 M PBS (pH 6.5) for repetitive

CV scans (0.4–1 V with a scan rate of 50 mV s−1) until a stable response was inlets (12-μm-thick PET, laser parameters: power 20%, speed 100%, PPI 1,000).
obtained. For LEG-non-imprinted polymer, the electrode was prepared following The third layer (channel layer), in contact with the inlets layer, was patterned with
the same procedure as LEG–MIP except that no template was added in the microfluidic channels (Adhesives Research 93049, laser parameters: power 45%,
polymerization solution. speed 100%, PPI 1,000). The fourth layer (reservoir layer), sandwiched between
For the indirect-detection MIP sensors, electrochemically synthesized RARs the channel layer and the electrode PI layer, was patterned with the reservoir and
(for example, PBNPs or AQCA) were first modified on the LEG electrode. The the outlet (3M 468MP, laser parameters: power 60%, speed 90%, PPI 1,000). The
PBNP RAR on the LEG was prepared with CV (20 cycles) (−0.2 to 0.6 V with reservoir is an ellipse with a 5.442 mm major axis and a 4.253 mm minor axis to
a scan rate of 50 mV s−1) in an aqueous solution containing 3 mM FeCl3, 3 mM fully enclose the active sensing area. The thickness of the channel layer is ~0.1 mm
K3Fe(CN)6, 0.1 M HCl and 0.1 M KCl. A PBNP layer with appropriate redox signal (Adhesives Research 93049), and the thickness of the reservoir layer is 0.13 mm
is necessary to produce a good sensitivity for the final MIP sensors; to achieve (3M 468MP). The reservoir area is 18.17 mm2, and thus the reservoir volume
this stable and suitable redox signal, the LEG electrode was rinsed with distilled can be calculated as the area multiplied by the thickness of the reservoir layer
water after the initial Prussian blue (PB) deposition, and the PB electrodeposition (0.13 mm), which totals 2.36 µl.
step was repeated two more times until a stable 70 µA LSV peak in 0.1 M KCl
solution was achieved. Subsequently, the LEG–PB was rinsed with distilled water Fabrication of agonist agent hydrogels. Hydrogels containing muscarinic agent
and immersed in a solution containing 0.1 M HCl and 0.1 M KCl for repetitive carbachol was prepared as follows. Briefly, for anode gel, agarose (3% w/w) was
CV scans (−0.2 to 0.6 V with a scan rate of 50 mV s−1) until a stable response was added into de-ionized water and then heated to 250 °C under constant stirring.
obtained. To prepare the AQCA RAR on the LEG, the LEG electrode was first After the mixture was fully boiled and became homogeneous without agarose
incubated in 50 µl PBS (pH 6.5) with 5 mM AQCA at 4 °C overnight. Subsequently, grains, the mixture was cooled down to 165 °C and 1% carbachol was added to the
the LEG–AQCA was rinsed with distilled water and immersed into a phosphate above mixture. Subsequently, the cooled mixture was slowly poured into pre-made
buffer solution for repetitive CV scans (−0.8 to 0 V with a scan rate of 50 mV s−1) cylindrical moulds or into assembled microfluidic patch and solidified for 10 min
until a stable response was obtained. For the indirect-detection LEG–PB–MIP at 4 °C. The cathode gel was prepared similarly except that NaCl (1% w/w) was
sensors, an additional PB activation process was conducted right after the template used instead of carbachol.
extraction (IT scan at 1 V in 0.5 M HCl for 600 s), followed by an LEG–PB–MIP
sensor stabilization process in 0.1 M KCl (CV scans at −0.2 to 0.6 V with a scan rate Signal conditioning, processing and wireless transmission for the wearable
of 50 mV s−1). It should be noted that, for the LEG–AQCA–MIP sensor, only three sensor. The block diagram of the electronic system (Fig. 1g and Supplementary
CV cycles of polymerization were used to prepare the MIP layer, and the sensor Fig. 4) represents both the wearable electronic patch and the smart watch that
was stabilized in 0.01 M PBS (pH 6.5) (CV scans at −0.8 to 0 V with a scan rate can (1) induce sweat via iontophoresis and (2) monitor sweat via electrochemical
of 50 mV s−1). methods. The sweat induction and the sweat sensing procedures are initiated and
The morphology of materials was characterized by scanning electron controlled by the microcontroller (STM32L432KC, STMicroelectronics) when it
microscopy (Nova Nano SEM 450) and transmission electron microscopy receives a user command from the Bluetooth module over universal asynchronous
(Talos S-FEG FEI, USA). The Raman spectrum of the electrodes with different receiver–transmitter (UART) communication.
modification was recorded using a 532.8 nm laser with an inVia Reflex (Renishaw).
Fourier-transform infrared spectra were measured using infrared spectrometry Sweat induction. Programmable iontophoretic current is generated by a
(Nicolet 6700). voltage-controlled current source that consists of a unity-gain difference amplifier
(AD8276, Analog Devices) and a boost transistor (BC846, ON Semiconductor).
Characterization of the LEG sensor performance. A set of electrochemical The circuit is supplied by the output of a boost converter (LMR64010) that
sensors were characterized in solutions of target analytes. All the in vitro sensor boosts the 3.7 V battery voltage to 36 V. The microcontroller controls the
characterizations were performed through CHI 832D. The response of the digital-to-analogue converter (DAC) (DAC8552, Texas Instruments) over a serial
Na+ sensor was characterized with open circuit potential measurements in the peripheral interface to set the control voltage of the current source. The current
solutions containing varied Na+ levels. DPV analysis was performed for all the source output is checked by a comparator (TS391, STMicroelectronics), and the
direct-detection LEG–MIP sensor characterizations in 0.01 M PBS (pH 6.5) or microcontroller is interrupted through its general-purpose input/output pin at
in raw sweat. The DPV conditions were as follows: range, 0.4–1 V; incremental output failure. The protection circuit consists of a current limiter (MMBF5457, ON
potential, 0.01 V; pulse amplitude, 0.05 V; pulse width, 0.05 s; pulse period, Semiconductor) and analogue switches (MAX4715, Maxim Integrated; ADG5401,
0.5 s; and sensitivity, 1 × 10−5 A V−1. For in vitro indirect detection of the target Analog Devices). The microcontroller’s general-purpose input/output is also used
molecules based on the LEG–PB–MIP sensors, LSV analysis (0.4–0 V) was to enable or disable the iontophoresis circuit. For the optimized design, a 100-µA
performed in 0.1 M KCl. The LSV conditions were as follows: range, 0.4–0 V; current (~2.6 µA mm−2) was applied for on-body iontophoresis sweat induction
scan rate, 0.005 V s−1; sample interval, 0.001 V; quiet time, 2 s; and sensitivity, using the flexible microfluidic patch.
1 × 10−4 A V−1. For in vitro indirect detection of the target molecules based on the
LEG–AQCA–MIP sensors, negative DPV analysis (0 to −0.8 V) was performed in Power analysis. When powered at 3.3 V, the electronic system consumes ~28 mA
0.01 M PBS. The negative DPV conditions were as follows: 0 to −0.8 V; incremental during an active electrochemical measurement and ~61 mA during iontophoresis.
potential, 0.01 V; pulse amplitude, 0.05 V; pulse width, 0.05 s; pulse period, 0.5 s; The microcontroller and Bluetooth module each consume ~12 mA; the sensor
and sensitivity, 1 × 10−5 A V−1. For in situ sweat analyte measurement, background interface consumes ~4 mA; the boost converter and iontophoresis module
and signal curves were recorded before and after incubation; the signal current consumes ~33 mA; and the display module consumes an additional ~8 mA when
was obtained as the difference of the peak amplitudes between the post-incubation refreshing its screen.
signal and the background current curves (Fig. 3b–d and Supplementary Fig. 29).
The temperature sensor characterization was carried out on a ceramic hot plate Sweat sensing. The sweat sensing circuitry can perform two-channel simultaneous
(Thermo Fisher Scientific) (Supplementary Fig. 28). The sensor response was DPV, as well as potentiometric and temperature measurements. A bipotentiostat
recorded using a parameter analyser (Keithley 4200A-SCS) and compared with the circuit is constructed by a control amplifier (AD8605) and two transimpedance
readings from an infra-red thermometer (LASERGRIP 800; Etekcity). amplifiers (AD8606). A series voltage reference (ISL60002, Renesas Electronics)
To evaluate the performance of the various electrode substrates for MIP-based and a DAC (DAC8552, Texas Instruments) is used to generate a dynamic potential
AA sensing, LEG, printed carbon electrode, Au electrode and glassy carbon bias across the reference and working electrodes. An instrumentation amplifier
electrode were chosen. The glassy carbon electrodes were purchased from CH (INA333, Texas Instruments) is used for potentiometric measurements, and a
Instruments. The printed carbon electrodes were printed on the PI substrate voltage divider is used for the resistive temperature sensor. All analogue voltage
using a Dimatix Materials Printer DMP-2850 (Fujifilm, Minato, Japan) with a signals are acquired by the microcontroller’s built-in analogue-to-digital converter
commercial carbon ink from NovaCentrix. The Au electrodes were fabricated (ADC) channels, processed and then transmitted over Bluetooth to a user device.
via E-beam evaporation: 20 nm of Cr and 100 nm of Au were deposited onto an
O2-plasma pre-treated PET substrate. MIP films were prepared with CV deposition Custom mobile application design. The custom mobile application was developed
(0–1 V for ten cycles, 50 mV s−1). with the cross-platform Flutter framework. The mobile application can wirelessly
communicate with the wearable devices via Bluetooth to send commands, and
Fabrication and characterization of microfluidic channels. The microfluidic to acquire, process and visualize the sweat biomarker levels. The application
module was fabricated using a 50 W CO2 laser cutter (Universal Laser System) establishes a secure Bluetooth connection to the wearable sensor. The home
(Supplementary Fig. 1). Briefly, layers of double-sided and single-sided medical page plots the user’s historical biomarker levels, and highlights the most recently
adhesives (3M) were patterned with channels, inlets, the iontophoresis gel measured analyte concentrations. When a sweat biomarker measurement is
outlines and reservoirs. For all microfluidic layers, the iontophoresis gel outlines prompted, the user can switch over to the measurement page that plots the sweat
were patterned to enable the current flow from the top PI electrode layer. The sensors’ voltammograms in real time. Following the voltammetric measurement,
bottom layer, which is the double-sided adhesive layer in contact with the skin the app extracts the voltammograms’ peak currents using a custom baseline
(accumulation layer), was patterned with a sweat accumulation well (3M 468MP, correction algorithm, then converts the peak currents to corresponding biomarker
laser parameters: power 60%, speed 90%, PPI 1,000). The second layer (the inlet concentrations. These measurement data are added to the list of historic analyte
layer), in contact with the accumulation layer, was patterned with the multiple levels on the home page.

Refreshing time analysis and simulations. The refreshing time analyses were Five-minute iontophoresis was applied on the subjects. The sensor data recording
performed using numerical simulations (COMSOL). Three-dimensional models process was the same as in exercise-based human trials.
of different microfluidic designs with same dimensions of the actual device were
created in Rhinoceros and imported into COMSOL Multiphysics. The mass Sensor evaluation with BCAA diet challenge. For the BCAA studies, the subjects
transport process was simulated by numerically solving the Stokes equation for an were asked to consume 5 g BCAAs (2:1:1 Leu:Ile:Val) or a standardized snack
incompressible flow coupled with convection–diffusion equation (Supplementary including a protein drink (Fairlife, Core Power Elite) and a CLIF energy bar. An
Note 3). iontophoresis session was implemented with carbachol gels for sweat induction.
Over the entire study period, the subject’s sweat was sampled periodically and
Human subject recruitment. The validation and evaluation of the sweat analysed by the sensor patch. Blood glucose level was recorded every 15 min with a
sensor were performed using human subjects in compliance with all the ethical commercial Care Touch Blood Glucose Meter. Fresh capillary blood samples were
regulations under protocols (ID 19-0892 and 21-1079) that were approved by the collected using a finger-prick approach during the human studies. After cleaning
institutional review board at California Institute of Technology. The participating the fingertip with alcohol wipe and allowing it to air dry, the skin was punctured
subjects (aged over 18 years) were recruited from the California Institute of with a CareTouch lancing device. Samples were collected with centrifuge tubes
Technology campus and the neighbouring communities through advertisement. after wiping off the first drop of blood with gauze. After the 90-min standardized
All subjects gave written informed consent before study participation. For wearable clotting procedure finished, serum was separated by centrifuging at 6,000 rpm for
sensor evaluation, healthy subjects with a body mass index (BMI) of 18.5– 15 min, and instantly stored at −20 °C for analysis with GC–MS, the LEG–MIP
24.9 kg m−2 with fasting serum glucose <100 mg dl−1 were recruited. For the BCAA sensors and the custom insulin assay.
study, inclusion criteria include: group I, individuals with normal weight who have
a BMI of 18.5–24.9 kg m−2 with fasting serum glucose <100 mg dl−1 (healthy); group Blood insulin analysis. For the BCAA diet challenge study, the collected serum
II, individuals with overweight/obesity who have a BMI of 25–35 kg m−2 and fasting samples were analysed using a custom insulin sandwich immunoassay. The MBs
serum glucose <6 mg dl−1 (overweight/obesity); group III, individuals with obesity were modified on the basis of a previous publication50. Briefly, 3 μl MBs were
who have a BMI of 25–35 kg m−2 and fasting serum glucose ≥126 mg dl−1 (obesity activated with 50 mg ml−1 EDC/sulfo-NHS in MES buffer (25 mM, pH 5) for
and T2DM). COVID-19-positive and COVID-19-negative serum samples were 35 min followed by capture antibody immobilization (25 μg ml−1 in MES buffer)
purchased from RayBiotech. for 15 min. After de-activation with 1 M ethanolamine in phosphate buffer (0.1 M,
pH 8), MBs were incubated in 25 μl standards prepared in 1% BSA or serum
GC–MS analysis for sensor validation. GC–MS analysis of the AAs in sweat samples diluted five times in 1% BSA for 15 min. From here, the beads were
and serum samples was performed using EZ:Faast kit from Phenomenex, which rinsed with 1% BSA twice after each binding step. Next, the MBs were incubated
enables sample preparation, derivatization and GC–MS analysis of free AAs. A in 25 μl of biotin-detector antibody (1.0 μg ml−1) in 1% BSA for 30 min followed
Varian Saturn 2000 was used for the GC–MS runs. One microlitre of prepared by 15 min in streptavidin–peroxidase conjugate (2,500×) prepared in 1% BSA.
sample solution was injected for GC in helium carrier gas at 1.0 ml min−1 constant The amperometric detection was carried out by applying a constant potential of
flow with a pulse pressure of 20 pounds per square inch for 0.2 min, with the oven −0.2 V to MBs resuspended in 45 μl 1 mM hydroquinone, and 5 μl 5 mM H2O2
programmed from 110 °C to 320 °C at 32 °C min−1. The mass chromatography was was pipetted onto the screen-printed carbon electrodes when background current
set with source at 240 °C, quad at 180 °C and auxiliary at 310 °C with a scan range stabilized.
of 45–450 m/z at a sampling rate of 3.5 scans s−1. Selected ion monitoring was
used, which records the ion current at selected masses that are characteristic of the Reporting summary. Further information on research design is available in the
certain AA in an expected retention time49. For example, after the derivatization Nature Research Reporting Summary linked to this article.
of the EZ:Faast kit, Trp has a characteristic mass at 130 with a retention time at
around 5.1 min, and peak height is recorded for Trp measurements at ion number Data availability
130 and at 5.1 min from the raw data spectrum. The internal standard The main data supporting the results in this study are available within the
(IS; norvaline) was added during the sample derivatization process to account paper and its Supplementary Information. Source data for Figs. 4 and 5 and for
for potential evaporation-induced increase in peak detection; the IS norvaline peak Supplementary Figs. 36 and 39–41 are provided with this paper. All raw and
height is recorded at its ion number 158 at 1.65 min (Supplementary analysed datasets generated during the study are available from the corresponding
Fig. 26). The Trp peak height recorded from raw data spectrum was calibrated with author on request. Source data are provided with this paper.
respect to the IS in the same run: normalized Trp peak height = Trp peak height/
IS peak height. With normalized peak heights of different levels of Trp standards, Received: 7 June 2021; Accepted: 19 June 2022;
calibration plots were constructed. For other samples, the normalized peak height
of Trp was used to calculate the concentration.
Published online: 15 August 2022
Integrated system validation in human subjects. System evaluation during References

exercise. To validate the wearable sensor system, we conducted constant-load 1. Trumbo, P., Schlicker, S., Yates, A. A. & Poos, M. Dietary reference intakes
cycling exercise on healthy subjects. The subjects reported to the lab after fasting for energy, carbohydrate, fiber, fat, fatty acids, cholesterol, protein and amino
overnight and were given a standardized protein drink (Fairlife, Core Power acids. J. Am. Diet. Assoc. 102, 1621–1630 (2002).
Elite). The subjects’ foreheads and necks were cleaned with alcohol swabs and 2. Wu, G. Amino acids: metabolism, functions, and nutrition. Amino Acids 37,
gauze before the sensor patches were placed on the body. A stationary exercise 1–17 (2009).
bike (Kettler Axos Cycle M-LA) was used for cycling trials. The subjects cycled 3. Lynch, C. J. & Adams, S. H. Branched-chain amino acids in metabolic
at 60 rpm for 60 min or until fatigue. During the on-body trial, the data from the signalling and insulin resistance. Nat. Rev. Endocrinol. 10, 723–736 (2014).
sensor patches were wirelessly sent to the user interface via Bluetooth. When the 4. Wang, T. J. et al. Metabolite profiles and the risk of developing diabetes. Nat.
subjects started biking, the sensor system continuously acquired and transmitted Med. 17, 448–453 (2011).
temperature and sodium sensor data. Every minute, the electronic system initiated 5. Mayers, J. R. et al. Elevation of circulating branched-chain amino acids is an
a transient voltage bias between the reference and working electrodes. When the early event in human pancreatic adenocarcinoma development. Nat. Med. 20,
bias triggered a current above an experimentally determined threshold, 1193–1198 (2014).
the system would start a CV cleaning cycle and then the first DPV scan as 6. Kau, A. L., Ahern, P. P., Griffin, N. W., Goodman, A. L. & Gordon, J. I.
the initial background without target incubation. The DPV scan was repeated Human nutrition, the gut microbiome and the immune system. Nature 474,
7 min later as the post-incubation curve. Between the two scans, sodium 327–336 (2011).
and temperature sensor data were continuously recorded. Right after the 7. Fernstrom, J. D. & Wurtman, R. J. Brain serotonin content: physiological
post-incubation DPV, another cycle started with an IT cleaning/regeneration dependence on plasma tryptophan levels. Science 173, 149–152 (1971).
step, followed by an initial background DPV scan. The collected temperature, 8. Fernstrom, J. D. & Fernstrom, M. H. Tyrosine, phenylalanine, and
sodium and DPV data were wirelessly transmitted to a user device via Bluetooth catecholamine synthesis and function in the brain. J. Nutr. 137,
in real time, where the molecular data were extracted, calibrated and converted to 1539S–1547S (2007).
concentration levels. Sweat samples were collected periodically from the subjects 9. Dierckx, T. et al. The metabolic fingerprint of COVID-19 severity. Preprint at
during the studies using centrifuge tubes. The sweat samples were then frozen medRxiv (2020) https://doi.org/10.1101/2020.11.09.20228221
at −20 °C for further testing and validation via electrochemical test with the 10. Grant, W. B. et al. Evidence that vitamin D supplementation could reduce
biosensors and GC–MS analysis. risk of influenza and COVID-19 infections and deaths. Nutrients 12,
988–1006 (2020).
System evaluation with Tyr/Trp supplement intake. The subjects reported to the 11. Belanger, M. J. et al. COVID-19 and disparities in nutrition and obesity.
lab after fasting overnight. The subjects’ arms were cleaned with alcohol swabs N. Engl. J. Med. 383, e69 (2020).
and gauze before the sensor patches were placed on the body. The subjects were 12. Stefan, N., Birkenfeld, A. L. & Schulze, M. B. Global pandemics
provided Tyr and Trp supplement (1 g each) for the intake study. In contrast, the interconnected—obesity, impaired metabolic health and COVID-19. Nat. Rev.
control study was performed on the subjects without any supplementary intake. Endocrinol. 17, 135–149 (2021).

13. Wishart, D. S. Emerging applications of metabolomics in drug discovery and 43. Parlak, O., Keene, S. T., Marais, A., Curto, V. F. & Salleo, A. Molecularly
precision medicine. Nat. Rev. Drug Discov. 15, 473–484 (2016). selective nanoporous membrane-based wearable organic electrochemical
14. Dettmer, K., Aronov, P. A. & Hammock, B. D. Mass spectrometry-based device for noninvasive cortisol sensing. Sci. Adv. 4, eaar2904 (2018).
metabolomics. Mass Spectrom. Rev. 26, 51–78 (2007). 44. Kaspar, H., Dettmer, K., Gronwald, W. & Oefner, P. J. Automated GC–MS
15. Ray, T. R. et al. Bio-integrated wearable systems: A comprehensive review. analysis of free amino acids in biological fluids. J. Chromatogr. B 870,
Chem. Rev. 119, 5461–5533 (2019). 222–232 (2008).
16. Hollander, J. E. & Carr, B. G. Virtually perfect? Telemedicine for COVID-19. 45. Riedl, B., Nischik, M., Birklein, F., Neundörfer, B. & Handwerker, H. O.
N. Engl. J. Med. 382, 1679–1681 (2020). Spatial extension of sudomotor axon reflex sweating in human skin. J. Auton.
17. Niu, S. et al. A wireless body area sensor network based on stretchable Nerv. Syst. 69, 83–88 (1998).
passive tags. Nat. Electron. 2, 361–368 (2019). 46. Newsholme, E. A. & Blomstrand, E. Branched-chain amino acids and central
18. Meng, K. et al. A wireless textile-based sensor system for self-powered fatigue. J. Nutr. 136, 274S–276S (2006).
personalized health care. Matter 2, 896–907 (2020). 47. Ahlqvist, E. et al. Novel subgroups of adult-onset diabetes and their
19. Wang, C. et al. Monitoring of the central blood pressure waveform via a association with outcomes: a data-driven cluster analysis of six variables.
conformal ultrasonic device. Nat. Biomed. Eng. 2, 687–695 (2018). Lancet Diabetes Endocrinol. 6, 361–369 (2018).
20. Ni, X. et al. Automated, multiparametric monitoring of respiratory 48. Layman, D. K. & Walker, D. A. Potential importance of leucine in treatment
biomarkers and vital signs in clinical and home settings for COVID-19 of obesity and the metabolic syndrome. J. Nutr. 136, 319S–323S (2006).
patients. Proc. Natl Acad. Sci. USA 118, e2026610118 (2021). 49. Kitson, F. G., Larsen, B. S. & McEwen, C. N. Mass Spectrometry: A Practical
21. Jeong, H. et al. Differential cardiopulmonary monitoring system for Guide, pp. 3–23 (Academic Press, San Diego, 1996).
artifact-canceled physiological tracking of athletes, workers, and COVID-19 50. Torrente-Rodríguez, R. M. et al. Rapid endoglin determination in serum
patients. Sci. Adv. 7, eabg3092 (2021). samples using an amperometric magneto-actuated disposable immunosensing
22. Sempionatto, J. R. et al. An epidermal patch for the simultaneous monitoring platform. J. Pharm. Biomed. 129, 288–293 (2016).
of haemodynamic and metabolic biomarkers. Nat. Biomed. Eng. 5,
737–748 (2021). Acknowledgements
23. Xu, C., Yang, Y. & Gao, W. Skin-interfaced sensors in digital medicine: from This project was supported by the National Institutes of Health grant R01HL155815,
materials to applications. Matter 2, 1414–1445 (2020). Office of Naval Research grants N00014-21-1-2483 and N00014-21-1-2845, the
24. Heikenfeld, J. et al. Accessing analytes in biofluids for peripheral biochemical Translational Research Institute for Space Health through NASA NNX16AO69A,
monitoring. Nat. Biotechnol. 37, 407–419 (2019). NASA Cooperative Agreement 80NSSC20M0167, High Impact Pilot Research Award
25. Bariya, M., Nyein, H. Y. Y. & Javey, A. Wearable sweat sensors. Nat. Electron. T31IP1666 and grant R01RG3746 from the Tobacco-Related Disease Research
1, 160–171 (2018). Program, Caltech-City of Hope Biomedical Initiative Pilot Grant and the Rothenberg
26. Yang, Y. & Gao, W. Wearable and flexible electronics for continuous Innovation Initiative Program at California Institute of Technology. J.T. was supported
molecular monitoring. Chem. Soc. Rev. 48, 1465–1491 (2019). by the National Science Scholarship (NSS) from the Agency of Science Technology
27. Kim, J., Campbell, A. S., de Ávila, B. E.-F. & Wang, J. Wearable biosensors for and Research (A*STAR) Singapore. We gratefully acknowledge critical support and
healthcare monitoring. Nat. Biotechnol. 37, 389–406 (2019). infrastructure provided for this work by the Kavli Nanoscience Institute at Caltech.
28. Gao, W. et al. Fully integrated wearable sensor arrays for multiplexed in situ This project benefitted from the use of instrumentation made available by the Caltech
perspiration analysis. Nature 529, 509–514 (2016). Environmental Analysis Center, and we gratefully acknowledge support on GC–MS from
29. Bandodkar, A. J. et al. Battery-free, skin-interfaced microfluidic/electronic N. Dalleska. We also gratefully acknowledge Z. Wang for the contribution to mobile app
systems for simultaneous electrochemical, colorimetric, and volumetric development, R. M. Torrente-Rodríguez for the insulin assay optimization and S. Bao
analysis of sweat. Sci. Adv. 5, eaav3294 (2019). for the valuable inputs.
30. Emaminejad, S. et al. Autonomous sweat extraction and analysis applied to
cystic fibrosis and glucose monitoring using a fully integrated wearable
platform. Proc. Natl Acad. Sci. USA 114, 4625–4630 (2017). Author contributions
31. Torrente-Rodríguez, R. M. et al. Investigation of cortisol dynamics in W.G., M.W., Y.Y. and J.M. initiated the concept and designed the studies; W.G. supervised
human sweat using a graphene-based wireless mHealth System. Matter 2, the work; M.W., Y.Y. and J.M. led the experiments and collected the overall data; Y.S.,
921–937 (2020). J.T., D.M., C.Y. and C.X. contributed to sensor characterization, validation and sample
32. Koh, A. et al. A soft, wearable microfluidic device for the capture, storage, analysis; N.H. contributed to the signal processing and app development. J.S.M., T.K.H.
and colorimetric sensing of sweat. Sci. Transl. Med. 8, 366ra165 (2016). and Z.L. contributed to the design of the human studies. W.G., M.W., Y.Y. and J.M.
33. Ray, T. R. et al. Soft, skin-interfaced sweat stickers for cystic fibrosis diagnosis co-wrote the paper. All authors contributed to the data analysis and provided feedback
and management. Sci. Transl. Med. 13, eabd8109 (2021). on the manuscript.
34. Lee, H. et al. A graphene-based electrochemical device with thermoresponsive
microneedles for diabetes monitoring and therapy. Nat. Nanotech. 11, Competing interests
566–572 (2016). The authors declare no competing interests.
35. Lee, H. et al. Wearable/disposable sweat-based glucose monitoring device
with multistage transdermal drug delivery module. Sci. Adv. 3,
e1601314 (2017). Additional information
36. Kim, J. et al. Noninvasive alcohol monitoring using a wearable tattoo-based Supplementary information The online version contains supplementary material
iontophoretic-biosensing system. ACS Sens. 1, 1011–1019 (2016). available at https://doi.org/10.1038/s41551-022-00916-z.
37. Kim, J. et al. A skin‐interfaced, miniaturized microfluidic analysis and Correspondence and requests for materials should be addressed to Wei Gao.
delivery system for colorimetric measurements of nutrients in sweat and
supply of vitamins through the skin. Adv. Sci. 9, 2103331 (2022). Peer review information Nature Biomedical Engineering thanks the anonymous
38. Sempionatto, J. R. et al. Wearable and mobile sensors for personalized reviewers for their contribution to the peer review of this work.
nutrition. ACS Sens. 6, 1745–1760 (2021). Reprints and permissions information is available at www.nature.com/reprints.
39. Yang, Y. et al. A laser-engraved wearable sensor for sensitive detection of uric Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
acid and tyrosine in sweat. Nat. Biotechnol. 38, 217–224 (2020). published maps and institutional affiliations.
40. Yu, Y. et al. Biofuel-powered soft electronic skin with multiplexed and wireless
sensing for human-machine interfaces. Sci. Robot. 5, eaaz7946 (2020). Springer Nature or its licensor holds exclusive rights to this article under a publishing
41. Wang, J. Electrochemical biosensors: towards point-of-care cancer agreement with the author(s) or other rightsholder(s); author self-archiving of the
diagnostics. Biosens. Bioelectron. 21, 1887–1892 (2006). accepted manuscript version of this article is solely governed by the terms of such
42. Uzun, L. & Turner, A. P. Molecularly-imprinted polymer sensors: realising publishing agreement and applicable law.
their potential. Biosens. Bioelectron. 76, 131–144 (2016). © The Author(s), under exclusive licence to Springer Nature Limited 2022

nature research | reporting summary
Corresponding author(s): Wei Gao
Last updated by author(s): May 30, 2022
Reporting Summary
Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Research policies, see our Editorial Policies and the Editorial Policy Checklist.
Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested

A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient)
AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)
For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.
Software and code

Policy information about availability of computer code
Data collection Mbed was used to program the microcontroller. CH Instrument was used for off-body sensor data collection. Comsol was used for the
simulations.
Data analysis Origin 2018 was used to analyse and plot all the data, and to calculate the statistical parameters.
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors and
reviewers. We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.
Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A list of figures that have associated raw data
April 2020
- A description of any restrictions on data availability
The main data supporting the results in this study are available within the paper and its Supplementary Information. Source data for Figs. 4 and 5 and for
Supplementary Figs. 36 and 39–41 are provided with this paper. All raw and analysed datasets generated during the study are available from the corresponding
author on request.
1
nature research | reporting summary
Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf
Life sciences study design

All studies must disclose on these points even when the disclosure is negative.
Sample size For the on-body evaluation of the wearable sensor, 13 healthy subjects were recruited; for the sweat BCAA study, 10 normal-weight healthy
subjects, 7 overweight or obese subjects, and 3 obese T2DM patients were involved; for COVID-19 metabolic fingerprint analysis, 8 COVID-19-
positive and 8 COVID-19-negative samples were used. Sample sizes were chosen on the basis of standards in the literature for proof-of-
concept experiments.
Data exclusions No data were excluded.
Replication All attempts at replication were successful when following the device-fabrication process described in the paper.
Randomization The devices were fabricated with the same process and tested in all participants under same conditions. Randomization was therefore not
relevant to the study.
Blinding Blinding was not relevant, because a blinding process wouldn't influence the sampling result.
Reporting for specific materials, systems and methods

We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.
Materials & experimental systems Methods

n/a Involved in the study n/a Involved in the study
Antibodies ChIP-seq
Eukaryotic cell lines Flow cytometry
Palaeontology and archaeology MRI-based neuroimaging
Animals and other organisms
Human research participants
Clinical data
Dual use research of concern
Human research participants

Policy information about studies involving human research participants
Population characteristics Normal-weight healthy individuals with a body mass index (BMI) of 18.5 to 24.9 kg m-2 with fasting serum glucose < 100 mg
dL-1 (Healthy); Overweight/obese individuals with a BMI of 25 to 35 kg m-2 and fasting serum glucose < 6 mg dL-1
(Overweight/Obesity); Obese individuals with a BMI of 25 to 35 kg m-2 and fasting serum glucose >= 126 mg dL-1 (Obesity &
T2DM). COVID-19-positive and COVID-19-negative serum samples were purchased from RayBiotech, Inc.
Recruitment The participating subjects were recruited from Caltech campus, UCLA hospital and the neighboring communities through
advertisement by posted notices, word of mouth, and email distribution. There were no self-selection biases or other biases.
Ethics oversight California Institute of Technology
Note that full information on the approval of the study protocol must also be provided in the manuscript.
April 2020

A Wearable Electrochemical Biosensor For The Monitoring of Metabolites and Nutrients

Uploaded by

Copyright:

Available Formats

A Wearable Electrochemical Biosensor For The Monitoring of Metabolites and Nutrients

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Wearable Electrochemical Biosensor For The Monitoring of Metabolites and Nutrients

Uploaded by

Copyright:

Available Formats

Articles

A wearable electrochemical biosensor for the

Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng 1225

COVID-19 Central fatigue Cardiac hypertrophy Hepatic lipid storage T2DM

Sweat Inlet layers

ADC 1 TIA 1 WEDPV1

ADC 4 Voltage divider T sensor

UART Bluetooth User interface

1226 Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng

Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng 1227

ΔJ (µA mm–2) Potential (V)

Current density (µA mm–2)

LEG 0.4 0.6 0.8 1.0 0.4 0.6 0.8 1.0 0 6 12 18

Indirect detection Lys Phe VC Creatine

RAR 0 0.1 0.2 0.3 1 1.5 2 2.5 1 1.5 2 2.5

Current density (µA mm–2)

Concentration by sensor (µM)

Raw sweat Leu sensor Sweat Tyr

1228 Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng

ΔI ' of STrp (µA)

Sweat rate (µl min–1)

Sweat rate (µl min–1)

120 Predicted Carbagel, S3 3 Carbagel, S3

Sweat Pilogel, S1 Pilogel, S1

Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng 1229

Tyr, Trp level

BCAA concentration (µM)

Trp concentration (µM)

Subject 2 Subject 2 Subject 2

Subject 1 Intake Tyr No intake Tyr

1230 Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng

Healthy Obesity and T2DM 0 35 70 105 140

Serum BCAA concentration (mM)

Serum Leu concentration (µM)

Leu concentration (µM)

BCAA concentration BG concentration

200 560 200 460

BCAA intake Protein intake

Serum Leu concentration (µM)

Obesity COVID-19 positive

Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng 1231

1232 Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng

Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng 1233

Integrated system validation in human subjects. System evaluation during References

1234 Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng

Nature Biomedical Engineering | VOL 6 | November 2022 | 1225–1235 | www.nature.com/natbiomedeng 1235

A description of all covariates tested

Software and code

- A description of any restrictions on data availability

Life sciences study design

Data exclusions No data were excluded.

Reporting for specific materials, systems and methods

Materials & experimental systems Methods

Human research participants

Ethics oversight California Institute of Technology

You might also like