Chronic liver disease features parenchymal damage and repair, gradually proceeds to fibrosis and finally to cirrhosis, accompanied by increasing hepatic impairment together with a markedly heightened risk for HCC, a pathological process in which cirrhosis and HCC constitute the life-threatening end state (Liu et al., 2013). Due to the widely different etiologies, including chronic infections with hepatotropic viruses, intoxications, especially as a result of alcohol abuse, autoimmune diseases, and metabolic disorder in the context of overweight or obesity, prevention as well as treatment differ depending on the proximate cause (Blachier et al., 2013; Giannitrapani et al., 2013). However, irrespective of the etiology, chronic hepatic inflammation is the response and further stimulates the progress of the disease (Sahasrabuddhe et al., 2012; Seidler et al., 2012; Giannitrapani et al., 2013). Consequently, the intake of NSAIDs for example, indicated a risk reduction of both deaths from CLD and the development of HCC, results that can be ascribed to the ability of NSAIDs to diminish chronic inflammation by blocking prostaglandin synthesis (Sahasrabuddhe et al., 2012) and - specifically regarding aspirin - by down-regulating pro-inflammatory cytokines (Imaeda et al., 2009).

Since controlling the inflammatory state within the liver directly impacts the progression of CLD, the question arises whether this could also be achieved naturally, supplementing standard approaches by inherent properties that are well-known to unleash potent anti-inflammatory effects on their own. Physical activity is principally suitable - beside yielding overall health-promoting benefits - to specifically generate a quantifiable anti-inflammatory activity (e.g., Weinhold et al., 2015) and, in contrast, the degree of physical inactivity commonly occurring on a global scale substantially contributes to chronic diseases and premature death (Booth et al., 2012; Lee et al., 2012). Moreover, the discernible necessity of the body to be active might at least in part be tied to the role that physical activity has played in the survival of our ancestors. Capabilities like running might have served as a tool to successfully search for food, thereby representing a selective advantage and it was hypothesized that endurance running might have been utilized as a hunting method by hominids (Carrier et al., 1984) or to maximize success in scavenging, in either case being of major importance for the evolution of the body shape (Bramble and Lieberman, 2004) and even present-day hunter-gatherers were observed to run after game until it dies from exhaustion (Liebenberg, 2006). Ultimately, both our environment and lifestyle have increasingly changed within a short period of time, whereas our genetic constitution virtually remained the same, creating a conflict that might find its expression in chronic diseases (Eaton et al., 1988), but might likewise be mitigated by reintegrating activities stabilized by evolution, including physical activity, as a permanent feature into our lives.

According to the WHO, physical inactivity is responsible for 6% of deaths worldwide, thus appearing in the fourth position of risk factors for mortality, and consequently, the WHO recommends a specified amount of physical activity to counteract the associated global burden of disease (World Health Organization [WHO], 2010). Additionally, physical activity or more specifically exercise has been included into practice guidelines for the prevention and treatment of NAFLD, a medical condition characterized by a fat mass greater than 5% related to the total liver and the most frequent manifestation of CLD in industrialized nations around the world (Rector and Thyfault, 2011; Neuman et al., 2015; Watanabe et al., 2015; Zhu et al., 2015; Oliveira et al., 2016). These recommendations implicate aerobic exercise of moderate intensity for at least 20 to 60 min on 5 days per week in combination with resistance training performed thrice weekly and for further benefits increasing practice beyond 250 min per week (Oliveira et al., 2016).

The goal of this review is to provide evidence for the therapeutic applicability of exercise in advanced states of CLD, including fibrosis, cirrhosis, and HCC, and to assess whether exercise might be of value as adjuvant therapy in the treatment of CLD.

It is obvious that regular endurance exercise promotes favorable structure and metabolism adaptations in contracting organs. However, regular sport, physical exercise, and resistance training have also many health-promoting effects on tissues and organs (Figure 1; Table 1). Due to the immense technological progress within the last decades more and more people display a predominantly sedentary way of life without the imperative to move adequately. Additionally, dietary habits have changed, processed foods are the standard, and often consuming convenience foods instead of wholesome food further compounds the situation. As a consequence, both physical inactivity and the absence of a healthy and varied diet contribute to a rise in overweight and obesity in the industrialized countries. Of particular importance is the buildup of visceral fat, which attracts pro-inflammatory immune cells, resulting in elevated levels of pro-inflammatory adipokines - factors originating from adipose tissue and unfolding either a pro-inflammatory or an anti-inflammatory action - and the emergence of a low-grade systemic inflammatory condition. This chronic low-grade inflammation is marked by two to threefold elevated systemic concentrations of TNF-α, IL-1, IL-6, IL-1ra, sTNFR, and CRP (Pedersen and Febbraio, 2008). Moreover, it is made responsible for the development of CNCDs, such as atherosclerosis, insulin resistance, and type 2 diabetes (Ouchi et al., 2011), neurodegenerative alterations like depression and dementia (Leonard, 2007), and some cancers (Mathur and Pedersen, 2008). Therefore, it is crucial to reverse this fatal trend by bringing down obesity and low-grade inflammation in order to prevent or ameliorate the associated morbidities. Regular exercise offers a potent instrument to tackle chronic inflammation (Mathur and Pedersen, 2008; Kawanishi et al., 2013b) and its use should not be limited to the prevention of CNCDs, but it can also be useful as a treatment option (Karsten et al., 2016), which is why exercise is meanwhile regarded as medicine (Gleeson et al., 2011). Most of the health-promoting effects of regular exercise are mediated by its anti-inflammatory properties, which have been thoroughly investigated and are now well-founded. In this chapter, we will outline the most relevant anti-inflammatory mechanisms of exercise on the molecular level.

In general, patients suffering from CLD are limited in performing extensive physical activities. This is due to the fact that the gradual loss in functionality impacts myriad functions of the liver that physiologically acts as the body’s chemical factory. As such, it is one of the major glycogen storage sites that supply the body with readily available glucose. Liver disease is associated with alterations in processing of glycogen synthesis or breakdown resulting in reduced capacity to provide quick energy during phases of exercise. Chronic ethanol consumption, for example results in a dramatic decrease in liver glycogen concentration that is due to direct effects of ethanol on glycogen metabolism that is exerted at several levels, including posttranslational modulation of enzyme activities (Van Horn et al., 2001). Moreover, the healthy liver is a central storage site for iron and blood, while liver disease is associated with reduced blood oxygenation, intrahepatic angiogenesis and other hemodynamic changes leading to portal hypertension, again preventing extensive exercise. Particularly, this is evident in cirrhotic, high-risk candidates for liver transplantation, in which the oxygen consumption at peak exercise is severely impaired (Dharancy et al., 2008). Similarly, the disease-associated accumulation of toxic products, such as physiological metabolites or xenobiotics (e.g., medication), that results from limited degradation, excretion and detoxification within the diseased liver is a hindrance preventing extensive movements. Moreover, the loss of the hepatic gland function (e.g., production of bile acids) and the decreased anabolic activity in the synthesis of proteins, fats, carbohydrates, and hormones impact the ability to perform intense physical activity. Therefore, in the past, patients with liver disease were often incorrectly advised to restrict their physical activity. The view that CLD is incompatible with physical exercise has dramatically changed during the last years.

Numerous studies in rodents and in humans focused on the influence of exercise on NAFLD, unsurprisingly, since this clinical syndrome occurs with increasing frequency and meanwhile its worldwide prevalence is estimated to be in the range of 4 to 46%, variations mainly depending on the diagnostic procedure being used (Sherif et al., 2016), and afflicting more than one out of three US citizens (Noureddin et al., 2015). The prevalence of nonalcoholic steatohepatitis (NASH), the severe subtype of NAFLD, which is marked by liver fat accumulation combined with inflammation for multiple years, is considered varying between 3 to 5% around the world (Sherif et al., 2016). The observed rise in both obesity and the metabolic syndrome, which in turn are closely linked to unwholesome dietary habits, together with an absence of exercise are conducive to NAFLD (Toplak et al., 2016), which has been identified as the leading CLD worldwide (Paradies et al., 2014).

Several experimental studies and clinical observations have shown that regular exercise beneficially impacts the risk of onset and progression of CLD (Berzigotti et al., 2016). Interestingly, under conditions that are associated with an increase of intrahepatic lipid content such as NASH, vigorous activity that allows reducing the overall hepatic lipid content was shown to be more favorable than moderate activity. This was exemplarily demonstrated in a high-fat diet-induced obese mouse model, in which vigorous-intensity and interval treadmill running was more effective in alleviating hepatic steatosis than moderate-intensity and continuous treadmill running (Cho et al., 2015). These results were also confirmed in patients with NAFLD determined by liver biopsy: While vigorous-intensity exercise inversely correlated with NAFLD severity, there was no such effect visible regarding either moderate-intensity training or the total amount of exercise (Kistler et al., 2011). In comparison with NAFLD severity, intensity appears to be of minor importance in terms of NAFLD prevalence. For instance, in more than 5,700 study participants increased amounts of physical activity indicated a decrease in the prevalence of NAFLD, revealing a dose-response relationship between physical activity and NAFLD (Oni et al., 2015). In addition, even a reduction in the time spent sitting together with a rise in physical activity was noticed to result in a lower prevalence of NAFLD in a large number of examined Koreans (Ryu et al., 2015) and also findings of a systematic review pointed at a beneficial impact of a decreased amount of sitting on NAFLD (Whitsett and VanWagner, 2015). Thus, the previous outcomes imply that the therapeutical effect might rather depend on the intensity of exercise, whereas the actual amount of activity might more likely be sufficient for the prevention of NAFLD.

However, also performing ordinary resistance training exclusively composed of pushups and squats three times a week for 12 weeks was accompanied by a gain in muscle mass, a concomitant drop in fat mass, and improved hepatic steatosis as well as traits of the metabolic syndrome in patients suffering from NAFLD (Takahashi et al., 2015).

Also, clinical studies suggested that already short-term exercise is sufficient to reduce hepatocyte apoptosis in obese individuals with NAFLD by improving insulin sensitivity and increasing the oxidative capacity (Fealy et al., 2012). This notion is also confirmed in many other studies showing that active lifestyles and caloric restriction decrease insulin resistance, body weight, and further result in decreased histological signs of liver injury (Gonçalves et al., 2013). Mechanistically, vigorous exercise prevents liver disease through mitochondrial adaptations that affect the content of cytochrome c and the activity of enzymes that are directly linked to fatty acid oxidations (Rector et al., 2011; Gonçalves et al., 2013). Recently, it was shown that the modifying role on lipogenic genes is mediated via activation of the AMP-activated protein kinase (AMPK) pathway that is central in the regulation of cellular energy metabolism by modulating the activity of the sterol regulatory element-binding protein family that has been established as a group of transcription factors regulating genes involved in cholesterol and fatty acid synthesis (Cintra et al., 2012).

Interestingly, the exercise-induced antioxidant response in patients who were classified as heavy drinkers was similar to that observed in patients who did not drink heavily, suggesting that most of the observed beneficial effects do not depend on the health status of the liver (Georgakouli et al., 2015). A protective role against hepatic oxidative stress induced by doxorubicin, an antibiotic drug frequently applied to cancer treatment, was reported by researchers who first subjected rats to daily treadmill running for a 6-week period five times per week and subsequently administered the drug: The aerobic training induced a significant rise in the levels of the antioxidant enzymes superoxide dismutase and glutathione peroxidase combined with a simultaneous significant drop in nitric oxide and malondialdehyde, both markers for oxidative stress (Zolfagharzadeh and Roshan, 2013). By extension, it is also conceivable that the hepatoprotective effect of exercise via antioxidant mechanisms might similarly contribute to an improved tolerance of drugs by the liver and thus reasonably support drug treatment.

Early studies regarding the effect of lifestyle modification on NAFLD found that a combination of caloric restriction and exercise training led to a reduction in body weight in the range of 5-10% and a significant improvement in or even a normalization of serum ALT levels (Palmer and Schaffner, 1990; Hickman et al., 2002). However, a major problem of these combined approaches was that it could not be clearly determined whether those beneficial effects had to be ascribed predominantly to weight loss or were directly instigated by exercise training. Not until later it has been shown that regular exercise and weight loss independently of one another were able to normalize ALT levels in patients with NAFLD (Suzuki et al., 2005). Subsequently, also other investigations confirmed marked improvements of aminotransferase levels caused by moderate or even low intensity exercise despite no weight loss (Sreenivasa Baba et al., 2006; St. George et al., 2009). Although more recent findings also substantiated these improvements of ALT levels independent of the form of exercise (aerobic versus resistance) training and the length of the study (Thompson et al., 2010; Fealy et al., 2012; Khaoshbaten et al., 2013; Shamsoddini et al., 2015), others did not (Johnson et al., 2009; Bacchi et al., 2013).

Similar to the frequently observed decrease in serum levels of ALT, the most diagnostically conclusive biomarker for liver injury, exercise performed on a regular basis has been reported to lower intrahepatic lipid content (Johnson et al., 2012; Loomba and Cortez-Pinto, 2015). A reduction of liver fat in obese individuals solely as a result of regularly performed aerobic exercise without any change in body weight was first documented by Johnson et al. (2009) and the beneficial effect of aerobic exercise training on hepatic fat in the absence of weight loss was then confirmed in several other studies (Finucane et al., 2010; van der Heijden et al., 2010; Sullivan et al., 2012). In addition, also resistance training has been shown to reduce liver fat (e.g., Zelber-Sagi et al., 2014) and there is evidence to suggest that both forms of exercise training are equally effective in decreasing intrahepatic fat independent of weight loss in people afflicted with NAFLD (Bacchi et al., 2013; Shamsoddini et al., 2015). Even though ALT levels approximately correlate with the extent of liver inflammation and heightened ALT levels largely match with the severity of NAFLD, there is only a relatively weak correlation between ALT levels and the intensity of hepatic fibrosis (Kim et al., 2008). Due to this, serum ALT levels are not sufficient as an indicator of liver fibrosis on their own, but can be combined with the serum levels of AST, another liver enzyme and also a surrogate marker of hepatic injury, to compute the AST/ALT ratio, which then can be applied to predict advanced fibrosis (Wilkins et al., 2013). Apart from the AST/ALT ratio also the AST/platelet count ratio, the FibroTest (FibroSure) or magnetic resonance elastography are in wide use as noninvasive biomarkers for advanced fibrosis, but liver biopsy constitutes the norm to accurately determine the grade of inflammation as well as the stage of fibrosis or to differentiate between simple hepatic steatosis and NASH (Wilkins et al., 2013). Tissue samples can serve to identify histological hepatic fibrosis for example with the help of staining: Kawanishi et al. (2013c) employed Sirius red and Masson trichrome to display collagen fibers and immunohistochemical staining of α-SMA to specifically flag activated hepatic stellate cells in murine liver tissue, since activation of hepatic stellate cells and excessive deposition of collagen are key events during hepatic fibrogenesis (Schon et al., 2016). Thus, Kawanishi et al. (2013c) succeeded in demonstrating that hepatic fibrosis was alleviated in the exercise group compared with the control group.

Furthermore, some studies suggest that exercise can limit hepatic inflammation and prevent the progression to states of advanced liver damage, such as fibrosis and cirrhosis. In a mutant form of mice, generating a hyperphagic phenotype and fed with atherogenic food, exercise subdued inflammation in adipose tissue and in the liver as well as the development of NASH and of hepatic fibrosis (Haczeyni et al., 2015). Similar results were obtained in another murine study, in which mice were fed with a high-fat diet supplemented with high-fructose water and performed regular exercise training on a treadmill: Likewise hepatic injury, inflammation, and fibrosis decreased in the animals that had undergone training compared with those of the control group (Kawanishi et al., 2012). And reviewing the impact of exercise in humans suffering from NAFLD, Oliveira et al. (2016) stated that exercise is capable of alleviating liver inflammation and oxidative stress by reducing levels of pro-inflammatory cytokines, what may also help to prevent fibrosis, because - beside lipid peroxidation and involvement of Fas ligands - pro-inflammatory cytokines contribute to cumulative liver injury.

Also, exercise has been shown to exert a beneficial influence on accompanying symptoms of NAFLD. For instance, endothelial dysfunction of elastic arteries - a condition that is frequently encountered in patients with NAFLD - indicates a risk for cardiovascular disease and constitutes an early sign of atherosclerosis (Pugh et al., 2014). In marked contrast to measures of conventional care, exercise training of moderate intensity reversed endothelial dysfunction, thereby also lowering the risk for cardiovascular disease characteristic of NAFLD, which is why the authors proposed to prescribe exercise as an essential part of treatment within this risk group (Pugh et al., 2014). The positive effects on endothelial performance did occur without a significant decline in either liver fat or visceral adipose tissue and were thought to be generated by recurrent surges of shear stress acting on endothelial cells (Pugh et al., 2014).

Already almost three decades ago, Ritland (1988) emphasized that individuals suffering from CLD due to viral infection can benefit from regular exercise, because it strengthens oxygen consumption as well as working capacity without any signs of impairing liver function. Even in those in an acute phase of infection, exercise can contribute to maintain physical working capacity, demonstrating that individuals with both acute and CLDs can profit by physical activity (Ritland, 1988). And the notion that exercise in general might be contraindicated during an existing hepatic infection seems to be unfounded as illustrated by the following rather extreme example: From 26 active participants in a 100 km ultra-marathon blood samples were taken one week beforehand, directly after the race, and once again 24 h upon completion and were subsequently analyzed regarding muscle damage, liver function, and oxidative stress parameters. There was a marked increase in all of these markers in the 8 hepatitis B virus carriers as well as in the 18 non-infected runners after the race, but no significant difference was detectable between the two groups themselves (Chiu et al., 2013). Moreover, there was also no indication of virus reactivation in carriers in the 24-h period following the race, altogether leading to the conclusion that neither the risk for inflammation nor for deficits in liver function is greater in hepatitis B virus-infected runners than in their non-infected co-runners upon participation in such a long distance event (Chiu et al., 2013).

Less research has focused on potential benefits derived from exercise in the terminal stage of CLD. This is particularly true for liver cirrhosis, which is known to be associated with a decline in exercise capacity as well as muscle strength and an accelerated start of anaerobic metabolism during practice (Jones et al., 2012). Furthermore, in individuals suffering from portal hypertension, there exist increased risks for hemorrhaging due to heightened portal pressure in response to exercise, a fact which must be balanced by workouts specifically adapted for the needs of these patients (Jones et al., 2012). Up to now there are two studies investigating the effect of exercise on individuals with cirrhosis and both documented exercise-induced enhancements in muscle strength as well as in exercise capacity.

In a study from 1983 five participants with liver biopsy-confirmed cirrhosis were examined at the start of a 12-week practice and then again twice until the end of the program. The maximal oxygen consumption or VO_2max - the amount of oxygen, which can maximally be taken up by the body despite an increasing physical effort (Hawkins et al., 2007), and constituting a measure of physical fitness and endurance capacity - initially rose by 19% in the first part and by 29% toward the end of the period without evidence of any complications (Ritland et al., 1983). And in two of four participants in the second study there was an increase in the VO_2peak, which characterizes the oxygen uptake maximally attained and not necessarily identical with the VO_2max (Day et al., 2003), by 21.2 and 27.5%, respectively, and a gain in muscle mass by 20 and 18%, whereas there was no change in the peak oxygen uptake of the other two patients (Campillo et al., 1990). These results admittedly hint at a positive impact of exercise even in patients with cirrhosis, but it is also obvious that further in-depth studies are needed to safely judge the effects in these high-risk patients.

Some recent studies in rodents and in humans also document a beneficial effect of exercise on the development of HCC and on tumor growth. For instance, to determine the influence of regular exercise on the incidence of liver cancer, genetically modified mice were employed that spontaneously develop steatohepatitis as well as HCC. These animals were involved in an exercise program consisting of running on a treadmill during a 32-week period for 1 h a day for 5 days per week. As a result of the exercise program both the quantity and the extent of the tumors within the examined livers were diminished, verifying that exercise is able to restrict tumor cell proliferation (Piguet et al., 2015).

Similar results were obtained in an experimental study with rats, that were split into four groups, one receiving only a low-fat diet, the second a low-fat diet together with swim training, the third a high-fat diet, and the last a high-fat diet combined with swim training (Aguiar e Silva et al., 2012). The swim training, which lasted for eight weeks with five training units per week, led to a reduction in body fat and weight in the two differently fed groups and to ameliorated cholesterol levels in the high-fat diet group, additionally indicating that regular exercise of that kind coupled with a reasonable, i.e., a reduced-fat diet, might be able to abate the development of liver cancer (Aguiar e Silva et al., 2012).

In another clinical study, including 97 patients with HCC, low levels of adiponectin, an adipokine with predominantly anti-inflammatory and anti-diabetic properties, were significantly related to worse histological grades of HCC (Sumie et al., 2011) and generally adiponectin is thought to unfold antitumor effects as well and to assist in offering protection against HCC (Wieser et al., 2012). In addition, adiponectin has been shown to inhibit the progression of HCC in vitro by promoting apoptosis via thioredoxin, an antioxidant protein (Xing et al., 2015).

Since adiponectin levels can be markedly increased after regular exercise (Markofski et al., 2014), properly dosed exercise should be beneficial for HCC patients. An example for the application of exercise in respective patients is presented by Kaibori et al. (2013), who investigated the impact of exercise on patients with hepatectomy due to HCC. The exercise program started one month prior to surgery, followed by a 1-week pause immediately after the surgical operation and was then resumed for another six months. At the end of this period fat mass as well as whole body mass were significantly reduced associated with a significant rise in peak oxygen consumption and anaerobic threshold (Kaibori et al., 2013).

Finally, according to a recent study a decline in both incidence and tumor growth by more than 60% was demonstrated in response to regular exercise in different mouse models, also including one model for DEN-induced liver cancer (Pedersen et al., 2016). Just 31% of those mice that had received a DEN injection and additionally exercised developed liver tumors as opposed to 75% of the animals in the control group and running also led to a significant decrease in tumor burden (Pedersen et al., 2016). These effects were mediated by NK cells - a crucial cell type of the innate immune system that is able to identify cancer cells without the help of antibodies. After mobilization of NK cells by adrenaline IL-6, which is secreted by muscle cells during exercise, is involved in the further distribution of the immune cells to tumors (Pedersen et al., 2016).

Interestingly, a further study reported the activation of NK cells induced by one bout of exercise. At least for 24 h after execution of a half marathon, both histone acetylation and the expression of a functional marker of NK cells remained elevated, suggesting that exercise might be helpful to reduce cancer risk as well as recurrence rates (Zimmer et al., 2015).

Based on the previous examples, exercise might even be of worth for advanced stages of CLD. However, from the clinical perspective, the beneficial effects of exercise have not been translated into positive outcomes and there is an urgent need to implement strategies to promote a behavior change, including the implementation of daily resistance training-based exercise regimens to reduce liver fat, and preventing and treatment of NAFLD (Johnson et al., 2012; Loomba and Cortez-Pinto, 2015). Personalized exercise programs dependent on the severity of hepatic disease might also be beneficial to enhance blood oxygenation, improve energy metabolism, reduce disease-associated atrophies, or to increase the patient’s mood by stimulating the release of endorphins and modulating neurotransmitter synthesis or activity.

Despite constant efforts to develop effective drugs, up to now there exist only a few agents, which are applicable to the medication of CLD. Therefore, the conventional approach focuses on prevention and early diagnosis, whereas at an advanced stage the only remaining possibility often consists in liver transplantation. Independent of the etiology, there is always an underlying chronic inflammation in the heart of CLD and dampening the inflammatory state can slow down the progression of the disease, thereby preserving and ameliorating liver function.

Exercise has repeatedly been shown to unleash potent anti-inflammatory effects and, conversely, a lack of physical activity is believed to be responsible for a wide spectrum of chronic diseases. Increased amounts of physical activity and exercise meanwhile constitute a classic in the prevention and treatment of diseases of the cardiovascular system, such as coronary heart disease or heart failure (Lavie et al., 2015). Therefore, it is plausible reasoning to make use of these benefits in the context of CLD, too, while keeping the patients’ safety in mind. There is no doubt that exercise can even be beneficial in an advanced stage of liver disease and - under the guidance of qualified healthcare professionals - can be an appropriate adjuvant therapy. In response to exercise, various endogenous agents are produced, which unfold a quantifiable pharmacological activity, thereby contributing to recovery. Since exercise has been shown to be effective in both prevention and therapy of certain chronic diseases, consequently the notion of exercise as medicine was coined (Gleeson et al., 2011). But it is also necessary to determine the optimal amount and intensity of exercise and to identify the most suitable workout to maximize its value. In order to obtain this knowledge, further studies are essential. As long as these questions are unanswered, it will yet be better to be physically active for at least 5 to 10 min every day than to stay inactive, since such small amounts are already connected with a clear drop in the risks of all-cause and cardiovascular mortality (Lee et al., 2014).

H-TS and RW have written this review and agree to be accountable for the content of the work.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.