Sweat Analysis: A Painless Alternative to Real-Time Vital Signs

Introduction

Sweating is an essential body function that occurs spontaneously during exercise and other types of physical activity. The core reason for sweating is to assist the body in regulating temperature through the cooling effect of the excreted water (González‐Alonso & Kalsi 2015). In addition to water, sweat contains other chemical constituents that can serve as indicators of an individual’s health or physiological status. These chemical components are referred to as biomarkers and include molecules such as glucose, sodium, and protein among others (Sekine et al. 2018). While blood also contains these biomarkers, which implies that blood samples are suitable for the evaluation of these molecules to determine an individual’s health standing, collecting blood is an invasive process often associated with additional risks such as infection at the puncture sites. The presence of important biomarkers in sweat offers the potential for its use as a painless alternative to the current procedure for the analysis of vital signs. Therefore, it is possible to use sweat in its natural state to identify specific biomarkers. Furthermore, recent technology has supported the creation of body sensors to facilitate the use of sweat as a biomarker directly and indirectly (Rose et al. 2015; Gao et al. 2016; Choi et al. 2017). The purpose of this dissertation is to investigate sweat analysis as a painless option for real-time analysis of vital signs.

State of the Art (Literature Review)

Importance of Vital Signs

Currently, more than 764 million people are living with chronic diseases all over the world (Vos et al. 2015). These disorders include diabetes, renal disease, and cardiovascular disease, among others, contributing to significant health care expenditure in developing and developed nations. Affected individuals are required to visit health care settings to have their vital signs checked to ascertain whether they fall within the normal range or reflect abnormalities. This process often involves vast resources in terms of transportation costs to and from hospitals in addition to time that could have been spent on other meaningful tasks. It is estimated that a quarter of these patients could realize better health outcomes if they were able to keep track of their health from home (Huang 2015). On the other hand, incorporating available medical resources into mobile phones and other portable gadgets promises to prove beneficial for half of these patients. In addition, such devices would help in tracking and documenting specific measurements over time (De la Iglesia et al. 2015). In most cases, taking vital signs involves simple procedures that patients can easily accomplish in the comfort of their homes with minimal assistance from health care providers. The convenience afforded by providing medical resources in the form of portable tools can be useful in the continuous monitoring of vital signs in pursuit of optimal health outcomes. Data collected by patients at home can then be conveyed to their health care providers through sensor networks and multiagent systems.

Monitoring vital signs during physical exercise is important from medical, health care, and sports physiology perspectives. In professional team sports, training athletes while having the capacity to draw on scientific factors such as vital signs and physical indications generates better outcomes than merely relying upon a trainer’s intuition and experience (Hara et al. 2016). A classic example is the training of the Japanese national rugby team using personalized locations as exhibited by Global Positioning System (GPS) systems to measure speed and bursts of speed. This technology-enhanced training has led to substantial improvements in athletes’ performance. Similar tactics are being used by professional football teams in Japan through the introduction of information and communications technology (ICT) applications into sports (Hara, Kawabata & Nakamura 2015).

The monitoring of vital signs is not limited to professional sports and medical practice. Physical education teachers in elementary and junior high schools are also urged to be cautious when involving schoolchildren in outdoor physical activities. A recent report in Japan shows that many schoolchildren have ended up in hospitals after suffering heatstroke while exercising (Hara, Kawabata & Nakamura 2015). The researchers attributed these observations to drastic daily temperature changes and the fact that children require more time to acclimatize to heat than adults. Such mishaps can be avoided by incorporating an ICT gadget that can detect and measure vital and physical signs for each child during and immediately following exercise. However, a wide gap exists between specialized sports and school sports. Furthermore, the weight, dimensions, and price of vital sign sensors as well as unsolved technical issues are currently limiting the use of these devices in school settings.

Sweat Analysis in Monitoring Vital Signs

Initially, sweat analysis was done by collecting sweat and subjecting it to conventional laboratory analyses similar to those involving other biofluids such as blood and urine. However, innovations in technology have steered the development of wearable sensor systems that facilitate determining the body’s physiological status and delivering immediate feedback that allows prompt intervention. Therefore, these gadgets play a crucial role in assessing and improving individual health and functioning. At the moment, wearable technologies can be obtained commercially and used to observe physical activities and monitor vital signs. However, the currently available commercial technology cannot discreetly avail molecular-level data linked with the body’s dynamic chemistry. Thus, the development of sweat-based wearable devices that monitor vital signs and offer biomonitoring capabilities promises to enhance health care.

Currently, several existing gadgets have the capability to use wireless innovations such as Bluetooth or Wi-Fi to establish connections that link to sensors when monitoring vital signs. Examples of such instruments are pulse sensors (Liu et al. 2018), glucometers, and blood oxygen sensors. Nonetheless, these devices suffer from a shortcoming in that they can only display information transmitted through the mobile tool because they lack inherent information-processing capabilities (De la Iglesia et al. 2015). Furthermore, although available wearables track indicators such as heart rate and physical activity, they are not designed to provide information at a deeper molecular level. This technological gap has encouraged a rapid advancement in chemical sensors that can non-invasively detect analytes in accessible biofluids, providing a window into the body’s overall dynamic biomolecular state (Bariya, Nyein & Javey 2018).

A few candidates exist for wearable sensing of physiological information. However, most of these have shortcomings that limit their application. For example, blood and interstitial fluid can be sampled continually by implantable devices but are not readily accessible through non-invasive means in a wearable format. As another example, obtaining tears can be painful or risky. Furthermore, reflex tears can be generated by irritation of the eyes, thereby interfering with sensor readings. Sensors that use urine cannot be executed on a wearable platform, and the constituents in saliva are largely influenced by the most recent meal and thus can provide a highly limited amount of physiological insight. Conversely, sweat promises the immense potential for wearable sensing because it can be produced whenever the need arises without the use of invasive procedures, for instance, through chemical stimulation at localized, convenient anatomical sites. Therefore, sweat is useful for the constant surveillance of physiological conditions. Furthermore, sweat sensors can be placed near the site of sweat production, enabling prompt detection before the analytes decompose. Even though sweat analysis offers challenges regarding the reliable quantification and interpretation of levels of biomarkers, its strengths over the analysis of other types of biofluids have made it the most ideal biofluid for inventions in wearable technology.

Various analytical procedures can be done using sweat to reveal physiological information. For example, electrochemical sensing can be accomplished during sweat analysis. In particular, skin conductance also referred to as galvanic skin response (GSR), is an analytical method that estimates skin electrical conductance. This parameter often changes depending on the wetness levels on the outward skin and is valuable given that sweat glands are controlled by the sympathetic nervous system (Vinik et al. 2017). Consequently, the electrical resistance of the skin changes whenever an individual experience a strong emotional change (Khan et al. 2016), making GSR a reliable sign of psychological or physiological arousal.

Rather than a diagnostic tool constrained to clinical application, skin sweat is a bodily indication used to evaluate human responses in various settings. Various life scenarios can lead to neurological responses from the autonomic nervous system and trigger a rise in skin perspiration. The ensuing moisture alters the skin’s electrical conductance, permitting estimation of the amount of sweat produced by the sweat glands in what is referred to as the galvanic skin response (GSR). Because the autonomic nervous system also regulates other bodily features such as respiration, blood pressure, and heart rate, GSR has been used to evaluate these life signs. For example, changes in the pace of an individual’s pulse together with skin sweat can guide the classification of a person’s mental state, assist in distinguishing different mental states and help in the diagnosis of mental stress (Dias & Cunha 2018).

In physical activities, continuous surveillance of skin sweat is a valuable physiologic indication, offering numerous uses in sports and investigations of human behavior. Sweat analysis also opens a novel area of inquiry into clinical situations such as dehydration. However, a knowledge of the milieu of physical activity should be a prerequisite to using sweat analysis findings to identify dehydration. These achievements are possible because skin perspiration lends itself to the acquisition of data about a participant’s physiological status as a result of the different ions and biomolecules found therein. Therefore, sweat is an ideal biological liquid for non-invasive chemical assessment to detect medical disorders via examination of ion concentrations in clinical settings. Components of sweat can provide indications of electrolyte imbalance in addition to pathological states such as osteoporosis, cystic fibrosis, physical stress, and bone mineral loss (Kim et al. 2018). Evaluation of physical stress is particularly valuable in the psycho-physiological assessment of military officers undergoing intensive training (Familoni et al. 2016).

Sensors in Sweat Monitoring

Sweat monitoring uses two major types of sensors: epidermal-based and fabric or flexible plastic sensors. The epidermal sensors provide a conformal connection between the surface of the electrodes and the biofluid that permits direct contact between the electrodes and the epidermis for uninterrupted monitoring. In contrast, plastic-based sensors are frequently used because of their ability to sustain continuous exchanges over a large skin surface area. Such sensors can be fixed into fabric or attached by screen-printing into cloth. Consequently, it is possible to collect precise measurements such as pH and levels of ions such as potassium, ammonium, and chloride.

Even though analyzing blood as a biofluid facilitates the direct detection of specific diseases, understanding regarding the levels of sweat analytes and health standing remains limited. In the endeavor to comprehend the correlation between sweat analytes and blood or interstitial fluid levels for the purposes of sweat examination for clinical or health checking, it is crucial to first appreciate the mechanism by which analytes are apportioned into a sweat (Bariya, Nyein & Javey 2018). Therefore, the following section describes the biology behind the formation and excretion of sweat.

The Physiology of Sweat

Production of Sweat

Sweat glands are accessories of the epidermal region of the skin and are normally distributed throughout the entire body except the lips, external genitalia, and nipples. Sweat glands function in the process of perspiration to help to eliminate waste products from excretory organs such as the kidneys. These glands are classified based on their physical structure and mode of secretion as eccrine and apocrine glands. The average human being has about 4 million sweat glands, of which 3 million are eccrine glands (Asahina, Poudel & Hirano 2015). Therefore, sweat glands permit the secretion of sweat through the cell membrane without pinching off the external cell parts. Postganglionic sympathetic fibers innervate the apocrine and eccrine glands (Hu et al. 2018).

The production of sweat is associated with fluctuations in body temperature. The major neurotransmitter for eccrine glands is acetylcholine, while catecholamines serve as the key neurotransmitters for apocrine glands. An increase in body temperature causes the sympathetic nervous system to activate the eccrine sweat glands to release water to the outer surface of the skin (Cui & Schlessinger 2015; Owen 2016). The water then evaporates and cools the body. Therefore, eccrine sweat is central to adjusting body temperature. When temperatures are extremely high, humans are capable of excreting more than a liter of sweat in an hour (Amano et al. 2015).

Following the formation of sweat, the dermal ducts in the sweat glands convey this substance to the surface of the skin. Throughout this process, analytes such as metabolites, ions, hormones, acids, and low molecular weight peptides are apportioned into a sweat. The predominant ions in sweat are sodium and chloride, which are conveyed through active transport from the blood and the secretory coil. Consequently, osmotic pressure builds up and propels water into the sweat gland. Sodium and chloride ions are then reabsorbed via openings situated in the walls of the dermal duct walls. The rate of reabsorption is relatively constant, leading to a corresponding increase in the levels of sodium and chloride ions in secreted sweat as the sweat rate increases.

Understanding of the actual mechanism involved in the secretion of other sweat analytes is limited. However, it has been hypothesized that other analytes are partitioned into sweat via active or passive transport from the blood or interstitial fluid in neighboring blood vessels. The ultimate concentration of the analytes in sweat is determined by the precise method of partitioning together with aspects such as molecular size, charge, and sweat rate.

Eccrine sweat in humans is basically a watery solution of sodium chloride with minimal quantities of other electrolytes that are usually found in plasma. Eccrine sweat may also exhibit a reddish pigment. People who do not sweat heavily can lose large quantities of sodium chloride when exposed to elevated temperatures or intensive labor, potentially leading to sodium deficiency (Hurley & Johnson 2015; Turner & Avolio 2016). This deficiency may be linked with inefficiencies of the eccrine gland (Cheshire 2016). However, the efficacy of the gland improves with continued use, minimizing salt loss. The secretory region of the eccrine glands secretes an ultrafiltrate, which is processed by cells that cover the duct area. Reabsorption of electrolytes such as sodium occurs in this area, leading to the production of hypotonic sweat and preserving electrolytes as a result.

The apocrine sweat glands in humans are found in areas of the body with hair, for example, the armpits, scalp, and genital region (Borowczyk-Michalowska et al. 2017). These glands constantly produce concentrated fatty sweat into the tubes within the glands. Emotional stress triggers the contraction of the glands, causing the glands to eject their contents. Normal flora (bacteria) on the exterior of the skin then act on the fat in the sweat to produce unsaturated fatty acids that have a pungent smell (Jobling 2015; Lübke & Pause 2015). In nonhuman species, an additional function of apocrine sweat is to provide pheromone signaling, which influences social activities such as parenting and mating (Mutic et al. 2016). The role of apocrine sweat in humans is unknown but is thought to contribute to sexual attractiveness (Motofei & Rowland 2016).

There exist mixed sweat glands that are referred to as apocrine glands, which are located in the perianal and axillary areas in humans. Apocrine glands mature during adolescence from the predecessors of eccrine glands. Several investigations using the hormones epinephrine and methacholine as sweat stimulants have shown that the apocrine glands generate sweat at a rate five times higher than that of the eccrine glands (Vary 2016).

The thermoregulatory hub in the hypothalamus, which is sensitive to temperature changes, is responsible for the regulation of body temperature by controlling the levels of eccrine sweat and the flow of blood to the skin. The thermoregulatory center is also sensitive to other factors such as physical exertion, hormones, emotions, and endogenous pyrogens. The impact of emotions and physical activity on the thermoregulatory center is achieved via the limbic system (Asahina, Poudel & Hirano 2015; Mogenson 2018).

Sweat glands found on the palms as well as those on the soles of the feet are triggered mainly by emotional stimuli (Hashmonai et al. 2017). In comparison, axillary sweating is prompted by thermoregulatory modifications in addition to emotional impetus. However, no significant differences are evident in the morphology, nervous organization, and neurotransmitter responses between the palmar and plantar glands and other sweat glands. Therefore, it is hypothesized that a unique thermoregulatory center is responsible for sweating in the palms, soles, and axillae (Kauffman 2018). This hub is different from the control center that influences sweating in other parts of the body and is believed to be directed by input from the cortex only, making it impervious to temperature alterations. The lack of sweat triggered by emotion during sleep as well as under the influence of sedatives is an observation that backs this theory.

Emotional sweating is believed to be a primitive human function of use while hunting or on the battlefield. Low quantities of sweat on the palms and soles are helpful in enhancing friction by regulating the humidity of the stratum corneum. Consequently, the ability to grip is enhanced. Apart from cooling the body during physical activity and under elevated temperatures, other benefits of sweating prompted by heat include lowering blood sugar levels (Emaminejad et al. 2017), reducing alkali stores (Reis 2017), augmenting the number of erythrocytes (Asoglu et al. 2016) and boosting the specific gravity of blood. Furthermore, during emotional sweating, natural smells from the apocrine glands become aerosols and liberate pheromone signals (Banner et al. 2018).

The Chemical Basis of Sweat

Sweat is a clear biological fluid that is slightly acidic at a pH of 6.3, which makes it more acidic than blood (Nyein et al. 2016). Typical concentrations of ions such as potassium and calcium in sweat are usually in the millimolar range. Similarly, weak acids or alkalis – for example, ammonia – can reach the sweat glands via diffusion and form ions because of the elevated pH of sweat, which entraps the ions in the secretory coil and generates millimolar concentrations exceeding those in blood. Other moieties such as urea and lactate can move into a sweat from blood or be produced all through the metabolic activities of the sweat glands at millimolar levels. Molecules with high molecular weight, for instance, glucose, are found in sweat in the micromolar range, which is lower than that found in blood by orders of magnitudes. Conversely, proteins such as neuropeptides or hormones can be found in the blood in the nanomolar or picomolar range (Bariya, Nyein & Javey 2018). In addition to naturally produced analytes, sweat may contain such extraneous molecules as drugs, alcohol, and heavy metals as the body try to eliminate toxins (Bariya, Nyein & Javey 2018).

Consequently, based on the pH partition theory, alkaline drugs are more likely to build up in sweat than in blood (Sonner et al. 2015). The conveyance of a water-insoluble drug in blood as well as in other types of biological fluids is influenced by the acidity or alkalinity of the liquid and the drug’s pKa. These strictures can be substituted into the Henderson-Hasselbalch equation and used in the computation of the hypothetical concentration ratio of biofluid to plasma (Jadoon et al. 2015). Drug levels are usually higher in plasma than in sweat, creating a concentration gradient that ultimately prompts the diffusion of the unbound drug from the plasma to perspiration via the lipid bilayer of the membranes (Nakata et al. 2017). Water is the main constituent of sweat, comprising approximately 99%. Other constituents include nitrogenous substances including urea and amino acids, metal ions such as sodium and potassium, non-metals such as chloride ions (Dam, Zevenbergen & Van Schaijk 2016), metabolites such as pyruvate and lactate as well as xenobiotics including drug molecules.

The components of sweat are influenced by the mechanisms of analyte partitioning and the method of sweat stimulation (Bariya, Nyein & Javey 2018). In the normal physiological state, urine contains these substances in specific quantities. However, in pathological or disease states, the levels of these components change. Such changes can serve as indicators of various disorders for diagnostic or prognostic purposes. Alternative biofluids include saliva, blood, and urine. However, other impurities found in these substances can complicate their use in clinical studies. Sweat, on the other hand, contains negligible impurities, easing its preparation and thus making it an ideal biofluid for use as a biomarker. Additionally, sweat is less susceptible to contamination, which implies that it is possible to store this substance for protracted periods.

Another benefit of using sweat as a biofluid is the non-invasive sampling associated with its collection. Sweat analysis is thus deemed a rapid and easy procedure compared to methods involving other biofluids, particularly blood, whose sampling is an intrusive process that necessitates surgery. Consequently, patients requiring frequent analyses are predisposed to infections. Furthermore, blood samples need additional processing to eliminate plasma proteins before analysis in comparison to sweat samples. For a practical example, the rapid processing of sweat samples is valuable in doping regulation where test outcomes are needed urgently (Liu et al. 2015).

Despite these benefits, the regular use of sweat samples for clinical purposes has been constricted by prohibitive costs, painstaking sampling, the risk of infection, and the requirement for volume normalization (Heikenfeld 2016). Moreover, evaluating the metabolic products contained in sweat is a rigorous endeavor. Other limitations of sweat as a biofluid encompass difficulties in generating sufficient sweat for test purposes, evaporation of samples, inadequate or inappropriate sampling gadgets, and the need for highly trained workers. The presence of pilocarpine in sweat also introduces errors in measurements. The small quantities of sweat also pose problems in the normalization of sampled volumes when handling quantitative measurements (Luque de Castro 2016).

Introduction to Biomarkers

A biomarker can be defined as a feature that can be quantified objectively and assessed to determine physiological and pathological processes as well as pharmacological reactions to a treatment modality. Examples of biomarkers include biomolecules such as proteins, carbohydrates, lipids, RNA, DNA, genes, hormones, platelets, and enzymes. Any physiological substance that facilitates the identification of a disease can be considered a biomarker. Therefore, alterations in biological morphology or distinct attributes can also be considered biomarkers.

Biomarkers are classified in three different ways, the first of which is based on approaches related to molecular biology and genetics. The first group under this type of classification consists of type 0 biomarkers, which are also referred to as natural history biomarkers (Oberg et al. 2015). These quantify the inherent history of a disease and show a relationship with chosen clinical signs over time. Type 1 biomarkers are drug activity biomarkers, which are used as evidence of therapeutic interventions. The three main types of Type 1 biomarkers are efficacy, mechanism, and toxicity biomarkers. Efficacy biomarkers determine whether drug treatment is effective, whereas toxicity biomarkers indicate whether a drug has adverse effects on the body. Mechanism biomarkers are important in elucidating the mode of action of drugs. Type 2 biomarkers function as alternatives for clinical outcomes of a disease and can be used to forecast the treatment impact of an intervention.

The second mode of classification groups biomarkers as predictive, prognostic, pharmacodynamic, screening, staging, recurrence monitoring, and surrogate endpoints. Predictive biomarkers are used to pinpoint individuals who are likely to demonstrate a positive response to treatment (Gibney, Weiner & Atkins 2016). Therefore, such markers are beneficial in tailoring treatment options for patients to yield optimal outcomes. Prognostic biomarkers, on the other hand, can forecast possible outcomes in patients who do not undergo treatment. Pharmacodynamic biomarkers have to do with identifying the pharmacological effects of drugs. The third classification approach places biomarkers into two broad groups: biomarkers of disease and biomarkers of exposure. The former class consists of diagnostic and monitoring indicators, while the latter group includes predecessor biomarkers used to envisage disease risk.

The usefulness of biomarkers is determined by a number of features that contribute to ideal biomarkers. An ideal biomarker should be precise for a specific disorder and facilitate distinguishing various disease states. It should not pose risks during collection and measurement and at the same time should be easily quantifiable. Other desired qualities of biomarkers include cost-effectiveness, the capacity to render accurate outcomes, rapidity to shorten the diagnostic time, and a specified baseline limit. Finally, for a biomarker to be considered ideal, it should yield similar outcomes across different ethnic groups and genders.

Useful aspects in the use of biomarkers for various functions include validation and qualification. Biomarker validation entails evaluating an assay or measurement of performance attributes (Emerson et al. 2018). Conversely, biomarker qualification comprises all steps taken to provide evidence connecting a given biomarker to biology and clinical endpoints.

Clinical Applications of Biomarkers

Biomarkers serve several functions in health care and clinical research. Their roles include extrapolation of reaction to therapy, forecasting clinical upshots (surrogate endpoints), screening, risk evaluation, pharmacogenetics, diagnosis, and monitoring of patients throughout treatment. These roles are evident in all disease processes.

Prognostic attributes are useful in estimating the possibility of developing a certain disease or the expected time before certain clinical outcomes such as disease progression or death will be observed. Prognostic biomarkers serve these purposes. Their development is usually based on the causal mechanisms of the disease process (Dalerba et al. 2016). However, biomarkers can simultaneously serve prognostic and effect modifier roles. For example, in oncology, excessive expression of the human epidermal growth factor receptor 2 is an indication of a poor prognosis in women with breast malignancy who fail to receive any form of adjuvant systemic therapy (Denkert et al. 2015). In addition, patients exhibiting this phenomenon often experience unsatisfactory outcomes when they receive endocrine therapy. However, these patients respond very well to a number of chemotherapies, including taxanes and anthracyclines as well as inhibitors of the receptor. As a result, it is important for studies dealing with the prognostic function of a biomarker to consider the anticipated clinical use of the biomarker and the possible confounding effects.

The development of targeted cancer therapies that are meant to address specified somatic molecular aberrance necessitates the use of appropriate indicators to guide therapy. Therefore, biomarkers can inform the choice of individual agents. Consequently, an effect modifier may be linked to a category of treatment such as chemotherapy or a particular drug within the class. For example, measuring the production of human epidermal growth factor receptor 2 is useful in deciding whether to use endocrine or anti-HER2 therapies in patients with breast cancer (Denkert et al. 2015), changes in the K-ras gene determine the use of antibody therapy against the epidermal growth factor receptor and mutations in the anaplastic lymphoma kinase gene influence the use of crizotinib as a treatment agent in lung cancer (Kim et al. 2017).

Surrogate endpoints are biomarkers that work as alternatives for clinical endpoints by forecasting any likely clinical advantage or shortcoming using treatment, epidemiologic, pathophysiologic, or other forms of scientific evidence (Robb, McInnes & Califf 2016). In conventional clinical practice, surrogate endpoints inform medical providers about the probability or actual occurrence of a phenomenon. For example, an upsurge in the concentration of tumor biomarkers in the bloodstream of a patient with a prior diagnosis of cancer may point towards relapse or metastasis. Consequently, such information can inform the most appropriate therapeutic options instead of waiting for a relapse to be identified through other means.

Biomarkers are beneficial for diagnostic purposes by aiding in the assessment of a patient’s likelihood of having a given disease. The benefits of risk assessment are often realized in cases where preventive approaches, prompt diagnosis, and timely treatment are associated with improved patient outcomes. For instance, in most cancers, prompt diagnosis and timely treatment lead to reduced morbidity and mortality. In particular, the use of prophylactic surgery together with chemoprevention using selective estrogen receptor modulators minimizes the need for surgery and the associated risks in patients with breast cancer. However, these strategies can only be applied to individuals with an elevated risk of breast malignancy, for example, people with inherent breast cancer risk (alterations in the BRCA1 and 2 genes). Biomarkers also make it possible to diagnose a disease in its initial stages when treatment options can yield optimal outcomes, often the case in different cancers.

Diagnostic biomarkers also ascertain whether a patient has a specific disease. For example, in an oncology situation, a patient may exhibit medical, radiographic, and tissue findings that are positive indicators of a cancer diagnosis. However, the main cause of the malignancy cannot be easily determined. Therefore, biomarkers can be employed in discriminating between malignant and benign tumors. Similarly, these biomarkers can point out the origin of malignancy, for example, distinguishing between the colon and cervical cancers or differentiating a solid malignancy from a lymphoma.

Even though drugs belonging to a specific class have known patterns of pharmacokinetics and pharmacodynamics, individual factors can also modify the observed effects during treatment. These include inherent variations in DNA sequences, for example, single nucleotide polymorphisms (SNPs), which can alter the dissemination of a drug in the body, its metabolism, and its action on target tissues. The investigation of these individual disparities is referred to as pharmacogenetics. Therefore, biomarkers such as SNPs can be used to forecast any unique predisposition to drug toxicity (Shirasaka et al. 2016).

Monitoring biomarkers are types of biomarkers that help medical providers to scrutinize patients for possible disease reappearance or other adverse effects that may not be easily discernible at a given time. While the use of tissue-based biomarkers is feasible, it constitutes an invasive process. Therefore, other forms of biomarkers such as those found in blood or other biofluids are ideal for monitoring purposes. As a result, the process of using monitoring biomarkers makes it possible to keep an eye on previously ill patients who have been declared free of disease or to discover a phenomenon more quickly than might be achievable with conventional clinical methodologies.

Sweat Biomarkers

As mentioned earlier, sweat is a mildly acidic solution with a pH ranging from 5.5 to 5. It consists primarily of water and electrolytes such as potassium, sodium, and chloride as well as other substances such as pyruvic acid, urea, lactic acid, and peptides (Hurley & Johnson 2015). Additional substances, found in lower quantities, include components of the immune system such as cytokines, antigens, and antibodies. Xenobiotics such as ethanol, cosmetics, and drugs are also present. All these constituents are held in the eccrine and apocrine glands and are released from these glands into the sweat and conveyed through a sweat pore to the external skin surface. These substances function as biomarkers of varying physiological states.

Researchers have evaluated the types of antibodies found in sweat from healthy human subjects. The three main antibody isotypes that have been characterized in human sweat are IgG, IgE, and IgA. Identified cytokines include interleukin (IL)‐1α, IL‐6, IL‐1β, IL‐31, IL‐8, transforming growth factor (TGF)‐β and tumour necrosis factor (TNF)‐α. The assessment of sweat components using mass spectrometry shows that sweat also contains serum albumin, clusterin, apolipoprotein B, and prolactin‐inducible protein (Katchman et al. 2018).

Sodium and chloride ions are usually reabsorbed in part during their movement in the reabsorptive sweat duct. Pathological states can alter the composition of sweat by changing the levels of various constituents or causing the inclusion of new substances that serve as biomarkers of disease. For instance, the readings of sodium and chloride ions in perspiration have been used to inform the identification of cystic fibrosis in infants and children through pilocarpine-powered iontophoresis. Cystic fibrosis is a consequence of an aberrant chloride membrane transporter, causing a decrease in the movement of sodium and chloride ions in the reabsorptive duct, which leads to the excretion of more concentrated sweat than is found in healthy individuals.

Proteomic analyses show that certain biomarkers in sweat correlate positively with specific disease states. For example, the P.O.W.E.R. study showed that elevated levels of proteins such as substance P, neuropeptide Y, and calcitonin gene‐related peptide were present in the sweat of premenopausal women suffering from the last stages of major depressive disorder (Oliveira et al. 2018). Conversely, a proteomic evaluation of merged samples of sweat from schizophrenic patients revealed that the levels of prostatic‐binding protein, kallikrein, and thioredoxin were two times higher in the test subjects than in the controls. In a different study conducted by Adewole et al. (2016), pooled sweat samples from patients suffering from active tuberculosis contained a total of 26 distinct proteins.

Sweat generated as a consequence of heat, exercise, chemical prompting or stress usually varies in the chemical makeup. Therefore, wearable sensing apparatuses ought to match the anticipated application. For example, during intensive physical activity, the body goes through rapid physiological adjustments. Consequently, the ensuing sweat will exhibit wide-ranging analyte profiles that point toward high metabolic activity, which may be beneficial in fitness monitoring. However, better outcomes are obtained in medical screenings using ‘equilibrium’ sweat collected when the body is at rest. Such sweat can be sampled with the aid of local chemical stimulation of the sweat glands in a process known as iontophoresis (Bariya, Nyein & Javey 2018).

Reliability of Sweat Concentration

Thermoregulatory sweating is responsible for the loss of water and electrolytes when sweating. In some cases, protracted exercise periods, involvement in high-intensity exercise, and working out in hot surroundings can lead to the loss of large quantities of sweat, which can cause dehydration and electrolyte imbalances, thus hampering physiological performance. In medical and sports settings, the sweating rate (SR) and the levels of electrolytes in sweat are known to differ significantly from one individual to the next due to numerous inherent or extraneous causes. Therefore, it is necessary to consider customized fluid replacement strategies as recommended by Baker (2017). Furthermore, before deciding to use sweat components as biomarkers for determining vital signs and disease diagnosis, it is imperative to bear in mind the dependability of sweat concentration. The quantity and concentration of sweat determine the testing outcomes. Therefore, any undue variations in sweat concentrations can interfere with the final outcomes.

Consequently, scientists and medical experts have performed sweat tests using different methods, which can also cause unwanted variability in sweat rate and concentration. For instance, sweat analyses can be done using sweat samples collected using whole-body methods or can be restricted to a particular body region. In addition, the analytical methods used to analyze sweat can vary in their modes of operation, scheduling and length of sweat collection, skin sanitizing processes, and collection and storage of sweat samples as well as the actual analytical procedure. Using unacceptable or uneven methods of sample collection and analysis can alter outcomes substantially by introducing errors, background noise, and misinterpretation of findings. Therefore, this section looks at the methodological considerations to guarantee the reliability of sweat concentrations as well as various studies conducted to validate different sampling and analytical approaches. Even though sweating leads to the loss of several electrolytes, sodium is lost in large quantities, significantly affecting the body’s fluid and electrolyte balance. Therefore, most studies reviewed herein tend to focus on sweat sodium content as an indicator of sweat concentration.

Influence of Analytical Methods on Sweat Concentration Validity

Different analytical procedures can be exploited to determine sweat sodium concentration, which could have varying implications for athletes and scientists. Therefore, the reliability of sweat concentration can be influenced by the laboratory methods used to analyze the sweat. A few studies have sought to determine the influence of this parameter on sweat concentration.

For example, Baker et al. (2014) weighed field methods of extracting and analyzing sodium and potassium in sweat against reference laboratory approaches for sweat samples collected using absorbent patches in a hot, moist environment. The field extraction method entailed using a syringe to suction sweat samples from the absorbent patch, whereas the laboratory approach involved centrifuging the patch to collect accumulated sweat. In contrast, the analytical method in the field used the HORIBA compact system, whereas laboratory analysis used ion chromatography and high-performance liquid chromatography (HPLC). Sweat samples for both conditions were collected from seven body sites, including the forehead, right anterior mid-thigh, left posterior mid-forearm, right posterior mid-forearm, and upper chest as well as the left and right scapula. A total of 30 athletes were involved in one-hour cycling activities in a heat chamber at 33°C. Skin areas at the specified anatomical sites were cleaned using deionized water and dried with gauze 10 minutes following the initiation of exercise. Each anatomical site had two sweat patches, each of which was analyzed in the heating chamber using the field system or laboratory approach.

The researchers noted that the sodium and potassium concentrations obtained using the syringe system fell within error measurements of ±15.4 and ±0.68 milliequivalents per liter, respectively (Baker et al. 2014). These results did not differ significantly from those obtained in a typical laboratory setting. The findings of this study showed the reliability of sweat concentration when extracted and analyzed using the field (syringe HORIBA) technique from local absorbent patches. Therefore, this method can be used successfully in instances where rapid estimates of sweat sodium are required in a hot, moist environment.

In a separate study, Goulet et al. (2017) compared the concentrations of sodium in sweat as measured using 5 analytical techniques: flame photometry, ion chromatography, ion conductivity, indirect ion-selective electrode, and direct ion-selective electrode. A total of 14 participants were involved in the study, leading to the collection of 70 sweat samples that were subjected to analysis using various techniques. Ion chromatography was employed as the reference investigative instrument. Excellent relative and absolute reliabilities were demonstrated among the instruments with an intraclass correlation coefficient of 0.999 and a coefficient of variation (CV) of 2.6%. A high relative validity was also achieved using the five analytical techniques.

When determining the inter-technique absolute validity (using ion chromatography as the reference standard), similar standard errors of estimates were noted among the methods, ranging between 2.8 and 3.8 mmol/L. However, the lowest CV was observed with the direct ion-selective electrode (3.9%), whereas the highest CV was recorded with ion conductivity at 12.3%. These findings led to the conclusion that sweat sodium concentration varies with the type of analytical technique used to determine it. Consequently, findings obtained using different techniques are not equivalent and should not be used as substitutes. Nonetheless, considering typical variations in sweat sodium levels, which Goulet et al. (2017) estimated at ±12%, the inexactness of the endorsements based on flame photometry, ion chromatography, ion conductivity, indirect ion-selective electrode, and the direct ion-selective electrode has negligible health and physiological effects. However, the impact of these differences in diagnostic situations is unknown.

Influence of Sampling Site on Sweat Concentration Validity

Different types of sweat glands are found in various body parts, which means that the levels of sweat excreted may differ from one body part to another. Therefore, the reliability of sweat concentration may be affected by the area of the body from which the sweat has been collected. In this regard, Baker et al. (2018b) determined the association between regional and whole-body sweating rate in addition to regional and whole-body sweat Na+ concentration over the course of exercise. A total of 26 participants were involved in the study, out of which 17 were male and 9 were female. The subjects were engaged in recreational cycling for 1 hour and 30 minutes. Whole-body sweat was collected using the washdown method. In contrast, the regional sweating rate and sodium concentration were measured using absorbent patches from nine anatomical sites. A meaningful agreement was observed between the rate of sweat production at various body regions, sodium concentration, and the whole-body measurements. Day-to-day variability had a noteworthy impact on the regression model to forecast whole-body sweat levels from regional sweat rates at most body sites. However, no effect was discernible on the regional and whole-body sweat sodium concentrations. The findings suggested that regional sweating responses cannot be handled as immediate substitutes for whole-body sweating rejoinders. However, using regression equations to forecast whole-body sweat sodium from regional sweat sodium can estimate whole-body sweat sodium with satisfactory accuracy rates, particularly when using the thigh or forearm. Nevertheless, traditional whole-body mass balance computations remain the recommended method for measuring the speed of whole-body sweat production. This decision is informed by the fact that using the regional sweat rate to project the whole-body sweat rate when absorbent patches are used to collect sweat fails to satisfy the precision or reliability requirements needed to guide fluid consumption endorsements.

Influence of Sample Handling and Storage on the Reliability of Sweat Concentration

As a biological fluid, sweat contains several molecules whose stability may be affected by storage conditions. These changes may affect the outcomes of sweat analysis. In this regard, Baker et al. (2018b) determined the impact of storage temperature on the concentrations of sodium, chloride, and potassium in sweat samples tested 7 days post sampling. The sweat samples were collected by way of the absorbent patch method. A total of 845 sweat samples were obtained from 39 participants with a mean age of 32 years and an average bodyweight of 72.9 kgs. Of the samples, 609 were tested on the same day (pre-storage) for sodium, potassium, and chloride by ion chromatography, while 236 were analyzed for sodium concentration using only a compact ion-selective electrode. The samples were subsequently stored at four different conditions: −20 °C, 8 °C, 23 °C or alternating between 8 °C and 23 °C for a week. The samples were then tested using the same techniques and labeled post-storage.

The researchers noted a high correspondence between pre-storage and post-storage sweat electrolyte concentrations. Mean differences between the two storage conditions were statistically significant. However, the difference did not have a substantial impact on the practical applications of sweat electrolyte concentrations. All storage conditions generated reliable outcomes that did not differ significantly in the levels of sweat electrolytes obtained when the samples were tested immediately after sampling versus those assessed after holding the samples for 7 days.

Influence of Biological Factors on the Reliability of Sweat Concentration

Sweat sodium chloride concentrations have been employed in the identification of cystic fibrosis for several decades. The disorder is attributed to a dysfunction of the CF transmembrane conductance regulator (CFTR), which is a protein channel that oversees the conveyance of chloride and bicarbonate ions (Collaco et al. 2016). Impaired CFTR performance in the sweat gland results in high chloride levels in perspiration. Therefore, changes in the CFTR function can also interfere with the validity of sweat concentration when sampling sweat for the measurement of vital signs as well as other diagnostic or prognostic purposes.

Collaco et al. (2016) investigated the causes of discrepancies in sweat chloride concentrations among patients suffering from cystic fibrosis. The researchers took into consideration a number of biological factors, including demographic, environmental, and distinct individual variations. Sweat chloride amounts were measured in 1,761 participants, including twins or family members. Transmutations in the CFTR gene were mainly responsible for disparities in the dilutions of sweat chloride, accounting for approximately 56.1% of the variation (Collaco et al. 2016). Other sources of variation included time (sweat testing on different days), which accounted for 13.8% of the differences; environmental attributes such as diet and climatic conditions contributed towards 13.5% of the variation, whereas other outstanding factors such as test inconsistency accounted for 9.9% of the variation. Distinct individual features such as genetic variations and exposure to specific environments led to 6.8% of the variability. The evaluation of information from identical siblings showed that modifier genes had no substantial influence on outcomes because the heritability approximation was insignificant. Therefore, for a person with cystic fibrosis, while changes in the CFTR gene influence most of the deviations in chloride levels in sweat, the rest of the discrepancies are linked to random factors. The authors concluded that sweat chloride quantities were reliable biomarkers for evaluating patients’ reaction to treatments meant to remedy a mutant CFTR gene if assay precision and exactitude can be augmented.

These studies show that sweat concentration is a stable biomarker for various medical purposes. However, its reliability can be affected by several biological and nonbiological factors as already described. These challenges can be circumvented by following a few best practices as recommended by Baker (2017). First, it is worth noting that to collect data that are true reflections of sweating during exercise, the application of sampling patches should be done following the initiation of physical activity. The rate of sweating usually increases gradually at the commencement of exercise until a stable rate is achieved. However, the most appropriate time to apply the patch has not been established and can differ based on various features such as the intensity of exercise, the surroundings, and heat adjustment status among others. However, it is proposed that patches should be applied approximately 20 to 30 minutes after the commencement of exercise to yield reliable exercise sweating rates and sweat concentrations. Nonetheless, if the purpose of sweat collection is for diagnostic purposes where sweat rate values are not required, the patches can be applied earlier. Still, the need remains for additional research to ascertain the effect of patch application scheduling on sweat sodium concentration to inform best practices in the testing of sweat.

When preparing to collect sweat, the athlete’s or patient’s skin should be cleaned using alcohol, cleansed with deionized or distilled water, and dried with a paper towel or gauze that is free of any electrolyte. These processes should be done just before the application of the patch to avoid contaminating the collected sweat. Other researchers suggest that forceful cleaning and thorough cleansing of the skin is needed to eliminate skin surface contamination attributed to mineral deposits such as zinc, iron, magnesium, copper, and calcium as well as skin desquamation. However, it has been demonstrated that these procedures are unnecessary when quantifying sodium and potassium levels (Buono, Stone & Cannon 2016).

Another notable challenge is the shedding of patches in the course of exercise. To avoid this problem, the sampling anatomical site can be shaved before applying the patch. An arm wrapper consisting of breathable material can also be used to cover forearm patches to preclude the loss of sticking power. Another best practice is to monitor patches and remove them after an adequate sweat sample has been absorbed but before the patch becomes saturated. Visual assessment is required in this case. The absorbent pad should then be detached from the adhesive dressing using a sterile pair of forceps and kept in an airtight container before analysis. Furthermore, the researcher should don clean gloves free of any electrolyte when applying and removing the patch to avoid introducing contaminants to the sweat sample. Sweat can be removed from the previous cushion by placing it in a sieve tube followed by centrifuging it at approximately 3000 rpm for about 10 minutes if the sample is to be taken to a test center for additional analysis. Another alternative is to place the pad in a syringe and squeeze the sweat out if the analysis is to be conducted in a field setting.

Regarding the storage of sweat samples, it is necessary to ensure that absorbent pads or sweat samples are sealed in airtight containers to avoid evaporation, which could lead to inaccurate assessment of electrolyte concentrations. Few investigations have been conducted to establish the influence of storage time and temperature on the integrity of sweat samples. In the case of sweat testing criteria meant for the identification of cystic fibrosis, samples are required to be stored at approximately -4°C for 72 hours at most to preserve the integrity of the sweat and avoid evaporation (Collie et al. 2014). However, the studies that informed the development of these guidelines did not consider the storage of sweat samples for longer durations, such as 7 days. Nonetheless, Baker et al. (2018b) filled this gap and ascertained that no practical differences were observable in sweat concentration values obtained from samples analyzed immediately and after 7 days. Nevertheless, appropriate storage conditions are needed to ensure the integrity of sweat volumes and composition.

For the purpose of sweat analysis, numerous analytical techniques have been established to measure sweat electrolytes. Examples include mass spectrometry, ion-selective electrode, ion chromatography, and flame photometry (Baker et al. 2018). Modern-day laboratory reference methods for the investigation of sweat electrolytes include inductively coupled plasma mass spectrometry and ion chromatography. These methods need minute quantities of sweat samples and are associated with high levels of accuracy, sensitivity, and reliability with CVs ranging from 1 to 5% (Doorn et al. 2015). In the case that the investigator cannot meet the required sample storage conditions and duration, it may be better to analyze samples in the field as opposed to transporting them to the laboratory for subsequent testing. Other benefits of field testing of sweat samples include reduced transport costs and delays in obtaining outcomes. During field analysis, commonly used techniques are ion-selective electrode and conductivity, which have demonstrated high reliability with CVs ranging from 1 to 4%; in addition, these techniques can generate sweat sodium concentrations within 2 to 4 mmol/L (Baker et al. 2014). However, more studies are needed that will compare different analytical methods to come up with best practices in sweat sodium concentration analysis in field and laboratory setups.

Local sweat concentration is not an acceptable direct determinant for whole-body sodium concentration as shown by Baker et al. (2018a). The researchers attributed this observation to the formation of a microenvironment following skin coverage by a patch, which enhances local humidity and wetness of the skin. Furthermore, it is possible for the skin stratum corneum to interact with sweat accumulated within occlusive layers. It has also been shown that sweat sodium levels differ across various anatomical sites (Baker 2017).

These three shortcomings and their confounding impact on sweat sodium concentration can be alleviated in several ways. For example, absorbent patches are made up of occlusive layers that enhance the accumulation of moisture on the skin. As a result, sweat ducts block gradually, leading to the withholding of sweat at the sweat sample collection site. This phenomenon is known as hydromeiosis (Baker et al. 2018b) and can be reduced by using patches that consist of substances with high absorbency. Reducing the duration that the patch rests on the skin can also lower this effect. Some authors propose that patches should be left on the skin for a maximum of 5 minutes, while others have left the patches on for as long as 90 minutes in field conditions. Prolonged patch times may be associated with the inability of the investigator to reach the subject during exercise or the need for large quantities of sweat samples. Information is limited regarding the effect of patch adherence time and how it affects the concentration of sodium in sweat. Therefore, future studies should focus on ascertaining best practices in the use of absorbent patches to collect sweat.

Another common problem in sweat analysis is obtaining falsely high electrolyte concentrations due to leakage of electrolytes to the samples from the skin using occlusive dressings. It is also possible for the skin to absorb the water from sweat. This issue can be avoided by using sweat potassium concentrations as a quality control check. Given the physiology involved in the movement of sodium and potassium, it is expected that sweat potassium levels should remain constant even in the face of fluctuating sweat rates. Consequently, having sweat potassium levels that are not within the expected limit of 2 to 10 mmol/L is an indication of possible leakage, evaporation, or contamination (Dziedzic et al. 2014).

Studies have shown that the rate of sweating and sweat sodium concentrations vary across various parts of the body (Baker et al. 2016; Baker 2017; Baker et al. 2017). Regional discrepancies in the sweating rate can be explained by anatomical differences. Similarly, interregional variations in the level of sweating and concentration of sodium follow a distinct pattern with the highest rate being observed on the forehead followed by chest, scapula, forearm and finally the thigh having the lowest proportion. The most frequently used sites in the quantification of sweat parameters usually overapproximate whole-body sodium concentration by 25 to 100%. These areas include the forearm, chest, forehead, and scapula. However, the levels of sodium in specific anatomical sites have a high correlation with whole-body sodium amounts. For these reasons, it is possible to employ mathematical regression equations to predict total sodium levels using sweat obtained from specific anatomical sites.

Overall, substantial inconsistencies are have been observed in the pace of sweating and sodium concentrations in perspiration during exercise. These variations can be explained by factors such as disparities in the intensity of exercises, the state of the surroundings, the capacity to acclimatize to the heat, and genetic disposition among others. Unexpected variations can also occur because of unreliable methodology. Furthermore, small variabilities in sweat testing outcomes can still be observed when an investigator adheres to recommended best practices. Changes in body mass prior to and the following exercise can inform the estimation of the whole-body sweating rate. However, relevant corrections for other factors contributing to body mass that are unrelated to sweat are necessary. These include fluid input and output in the form of urine.

Real-Time Sweat Analysis

Real-time analysis can be defined as the process of evaluating data as soon as it is made available. Real-time sweat analysis refers to the testing of sweat samples for specific biomarkers as soon as sweat is produced. Consequently, it is possible to obtain pertinent data immediately and act on them promptly. Advances in health monitoring through the collection and quantifiable chemical analysis of sweat can supplement or hypothetically do away with the need for methods that require the random measurement of blood samples. Commonly used sweat monitoring approaches make use of uncomplicated fabric strips and are confined to rudimentary analysis in a regulated hospital or laboratory settings.

Bendable, wearable sensing devices can produce valuable information regarding the basic functioning of a human subject for use in real-time wellbeing and fitness surveillance. Significant progress has been made in the design and creation of flexible biosensors that conform naturally to the epidermis. However, most inventions can only measure a few physical or electrophysiological strictures and overlook the valuable chemical data found in sweat biomarkers (Imani et al. 2016). Given the intricacy of the sweat secretion, instantaneous and multiplexed assessment of specific biomarkers is important and needs full system amalgamation to guarantee precise measurements (Gao et al. 2016).

Sweat is a representative biofluid, and interest in physiological monitoring using this resource has increased because it is easy to collect without the need for invasive procedures. It is also rich in valuable biomarkers, for instance, proteins, electrolytes, and small molecules (Koh et al. 2016). However, despite the value of sweat analysis in medicine, numerous challenges exist in the interpretation of sweat data because of doubts about its association with other biofluids such as blood and interstitial fluid. Furthermore, biomedical gadgets to facilitate the direct collection and identification of more than one biomarker devoid of evaporation are currently inadequate.

The quantitative analysis of sweat in situ is beneficial for observing an individual’s state of health such as hydration state and disease diagnostics as in cystic fibrosis. However, accessible schemes for whole-body sweat sampling are limited to laboratory settings where conventional chemical analysis approaches are used to identify the composition of sweat. Contemporary endeavors to collect and concurrently analyze sweat entail creating direct contact between sensors and the skin where cloth or paper strips can collect sweat for optical and electrochemical evaluation (Koh et al. 2016). For example, it is possible for electrochemical sensors bonded directly on the epidermis of the skin to identify chemical constituents, including lactate and sodium ions, as soon as they are produced.

Colorimetric rejoinders in modified permeable substrates can generate chemical data such as sweat pH and permit uncomplicated quantitative analyses using gadgets such as smartphones, which can produce high-quality digital images (Jadoon et al. 2015). Further functionality can be added to the system by the inclusion of radio frequency identification systems on the surface of the permeable material to facilitate the wireless transfer of data. The sweat production rate can thus be calculated using these technologies. However, due to limited facts regarding the dissemination of sweat glands in several body parts, it may not be possible to find the overall sweat rate and estimate whole-body sweat loss. Furthermore, initial systems meant for the instantaneous analysis of sweat do not reveal the concentrations of various sweat constituents concurrently. These systems are incompatible with more recent technologies such as radio technologies, skin-fitted electronics, physical sensors, and energy storage gadgets.

As a result, several scientists have devised innovations to circumvent the challenges with laboratory testing of sweat as well as early real-time sweat analysis systems. For example, Liu et al. (2015) developed a real-time sweat analysis system to measure levels of dehydration in humans during physical activity, which has been identified as a serious physiological challenge inflicting adverse effects on affected individuals if not identified in a timely fashion. Sweat is a useful biomarker for the determination of an individual’s hydration status, thus enabling the identification of dehydration. Although previous innovations toward this endeavor did not consider the electrical and chemical attributes of sweat in the validation of the tools, Liu et al. (2015) developed a straightforward test setup to analyze synthetic sweat containing the key constituents of human sweat. Through this setup, the electrical and chemical behavior of the man-made sweat were characterized at temperatures ranging from 5 °C to 50 °C. Furthermore, the authors used three-dimensional printing to create a cost-effective and competent sweat collecting and analysis system. The procedure was substantiated using human subjects. It was shown that dehydration occurred after 40 minutes of physical activity, revealing a need to initiate fluid intake during exercise. The creation of this device was a bold move towards the development of more advanced real-time sweat analysis systems.

Liu et al. (2016) planned, created, and evaluated a sweat-based conductivity sensor gadget to facilitate real-time monitoring of human physiological conditions without invasive procedures. A wearable device was formed by combining a sweat amasser, conductivity feeler, and an interfacing circuit. The sweat collector was made of polydimethylsiloxane (PDMS) and took advantage of the hydraulic pumping act of sweat glands to accumulate sweat from the skin. Three-dimensional plastic printed molds as previously described by Liu et al. (2015) were used to create the PDMS sweat collectors. Electrochemical impedance spectroscopy was used to characterize the conductivity sensor, which was then used to form the interfacing circuit. The system was tested using human subjects to validate its functioning in real-time analysis of human sweat. The initial reading was obtained between 7 and 20 minutes. The precise time was determined by the subject under investigation and the position of the electrodes. As expected, the sweat rate became constant following steady exercise. Sweat conductivity readings declined following the initial reading, which was ascribed to the availability of mineral elements on the skin. However, the seat conductivity readings increased again as a result of changes in the subjects’ hydration statuses.

Further advances in real-time sweat testing were made by Gao et al. (2016) through the development of a mechanically supple and fully assimilated sensor system for complex in situ sweat analysis. Using this system negated the need for other external analyses. The system could quantify distinct sweat metabolites such as lactate and glucose as well as specific electrolytes, particularly sodium and potassium, in addition to body temperature. The purpose of the skin temperature readings was to standardize the reactions of the sensors.

Through this work, Gao et al. (2016) close the technical interlude ‘between signal transduction, conditioning (amplification and filtering), processing and wireless transmission in wearable biosensors by merging plastic-based sensors that interface with the skin with silicon integrated circuits consolidated on a flexible circuit board for complex signal processing’ (p. 2). The separate use of each of these technologies would not have led to these achievements due to the inherent shortcomings of each method. The wearable gadget can generate a comprehensive sweat profile for human participants who have been involved in protracted physical activities in indoor and outdoor settings, thereby facilitating real-time evaluation of the biological state of the subjects. A broad range of customized diagnostic and physiological surveillance applications can be achieved with this system.

System Design

The real-time analysis of sweat for the measurement of vital signs requires that sweat samples are evaluated as soon as they are produced. The sweat analysis system should also be made in such a way that adequate quantities of sweat can be collected before measurement. Given the physiology of sweat production, the system should also ensure that sweat samples will be produced even in the absence of sweat-stimulating triggers in the patient’s surroundings. This section explains the design of a real-time sweat analysis system. Important features such as induction of sweating, sampling, and sweat analysis are also provided.

Induction of Perspiration

Sweat differs from other biofluids that can be collected directly because its production is dependent on several physiological factors. Therefore, it is necessary to stimulate the body to produce sweat before this substance can be collected and analyzed. Biological factors that promote perspiration and ensure the collection of specific quantities of sweat include aerobics and anxiety. In contrast, cold lowers perspiration. To generate sweat volumes that will be adequate for subsequent analyses, perspiration can be induced by modifying environmental factors such as relative humidity and temperature to tweak the body’s regulatory systems, for example, the hormonal and sympathetic nervous system, or diet modifications. The administration of sweat-stimulating chemical compounds such as pilocarpine is also beneficial. The most reliable method of perspiration induction is the application of an electrical current of approximately 3.0 mA for about five minutes simultaneously with pilocarpine on a small area of the leg or arm.

Sampling of Sweat

A perfect sampler should be easy to use without posing any danger to the skin. It should also facilitate the quick collection of suitable quantities of sweat. The simplest sampling system in the studies considered here consisted of an occlusive patch made up of 2 to 3 sheets of gauze or filter paper. However, the major shortcomings of this approach were the inability to adjust the patches according to the patient’s skin, skin irritation, and alteration of the sample pH. Adopting the method was complicated by the necessarily large size of the patch that was required. A nonocclusive contrivance was adopted to circumvent the issue of skin irritation. Hooton and Li (2017) proposed the use of a patch made of Whatman filter paper attached to a surgical dressing sheet covered with a layer containing a bonding agent to facilitate the adjustment of the patch to the skin on the arm or leg. This patch permitted the selective conveyance of oxygen, water, and carbon dioxide through the semipermeable film, enhancing the safety of the patch on the skin. The film also prevented the infiltration of non-volatile substances into the film.

However, the system permitted the vaporization of water from concentrated sweat samples, thereby leaving behind only the solid constituents of sweat. Therefore, it was impossible to tell the total amount of sweat, which in turn affected the accuracy of the subsequent analyses in the identification of cystic fibrosis. Later, the use of sweat collection bags that precluded the free flow of water and air in and out helped to solve the problem of water evaporation. These bags were attached to the patch using adhesive rubber. The entire volume of collected sweat was retained in the bags. Thereafter, this design was enhanced by attaching pipettes, glass rollers, and holders to facilitate the collection of sweat gathered in small quantities. This technology has been commercialized, and currently, Macroduct and Megaduct are readily available commercial samplers (Luque de Castro 2016). Furthermore, connecting sweat samplers with analytical instruments promotes efficiency.

Sweat Analysis

The efficacy of sample collection, as well as the accuracy and responsiveness of analytical techniques, can determine the excellence of sweat analysis. At present, several techniques can be used to test sweat that has not yet undergone metabolism. Nonetheless, a need to investigate drug metabolites in sweat remains. The principles that guide sweat analyzers are colorimetry, conductivity, potentiometry, and osmolality.

Colorimetry is an analytical technique applied in gauging the concentration of colored substances. The method uses the Beer-Lambert law, which asserts that the concentration of a substance has a direct relationship with the quantity of light absorbed (measured as absorbance or optical density). The main instrument used for this purpose is a colorimeter, which comprises a sample cuvette, a source of light of varying wavelengths, a light sensor, and a way of regulating the light source and interpreting the intensity of the transmitted light. Colorimetry has been applied in the creation of a microfluidic instrument for sweat analysis during fitness exercises in a regulated environment and long-distance bicycle contests in dry outdoor settings (Koh et al. 2016).

Conductivity is the degree to which an electric current, charge, or heat can pass through a substance. Materials that allow electricity or heat to pass through them devoid of resistance are known as conductors. Two types of conductivity exist: thermal and hydraulic. Thermal conductivity is a measure of a substance’s capacity to transmit heat, whereas hydraulic conductivity refers to the ability of absorbent materials to convey water. Conductance has been applied successfully in the diagnosis of cystic fibrosis (Mattar et al. 2014; Liu et al. 2015). In a study performed by Mattar et al. (2014), cystic fibrosis diagnoses were eliminated in 714 out of 738 subjects using conductivity assays and chloride tests. The researchers also noted that conductivity values greater than 90 mmol/L corresponded to a sensitivity of 83.3% and 99.7% specificity and concluded that sweat conductivity tests provided high levels of diagnostic accuracy, matching those produced by sweat chloride. Therefore, sweat conductivity can be employed as an indicative exam for cystic fibrosis on its own.

Potentiometry is an analytical method used to determine the concentration of a substance in a solution by measuring the electrical potential between two electrodes that have been immersed in a solution containing the substance under investigation. A high impedance voltmeter is useful for this purpose. Under static conditions, negligible current passes through an electrochemical cell. However, as the level of the electrolyte in the electrochemical cell changes, there is the flow of electrons from the solution to the electrodes following the application of a current. The degree of the electrical potential difference is related to the concentration of the electrolyte. This relationship can inform the deduction of analyte levels. Potentiometry has been applied in sweat analysis to determine physiological stress through the measurement of sodium ion concentration (Cazalé et al. 2016). This technique has also been employed in the development of a wearable chloride ion instrument (Choi et al. 2016).

Currently, different investigative methods have facilitated the analysis of drugs that can be excreted through sweat before undergoing metabolism. However, further investigations are needed to study drug metabolites that may be present in normalized sweat. Most sweat analyzers employ the rules of colorimetry, potentiometry, osmolarity, or conductivity. Capillary electrophoresis and chromatographic techniques – for example, liquid chromatography (LC) and gas chromatography (GC) – can be coupled with mass spectrometers (MS) to enable the precise separation of compound metabolites or drug molecules in sweat (Jadoon et al. 2015). GC-MS linked to electron impact ionization is commonly used for the study of drug content and concentration in sweat (Gentili et al. 2016). In addition, electrospray ionization can be coupled to LC-MS/MS and used to determine drug levels in sweat. Other pertinent techniques include immunoassay methods such as radioimmunoassays and enzyme-linked immunosorbent assays. Overall, analytical techniques in sweat analysis are chosen based on the target analytes in sweat. Potentiometric ion-selective electrodes can be used for single moiety analysis, for example, sodium ions. Other alternatives for sodium analysis include a sweat osmolarity analyzer and the colorimetric chloride patch (Jadoon et al. 2015).

The variability of sweat sample volumes necessitates the normalization of sweat volume to obtain reliable outcomes. Internal standards can be used to normalize sweat samples by measuring the quantities of sodium and potassium ions. However, it has been shown that measuring levels of sodium is a better method of normalizing the sampled volume compared to an estimation of potassium concentration (Jadoon et al. 2015).

A Prototype for a Wearable Sweat Analysis System

A typical sweat analysis system should have all three parameters explained previously: a means of sweat induction, a sampling system, and an analytical system. However, for a wearable device that facilitates real-time analysis of sweat samples, additional constituents are necessary. The sensing constituent should be chosen based on the type of analysis involved. In any case, it should be made of flexible material with electrodes that are suitable for the target analytes to permit incessant multiplexed measurements. To improve the performance of the sensor, a supple polymeric shaft should encase the sensing electrodes to contain changes in pressure while forming a chamber for sweat collection. Such a design minimizes sample losses by evaporation and safeguards the skin from abrasion through direct contact with the skin. A printed circuit board (PCB) should also be attached to the sensor to aid in the calibration of unprocessed analyte signals into consequential concentrations and convey the data to a customized receiver, printout, or phone application for easy read-outs.

Another alternative for signal transmittance is a radio frequency identification (RFID) chip customized for electrochemical detection of ions in sweat. In such an arrangement, sensing electrodes are attached to the same material as the wireless broadcasting constituent via electroplating, making it possible to shrink the entire setup into a wearable instrument. The RFID aerial conveys analyte readings to a smartphone for nonstop surveillance throughout the exercise period. Nevertheless, the efficiency of the sensor depends on near-field transmission between the patch constituents and smartphone to initiate sensing and data broadcasting. The patch also contains fundamental charge-storage modules to permit temporary, low-power procedures to carry on even if the distance between the smartphone and patch should temporarily increase beyond the recommended range. Overall, this system requires the smartphone to be held near the patch for optimal performance. Otherwise, the investigator risks disconnection and loss of data.

In cases where sweat monitoring is needed for clinical purposes, wearable sensors can be made to trigger local sweating via iontophoresis (Emaminejad et al. 2017). The result is equilibrium sweat in inactive scenarios. In such a system, iontophoretic gadgets should have extra electrodes to apply local current in addition to the conventional pair of electrodes for sweat sensing. This extra set of electrodes is impregnated with a hydrogel that holds a sweat-stimulating drug that is taken under the skin by the flow of current. Sweat glands in the area surrounding the site of drug application are prompted to secrete sweat that can then be analyzed to determine sweat analyte concentrations at equilibrium.

Discussion of Current Research

Applications of Sweat Analysis

The first major application of sweat analysis is in the area of disease diagnostics. In the last three decades, much emphasis has been laid on the use of sweat in diagnosing diseases. Cystic fibrosis provides the best example of a disease whose diagnosis depends on sweat analysis. The rationale for this choice is that the disorder arises due to transformations in the CFTR proteins, interfering with the formation and production of sweat, leading to high levels of chloride in sweat (Emaminejad et al. 2017). Therefore, the sodium to potassium ratio is an invaluable biomarker in the identification of cystic fibrosis. Dehydration is another problem that afflicts cystic fibrosis patients. As sweat analysis can also detect dehydration in patients, this procedure can help to confirm a diagnosis. Extensive research has been done to establish best practices in the use of sweat analysis to diagnose cystic fibrosis. Currently, it has been established that two cystic fibrosis diagnoses are possible based on sweat chloride levels: typical and atypical (Farrell et al. 2017). When a patient presents with at least one physical symptom and lab tests yield chloride levels that exceed 60 mmol per liter of sweat, then a real positive diagnosis, also referred to as typical cystic fibrosis, is made. On the other hand, atypical cystic fibrosis is diagnosed when the sweat chloride levels range between 30 and 60 mmol per liter of sweat. These cut-off values have been established based on the normal range of sweat chloride in infants and in the elderly, which is 30 to 59 mmol/L and 40 to 59 mmol/L, respectively. Contemporary studies have also looked into sweat potassium as a prospective biomarker for the primary diagnosis of cystic fibrosis to inform treatment efforts (Farrell et al. 2017). Nonetheless, the clinical use of this alternative biomarker is still undergoing research.

Recent studies have also shown that sweat offers valuable biomarkers for the diagnosis of diabetes. Diabetic biomarkers in sweat include the mean change in sweat rate, sweat constituents, and the relationship between glucose levels in sweat and blood (Oh et al. 2018). Comparing glucose levels in sweat and blood provides more reliable outcomes as long as no extraneous glucose is introduced into sweat samples. The most commonly used anatomical site for sweat glucose sampling is the foot. An uncomplicated indicator test that uses color change is employed. The patch is expected to change from blue to pink within 10 minutes of the addition of 6 drops of water. This color change is attributed to the conversion of anhydrous cobalt II chloride, which is blue in color, to the hydrated version of the salt that exhibits as pink.

For the detection of lung cancer, a sweat-based diagnostic procedure has been proposed. The essay distinguishes the metabolomics of healthy and sick individuals and involves the dilution of sweat samples with 0.1% formic acid. The mixture is then analyzed using LC-TOF/MS, which requires only 10 microlitres of sweat (Jadoon et al. 2015).

Sweat analysis can also be applied in drug testing, which is achieved via two methods: early and late testing (Jadoon et al. 2015). Early testing entails the use of immunochromatographic approaches for the qualitative identification of drugs used within the last 24 hours. In comparison, late testing involves the patch technique for the qualitative identification of drugs used in the last 7 days. This method is commonly employed to follow up on previously identified drug users to confirm abstinence. Therefore, sweat comprises a suitable sample for doping control. The whole human body excretes an average of 300 to 700 milliliters of sweat per day. This volume can potentially contain a small but measurable proportion of drugs that are eliminated via paracellular and transcellular pathways in the skin (Jadoon et al. 2015). Examples of drugs that are known to be excreted through sweat in measurable levels include amphetamines, opiates, gamma hydroxybutyrates, buprenorphine, cannabinoids, and cocaine. Moreover, the levels of ethanol produced in a sweat over time can be determined.

Novel Biomarkers in Sweat

Another notable advance in sweat analysis is the use of genomics and proteomics to identify novel biomarkers in sweat. One such biomarker is dermcidin (DCD), a peptide made up of 47 amino acids. In elevated salt concentrations as well as over a wide pH range, this peptide demonstrates antimicrobial activity against various microbes. Therefore, sweat is thought to play a vital role in the populations of human skin microflora. It has also been observed that aggressive breast carcinomas express DCD and its receptors in significant quantities. Such features are also seen in cells that metastasize to lymph nodes and brain neurons. These observations indicate that DCD plays a vital role in tumorigenesis by encouraging the growth and persistence of cells that form breast carcinomas (Yu et al. 2017).

Prolactin inducible protein (PIP) is a new prognostic biomarker that is generated in some exocrine tissues such as sweat glands. PIP is also expressed excessively in breast and prostate malignancies that have metastasized to other organs (Kachman et al. 2018). Moreover, a study performed on healthy and schizophrenic patients has led to the identification of prognostic biomarkers in sweat. Eccrine sweat has numerous proteins and peptides compared to serum, which shows that eccrine sweat can provide unique disease-associated biomolecules.

Hybrid Sensing Systems

Imani et al. (2016) created a wearable hybrid sensing system that enables concurrent real-time checking of biochemical data in the form of lactate and an electrophysiological signal. This hybrid system facilitates more in-depth fitness monitoring as opposed to either sensor in isolation. The two sensing technologies consisted of modalities, an amperometric lactate sensor made of three electrodes and a bipolar electrocardiogram sensor, which were co-attached to a flexible material to be affixed to the skin. The instrument, authenticated on human subjects, demonstrated the ability to obtain physiochemical and electrophysiological data simultaneously with minimal interference. This work paved the way for the creation of an innovative group of hybrid sensing appliances.

Even with the creation of wearable sweat sensors, few gadgets can gauge biochemical data and biological indications at the same time, which has constrained combined data analysis and extensive medical use. Hong et al. (2018) developed a multifunctional wearable system that combined sweat‐based sensing and monitoring of vital signs to quantify glucose levels before and after exercise. The system comprised a disposable glucose-sensing strip that used sweat as the test fluid and a smart band. The combined system used a single control software that evaluated glucose concentrations in sweat and checked vital signs continuously. The vital signs recorded by the system included blood oxygen concentration, heart rate, and physical activity. Hong et al. (2018) evaluated the efficacy of the system using human sweat samples collected from various anatomical sites as well as samples collected using different techniques in long- and short-term investigations. Consequently, the protocol was optimized for health surveillance. The blending of sweat glucose data and vital signs before and after exercise made it possible to determine blood glucose changes during physical activity. As a result, it was possible to acquire valuable data to help in precluding hypoglycaemic shock during intensive exercise. Furthermore, the amalgamated wearable system paved the way for the development of new, all-inclusive tailored health management strategies by merging the analysis of crucial metabolic and physiological health pointers.

Novel Wearable Sweat Sensor Devices

In most cases, patients often make contact with their medical providers after they have acquired diseases with pronounced symptoms. Thereafter, the patients become receivers of passive care and monitoring by health experts. This strategy is not helpful in the prevention of disease initiation because it places the emphasis on diagnostics and treatment instead of prevention and proactive health care. This approach also prevents individuals from taking charge of the monitoring of their health. The advancing field of wearables intends to address these shortcomings of consolidated, reactive health care by allowing individuals to have an insight into the functioning of their bodies. The longstanding foresight is to create sensors that can be integrated into wearable formats such as wristbands, clothes, patches, or tattoos to investigate various body indicators continuously. Through the transmitting of physiological data as the body changes from a healthy to a diseased state, people will be capable of keeping track of their health in the absence of trained experts and costly equipment (Lee at al. 2017). Recent advances in wearable sweat-sampling instruments have circumvented most of the conventional issues in sweat detection by enabling molecular-level discernment of the subtleties of human bodies.

Numerous wearable sweat sensors that have been developed recently blend varying form factors, substrates, and identification mechanisms. To facilitate ongoing fitness checking, these sensors have been incorporated into everyday athletic fittings such as wristbands or headbands, which can be easily assumed without interfering with or hampering movement. Patch-style platforms are desirable for medical uses due to their ability to stick to the skin inconspicuously and because of their ease of application on various parts of the body. Furthermore, integrated iontophoresis capabilities can be incorporated to facilitate the extraction of equilibrium sweat from specific anatomical sites. Several substrates have been created to execute the various forms of sensors, for example, momentary tattoos, pliable polymers, and fusion systems that merge malleable plastics with conventional silicon integrated circuits.

Different sensing instruments exist for the identification of analytes in sweat. Electrochemical detection is a widely used approach because of its adaptability. This technique quantifies electric currents or potentials at specific electrodes to determine the concentration of analytes. Colorimetric detection is another well-known method that depends on quantifiable color changes following the exposure of specific reagents to the target analytes. Other schemes consist of impedance-centered and optical detection. Even though most of these methodologies have been used in the identification of non-complex ions and metabolites, they can still be modified or implemented alongside other technologies such as synthetic polymer models or affinity-based aptamers for the discriminatory identification of multifaceted molecules. However, a need remains to examine the sturdiness of these techniques for on-body sensing.

Wang et al. (2018) developed a new kind of elastic and electrochemical sweat system made by putting copper submicron pieces on self-supporting graphene paper containing a single layer of MoS2 nanocrystals. The system is aimed at detecting glucose and lactate in biological samples. An amperometric i-t technique was used to quantify glucose and lactate. Glucose levels ranging from 5 and 1775 μM were detected, while the lactate levels ranged from 0.01 to 18.4 mM (Wang et al. 2018). In comparison, the limits of detection for glucose and lactate were 500 nM and 0.1 μM, in that order. The system exhibited a prompt response, acceptable selectivity, excellent reproducibility, and exceptional adaptability, making it an ideal model for checking glucose and lactate levels in human sweat. The structural incorporation of a three-dimensional transition metal, zero-dimensional transition metal sulfide, and two-dimensional graphene provided a novel intuition into the creation of bendable electrodes for the checking of sweat glucose and lactate as well as a broad range of uses in bioelectronics, biosensing, and lab-on-a-chip contraptions. A similar study was done by Lee et al. (2016) with similar outcomes.

Conclusion

Sweat is a biological fluid that offers a wealth of physiological and clinical data due to the presence of biomarkers. Important biomarkers in sweat include sodium, potassium, and chloride ions, glucose, proteins, and lactate. These components can provide information about the well-being of an individual following laboratory or field evaluation of sweat through analytical techniques such as the direct and indirect ion-selective electrode, ion conductivity, and mass spectroscopy among others. However, advances in technology have facilitated the creation of wearable sweat sensors that can permit the instantaneous analysis of sweat, thus allowing prompt feedback based on the available data. Apart from enabling the non-invasive diagnosis of diseases such as cystic fibrosis and diabetes, instantaneous sweat evaluation can measure vital signs, particularly body temperature and dehydration status. It is not possible to measure other vital signs, for example, heart rate, blood pressure, and respiratory rate, because the mechanism of sweat production differs significantly from other physiological processes that lead to respiration. Therefore, it is possible to combine other sensors that can measure vital signs such as blood pressure and heart rate with sweat sensors into flexible wearable devices to enable the comprehensive measurement of vital signs. Future studies should look into ways of enhancing the accuracy and precision of these technologies to ensure the generation and transmission of valid data.

References

Adewole, OO, Erhabor, GE, Adewole, TO, Ojo, AO, Oshokoya, H, Wolfe, LM & Prenni, JE 2016, ‘Proteomic profiling of eccrine sweat reveals its potential as a diagnostic biofluid for active tuberculosis’, Proteomics–Clinical Applications, vol. 10, no. 5, pp. 547-553.

Amano, T, Ichinose, M, Inoue, Y, Nishiyasu, T, Koga, S & Kondo, N 2015, ‘Modulation of muscle metaboreceptor activation upon sweating and cutaneous vascular responses to rising core temperature in humans’, American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, vol. 308, no. 12, pp. R990-R997.

Asahina, M, Poudel, A & Hirano, S 2015, ‘Sweating on the palm and sole: physiological and clinical relevance’, Clinical Autonomic Research, vol. 25, no. 3, pp. 153-159.

Asoglu, M, Aslan, M, Imre, O, Kivrak, Y, Akil, O, Savik, E, Buyukaslan, H, Fedai, U & Altındag, A 2016, ‘Mean platelet volume and red cell distribution width levels in initial evaluation of panic disorder’, Neuropsychiatric Disease and Treatment, vol. 12, pp. 2435-2438.

Baker, LB 2017, ‘Sweating rate and sweat sodium concentration in athletes: a review of methodology and intra/interindividual variability’, Sports Medicine, vol. 47, no. 1, pp. 111-128.

Baker, LB, Barnes, KA, Anderson, ML, Passe, DH & Stofan, JR 2016, ‘Normative data for regional sweat sodium concentration and whole-body sweating rate in athletes’, Journal of Sports Sciences, vol. 34, no. 4, pp. 358-368.

Baker, LB, Barnes, KA, Sopeña, BC, Nuccio, RP, Reimel, AJ & Ungaro, CT 2018a, ‘Sweat sodium, potassium, and chloride concentrations analyzed same day as collection versus after 7 days storage in a range of temperatures’, International Journal of Sport Nutrition and Exercise Metabolism, vol. 28, no. 3, pp. 238-245.

Baker, LB, Barnes, KA, Ungaro, CT, Sopeña, BC, Nuccio, RP, Reimel, AJ, Stofan, JR & Carter, JM 2017, ‘Body mapping of sweating rate and sweat sodium concentration in athletes during moderate intensity exercise: relation between local and whole body’, The FASEB Journal, vol. 31, no. 1_supplement, pp. lb741-lb741.

Baker, LB, Ungaro, CT, Barnes, KA, Nuccio, RP, Reimel, AJ & Stofan, JR 2014, ‘Validity and reliability of a field technique for sweat Na+ and K+ analysis during exercise in a hot‐humid environment’, Physiological Reports, vol. 2, no. 5, pp. 1-11.

Baker, LB, Ungaro, CT, Sopeña, BC, Nuccio, RP, Reimel, AJ, Carter, JM, Stofan, JR & Barnes, KA 2018b, ‘Body map of regional vs. whole body sweating rate and sweat electrolyte concentrations in men and women during moderate exercise-heat stress’, Journal of Applied Physiology, vol. 124, no. 5, pp. 1304-1318.

Banner, A, Frumin, I & Shamay-Tsoory, SG 2018, ‘Androstadienone, a chemosignal found in human sweat, increases individualistic behavior and decreases cooperative responses in men’, Chemical Senses, vol. 43, no. 3, pp. 189-196.

Bariya, M, Nyein, HYY & Javey, A 2018, ‘Wearable sweat sensors’, Nature Electronics, vol. 1, no. 3, pp. 160-171.

Borowczyk-Michalowska, J, Zimolag, E, Konieczny, P, Chrapusta, A, Madeja, Z & Drukala, J 2017, ‘Stage-Specific Embryonic Antigen-4 (SSEA-4) as a distinguishing marker between eccrine and apocrine origin of ducts of sweat glands’, Journal of Investigative Dermatology, vol. 137, no. 11, pp. 2437-2440.

Buono, MJ, Stone, M & Cannon, DT 2016, ‘Leaching from the stratum corneum does not explain the previously reported elevated potassium ion concentration in sweat’, Journal of Basic and Clinical Physiology and Pharmacology, vol. 27, no. 2, pp. 171-173.

Cazalé, A, Sant, W, Ginot, F, Launay, JC, Savourey, G, Revol-Cavalier, F, Lagarde, JM, Heinry, D, Launay, J & Temple-Boyer, P 2016, ‘Physiological stress monitoring using sodium ion potentiometric microsensors for sweat analysis’, Sensors and Actuators B: Chemical, vol. 225, pp. 1-9.

Cheshire Jr, WP 2016, ‘Thermoregulatory disorders and illness related to heat and cold stress’, Autonomic Neuroscience, vol. 196, pp. 91-104.

Choi, DH, Kim, JS, Cutting, GR & Searson, PC 2016, ‘Wearable potentiometric chloride sweat sensor: the critical role of the salt bridge’, Analytical Chemistry, vol. 88, no. 24, pp. 12241-12247.

Choi, DH, Li, Y, Cutting, GR & Searson, PC 2017, ‘A wearable potentiometric sensor with integrated salt bridge for sweat chloride measurement’, Sensors and Actuators B: Chemical, vol. 250, pp. 673-678.

Collaco, JM, Blackman, SM, Raraigh, KS, Corvol, H, Rommens, JM, Pace, RG, Boelle, PY, McGready, J, Sosnay, PR, Strug, LJ & Knowles, MR 2016, ‘Sources of variation in sweat chloride measurements in cystic fibrosis’, American Journal of Respiratory and Critical Care Medicine, vol. 194, no. 11, pp. 1375-1382.

Collie, JT, Massie, RJ, Jones, OA, LeGrys, VA & Greaves, RF 2014, ‘Sixty‐five years since the New York heat wave: advances in sweat testing for cystic fibrosis’, Pediatric Pulmonology, vol. 49, no. 2, pp. 106-117.

Cui, CY & Schlessinger, D 2015, ‘Eccrine sweat gland development and sweat secretion’, Experimental Dermatology, vol. 24, no. 9, pp. 644-650.

Dalerba, P, Sahoo, D, Paik, S, Guo, X, Yothers, G, Song, N, Wilcox-Fogel, N, Forgó, E, Rajendran, PS, Miranda, SP & Hisamori, S 2016, ‘CDX2 as a prognostic biomarker in stage II and stage III colon cancer’, New England Journal of Medicine, vol. 374, no. 3, pp. 211-222.

Dam, VAT, Zevenbergen, MAG & Van Schaijk, R, 2016, ‘Toward wearable patch for sweat analysis’, Sensors and Actuators B: Chemical, vol. 236, pp. 834-838.

De la Iglesia, DH, González, GV, Barriuso, AL, Murciego, ÁL & Herrero, JR 2015, ‘Monitoring and analysis of vital signs of a patient through a multi-agent application system’, ADCAIJ: Advances in Distributed Computing and Artificial Intelligence Journal, vol. 4, no. 3, pp. 19-30.

Denkert, C, Von Minckwitz, G, Brase, JC, Sinn, BV, Gade, S, Kronenwett, R, Pfitzner, BM, Salat, C, Loi, S, Schmitt, WD & Schem, C 2015, ‘Tumor-infiltrating lymphocytes and response to neoadjuvant chemotherapy with or without carboplatin in human epidermal growth factor receptor 2-positive and triple-negative primary breast cancers’, Journal of Clinical Oncology, vol. 33, no. 9, pp. 983-991.

Dias, D & Cunha, PSJ 2018, ‘Wearable health devices—vital sign monitoring, systems and technologies’, Sensors, vol. 18, no. 8, pp. 1-28.

Doorn, J, Storteboom, TTR, Mulder, AM, de Jong, WHA, Rottier, BL & Kema, IP 2015, ‘Ion chromatography for the precise analysis of chloride and sodium in sweat for the diagnosis of cystic fibrosis’, Annals of Clinical Biochemistry, vol. 52, no. 4, pp. 421-427.

Dziedzic, CE, Ross, ML, Slater, GJ & Burke, LM 2014, ‘Variability of measurements of sweat sodium using the regional absorbent-patch method’, International Journal of Sports Physiology and Performance, vol. 9, no. 5, pp. 832-838.

Emaminejad, S, Gao, W, Wu, E, Davies, ZA, Nyein, HYY, Challa, S, Ryan, SP, Fahad, HM, Chen, K, Shahpar, Z & Talebi, S 2017, ‘Autonomous sweat extraction and analysis applied to cystic fibrosis and glucose monitoring using a fully integrated wearable platform’, Proceedings of the National Academy of Sciences, vol. 114, no. 18, pp. 4625-4630.

Emerson, SC, Waikar, SS, Fuentes, C, Bonventre, JV & Betensky, RA 2018, ‘Biomarker validation with an imperfect reference: issues and bounds’, Statistical Methods in Medical Research, vol. 27, no. 10, pp. 2933-2945.

Familoni, BO, Gregor, KL, Dodson, TS, Krzywicki, AT, Lowery Jr, BN, Orr, SP, Suvak, MK & Rasmusson, AM 2016, ‘Sweat pore reactivity as a surrogate measure of sympathetic nervous system activity in trauma‐exposed individuals with and without posttraumatic stress disorder’, Psychophysiology, vol. 53, no. 9, pp. 1417-1428.

Farrell, PM, White, TB, Ren, CL, Hempstead, SE, Accurso, F, Derichs, N, Howenstine, M, McColley, SA, Rock, M, Rosenfeld, M & Sermet-Gaudelus, I 2017, ‘Diagnosis of cystic fibrosis: consensus guidelines from the Cystic Fibrosis Foundation’, The Journal of Pediatrics, vol. 181, pp. S4-S15.

Gao, W, Emaminejad, S, Nyein, HYY, Challa, S, Chen, K, Peck, A, Fahad, HM, Ota, H, Shiraki, H, Kiriya, D & Lien, DH 2016, ‘Fully integrated wearable sensor arrays for multiplexed in situ perspiration analysis’, Nature, vol. 529, no. 7587, pp. 509-514.

Gentili, S, Mortali, C, Mastrobattista, L, Berretta, P & Zaami, S 2016, ‘Determination of different recreational drugs in sweat by headspace solid-phase microextraction gas chromatography mass spectrometry (HS-SPME GC/MS): application to drugged drivers’, Journal of Pharmaceutical and Biomedical Analysis, vol. 129, pp. 282-287.

Gibney, GT, Weiner, LM & Atkins, MB 2016, ‘Predictive biomarkers for checkpoint inhibitor-based immunotherapy’, The Lancet Oncology, vol. 17, no. 17, pp. e542-e551.

González‐Alonso, J & Kalsi, KK 2015, ‘The ubiquitous ATP molecule: could it be the elusive thermal mediator igniting skin perfusion and sweating in the heat‐stressed human?’, The Journal of Physiology, vol. 593, no. 11, pp. 2399-2399.

Goulet, ED, Asselin, A, Gosselin, J & Baker, LB 2017, ‘Measurement of sodium concentration in sweat samples: comparison of 5 analytical techniques’, Applied Physiology, Nutrition, and Metabolism, no. 42, no. 8, pp. 861-868.

Hara, S, Kawabata, T & Nakamura, H 2015, ‘Real-time sensing, transmission and analysis for vital signs of persons during exercises’, in Engineering in Medicine and Biology Society (EMBC), 2015 37th Annual International Conference of the IEEE, IEEE, Milano, pp. 4017-4020.

Hara, S, Okuhata, H, Kawabata, T, Nakamura, H & Yomo, H 2016, ‘Real-time vital monitoring for persons during exercises—solutions and challenges—’, IEICE Transactions on Communications, vol. 99, no. 3, pp. 556-564.

Hashmonai, M, Cameron, AE, Connery, CP, Perin, N & Licht, PB 2017, ‘The etiology of primary hyperhidrosis: a systematic review’, Clinical Autonomic Research, vol. 27, no. 6, pp. 379-383.

Heikenfeld, J 2016, ‘Non‐invasive analyte access and sensing through eccrine sweat: challenges and outlook circa 2016’, Electroanalysis, vol. 28, no. 6, pp. 1242-1249.

Hong, YJ, Lee, H, Kim, J, Lee, M, Choi, HJ, Hyeon, T & Kim, DH 2018, ‘Multifunctional wearable system that integrates sweat‐based sensing and vital‐sign monitoring to estimate pre‐/post‐exercise glucose levels’, Advanced Functional Materials, vol. 28, no. 47, p. 1805754.

Hooton, K & Li, L 2017, ‘Nonocclusive sweat collection combined with chemical isotope labeling LC–MS for human sweat metabolomics and mapping the sweat metabolomes at different skin locations’, Analytical Chemistry, vol. 89, no. 15, pp. 7847-7851.

Hu, Y, Converse, C, Lyons, MC & Hsu, WH 2018, ‘Neural control of sweat secretion: a review’, British Journal of Dermatology, vol. 178, no. 6, pp. 1246-1256.

Hurley, SW & Johnson, AK 2015, ‘The biopsychology of salt hunger and sodium deficiency’, Pflügers Archiv-European Journal of Physiology, vol. 467, no. 3, pp. 445-456.

Imani, S, Bandodkar, AJ, Mohan, AV, Kumar, R, Yu, S, Wang, J & Mercier, PP 2016, ‘A wearable chemical–electrophysiological hybrid biosensing system for real-time health and fitness monitoring’, Nature Communications, vol. 7, pp. 1-7.

Jadoon, S, Karim, S, Akram, MR, Kalsoom, KA, Zia, MA, Siddiqi, AR & Murtaza, G 2015, ‘Recent developments in sweat analysis and its applications, ‘International Journal of Analytical Chemistry, vol. 2015, no. 164974, pp. 1-7.

Jobling, MA 2015, ‘On the nose: genetic and evolutionary aspects of smell’, Investigative Genetics, vol. 6, no. 1, p. 2.

Katchman, BA, Zhu, M, Christen, JB & Anderson, KS 2018, ‘Eccrine sweat as a biofluid for profiling immune biomarkers’, Proteomics–Clinical Applications, vol. 12, no. 6, pp. 1-8.

Kauffman, P 2018, ‘Primary hyperhidrosis’, in MP Loureiro, De Campos JRM, Wolosker N, & Kauffman P (eds), Hyperhidrosis: a complete guide to diagnosis and management, Springer, Cham, pp. 27-32.

Khan, Y, Ostfeld, AE, Lochner, CM, Pierre, A & Arias, AC 2016, ‘Monitoring of vital signs with flexible and wearable medical devices’, Advanced Materials, vol. 28, no. 22, pp. 4373-4395.

Kim, BJ, Kwak, MK, Kim, JS, Lee, SH & Koh, JM 2018, ‘Higher sympathetic activity as a risk factor for skeletal deterioration in pheochromocytoma’, Bone, vol. 116, pp. 1-7.

Kim, DW, Tiseo, M, Ahn, MJ, Reckamp, KL, Hansen, KH, Kim, SW, Huber, RM, West, HL, Groen, HJ, Hochmair, MJ & Leighl, NB 2017, ‘Brigatinib in patients with crizotinib-refractory anaplastic lymphoma kinase-positive non-small-cell lung cancer: a randomized, multicenter phase II trial’, Journal of Clinical Oncology, vol.35, no. 22, pp. 2490-2498.

Koh, A, Kang, D, Xue, Y, Lee, S, Pielak, RM, Kim, J, Hwang, T, Min, S, Banks, A, Bastien, P & Manco, MC 2016, ‘A soft, wearable microfluidic device for the capture, storage, and colorimetric sensing of sweat’, Science Translational Medicine, vol. 8, no. 366, pp. 366ra165-366ra165.

Lee, H, Choi, TK, Lee, YB, Cho, HR, Ghaffari, R, Wang, L, Choi, HJ, Chung, TD, Lu, N, Hyeon, T & Choi, SH 2016, ‘A graphene-based electrochemical device with thermoresponsive microneedles for diabetes monitoring and therapy’, Nature Nanotechnology, vol. 11, no. 6, pp. 566-572.

Lee, H, Song, C, Hong, YS, Kim, MS, Cho, HR, Kang, T, Shin, K, Choi, SH, Hyeon, T & Kim, DH 2017, ‘Wearable/disposable sweat-based glucose monitoring device with multistage transdermal drug delivery module’, Science Advances, vol. 3, no. 3, pp. 1-8.

Liu, G, Alomari, M, Sahin, B, Snelgrove, SE, Edwards, J, Mellinger, A & Kaya, T 2015, ‘Real-time sweat analysis via alternating current conductivity of artificial and human sweat’, Applied Physics Letters, vol. 106, no. 13, pp. 1-6.

Liu, G, Ho, C, Slappey, N, Zhou, Z, Snelgrove, SE, Brown, M, Grabinski, A, Guo, X, Chen, Y, Miller, K & Edwards, J 2016, ‘A wearable conductivity sensor for wireless real-time sweat monitoring’, Sensors and Actuators B: Chemical, vol. 227, pp. 35-42.

Liu, J, Chen, Y, Wang, Y, Chen, X, Cheng, J & Yang, J 2018, ‘Monitoring vital signs and postures during sleep using Wi-Fi signals’, IEEE Internet of Things Journal, vol. 5, no. 3, pp. 2071-2084.

Lübke, KT & Pause, BM 2015, ‘Always follow your nose: the functional significance of social chemosignals in human reproduction and survival’, Hormones and Behavior, vol. 68, pp. 134-144.

Luque de Castro, MD 2016, ‘Sweat as a clinical sample: what is done and what should be done’, Bioanalysis, vol. 8, no. 2, pp. 85-88.

Mattar, ACV, Leone, C, Rodrigues, JC & Adde, FV 2014, ‘Sweat conductivity: an accurate diagnostic test for cystic fibrosis?’, Journal of Cystic Fibrosis, vol. 13, no. 5, pp. 528-533.

Mogenson, GJ 2018, The neurobiology of Behavior: an introduction, Routledge, Abingdon.

Motofei, IG & Rowland, DL 2016, ‘Structural dichotomy of the mind; the role of sexual neuromodulators’, Journal of Mind and Medical Sciences, vol. 3, no. 2, pp. 131-140.

Mutic, S, Moellers, EM, Wiesmann, M & Freiherr, J 2016, ‘Chemosensory communication of gender information: masculinity bias in body odor perception and femininity bias introduced by chemosignals during social perception’, Frontiers in Psychology, vol. 6, pp. 1-11.

Nakata, S, Arie, T, Akita, S & Takei, K 2017, ‘Wearable, flexible, and multifunctional healthcare device with an ISFET chemical sensor for simultaneous sweat pH and skin temperature monitoring’, ACS Sensors, vol. 2, no. 3, pp. 443-448.

Nyein, HYY, Gao, W, Shahpar, Z, Emaminejad, S, Challa, S, Chen, K, Fahad, HM, Tai, LC, Ota, H, Davis, RW & Javey, A 2016, ‘A wearable electrochemical platform for noninvasive simultaneous monitoring of Ca2+ and pH’, ACS Nano, vol. 10, no. 7, pp. 7216-7224.

Oberg, K, Modlin, IM, De Herder, W, Pavel, M, Klimstra, D, Frilling, A, Metz, DC, Heaney, A, Kwekkeboom, D, Strosberg, J & Meyer, T 2015, ‘Consensus on biomarkers for neuroendocrine tumour disease’, The Lancet Oncology, vol. 16, no. 9, pp. e435-e446.

Oh, SY, Hong, SY, Jeong, YR, Yun, J, Park, H, Jin, SW, Lee, G, Oh, JH, Lee, H, Lee, SS & Ha, JS 2018, ‘Skin-attachable, stretchable electrochemical sweat sensor for glucose and pH detection’, ACS Applied Materials & Interfaces, vol. 10, no. 16, pp. 13729-13740.

Oliveira, MA, Lima, WG, Schettini, DA, Tilelli, CQ & Chaves, VE 2018, ‘Is calcitonin gene-related peptide a modulator of menopausal vasomotor symptoms?’, Endocrine, pp. 1-11, Web.

Owen, K 2016, ‘Excessive sweating: are patients suffering unnecessarily?’, The Journal for Nurse Practitioners, vol. 12, no. 1, pp. 35-40.

Reis, AH 2017, ‘Acidemia and blood free fatty acids: analysis of cardiovascular risk factors in a new context’, Discovery Medicine, vol. 23, no. 126, pp. 183-188.

Robb, MA, McInnes, PM & Califf, RM 2016, ‘Biomarkers and surrogate endpoints: developing common terminology and definitions’, JAMA, vol. 315, no. 11, pp. 1107-1108.

Rose, DP, Ratterman, ME, Griffin, DK, Hou, L, Kelley-Loughnane, N, Naik, RR, Hagen, JA, Papautsky, I & Heikenfeld, JC 2015, ‘Adhesive RFID sensor patch for monitoring of sweat electrolytes’, IEEE Transactions on Biomedical Engineering, vol. 62, no. 6, pp. 1457-1465.

Sekine, Y, Kim, SB, Zhang, Y, Bandodkar, AJ, Xu, S, Choi, J, Irie, M, Ray, TR, Kohli, P, Kozai, N & Sugita, T 2018, ‘A fluorometric skin-interfaced microfluidic device and smartphone imaging module for in situ quantitative analysis of sweat chemistry’, Lab on a Chip, vol. 18, no. 15, pp. 2178-2186.

Shirasaka, Y, Chaudhry, AS, McDonald, M, Prasad, B, Wong, T, Calamia, JC, Fohner, A, Thornton, TA, Isoherranen, N, Unadkat, JD & Rettie, AE 2016, ‘Interindividual variability of CYP2C19-catalyzed drug metabolism due to differences in gene diplotypes and cytochrome P450 oxidoreductase content’, The Pharmacogenomics Journal, vol. 16, no. 4, pp. 375-387.

Sonner, Z, Wilder, E, Heikenfeld, J, Kasting, G, Beyette, F, Swaile, D, Sherman, F, Joyce, J, Hagen, J, Kelley-Loughnane, N & Naik, R 2015, ‘The microfluidics of the eccrine sweat gland, including biomarker partitioning, transport, and biosensing implications’, Biomicrofluidics, vol. 9, no. 3, pp. 1-19.

Turner, MJ & Avolio, AP 2016, ‘Does replacing sodium excreted in sweat attenuate the health benefits of physical activity?’, International Journal of Sport Nutrition and Exercise Metabolism, vol. 26, no. 4, pp. 377-389.

Vary Jr, JC 2016, ‘Selected disorders of skin appendages-acne, alopecia, hyperhidrosis’, The Medical Clinics of North America, vol. 99, no. 6, pp. 1195-1211.

Vinik, AI, Camacho, PM, Davidson, JA, Handelsman, Y, Lando, HM, Leddy, AL, Reddy, SK, Cook, R, Spallone, V, Tesfaye, S & Ziegler, D 2017, ‘American Association of Clinical Endocrinologists and American College of Endocrinology position statement on testing for autonomic and somatic nerve dysfunction’, Endocrine Practice, vol. 23, no. 12, pp. 1472-1478.

Vos, T, Barber, RM, Bell, B, Bertozzi-Villa, A, Biryukov, S, Bolliger, I, Charlson, F, Davis, A, Degenhardt, L, Dicker, D & Duan, L 2015, ‘Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013’, The Lancet, vol. 386, no. 9995, pp. 743-800.

Wang, Z, Dong, S, Gui, M, Asif, M, Wang, W, Wang, F & Liu, H 2018, ‘Graphene paper supported MoS2 nanocrystals monolayer with Cu submicron-buds: high-performance flexible platform for sensing in sweat’, Analytical Biochemistry, vol. 543, pp. 82-89.

Yu, Y, Prassas, I, Muytjens, CM & Diamandis, EP 2017, ‘Proteomic and peptidomic analysis of human sweat with emphasis on proteolysis’, Journal of Proteomics, vol. 155, pp. 40-48.