Why do some glucose‐lowering agents improve non‐alcoholic fatty liver disease whereas others do not? A narrative review in search of a unifying hypothesis

Non‐alcoholic fatty liver disease (NAFLD) and type 2 diabetes (T2D) are metabolic disorders connected by common pathophysiological mechanisms. Since insulin resistance (IR) and metabolic alterations are common to both conditions, almost all glucose‐lowering agents which improve IR have also been studied in patients with NAFLD. Some have shown great efficacy, others none. Thus, the mechanisms behind the efficacy of these drugs in improving hepatic steatosis, steatohepatitis, and eventually fibrosis remain controversial. Glycaemic control improves T2D, but probably has limited effects on NAFLD, as all glucose‐lowering agents ameliorate glucose control but only a few improve NAFLD features. In contrast, drugs that either improve adipose tissue function, reduce lipid ingestion, or increase lipid oxidation are particularly effective in NAFLD. We therefore hypothesise that improved free fatty acid metabolism may be the unifying mechanism behind the efficacy of some glucose‐lowering agents on NAFLD and may represent the key to NAFLD treatment.


| INTRODUCTION
The strong connection between type 2 diabetes (T2D) and nonalcoholic fatty liver disease (NAFLD) is demonstrated by their epidemiology: the prevalence of NAFLD in T2D patients is about 60% (vs.25% in the general population) 1 and people with NAFLD have a 2-fold higher risk of developing diabetes. 2This risk is related to liver fat accumulation, inflammation and fibrosis, which impair hepatic glucose and lipid metabolism.][5] Under normal physiological conditions, the direct action of insulin on hepatocytes primarily facilitates net hepatic glycogen synthesis and inhibition of glycogenolysis, while in white adipose tissue (WAT) insulin suppresses lipolysis, consequently decreasing circulating free fatty acids (FFA) and glycerol.
In NAFLD, increased hepatic lipid storage worsens IR but is not sufficient alone to cause hyperglycemia and thus T2D.At the same time, high fat diet and obesity also lead to inflammation and macrophage activation in WAT, contributing to increased lipolysis and adipocyte IR, resulting in the continued delivery of FFA and glycerol to the liver despite high plasma concentrations of insulin. 6,7 has therefore been hypothesised that, in hepatic steatosis, the acute response to insulin through suppression of HGP is also impaired by WAT dysfunction. 8Thus, hepatic steatosis alone is not sufficient to cause hyperglycemia, but in conditions of already partially altered insulin secretion, impaired WAT function compromises the regulation of HGP. 9 On the other hand, IR and hyperinsulinemia in T2D patients increase fat storage in the liver, initiating a vicious cycle which worsens NAFLD. 10Indeed, in T2D, the ability of insulin to suppress adipose tissue lipolysis is impaired, leading to increased delivery of FFA to the liver.Moreover, the insulin resistant-state enhances de novo lipogenesis (DNL) because of the stimulation of lipogenic enzymes via sterol receptor-binding protein-C (SREBP-c).Actually, evolution to NASH and cirrhosis occurs more frequently in T2D patients than in persons without T2D diagnosis, and conversely, the presence of NAFLD in T2D increases cardiovascular risk in these patients. 11The above is evidence of the strong pathophysiological connection between IR, diabetes and NAFLD.
Considering the evidence supporting the association between IR, T2D and NAFLD, several trials have been designed to prove the efficacy of glucose-lowering agents on hepatic steatosis in all stages.Targeting IR, WAT dysfunction, incretin system or renal glucose secretion, glucose-lowering agents have been hypothesised to be effective drugs for NAFLD, but only some have shown efficacy in treating NAFLD in patients with and without a diagnosis of T2D.However, although these drugs have been widely studied to understand their mechanisms of action, how they work to improve hepatic steatosis, steatohepatitis and fibrosis, is still controversial.
Our aim here is to understand how different drugs acting via different mechanisms lead to NAFLD amelioration or resolution and whether they have a common mechanism of action.

| METFORMIN
Metformin action on hepatic steatosis is currently controversial.Its primary role is apparently based on the suppression of HGP. 12 This is exerted mainly by inhibition of mitochondrial respiratory chain complex 1 and activation of AMP-activated protein kinase (AMPK). 13,14However, the role of AMPK has been challenged by many studies and might only account for indirect variations in hepatic insulin sensitivity. 15Various mechanisms have been proposed for metformin's acute inhibition of gluconeogenesis, for example, alterations in cellular energy charge, AMP-mediated inhibition of adenylate cyclase and modulation of the cellular redox state. 16,17In contrast, recent work has pointed out that metformin increases HGP in individuals with recent onset T2D and in subjects without T2D. 18deed, metformin administration causes an increase in lactate levels, which in the liver can be converted into glucose, increasing gluconeogenesis and glucose turnover. 19atever the mechanism of action, it has been hypothesised that metformin could improve NAFLD by reducing de novo lipogenesis and consequently reducing intrahepatic TG, 20,21 but DNL, as explained above, is not the only mechanism implicated in hepatic fat storage.Moreover, Bugianesi et al. have suggested that metformin could ameliorate WAT dysfunction by changing tissue resident macrophage polarisation, thus reducing obesity-related inflammation. 22Even if WAT dysfunction is a major contributor to NAFLD pathogenesis, this effect does not seem to have a strong impact on NAFLD patients treated with metformin.[25][26] This could be partially explained by the fact that metformin improves lipid metabolism secondary to long-term alterations in body composition and glucose homoeostasis but does not affect FFA mobilisation and storage.Indeed, it has been demonstrated that metformin has no effect on the modulation of intrahepatic partitioning of fatty acids towards oxidation, re-esterification and resecretion as TG. 27en though some studies have pointed out that metformin treatment can lead to weight loss, especially in obese insulinresistant patients, nowadays metformin is not considered a real weight loss agent.Its effect on weight varies widely and is weaker than that of other drugs. 28,29 conclusion, even though the effects of metformin have not yet been completely clarified and consequently it is not clear how metformin works on the liver in NAFLD, it does not seem to exert any major effect.Since metformin does not seem to exert any role in the utilisation and redistribution of FFA (Table 1), it might be speculated that this drug has no effect on NAFLD because it does not improve the management of FFA.Further studies are warranted to fully understand why metformin has no effect on NAFLD.

| PIOGLITAZONE
Pioglitazone is an insulin-sensitising agent that acts through activation of the nuclear hormone receptor peroxisome proliferatoractivated receptor γ (PPARγ).It is one of the most effective drugs in NAFLD treatment -at all stages -; it reverses IR, especially in WAT, and improves hepatocyte steatosis, necrosis, and lobular inflammation. 30It may reduce fibrosis, 31 to a greater extent in subjects with severe fibrosis, 32 but its greatest value may be in preventing fibrosis progression (rather than promoting regression).
4][35] It stimulates lipoprotein lipase action and a large set of lipogenic genes, enhancing fatty acid uptake and synthesis in adipose tissue, which diverts FFA load towards adipocytes instead of other organs such as the liver and muscles. 36timately, inappropriate fat storage in organs other than subcutaneous adipose tissue is reduced during pioglitazone treatment, with subsequent improvement in insulin sensitivity despite the expansion of fat mass. 37Pioglitazone also upregulates adiponectin, an insulinsensitising and anti-steatogenic adipokine, which increases fatty acid beta-oxidation in liver and muscles. 38,39This drug also exerts anti-inflammatory effects on Kupffer cells, which might be indicative of direct hepatoprotective effects. 40veral randomised controlled trials (RCTs) have proven the efficacy of pioglitazone in NAFLD 31,[41][42][43] ; it has been demonstrated that pioglitazone therapy leads to biopsy-proven NASH resolution in 47% of patients without T2D and in 60% of T2D patients. 43,44en though pioglitazone is effective in reducing hyperglycemia, as shown in Table 1, it is not usually the glucose-lowering agent of choice due to its well-known side effects.Firstly, it causes weight gain, especially if combined with insulin, [45][46][47] secondly, studies have demonstrated an association between pioglitazone treatment and lumbar spine/femoral neck fractures, both in men and women 48 ; finally, it increases water retention with consequent increased risk of heart failure.
In conclusion, pioglitazone exerts an important effect on WAT remodelling, increasing fat deposition in WAT and reducing fat in ectopic organs like the liver, pancreas and heart (Figure 1).These effects are mediated by the reversal of the pathological FFA flux from WAT to other organs that occurs in obesity and WAT dysfunction, consequently promoting lipid storage in adipose tissue and improving NAFLD features, thereby underlining the importance of improving FFA utilisation to reduce NAFLD.
Semaglutide is the only GLP-1 RA with both oral and injectable formulations 53 and, as these drugs have a pronounced weight loss efficacy, both liraglutide and semaglutide have been approved for the treatment of obesity, independently of T2D. 54,557][58][59][60] All these drugs improve the clinical features of NAFLD, and all reduce plasma levels of aminotransferases.2][63][64][65] A trial has been designed to evaluate oral semaglutide in hepatic steatosis (NCT03884075) and a pilot study has shown that this new drug formulation improves indexes of hepatic steatosis and fibrosis in T2D subjects. 66 has been shown that GLP-1 RA enhances insulin sensitivity in the liver and WAT and reduces DNL and WAT lipolysis, possibly with a direct effect on the lipolysis pathway, consequently reducing FFA flux to other tissues, [67][68][69] as shown in Table 1.These drugs improve IR in WAT and increase adiponectin levels, improving adipose tissue function and redistributing lipids. 69 intestinal lipoprotein production. 70GLP-1 RA seems to have a direct effect on hepatocytes, which would seem to explain their effects on NAFLD 71,72 ; in murine models, it has been shown that GLP-1 directly reduces lipogenesis via the cAMP/AMPK pathway. 72In human hepatocytes, however, GLP-1 receptor expression and activity are still controversial. 11,73nsidering all this evidence, it has been hypothesised that the massive effect of GLP-1 RA on NAFLD is indirect and mediated by weight loss.Indeed, weight loss, induced by proper diet and physical activity, is the only treatment shown to have an impact on NAFLD and its evolution. 74Most GLP-1 RA have been shown to cause moderate (3%-5%) to consistent (>5%) weight loss in a recent metanalysis. 51They increase satiety and reduce caloric intake by delaying gastrointestinal emptying and directly acting on the central nervous system to inhibit appetite. 75 conclusion, weight loss-due to a reduction in caloric intake -

| DUAL AND TRIPLE AGONISTS
The only dual agonist approved by the FDA is tirzepatide: a glucosedependent insulinotropic peptide (GIP) and GLP-1 receptor agonist that has been shown to improve the biochemical and radiological markers of NASH and hepatic fibrosis in patients with T2D. 76,77udies are ongoing to evaluate this drug in the treatment of NASH without T2D.Moreover, tirzepatide is highly effective in inducing weight loss and is being proposed as a treatment for obesity.
As expected from its impressive effect on weight, in SURPASS-3 MRI, tirzepatide was shown to not only reduce liver fat content at magnetic resonance but also to reduce visceral adipose tissue (VAT) and abdominal subcutaneous adipose tissue (ASAT) compared to insulin treatment.
Effects on GLP-1 receptor are probably similar to other GLP-1 RAs, but effects mediated by GIP receptor agonists are not completely understood.GIP is physiologically produced in K cells, located in the mucosa of the duodenum and upper jejunum, and it is physiologically secreted in response to nutrient ingestion. 78 WAT.Moreover, GIP enhances the postprandial lipid-buffering capacity of WAT and the sensitivity of WAT to insulin, possibly preventing ectopic fat deposition. 79,80All these effects may account for NAFLD amelioration with tirzepatide treatment.However, some studies in animal models suggest that GIP enhances intra-hepatic lipid deposition and that inhibition of GIP signal by genetic manipulation prevents this process 81 ; therefore, it is not clear how the double GLP-1/GIP agonism exerts its function on the liver.In conclusion, the effect of GIP on hepatic lipid metabolism is still unclear, but GIP seems to increase metabolic flexibility since it reverses its function in conditions of hyperglycemia/euglycemia and fasting/ fed states. 80This ability, added to the agonism of GLP-1 effects, seems to favour the hypothesis of a metabolic switch in the preferred substrate, leading to increased beta oxidation of FFA and reduction of intra-hepatic lipid content.
Another dual agonist consisting of GLP-1 and glucagon (GCG) receptor agonist, named cotadutide, yet not approved by the FDA, has been shown to reverse hepatic steatosis in murine models and there are currently various ongoing trials to evaluate its efficacy on the metabolic consequences of diabetes. 82The weight-lowering effects of cotadutide appear to be less impressive than those of tirzepatide and some GLP-1 RA 80 , endorsing the hypothesis that effects on NAFLD are not exclusively related to weight loss.Apart from cotadutide, another dual GLP-1/GCGRs, for example, AR425899, pegapamodutide and efinopegdutide, have been studied in phase 1 and 2 trials.
Glucagon, along with insulin, controls glucose homoeostasis and lipid metabolism in the liver and adipose tissue reduces DNL and enhances lipolysis, resulting in the release of FFA.Indeed, glucagon stimulates intrahepatic lipolysis, increasing intrahepatic acetyl-CoA content and leading to gluconeogenesis.In addition, glucagon stimulates hepatic mitochondrial oxidation, promoting the use of lipids as a source of energy. 83By enhancing the utilisation of FFA as the principal energy substrate, GCG receptor agonists may lead to the resolution of hepatic steatosis.These effects may account for evidence of NAFLD amelioration in animal models treated with GLP-1/GCGR, which needs to be confirmed in ongoing and future studies.In conclusion, dual and triple agonists seem to be effective in NAFLD treatment, even if ongoing and future studies are needed to clarify their usefulness.These drugs seem to have an important effect on weight, which is similar or better than that of GLP-1 RA, which could account for their effect on NAFLD.However, GIP and GCG seem to have some direct effects on hepatic steatosis and its evolution, related to metabolic changes, such as the preferential use of FFA as the principal substrate.

| DIPEPTIDYL PEPTIDASE-4 INHIBITORS
Dipeptidyl peptidase-4 inhibitors (DPP-4i) are a class of glucoselowering agents that exert their function on the incretin system, like GLP-1 RA, by increasing the half-life of endogenous incretins GLP-1 and GIP, that are physiologically inhibited by dipeptidyl peptidase-4. 85tagliptin, vildagliptin, saxagliptin, and linagliptin (and others) are oral drugs with a good safety profile and tolerability. 86They do not exert any cardiovascular protection in T2D patients, but some can be used in patients with impaired renal function. 87P-4i are neutral on weight as they only cause a limited increase in GLP-1 concentration. 88,89These drugs have been actively studied in 7 trials in patients at risk of or with NAFLD, but most have not been shown to improve biopsy-proven or imaging detected NAFLD. 90[95][96][97][98][99] Considering the weight neutrality of DPP-4i, we can comprehend why these drugs have no effect on NAFLD treatment.This strengthens the hypothesis of the fundamental role of fat mass reduction, consequent lipid redistribution and improved FFA utilisation as keys to NAFLD improvement.
SGLT-2i has been shown to reduce fat liver content and to improve NAFLD 100 in all stages. 101In humans, several RCTs have been carried out to evaluate the effects of dapagliflozin, [102][103][104][105] canagliflozin, 106-109 empagliflozin [110][111][112] and ertugliflozin 113 in T2D patients with NAFLD.Dapagliflozin, empagliflozin and canagliflozin improve hepatic steatosis indexes and reduce intrahepatic lipid content, but only canagliflozin has shown a biopsy-proven improvement of NAFLD in a trial.Ertugliflozin has been shown to improve hepatic steatosis and fibrosis indexes. 101,114rrannini et al demonstrated that glycosuria, mediated by empagliflozin, not only improves fasting and post prandial glycaemia but also results in a compensatory increase in endogenous glucose release-probably by the liver -and lower tissue glucose uptake due to lower glucose and insulin levels and higher glucagon levels.They demonstrated that, as a consequence of decremented tissue glucose disposal and glucose oxidation, energy expenditure was maintained at the cost of a rise in lipid oxidation. 115Under conditions of reduced portal insulin-to-glucagon ratio, the increased delivery of FFA to the liver stimulates ketogenesis, resulting in a metabolic condition resembling a prolonged fast.Indeed, SGLT-2i causes a CICCARELLI ET AL.
-5 of 11 reduction of circulating glucose, usually available for all tissues, without affecting insulinemia -thus reducing the amount of glucose taken up.In this condition, tissues are forced to switch the main energy substrate from glucose to FFA, as shown in Table 1.
Moreover, it has been demonstrated that 5-week treatment with dapagliflozin leads to higher fat oxidation, lower carbohydrate oxidation and to negative 24h energy balance, mimicking caloric restriction. 116During dapagliflozin treatment, levels of FFA and βhydroxybutyrate were higher and there was greater endogenous (hepatic) glucose production.Due to the glycosuric effect of dapagliflozin, higher HGP was not associated with higher plasma glucose or worse glycaemic control but with better insulin sensitivity.
Moreover, it has been hypothesised that when adipose tissue sensitivity is improved, likely by negative energy balance, insulin succeeds in suppressing WAT lipolysis and, due to higher fat oxidation, other lipid deposits are used, increasing FFA turnover.
This mechanism may explain the reduction of intrahepatic lipid content and NAFLD amelioration.
The above recent study by Op den Kamp et al demonstrates the switch of substrates occurring in SGLT-2i treatment and supports the idea of 24h-substrate flexibility mediated by liver response to reduced carbohydrate oxidation and use of FFA as a principal substrate. 116en if SGLT-2i causes mild-to-moderate weight loss, this effect seems to be attenuated by the duration of treatment.This could be explained by higher caloric intake by T2D patients in active treatment or other metabolic arrangements that need to be investigated.
In any case, the mild-to-moderate weight loss caused by SGLT-2i is not sufficient to explain its effects on NAFLD.
In conclusion, it has been demonstrated that SGLT-2i causes a substrate shift from glucose to FFA oxidation.Thus, these glycosuric agents may exert their effect on NAFLD by stimulating FFA use, consequently reversing pathological lipid storage in the liver and its consequences.

| CONCLUSIONS
An analysis of our current knowledge of glucose-lowering agents and their mechanisms of action yields no clear explanation of their activity on hepatic steatosis and its evolution.In fact, it is unclear what mechanisms determine the improvement or resolution of NAFLD.
The only treatment approved for this condition is lifestyle intervention -healthy diet and physical activity 117,118 -to improve glucose and lipid metabolism.
As NAFLD is strongly associated with IR and metabolic syndrome, the opinion was that reversing this condition could be the key to NAFLD resolution.Even though IR certainly plays a role in NAFLD pathophysiology, reducing IR with metformin treatment seems to have no effect on NAFLD.Though it is not completely clear how metformin works, its effect on IR is well known and evidence of its inefficacy on NAFLD has led to the hypothesis that modifying IR is not the key to NAFLD treatment.
We have also seen that NAFLD is strongly associated with T2D; thus, it has been hypothesised that restoring euglycemia and ameliorating HbA1c could be useful in treating NAFLD.Indeed, as we have shown above, several glucose-lowering agents have demonstrated good efficacy in treating NAFLD.However, metformin and DPP-4i, for example, have no positive effects on NAFLD even though they have significant glucose efficacy, as shown in Table 1.Therefore, the fact that not all glucose-lowering agents are useful in NAFLD excludes the possibility that NAFLD is affected only by an improvement in glucose metabolism.SGLT-2i stimulate FFA oxidation in tissues due to glucose depletion without modifying insulin plasma levels.Therefore, we hypothesise that improvement in FFA utilisation may be the common key to successful NAFLD therapy and may represent the unifying mechanism for NAFLD amelioration in subjects treated with specific glucose-lowering agents (Table 1).Further studies are warranted to demonstrate this hypothesis and to find drugs that specifically reverse this physio-pathological mechanism.
likely to mediate the metabolic effects of GLP-1 RA.Caloric restriction continuously mimics the fasting state, reducing substrate availability and consequently stimulating the re-distribution of FFAs and their utilisation as principal cellular substrates.This effect ultimately causes the mobilisation of intrahepatic lipids and restores liver functions, likely explaining how GLP-1 RA leads to NAFLD amelioration.

F I G U R E 1
GIP not only stimulates insulin secretion but also simultaneously stimulates glucagon secretion.It is also implicated in the carbohydrate and lipid metabolism of adipocytes: in high insulin and modest hyperglycemia conditions, it increases glucose and FFA uptake, promoting DNL in Effects of glucose lowering-agents on non-alcoholic fatty liver disease (NAFLD).Metformin primarily suppresses hepatic glucose production (HGP), but this effect does not seem to improve NAFLD.Pioglitazone improves NAFLD by enhancing the capacity of subcutaneous adipose tissue to store lipids and reducing free fatty acids (FFA) flux to other organs such as the liver.The effect of glucagon-like peptide-1 receptor agonists (GLP-1 RA) on NAFLD is likely mediated by caloric restriction and consistent weight loss, which ultimately cause utilisation of stored fat.Dual and triple agonists seem to improve NAFLD, exerting the same effects as GLP1-RA and likely adding the effects of glucose-dependent insulinotropic peptide (GIP) and glucagon (GCG) on lipid utilisation.Dipeptidyl peptidase-4 inhibitors (DPP-4i) increase the half-life of physiological GLP-1 and GIP, likely not exerting any role in FFA mobilisation and utilisation and thus not improving NAFLD.Finally, sodium-glucose cotransporter-2 inhibitors (SGLT-2i) cause a reduction of glucose available as a cellular substrate, stimulating the shift to another source of energy, FFA, with consequent reduced deposit in the liver.

Finally, in phase 1
and 2 trials, triple agonists-GLP-1/GIP/GCGR -have demonstrated efficacy on glucose control in T2D and weight loss.Their effects on NAFLD have been studied in pre-clinical settings and they have been shown to decrease liver fat content, 84 likely mediated by the effects of incretins and GCG on weight loss and substrate switching.

Finally, NAFLD hasof 11 -
been associated with WAT dysfunction, alteration in the utilisation of FFA and evidence of ectopic fat stores.Thus, it has been hypothesised that modifying WAT function and FFA utilisation and distribution could be the main mechanism to improve or even resolve NAFLD.This hypothesis is supported by the fact that pioglitazone has shown the greatest efficacy in NAFLD treatment since it exerts a well-known and important effect on WAT remodelling, ultimately increasing fat deposition in WAT and reducing fat in ectopic organs like the liver, pancreas and heart.These effects are mediated by reversing the pathological FFA flux from WAT to other organs that occurs in obesity and WAT dysfunction, consequently promoting lipid storage in adipose tissue and improving NAFLD features.While GLP-1 RA and SGLT-2i do not exert a direct role in lipid metabolism or WAT, we hypothesised that they could ameliorate NAFLD via similar mechanisms of FFA management.Indeed, some GLP-1 RAs have a pronounced effect on weight loss that is mediated, as explained above, by different mechanisms, that could all account for NAFLD improvement using GLP-1 RAs.Weight loss causes utilisation and depletion of substrate stores, determining a greater request for energy, which stimulates the utilisation of FFA stored as ectopic fat.Thus, reduced substrate availability due to drug-induced food restriction and subsequent glycogen depletion during GLP-1 RA treatment stimulates FFA utilisation as a cellular substrate.Moreover, new therapies based on GLP-1/GIP receptor agonists, GLP-1/GCG receptor agonists and triple agonists seem to exert some effects on NAFLD likely mediated by the abovementioned mechanisms of GLP-1 RA and the GIP and GCG effects on glucose and lipid metabolism.SGLT-2i treatment causes caloric depletion due to an active glycosuric effect that mimics the fasting state.The reduction in blood glucose levels, while blood insulin levels remain the same, stimulates cells to find another energy source.SGLT-2i thus enhances the utilisation of lipid stores-circulant and stored FFA.This effect blocks FFA accumulation in ectopic organs and stimulates the utilisation of fat stores, likely leading to an improvement in NAFLD.In conclusion, metformin and DPP-4i, although they exert an optimal glycaemic control, do not impact NAFLD, likely because they do not improve FFA utilisation.Pioglitazone, some GLP-1 RA, tirzepatide and SGLT-2i, on the other hand, ameliorate NAFLD, and all modify FFA utilisation, although via different pathways (Figure1).Pioglitazone enhances fat accumulation in adipose tissue reducing FFA flux to organs like the liver, GLP-1 RA and tirzepatide strongly reduce caloric intake thus stimulating fat store utilisation, and finally 6 CICCARELLI ET AL.
Effects of glucose-lowering agents on pathophysiological mechanisms that could impact non-alcoholic fatty liver disease (NAFLD).All classes of glucose-lowering agents improve glycaemic control.Metformin has some important functions but does not play a role in free fatty acids (FFA) management and does not improve NAFLD.Pioglitazone strongly impacts on adipose tissue function, improving FFA utilisation.Glucagon-like peptide-1 receptor agonists (GLP-1 RA) and dual/triple agonists cause significant weight loss and consequent improved FFA use, while dipeptidyl peptidase-4 inhibitors (DPP-4i) are neutral on adipose tissue, weight loss and metabolic flexibility.Finally, sodium-glucose cotransporter-2 inhibitors (SGLT-2i) cause a switch in substrate utilisation from glucose to FFA.Enhanced metabolic flexibility is the common mechanism of action that improves NAFLD.
Moreover, they reduce postprandial hyperlipaemia via different mechanisms: delayed gastric emptying, inhibition of gastric lipase secretion and reduction of T A B L E 1 CICCARELLI ET AL.