Research advances in metabolism 2016

1. Introduction

Metabolic-related research is continuously evolving, since obesity, type 2 diabetes mellitus (T2DM) and metabolic syndrome (MetS) are the new epidemics of modern societies. Here, we summarize a selection of exceptional experimental and clinical studies published in Metabolism in 2016. We begin with very promising basic research studies and continue with important non-interventional as well as interventional clinical research studies. The aim of this work is to effectively summarize important advances in Metabolism as well as to present directions for future research.

2. Basic Research Studies

2.1. Dietary Changes in Fat Content and Composition Do Not Affect Obesity in Neuron-Specific Lipoprotein Lipase Deficient Mice [1]

The brain consists of a high amount of lipids, which are either synthesized de novo within the brain (saturated and monoun-saturated fatty acids), or are supplied to the brain via the blood (polyunsaturated fatty acids, PUFAs) [2]. Wang et al. have demonstrated in their first report that the enzyme lipoprotein lipase (LPL), which cleaves fatty acids from triglyceride-rich glycoproteins, is important for energy homeostasis, as it facilitates the entry of the cleaved lipids in the brain. In that report, mice with neuron-specific LPL-deficiency (NEXLPL −/− mice) became obese on a chow diet by 16 weeks of age due to reduced uptake of triglyceride-rich lipoprotein-derived fatty acids and lower levels of n-3 long chain polyunsaturated fatty acids (n-3 PUFAs) in the hypothalamus [3]. Now, in their follow-up study published in Metabolism, Wang et al. have investigated whether diet can affect the obesity phenotype of the LPL-deficient mice. For this aim, NEXLPL −/− and wild type (WT) mice have been randomly assigned to one of the three following diets: a) a high carbohydrate diet (HC), b) a high-fat diet (HF), c) HC diet supplemented with n-3 PUFAs (HCn-3). NEXLPL−/− mice on an HC diet, developed obesity, consistent with the previous findings using chow diet. On the contrary, the HF diet failed to increase more weight gain in NEXLPL−/− mice compared to wild type (WT) mice. Similarly, despite the deficiency of n-3 PUFAS in NEXLPL−/− mice, n-3 PUFA supplementation through the HCn-3 diet did not alter the obesity phenotype of the NEXLPL−/− mice; although the orexigenic hormone agouti-related peptide (AgRP) was reduced and the hypothalamic n-3 PUFA content was increased. The findings of this study show that regulation of energy homeostasis from lipids in the brain involves sophisticated still unidentified mechanisms. LPL seems to have an important role in body weight control, however further studies should aim to clarify the mechanism related to the LPL-dependent regulation of energy homeostasis in the central nervous system.

2.2. Eating at Inappropriate Times can Induce Obesity and Metabolic Disorders within a Short Period in Mice [4]

The circadian timing system, located in the suprachiasmatic nucleus of the anterior hypothalamus, affects important body functions, including sleep/wake cycles, body temperature, metabolism and hormone secretion [5]. Epidemiological studies suggest that the disruption of circadian rhythms is associated with obesity and MetS in humans [6]. Previous studies in mice have demonstrated that high-fat diet exposure during the active phase leads to significantly less weight gain compared to mice fed ad libitum, despite identical caloric intake. On the contrary, feeding during the inactive phase leads to weight gain within a few weeks [7]. In this study, Yasumoto et al. aimed to investigate the metabolic parameters and the underlying mechanism involved in these effects. For this purpose, mice were fed either in the inactive phase (day phase, DF mice) or in the active phase (night phase, NF mice) with a high-fat high carbohydrate (sucrose) diet, which is more close to the diet followed by obese humans [4]. Moreover, the mice had free access to run on wheels, which is a voluntary behavior in rodents and affects their metabolic outcome. After one week, DF mice had already gained more weight and fat, and the daily amount of wheel-running was reduced compared with NF mice. Additionally, DF mice demonstrated higher food consumption, which was accompanied by an increase in the expression levels of orexigenic genes (Agouti related peptide, AgRP and Neuropeptide Y, Npy) as well as in the circulating levels of leptin, suggesting leptin resistance. Furthermore, the mice developed hyperinsulinemia, hypercorticosteronemia and increased hepatic lipid accumulation. These results demonstrate that eating at inappropriate times can induce obesity and metabolic disorders within a short period in mice. Thus, conditions that demand changes in circadian rhythms in humans, such as jet lag and shift work, may therefore be associated with increased risk for obesity and diabetes.

2.3. FGF23 is Involved in Muscle Growth and Exercise Endurance in Mice [8]

Fibroblast growth-factor 23 (FGF23) is a hormone belonging to the FGF family, which consists of at least 23 members, with many of them involved in major metabolic and endocrine functions [9]. FGF23 is considered an important regulator of phosphate homeostasis and bone biology [10]. Moreover, high levels of FGF23 have been associated with cardiovascular and renal disease [11,12]. Despite the expression of FGF23 in skeletal muscle, its role in the function of muscle tissue remained unknown. Li et al. aimed to investigate the effect of exercise on circulating FGF23 concentrations as well as on FGF23 expression in skeletal muscle in mice. Moreover, they evaluated the effects of FGF23 treatment on exercise endurance and muscle function. They found that all types of exercise (acute, exhaustive, chronic) increased circulating FGF23 levels in mice. However, FGF23 mRNA and protein expression in skeletal muscle was upregulated only in chronic exercise. FGF23 protein was located in the cytoplasm of muscle cells and the localization was not altered by exercise. Administration of recombinant FGF23 for 3 days promoted exercise endurance by enhancing mitochondrial function, while it reduced reactive oxygen species and H₂O₂ production in skeletal muscle. These results reveal a novel and important role of FGF23 in muscle biology and exercise performance.

2.4. CNTF Improves Metabolism and Insulin Sensitivity in Muscle in Mice [13]

The Ciliary Neurotrophic Factor (CNTF) is a hormone that improves insulin sensitivity and stimulates weight loss by increasing energy expenditure and by reducing appetite through actions in the hypothalamus [14]. Tsompanidis et al. investigated whether CNTF acts also on muscle in order to mediate its important metabolic effects [13]. For this aim, diet-induced obese mice were administrated CNTF for 7 days. Whole transcriptome analysis of dissected soleus muscles revealed that CNTF is involved in pathways improving metabolism, energy expenditure and insulin sensitivity in muscle. Specifically, CNTF downregulated leptin and upregulated follistatin and Pak1 (insulin sensitizer in muscle) expression. Moreover, CNTF affected the expression of genes involved in all stages of the myogenic process, such as cell cycle regulation (H19, Bex1), myoblast differentiation (Mamstr, Mlp, Ankrd1, Ankrd2), structural adaptations (S100a8, Tubb6, Mylk2, Mylk4, Rhoc, Emb, Col1a1, Col3a1) and muscle innervation (Chrna1). These changes in gene expression indicate increased muscle differentiation and decreased muscle atrophy. Same genes and pathways (e.g. Akt pathway) are activated by follistatin, so that it is difficult to distinguish between direct effects of CNTF on muscle and indirect through the CNTF-mediated upregulation of follistatin. Nevertheless, these changes by CNTF were independent of the already established anorexigenic role of the hormone and point towards improved metabolism by stimulation of muscle growth.

2.5. Noninvasive Peripheral Electrical Stimulation Regulates Glucose in Rats [15]

Peripheral electrical stimulation (PES) is a therapeutic alternative that has demonstrated some promising glucose regulatory effects in rodents. Several studies have reported that 30–90 min of electro-acupuncture (EA) in anesthetized rodents improves glucose uptake and tolerance [16–18]. However, such a long duration of treatment can be poorly translated to humans, making EA a rather impracticable therapeutic option. Catalogna et al. have therefore investigated if PES can affect glucose and energy metabolism after a very short-duration of treatment in conscious, obese and insulin resistant rats [15]. Their results demonstrate that rats treated with PES for three minutes three times a week had significantly lower energy consumption, weight gain and visceral adiposity compared to control group. Most importantly, the PES-treated mice demonstrated lower glucose levels after intraperitoneal glucose tolerance test due to lower insulin resistance. Hyperinsulinemic euglycemic clamp after PES demonstrated a significant improvement of insulin sensitivity with an accompanied decrease of hepatic glucose output and increase in glycolysis and glycogen synthesis in both muscle and liver. Although further studies are necessary to define the mechanism behind these effects, this study provides proof of concept for a possible use of noninvasive PES treatment for glycemic control, justifying the evaluation of PES in humans.

2.6. Perilipin 1 Binding to Aquaporin 7 Affects Glycerol Release in Adipocytes [19]

Triacylglycerol (TAG) is the lipid which is primarily stored in a single large lipid droplet in adipocytes. Perilipin 1 (PLIN1) is a protein present on the surface of the lipid droplet that activates lipolysis during fasting via its phosphorylation by protein kinase A (PKA). The free fatty acids (FFAs) and glycerol which derive from lipolysis, are released from the cell in order to be used from other tissues for energy production. The efflux of glycerol is performed in adipocytes by aquaglyceroporin AQP7. In human adipose tissue, AQP7 translocates from the lipid droplet to the plasma membrane after catecholamine stimulation, while on the contrary AQP7 remains around the lipid droplet after insulin treatment. Hansen, Krintel et al. investigated the exact mechanism controlling the AQP7 trafficking in human adipocytes. They managed to demonstrate that PLIN1 is in physical contact with AQP7 through the cytosolic termini of AQP7. The proximity between the two molecules is increased under lipogenic conditions and reduced under lipolysis. PKA-dependent phosphorylation of the N-Terminus of AQP7 reduces PLIN1 binding. Altogether, these findings describe the mechanisms involved in glycerol release by adipocytes, revealing possible targets for future drugs against metabolic abnormalities.

2.7. Atorvastatin Prevents Cardiac Fibrosis by Blocking the AGE-RAGE System in Rats [20]

Cardiac fibrosis is a condition frequently observed in diabetic cardiomyopathy, which is characterized by impaired cardiac elasticity and contractile dysfunction due to increased myocardial fibroblast proliferation and differentiation [21]. Advanced glycation end products (AGEs) accumulate in the cardiovascular tissue, bind to their receptor (RAGE) and induce fibroblast proliferation [21]. Peroxisome proliferator-activated receptor gamma (PPAR-γ) is widely expressed in the cardiovascular system and is an important inhibitor of RAGE [22]. Atorvastastin is a statin and besides inhibiting cholesterol synthesis, it can activate PPAR-γ. Given the relation between atorvastatin and PPAR-γ, as well as PPAR-γ and AGE-RAGE axis, Chen et al. investigated in vitro and in vivo, if atorvastatin can affect cardiac fibrosis by regulating cardiac effects of AGEs. Administration of AGEs in rats induced fibroblast proliferation and differentiation by activating the AGEs-RAGE-ERK1/2 pathway. Treatment of rats with atorvastatin blocked this pathway through activation of PPAR-γ and consequently reduced fibroblast proliferation and cardiac fibrosis. These results expand the current knowledge regarding the cardioprotective effects of atorvastatin, suggesting beneficial effects of atorvastatin in diabetic patients with diabetic cardiomyopathy.

2.8. Effects of Testosterone Deprivation on Cognitive Impairment is Independent to Obesity in Rats [23]

Testosterone deprivation is associated with cognitive decline [24]. Similarly, obesity has been associated with reduced cognitive status [25]. Pintana et al. investigated whether obesity combined with testosterone deprivation could affect cognitive function, brain insulin sensitivity, brain mitochondrial function and hippocampal synaptic plasticity. For this purpose, male Wistar rats were bilaterally orchiectomized (ORX mice) or sham-operated (control) and fed with either normal or high fat diet for 12 weeks. ORX and control-mice following normal diet did not develop brain insulin resistance, while both ORX and control mice fed high-fat diet developed initially peripheral (week 8) and later brain insulin resistance (week 12) with mitochondrial dysfunction. Interestingly, both lean and obese-ORX mice demonstrated cognitive impairment and decreased hippocampal synaptic plasticity at the 8th week, in contrast to obese control mice which demonstrate the same changes much later at the 12th week. Altogether, these data demonstrate that testosterone deprivation acts neither additively nor synergistically with obesity to induce cognitive dysfunction.

2.9. Characteristics of GLP-1 Producing L-cells in Ileal Mucosa in Type 2 Diabetic Populations [26]

Glucagon-like peptide 1 (GLP-1) is an incretin secreted by L-cells of the gut mucosa that plays a major role in enhancing post prandial insulin responses, termed as “incretin effect” [27]. This incretin effect seems to be blunted in patients with T2DM [28]. Furthermore, GLP-1 concentration is reduced in the blood of patients with T2DM, while high glucose and glucagon decrease postprandial GLP-1 secretion [29,30]. Based on these findings, Kampmann et al. investigated whether patients with diabetes have less GLP-1 secreting cells (L-cells) compared to healthy individuals [26]. For this aim, they collected tissue samples from the ileum of 10 patients with and without diabetes that were age- and gender-matched. Authors found no difference in the number of L-cells in enteral mucosa between the two groups. Crypts had a significantly higher number of L-cells as compared to the villi in both groups. Proliferation was reduced in L-cells in ileal villi but apoptosis did not differ between the 2 groups. Kampmann et al. demonstrated that patients with diabetes have equal number of L-cells, so the reduction of GLP-1 levels observed in diabetes is probably the result of a defect in GLP-1 secretion. Being retrospective, this study does have its own limitations including uncertainty in exact anatomic location and size of the tissue samples. However, this finding is crucial in encouraging the ongoing research efforts of increasing GLP-1 secretion in patients with diabetes.

2.10. GLP-1 Receptor Agonists Demonstrate Antiatherogenic Effects in Normal-Weight but not in Obese subjects [31]

Certain GLP-1 receptor (GLP-1R) agonists have been associated with positive cardiovascular outcome in human studies, while dipeptidyl peptidase-4 (DPP-4) inhibitors have not demonstrated such positive effects. It is, therefore, still debated whether and how incretins are involved in atherosclerosis-atherogenesis and consequently in cardiovascular outcome. Previous studies have shown that GLP-1R agonists and/or DPP-4 inhibitors shift macrophages’ phenotype and cytokine production from pro-inflammatory (M1 phenotype) to anti-inflammatory (M2 phenotype) state, reducing in this way the risk for atherogenesis in obesity [32,33]. Another process involved in atherogenesis is autophagy. Autophagy is characterized by breakdown and recycling of intracellular components [34]. Stimulation of autophagy reduces the formation of foam cells (macrophages which accumulate lipids) preventing in this way the progression of atherosclerosis in obesity and diabetes [35]. Tanaka et al. investigated how GLP-1R agonist treatment can affect autophagy, foam cell formation and macrophage phenotype in control and in obese patients [31]. Their results demonstrate that, in normal-weight people, treatment with GLP-1R agonists has significant antiatherogenic effects by reducing foam cell formation and M1 proinflammatory cytokine expression and by inducing indicators of autophagy in monocytes. On the other hand, in obese individuals, GLP-1R agonist treatment induces foam cell formation and M1 cytokine expression, while it does not affect autophagy. These differential effects between obese and normal-weight were observed despite the similar expression levels of GLP-1R on the monocytes of both populations. In conclusion, the results from this important study indicate that GLP-1 signaling is altered in monocytes of obese patients and may therefore be involved in atherosclerosis.

2.11. DPP-4 Regulates Inflammatory Reactions in Macrophages via Toll-Like Receptors [36]

DPP-4, a well-known anti-diabetic target is also a T-cell co-stimulator that binds to adenosine deaminase and interacts with CD3/T cell receptor complex [37,38]. Stimulation of T cells is known to increase DPP4 expression and release of soluble DPP4 (sDPP4) which is found in much higher concentrations in patients with diabetes [39,40]. sDPP4 has been shown to demonstrate multiple cellular effects including T cell activation, induction of insulin resistance, inflammation in smooth muscle cells, reactive oxygen species (ROS) generation and CD86 up-regulation in antigen presenting cells [41–44]. Lee et al. investigated the effects of sDPP-4 and DPP-4 inhibition with vildagliptin on inflammatory reactions in a murine macrophage cell line (RAW264.7), as well as the in vivo effects of intravenously administered sDPP4 in mice [36]. sDPP4 increased the expression of iNOS, NO and the production of proinflammatory cytokines in lipopolysaccharide (LPS)-stimulated RAW264.7. These effects were probably mediated through an upregulation of toll-like receptor 4 and 2 (TLR4, TLR2) in these cells. Both receptors have an important role in recognition of pathogens and activation of innate immunity. The inflammatory effects were suppressed after treatment with vildagliptin. Injection of sDPP4 in mice increased the expression of TLR4 and TLR2 in kidney and white adipose tissue. These results indicate that sDPP4 exerts inflammatory effects through the TLR pathway, which can be suppressed by DPP4-inhibitors.

2.12. Novel DPP-4 Inhibitor Teneligliptin Scavenges –OH Radicals in Rats [45]

DPP-4 inhibitors including alogliptin, vildagliptin, and linagliptin have been shown to decrease oxidative stress by reducing reactive oxygen species (ROS) independent of glucose lowering [46–48]. Kimura et al. investigated the antioxidant properties of teneligliptin; especially whether its chemical structure demonstrates radical scavenging properties. With the use of electron spin resonance spectroscopy, they demonstrated that teneligliptin does not scavenge ${\mathrm{O}}_2^{-}$ but OH. Among ROS, OH demonstrates strong oxidizing properties and is therefore considered one of the most cytotoxic oxygen species. Teneligliptin maintained its antioxidant characteristics, depicted by the increased levels of 8-hydroxy-2′-deoxyguanosine in urine, kidney and aorta, in DPP-4 deficient rats, which show that its antioxidant effects are independent of the DPP-4 inhibition. In summary, this study shows that teneligliptin is a potent OH scavenger and it may therefore be useful for the prevention of diabetic complications.

2.13. Empagliflozin Improves Insulin Sensitivity Either as Monotherapy or in Combination with Linagliptin in db/db Mice [49]

DPP-4 inhibitors and sodium-glucose linked transporter 2 (SGLT-2) inhibitors are widely used potent drugs in management of diabetes. Empagliflozin (SGLT-2 inhibitor) has been shown to reduce fasting blood glucose, hemoglobin A1c (HbA1c) and improve insulin sensitivity [50,51]. Similarly, linagliptin (DPP-4 inhibitor, cleared by liver unlike other DPP-4 cleared by kidneys) also improves insulin sensitivity, blood glucose and HbA1c, as well as reduces triglycerides and improves hepatic steatosis [52–54]. However, combination effect of both these drugs has not been studied. Kern et al. investigated the effect of 8 week monotherapy with linagliptin (3 mg/kg/day) and empagliflozin (10 mg/kg/day) and combination (3 mg/kg/day of linagliptin and 10 mg/kg/day of empagliflozin) of both these drugs in female db/db mice. Euglycemic hyperinsulinemic clamp studies were performed to study the effect on glucose dynamics, hepatic triglyceride was measured from liver tissue of the mice and mRNA expression was measured in adipose tissue and liver. After 8 weeks of treatment, combination therapy had an additive effect on glucose metabolism with significant reduction in fasting plasma glucose levels, in serum triglycerides and HbA1c levels as compared to monotherapy. Fasting plasma insulin, total cholesterol and free fatty acids did not show any significant difference between the 3 groups. Liver lipid content decreased with empagliflozin and was further reduced with the combination treatment (p < 0.001). Glucose disposal rate and suppression of hepatic glucose production were also much higher (p < 0.001) in combination group as compared to monotherapy. Glucose tolerance was significantly improved and fasting blood glucose levels were much lower with the combination therapy and empagliflozin as compared to vehicle treated groups. Linagliptin also improved glucose tolerance but was less effective than empagliflozin. Muscle tissue specific glucose uptake was seen in the combination therapy group although significant glucose uptake was seen in liver and kidney as well in all the treatment groups. Expression of several key genes (FAS, PTP1B, SOCS-2) that are involved in inflammation was decreased significantly in all the treatment groups. Altogether, Kern et al. demonstrated beneficial effects of individual drugs and additive effects of combination therapy on glucose metabolism. Significant improvements in glucose tolerance, insulin sensitivity and lipid metabolism observed especially with combination therapy support its use in clinical practice.

3. Non-Interventional Clinical Research Studies

3.1. Baseline Glycemic Status and Mortality in 241,499 Korean Metropolitan Subjects [55]

Diabetes has been linked to increased mortality from cardiovascular disease and cancer. To investigate the relationships between diabetes and/or prediabetes with mortality risk, Rhee et al. [55] examined 241,499 participants from a health-screening program with normoglycemia, prediabetes, and diabetes (uncontrolled defined as HbA1c ≥ 7%). Of those participants, 877 died. Patients with diabetes showed an all-cause mortality hazard ratio of 1.58 if controlled or 2.26 if uncontrolled and a cancer mortality hazard ratio of 1.75 if controlled and 1.67 if uncontrolled, compared with normoglycemic patients. They were not more likely to die from cardiovascular disease. All-cause mortality and cancer-related mortality were higher with diabetic as compared to normoglycemic patients. Future studies should confirm these findings in longer-term studies as well as investigate mechanisms and potential treatments.

3.2. Synergistic Association of Elevated Serum Free Fatty acid and Glucose Levels with Large Arterial Stiffness in a General Population [56]

In this cross-sectional study of 9396 Japanese subjects, Tabara et al. [56] showed an association between elevated serum free fatty acids levels and large arterial stiffness, measured through brachial-to-ankle pulse wave velocity (baPWV) in a large-scale epidemiological setting. Individuals were subdivided according to clinical characteristics and fasting status. Fasting condition influenced intensely FFA and insulin levels variation and the pattern of variation in FFA was opposing to that of insulin. The authors found that a significant association between FFA levels and baPWV was observed only in subjects with fasting blood samples (≥12 h, P < 0.001) or near-fasting (5–11 h, P < 0.001) conditions, but not in non-fasting (<5 h, P = 0.307) condition. Therefore, whereas both HbA1c and T2DM were factors that increased baPWV, the positive relationship between FFA and baPWV was independent of plasma levels of insulin and glucose; in fact, the association between FFA level and baPWV remained significant (β = 0.052, P < 0.001) when adjusted for glucose levels. Moreover, in addition to their direct relationship, FFA and glucose levels were synergistically related with baPWV (FFA*glucose; β = 0.036, P < 0.001). These results support previous findings that suggest that FFAs cause a direct impairment of endothelium function through an insulin-independent pathophysiological pathway [57] and might explain the mechanism by which glucose levels synergize with elevated FFA levels. The authors conclude that physiological variation in FFA levels might raise the risk of large arterial stiffness in a general population and that FFAs and hyperglycemia might have a synergistic adverse effect on the vessels.

3.3. Association of Circulating Adiponectin and Carotid Intima–Media Thickness [58]

Adiponectin (APN) is the most abundant hormone secreted by adipose tissue. APN demonstrates insulin sensitizing, anti-inflammatory, and anti-atherogenic characteristics. Low levels of APN have been associated with obesity, T2DM and cardiovascular disease [59–61]. Carotid intima–media thickness (cIMT) is a robust independent predictor of macroangiopathy, especially of coronary events and stroke. Several studies have investigated the relation of cIMT and the circulating adiponectin levels, reporting however contradictory results [62,63]. Gasbarrino et al. performed a meta-analysis of the studies which investigated the relationship between APN and cIMT. Fifty-five articles consisting of only cross-sectional studies were analyzed. Authors found a weak inverse association between circulating APN levels and cIMT measurements in healthy subjects, in subjects with metabolic diseases and in other chronic diseases. On the contrary, there was no association between APN and cIMT in patients with inflammatory diseases. The conflicting results among the different studies may be due to their small sample size, their cross-sectional character and the lack of adjustments for confounding factors. Prospective human studies as well as animal studies are therefore needed in order to reach definitive conclusions about the association between circulating APN and cIMT.

3.4. Follistatin displays a Day–Night Rhythm and is Associated with Muscle Mass and Circulating Leptin Levels in a Healthy Population [64]

Follistatin is a glycoprotein that has been identified as an inhibitor of TGF-β family and is known to have a major role in the reproductive system [65]. Recent studies have demonstrated production of follistatin in the liver, especially in response to exercise and feeding status and thus this molecule has raised interest in the endocrinology world [66,67]. Several different groups have now reported elevated levels of follistatin in T2DM [68], nonalcoholic fatty liver disease [69] and PCOS [70]. Anastasilakis et al. investigated the physiology of follistatin circadian and seasonal variation and its response to meal ingestion. The authors studied 122 subjects (50% males) who had similar baseline characteristics. Other adipokines like leptin, irisin and adiponectin were also measured to see correlations. No statistically significant difference in follistatin levels was seen between males and females but after adjustment for leptin, males were found to have higher follistatin levels. This observation remained significant after addition of other factors like BMI, total body fat mass and waist circumference. Follistatin levels were found to be lower in follicular phase as compared to luteal phase in females. Follistatin levels did not vary with different types of diet and different levels of activity. Leptin was found to be a major determinant of follistatin levels and higher leptin levels were associated with higher follistatin levels independent of age, sex, waist circumference, insulin resistance, adiponectin, creatinine, GGT and CK levels. Furthermore, authors conducted a mixed meal study on 2 groups of 18 individuals who received 125 ml and 250 ml of a standardized meal respectively. No difference in follistatin level was seen in the group that received 125 ml of standardized meal but significantly high levels of follistatin were seen in the group that received 250 ml of standardized meal. Circadian rhythm was studied in 20 participants (50% males) and circulating follistatin was found to have two nadir points at 3 pm and midnight and two peak points at 9 pm and 3 am. Seasonal study was done on the same individuals and no difference in circulating follistatin was noted throughout the year. Summarizing, this study made some important observations regarding the physiology of follistatin: a) Circulating follistatin positively correlates with muscle mass and leptin with higher concentrations found in males than females. b) Positive correlation in follicular phase of menstrual cycle and circadian variation was also observed which provides crucial information in the physiology of the molecule.

3.5. Correlations Between Circulating metabolites and Parameters of Abdominal Obesity [71]

Obesity often leads to insulin resistance, dyslipidemia and chronic low-grade inflammation. All these conditions increase the risk for cardiovascular diseases and T2DM. Although BMI and waist circumference (WC) as well as serum lipids, glucose and insulin have been considered reliable markers of cardiovascular and T2DM risk, it is still debatable: i) in which extent WC and serum metabolome are defined by common genetic or environmental factors, ii) which adiposity and insulin resistance parameters correlate stronger with the serum metabolic profile, iii) whether these correlations are independent of genetic and familial background. Bogl et al. aimed to answer all these questions by analyzing circulating serum metabolites as well as obesity and insulin resistance measures in large cohorts of monozygotic (n = 531) and dizygotic (n = 837) twins. Their results demonstrate that abdominal fat (i.e. WC, android fat% and android/gynoid fat), HOMA-IR and CRP are associated with an atherogenic lipoprotein profile, with increased levels of branched-chain (BCAA) and aromatic amino acids, as well as with increased levels of glycoprotein. Additionally, genes associated with variations in abdominal obesity, especially in WC, overlap with genes affecting variation in serum metabolites. On the contrary, genetically identical twins (monozygotic twins) that differ in obesity measures, also differ in the serum lipoprotein profile. This indicates that obesity should be considered not only a genetically predetermined condition, but also an environmental factor leading to lipid disturbances. Consequently, health efforts aiming to prevent obesity and metabolic diseases should be intensified.

3.6. ANP System is a Predictor of Fat Mass Reduction and Insulin Sensitivity During Weight Loss in Humans [72]

Mature atrial natriuretic peptide (ANP) and the stable inactive fragment proANP derive from the precursor proANP, which is secreted by the atrial myocardium. Several studies have indicated an involvement of ANP system in metabolism [73]. ANP binds to natriuretic receptor A (NPR-A) to stimulate lipolysis in adipose tissue and mitochondrial biogenesis in muscle. In contrast, natriuretic receptor C (NPR-C) is primarily involved in ANP degradation acting as a clearing receptor [74]. Brachs et al. investigated with a clinical interventional study (n = 143, 31 male and 112 female), whether the ANP system was associated with the variability of fat mass reduction and metabolic improvement before and after a 3 month-standardized weight loss program. Their data indicated that the weight loss-mediated reduction of fat mass is associated with local NPR balance, including downregulation of NPR-C and upregulation of NPR-A in adipose tissue. This change in NPR-A/NPR-C ratio may even be more important in the regulation of fat mass than circulating ANP levels itself. On the contrary, no changes in the expression of NPRs were observed in muscle tissue, suggesting an adipose tissue specific effect during weight loss. Finally, Brachs et al. demonstrated that changes in proANP levels were independently associated with improvement of insulin sensitivity. Altogether, these data demonstrate that the ANP system is involved in both fat mass reduction and insulin sensitivity during weight loss.

3.7. Serum Uric Acid is a Strong Predictor of MetS [75]

MetS is a growing problem which includes central obesity, dyslipidemia, elevated blood pressure, and insulin resistance. There is some evidence that MetS may be related to serum uric acid levels. To determine if serum uric acid may predict the development of MetS, Yu et al. [75] examined 14,462 adults without MetS at baseline and who had 4 or more follow-up visits between 2006 and 2012 for the development of MetS. Of the 14,442 participants, 4215 developed MetS. Controlling for potential covariates, they found that higher levels of serum uric acid corresponded to higher hazard ratios for the development of MetS (1.25 in men and 1.32 in women for the fourth quartile of uric acid levels). For each 1 mg/dL increase of serum uric acid level, there was a hazard ratio of 1.09 for men and 1.15 for women towards developing MetS. These results demonstrate the use of serum uric acid levels as a biomarker for the development of MetS in adults. Future studies should investigate causal relationships between uric acid and MetS.

3.8. Circulating Level of Hepatocyte Growth Factor Predicts Incidence of T2DM [76]

Hepatocyte growth factor (HGF) is a paracrine factor involved in cellular growth, motility and morphogenesis. In vitro and animal studies indicate that HGF is involved in glucose homeostasis by positively affecting glucose uptake and utilization. On the contrary, epidemiologic studies demonstrate an association of HGF levels with obesity, insulin resistance, type 1 diabetes mellitus (T1DM) and T2DM, as well as MetS. However, prospective studies focusing on these associations do not exist. Bancks et al. investigated therefore the relation between serum HGF levels and incident T2DM in a prospective cohort study consisting of 6814 U.S. adult men and women without clinical cardiovascular disease at the time of the initial examination. The participants were from four racial ethnic groups (non-Hispanic white, African American, Hispanic, Chinese American) and were followed for over 12 years. The results of the study show that higher levels of HGF are positively associated with incidence of diabetes and this association is independent of race/ethnicity. The increase in HGF corresponds to higher hazard ratio for diabetes in men compared to women. However, a two-year increase in levels of HGF only slightly affects the incidence of diabetes. This study is the first to show prospectively that HGF is associated with incident diabetes and impaired metabolism.

3.9. Muscle Grip Strength is a Predictor of T2DM [77]

T2DM has been associated to reduced skeletal muscle mass [78] and strength [79], as well as to lower testosterone — a hormone that demonstrates in humans important anabolic muscle effects [49,80]. Li et al. investigated the relation of skeletal muscle mass and strength with incident T2DM in a prospective cohort of men, the Men Androgen Inflammation Lifestyle Environment and Stress (MAILES) Study. Moreover, they investigated whether the levels of testosterone or inflammatory markers are involved in any observed associations. Hand grip strength (dynamometer), body composition (dual-energy X-ray absorptiometry), testosterone, and inflammatory markers were measured. Of the 1632 men, incidence of T2DM occurred in 146 (8.9%). Muscle mass did not correlate with T2DM, but muscle grip strength was inversely related to incident T2DM [unadjusted odds ratio (OR) per 5 kg: 0.87, 95% confidence interval (CI): 0.80–0.95; adjusted OR, 95% CI: 0.87, 0.78–0.97]. Neither testosterone, nor the inflammatory markers IL-6 and TNF-α were associated with the observed relation between muscle grip strength and incident T2DM. Given that up to 29% of incidence of T2DM can be contributed to low grip strength, resistance training may be very important for preventing T2DM.

3.10. Non-alcoholic Fatty Liver Disease without MetS does not Increase the Risk for Cardiovascular Disease or T2DM [81]

Non-alcoholic fatty liver disease (NAFLD) and MetS have been associated with higher risk of cardiovascular diseases and T2DM. However, many patients with MetS also have NAFLD, so it is unclear whether NAFLD is an independent, additive or synergistic factor for cardiovascular diseases and/or T2DM. In order to address this point, Käräjämäki AJ et al. studied 958 middle-aged subjects that were classified into four subgroups: subjects with NAFLD and MetS, with NAFLD or MetS, and healthy controls [81]. Subjects with NAFLD and MetS demonstrated the highest risk for cardiovascular events (p < 0.001). Additionally, subjects with MetS, independently of the presence of NAFLD, showed an increased risk for T2DM (p < 0.001). On the contrary, subjects with only NAFLD, without MetS had similar cardiovascular and metabolic risk compared to healthy controls. These findings demonstrate that NAFLD increases the risk for cardiovascular diseases and T2DM, only with the additional presence of MetS.

3.11. Fatty Acid Metabolism is Altered in Non-Alcoholic Steatohepatitis Independent of Obesity [82]

Non-alcoholic fatty liver disease is an increasing obesity-related problem which can lead from simple steatosis to non-alcoholic steatohepatitis (NASH) and ultimately liver failure. The development of NASH from steatosis is particularly important to understand with the ultimate goal of treatment. Walle et al. [82] sought to understand whether fatty acid metabolism may be altered in NASH compared to normal liver or steatosis using 92 adults from the Kuopio Obesity Surgery Study. They found altered enzyme activity for fatty acid metabolism markers, i.e. increased delta-6 desaturase and stearoyl-CoA desaturase 1, and decreased delta-5 desaturase activity, in addition to altered mRNA expression of related genes in individuals with NASH as compared to those with normal liver or steatosis. These changes were further related to levels of alanine aminotransferase (ALT) indicating overall liver health. Future studies should investigate whether molecules that could alter the function of these enzymes may help to treat or prevent the development of NASH.

3.12. Activin A as a Potential Marker for NAFLD [83]

NAFLD is a growing problem that occurs as a result of fat accumulation in the liver and can lead to cirrhosis. The best diagnostic tool for NAFLD currently is a liver biopsy, which is invasive. Polyzos et al. [83] investigated if hormones activin A and follistatin can serve as potential biomarkers for NAFLD. For this purpose, they compared the levels of these hormones in patients with NAFLD confirmed by biopsy (15 with simple steatosis and 16 with NASH), with lean (n = 24) and obese (n = 28) controls in a case–control design. Activin A showed a trend towards an increase across lean controls, obese controls, simple steatosis (approximately 13% higher than obese controls) and NASH (approximately 43% higher than obese controls), which became significant after controlling for fat mass. Follistatin was not different between the groups, but was associated with NASH, perhaps due to the higher amount of fat mass in these patients. Altogether, activin A is a potential marker for NAFLD, which should be explored in larger studies.

3.13. Plasma Osteoprotegerin Levels are Inversely Associated with NAFLD in Patients with T2DM [84]

Osteoprotegerin is a soluble receptor in the tumor necrosis factor family which seems to be related to glucose control. To investigate if osteoprotegerin may be related to NAFLD in T2DM, Niu et al. [84] examined 367 patients with ultrasound-diagnosed NAFLD and 379 controls with T2DM but without NAFLD for this case–control study in China. Osteoprotegerin was approximately 18% lower in patients with NAFLD versus diabetic controls and were correlated with age and systolic blood pressure. Osteoprotegerin could be a potential marker related to NAFLD and should be confirmed in other ethnicities and in patients without T2DM.

3.14. Hepatic Insulin Resistance is a Distinct Metabolic Feature of Impaired Fasting Glucose in Humans [85]

Hyperglycemia in T2DM is the result of an increase in basal endogenous glucose production (EGP), combined with a defect in insulin-mediated glucose uptake as well as with unsuccessful suppression of EGP by insulin [86]. To determine the contribution of these processes to impaired fasting glucose (IFG) levels, Ter Horst et al. enrolled 131 obese non-diabetic adults with normal fasting glucose levels or IFG and studied basal EGP and hepatic, adipose tissue and peripheral insulin sensitivity by two-step euglycemic and hyperinsulinemic clamp studies with [6,6-(2)H2] glucose infusion. The authors demonstrate that IFG, in the context of human obesity, is characterized by distinct metabolic disorders. When compared with obese adults with normal fasting glucose (NFG), equally obese subjects with IFG had significantly and markedly reduced insulin-mediated suppression of EGP (p < 0.001), whereas there was no difference in basal EGP, insulin-mediated suppression of circulating FFA and insulin-stimulated glucose disposal [85]. These data suggest that hepatic insulin resistance is a distinct metabolic feature of IFG in obesity and that impaired insulin action in the liver can play an important role in the progression from compensated insulin resistance to prediabetes.

3.15. Associations of Insulin Resistance, Inflammation and Liver Synthetic Function with Very Low-Density Lipoprotein [87]

The authors of this cross-sectional study of over 1850 older subjects from the Cardiovascular Health Study (CHS), aimed to evaluate the associations between very low-density lipoprotein (VLDL) profiles and insulin sensitivity, liver synthetic function and inflammation. Previous studies have shown that insulin resistance, impaired glucose tolerance (IGT) and the incidence of T2DM are associated with higher concentrations of large VLDL particles (VLDL-P) and also larger mean VLDL size [88]. Furthermore, the secretion of VLDL is influenced by other pathophysiological processes associated with insulin resistance, like impaired liver synthetic function and inflammation. Albumin, lipoproteins (VLDL and high density lipoprotein) and coagulation factors production is decreased by impaired liver synthetic function. Besides, lower levels of both LDL cholesterol and VLDL have been associated with reduced levels of apolipoprotein B in NASH and cirrhosis [89,90]. Jiang et al. showed in their study that indices of insulin sensitivity (1/HOMA-IR and Matsuda index) were negatively associated with concentration of total VLDL-P. Moreover, they observed the same and even stronger association with large VLDL-P. Concentrations of the subclasses of VLDL-P were positively associated with markers for liver synthetic function. Inflammatory markers such as CRP and IL6 had, when tested alone, non-linear associations with VLDL-P concentration and when these markers were adjusted for insulin sensitivity and liver synthetic function they showed a negative association. In conclusion, this study demonstrates that VLDL profile can potentially be used as a predictive marker for steatohepatitis and liver fibrosis.

3.16. Inappropriate Insulin Secretion After Roux-en-Y Gastric Bypass is Associated with Post Prandial Hypoglycemia [91]

Bariatric surgery, such as Roux-en-Y gastric bypass (RYGB), is an effective treatment for obesity and its comorbidities. However, there is a major concern of hypoglycemia following bariatric surgery, which can lead to further complications. To investigate whether those patients who will develop hypoglycemia may show altered glucose and insulin secretion early after surgery, Vaurs et al. [91] studied 46 patients prospectively, measuring continuous glucose and oral glucose tolerance tests (OGTT). Of these 46 patients, 25 developed hypoglycemia, and these patients were characterized by younger age, higher weight loss, and less incidence of pre-surgical diabetes. They also had higher early (15 min after glucose load) and lower late (90 and 120 min after glucose load) insulin secretion rates in the OGTT which were reflected by a steeper slope (−13.3 in the hypoglycemia group vs. −7.1 in the non-hypoglycemia group). These results suggest that patients who will develop hypoglycemia show early changes in insulin secretion rates that can identify them for early treatment and monitoring. Future studies may want to examine mechanisms of these changes as well as possible therapies.

3.17. Weight Loss Induced by Very Low Calorie Diet is Associated with a More Beneficial Systemic Inflammatory Profile than by Roux-en-Y Gastric Bypass [92]

Roux-en-Y gastric bypass (RYGB) and very low calorie diets (VLCD) can lead to improved metabolic biomarkers and insulin sensitivity in obese patients with diabetes. To investigate whether this may be related to changes in inflammatory pathways, Lips et al. [92] examined 27 age- and BMI-matched women with diabetes who underwent RYGB or VLCD, before and 3 months after intervention. CRP and leptin decreased by approximately −60% and −50% respectively with RYGB and by approximately −38% and −44% respectively with VLCD. Adiponectin increased by approximately 15% and 32% with RYGB and VLCD respectively. TNFα and activated peripheral cytotoxic T and B cells were increased after RYGB by approximately 23%, 88%, and 8% respectively, while TNFα was decreased by approximately −25% with VLCD. IL-2, IL-6 and IL-4 were altered by VLCD but not by RYGB. Thus, these two mechanisms of weight loss have different impacts on inflammatory factors, which may be related to differences of invasiveness. Longer term studies would be needed to explore the length of these effects and relation to the surgical event.

3.18. Serum Anti-Müllerian Hormone Concentration in Women with Polycystic Ovary Syndrome and T1DM [93]

The prevalence of PCOS in T1DM is increasing, while the reason for this increase remains unknown [94–96]. Use of high amounts of insulin in this population has been speculated to enhance the effect of LH on theca cells resulting in high androgen levels and accelerated growth and recruitment of follicles [97,98]. Anti-Müllerian hormone (AMH) is a member of TGF-β family, is secreted by granulosa cells and plays an important role in follicle recruitment and growth. Levels of AMH are high in patients with PCOS. Lebkowska et al. did a cross-sectional study on 4 groups of women; PCOS + T1DM (n = 16), T1DM and no PCOS (n = 21), PCOS (n = 36), control (n = 16) and correlated their AMH levels with other hormones like LH, testosterone and with ovarian ultrasound characteristics. All the groups had comparable baseline characteristics including anthropometric measurements except waist hip ratio which was high in the patients with PCOS. The authors found that serum AMH levels were significantly higher in patient with PCOS and PCOS +T1DM as compared to the control and only T1DM group. No difference was noted in the AMH levels between PCOS and PCOS +T1DM. Ovarian follicular count and ovarian volume were also very similar in PCOS and PCOS +T1DM and were higher as compared to the control and T1DM group. Authors also found a positive correlation between AMH, LH and testosterone as described in previous studies. Logistic regression analysis further revealed an independent association between AMH, total dose of long acting insulin, ovarian follicle number and volume with the diagnosis of PCOS. The study confirms the findings of high AMH concentrations in PCOS including patients with T1DM and PCOS. Interestingly, high basal doses of insulin correlated with PCOS. Further studies to confirm this finding and to study the pathophysiology of PCOS in T1DM are needed.

4. Intervention Clinical Research Studies

4.1. Changes in Hormones Related to Energy Homeostasis Affect Central Nervous System Response to Food Cues During Treatment with Liraglutide in Humans [99]

Analogs of glucagon-like peptide 1 (GLP-1) such as liraglutide have been increasingly recognized as important therapies for obesity and T2DM, although the mechanisms – whether central or peripheral – are not yet well-defined in humans. Farr et al. [99] enrolled 20 patients with T2DM in a randomized, placebo-controlled, cross-over trial who were studied after placebo and liraglutide for 17 days (3 days at 1.8 mg) and before weight loss. While on liraglutide, participants showed approximately a 10% decrease in fasting leptin levels (p < .01) and a 53% increase in GIP (p < .03). Changes in GIP corresponded to deactivation of the insula, an area involved in attention and reward, while changes in leptin levels corresponded to increased activations in reward-related areas, including the midbrain, precuneus, dorsolateral prefrontal and motor cortex as well as decreased activations in the attention and cognitive control related areas including the parietal cortex, thalamus and pre-supplementary motor areas. These results indicate that compensatory changes in hormones related to energy homeostasis may impact CNS responses to food cues and consequent consumption during the use of GLP-1 analogs. Future studies may wish to examine how combined therapies may help to prevent the plateau in weight loss by acting on these compensatory hormone changes.

4.2. Cold-Exposure Affects Energy Expenditure and Supraclavicular Brown Adipose Tissue Volume in Humans

Brown adipose tissue (BAT) increases energy expenditure by inducing thermogenesis [100]. In order to study whether repeated cold-exposure increases metabolic rate and/or BAT volume in humans, Romu et al. enrolled 28 healthy participants in a randomized, open, parallel-group trial. These subjects were randomized to achieve cold-exposure 1 h/day, or to avoid any sense of feeling cold, for 6 weeks (12 were allocated to controls and 16 to cold-exposure). Metabolic rate (MR) was measured by indirect calorimetry measurements at baseline and at the end of the study, in the fasting state before and after acute cold-exposure (by wearing cold vests or drinking cold water). The amount of BAT in the supraclavicular region was measured with magnetic resonance imaging (MRI). In the group randomized to avoid cold for 6 weeks both the non-stimulated and the cold-activated MRs were significantly lowered (p = 0.047 and p = 0.028 respectively). In the group exposed to cold, there was only a trend towards increased MR at room temperature (p = 0.052). The group exposed to cold had higher MR (both non-stimulated and cold-activated) compared to the group avoiding cold. Supraclavicular BAT volume increased in the group exposed to cold (from 0.0175 ± 0.015 l to 0.0216 ± 0.014 l, p = 0.049) [4]. These findings demonstrate that exposure or not to cold can affect metabolic rate probably through the regulation of BAT volume in humans. Future studies should focus on the mechanisms involved in these effects, as well as whether specific hormones can have similar effects.

4.3. Energy Deprivation Regulates Appetite by Changing Metabolic Homeostasis [101]

Factors contributing to reduced energy intake and the subsequent overeat are not fully characterized. Karl et al. aimed to investigate associations between appetite, metabolic and endocrine parameters in 23 young military personnel during and after severe, short-term energy deprivation. In this randomized, controlled and crossover trial, the authors reported that appetite increased during severe short-term energy deprivation despite maintenance of diet volume. Circulating FFAs, BCAA, beta-hydroxybutyrate (BHB) and cortisol concentrations were increased, while glucose, triglycerides, insulin and leptin concentrations, and net protein balance were decreased. Markers of fat mass (leptin) and body protein losses (i.e., lean mass) were both inversely associated with ad libitum energy intake following energy deprivation [101]. Altogether, metabolic factors increase appetite during energy deprivation independent of diet volume. Following energy deprivation, signals related to the observed adipose and lean tissue loss may lead to energy restoration. Future studies should aim to define these signals involved in energy restoration.

4.4. Energy Replacement Diminishes the Effect of Exercise on Postprandial Lipemia in Boys [102]

TAG levels predict the development of atherosclerosis and cardiovascular disease. Exercise has been shown to lower these levels, but it is not yet clear whether and how the length of time between exercise and food consumption may impact TAG reductions. Thackray, Barrett, and Tolfrey [102] investigated the effects of exercise on TAG levels with an immediate or delayed energy replacement meal in 18 healthy 11–13 year old boys. They discovered that immediate replacement of energy lost during exercise prevents the reduction of TAG levels where the delayed replacement condition showed 16% lower TAG levels than the immediate and 15% lower than resting/no exercise conditions. There was no difference between levels when resting and in the immediate replacement conditions. This suggests that meals should be delayed after exercise for maximal benefits on TAG levels. Future studies should confirm these results in adults and girls whose metabolisms may be different.

4.5. Effects of Sprint Training on Resting Circulating Irisin Levels [103]

Irisin is a hormone secreted by muscle that stimulates energy expenditure by browning of adipose tissue in mice [104,105]. Several previous studies have investigated the effect of exercise on circulating levels of irisin, reporting however contradictory results [106–108]. Tsuchiya et al. aimed in their study to investigate the effect of a 4 week sprint training program (once vs twice every other day) on circulating irisin levels in 20 healthy males. Their results showed that resting irisin concentration after 4 weeks was significantly reduced in the subjects following the sprint training program (independently of the times trained every other day). These findings are in contrast with previous results demonstrating either an increase of irisin levels after acute training or no changes/small increase after chronic exercise. Further studies with well validated ELISA kits [109] are necessary to define the role of exercise in irisin secretion.

4.6. Associations of HDL Changes and Adiponectin Levels by Lifestyle Intervention and Metformin Treatment in Humans [110]

In order to investigate the mechanisms through which high density lipoproteins (HDL) may be impacted by lifestyle interventions and metformin treatment, Goldberg et al. [110] studied 1645 participants in the year-long Diabetes Prevention Program (DPP). They observed that changes in adiponectin explained the positive effects of lifestyle modifications on HDL cholesterol possibly through increases in HDL particle concentration when controlling for potential confounders. However, change in BMI and plasminogen activator levels may mediate the effects of metformin on HDL cholesterol. Altogether this may suggest that lifestyle modifications and metformin may have different mechanisms of actions to impact cholesterols and cardiovascular health but this will need to be studied in more detail in the future.

4.7. Extended-Release Niacin Reduces Plasma Lipoprotein(a) Levels [111]

Lipoprotein (a) [Lp(a)] is a lipoprotein similar in protein and lipid composition to LDL, which demonstrates atherogenic and prothrombotic effects. Increasing Lp(a) levels are associated with a continuous increase in the risk of cardiovascular disease. Niacin has been reported as a possible treatment for the reduction of Lp(a) levels, while it can reduce coronary atherosclerosis and the risk for cardiovascular events. Sahekbar et al. investigated in a meta-analysis of randomized placebo-controlled trials the effect of extended-release niacin on plasma Lp(a) levels. Their results show that Lp(a) levels were significantly reduced after extended release niacin treatment. This shows that niacin may be a potential effective antihyperlipidemic drug, especially in populations with high Lp(a) levels.

5. Conclusion

A great variety of studies were published in 2016 in Metabolism that covered a wide spectrum of important areas in the complex field of metabolic research. From appetite regulation, energy homeostasis and obesity to glucose regulation, T2DM and MetS, the new insights presented in this article are poised not only to advance and stimulate further research in the field, but also to change clinical practice in the near future.