Non-alcoholic fatty liver disease: Update on treatment options and translational implications of sleep disruption

Non-alcoholic fatty liver disease (NAFLD) is a condition that is associated with cirrhosis and hepatocellular carcinoma, and is increasing in prevalence worldwide. Sleep disruptions are commonly seen in NAFLD, and the disease process is associated with sleep disorders, including obstructive sleep apnea, circadian rhythm disorders, and insuﬃ cient sleep. The intermittent hypoxia seen in obstructive sleep apnea may contribute to ﬁ brotic changes in the liver. A major component of this linkage may be related to gut microbiome changes. One notable change is increase in Bacteroidetes/Firmicutes ratio, and decrease in ﬂ ora that ferment ﬁ ber into anti-inﬂ ammatory short-chain fatty acids. Several therapeutic options exist for NAFLD that target both sleep and NAFLD, including non-pharmacological factors, such as lifestyle modiﬁ cation (mainly diet and exercise). Pharmacological options include melatonin, Vitamin E, thiazolidinediones, and fecal microbiota transplantation. The pathogenesis of non-alcoholic fatty liver disease is closely tied to sleep and circadian rhythm abnormalities, through shared inﬂ ammatory pathways and altered metabolism. This review explores the pathogenesis of NAFLD in the context of sleep and circadian abnormalities. The associated inﬂ ammatory response is linked to changes in gut-microbiome interactions that contribute to the disease process. Understanding of this linkage has implications for various therapies for disease mitigation.


Introduction
Fatty liver disease is a condition involving lipid accumulation within hepatocytes and has become the most common liver disease worldwide [1]. Non-alcoholic fatty liver disease (NAFLD) is a subset of fatty liver disease and is de ined as hepatic steatosis without a cause of secondary fat accumulation such as excessive alcohol consumption, druginduced etiologies, or hereditary disorders [2]. This subset of fatty liver disease affects an estimated 30% of adults in the United States and a median of 20% adults worldwide [1].

Pathogenesis of NAFLD
Non-alcoholic fatty liver disease is further subdivided into non-alcoholic fatty liver (steatosis) and non-alcoholic others progress to NASH and have more progressive disease [6]. Current understanding involves alterations in metabolic pathways of hepatic lipid metabolism resulting from systemic insulin resistance, oxidative stress, and lipotoxicity [6]. A two-hit hypothesis helps to contextualize the current pathogenesis of NASH: the ' irst-hit' consisting of excess hepatic triglyceride accumulation due to dysregulation of fatty acids and insulin resistance in the absence of other causes and the 'second-hit' consisting of oxidative stress and cytokine expression leading to a shift from steatosis to NASH [7,8].
Sleep disruption has been a point of interest in regard to NAFLD, although the exact mechanisms have been poorly understood. Recent studies into NAFLD and various forms of sleep disruption in animal models and the transition to clinical practice will be explored in this article.

Normal circadian function
The human circadian rhythm is marked by the diurnal oscillation of function, affecting sleep, metabolism, and immune response [9][10][11][12]. The central regulator of circadian function is the suprachiasmatic nucleus (SCN) of the hypothalamus, which receives light-dark input from the retinohypothalamic tract to regulate circadian function through inhibition/activation of pineal gland production of melatonin, and through production and downstream effects of Circadian Locomotor Output Cycles Kaput (CLOCK) protein [13]. In addition, various tissues, including the liver and pancreas, display circadian automaticity in function, and are referred to as peripheral tissue oscillators [13]. Gastrointestinal circadian function is regulated by timing and frequency of food intake, and can affect the ef iciency of nutrient absorption and processing, daily oscillation of gut microbial composition, and the related metabolic and immune consequences [14,15]. Clock protein dimerizes with Bmal1, and binds to an E-box, a DNA-response element site, to increase production of proteins including PER-1, PER-2, CRY-1, and CRY-2, which regulate various metabolic processes as well as provide periodic CLOCK gene inhibition [16,17]. Misalignment between central and peripheral rhythms, such as that seen in night-shift workers, are referred to as circadian dysrhythmia, and can be associated with loss of normal circadian function and metabolic derangements.

OSA and Sleep Disruption in NAFLD
Obstructive sleep apnea (OSA) has been well studied in its relationship to NAFLD. OSA is a sleep disorder characterized by repetitive episodes of non-breathing throughout the night, typically due to an upper airway obstruction or other anatomical abnormality and resulting in either hypopnea or apnea [18]. These episodes lead to wide ranging effects including chronic hypoxia during sleep or activation of the sympathetic nervous system and can have profound effects on the cardiovascular system, endocrine system, central nervous system [18]. Furthermore, this chronic hypoxia may lead to induction of oxidative stress and subsequent lipid peroxidation which could lead to hepatocyte necrosis and apoptosis, a has been postulated to be a key factor in the "second hit" in the pathogenesis of NAFLD [19].
Evidence has shown OSA to be a potential risk factor for the development and progression of NAFLD due to a hypoxia-induced mechanism [20,21]. Chronic hypoxia leads to oxidative stress, lipid peroxidation, and systemic in lammation leading to an imbalance between oxidant and antioxidant agents and a progression of NASH through liver ibrosis. One study evaluated patients with NAFLD through polysomnography and serum analysis [20]. The participants were divided into control, moderate OSA, and severe OSA. The results revealed a signi icant increase in alanine aminotransferase (ALT), aspartate aminotransferase (AST), total cholesterol, low density lipoprotein-cholesterol, and high-sensitivity C-reactive protein with increased severity of OSA. These indings were independent of factors such as age, gender, obesity, in lammation, blood pressure, serum glucose, and lipid pro ile suggesting nocturnal hypoxia as a risk factor in progression of NAFLD.
While the exact pathophysiologic mechanisms involved in the relationship of OSA and NAFLD are still being investigated, there have been several factors that seem to be at play in this multifactorial process. There is evidence that the chronic intermittent hypoxia may trigger liver injury, induce a systemic in lammatory state, and lead to ibrogenesis [22].
A key factor may be hypoxia-inducible factor 1-alpha (HIF-1a), a transcription factor involved in the cellular response to hypoxia. One study comparing HIF-1a knockout to wild-type mice, after inducing NAFLD through a high trans-fat diet for 6 months, the wild-type mice had 80% more hepatic collagen [23]. Furthermore, this study induced the hepatocytes to a sustained hypoxic state vs normoxia and found that those under hypoxic conditions allowed for avid collagen crosslinking whereas knockout mice did not have signi icant cross-linking. Thus, hepatocyte HIF-1a appears to be a key mediator in the increase in hypoxia-induced liver ibrosis, and decrease may be protective from ibrosis [24]. HIF-1a activation seems to also have a close relationship to lectinlike oxidized low-density lipoprotein receptor-1 (LOX-1), a cell surface protein that is involved in endocytosis of lipoproteins into endothelial cells. It has been shown that LOX-1 levels are higher in patients with OSA compared to the normal population and this is important in regard to NAFLD pathogenesis because oxidative-LDL stimulates production of LOX-1 through a ROS/NF-κB signaling pathway and mediates ox-LDL induced endothelial injury and defenestration of human liver sinusoidal cells [25]. Thus it is reasonable to postulate the progression of NAFLD is mediated by a HIF-1a and its downstream effect on LOX-1 expression causing hepatocyte injury. https://doi.org/10.29328/journal.acgh.1001030 Another study analyzed the association between chronic intermittent hypoxia (CIH), morbid obesity, and the effect on liver and adipose tissue [21]. Across 101 morbidly obese subjects, histopathologic analysis of liver biopsies showed NAFLD lesions, NAFLD activity score, and ibrosis were signi icantly more severe in patients with higher oxygen desaturation index (a quanti ied severity of nocturnal CIH). This suggests that there is a dose-response relationship between the amount of nocturnal oxygen desaturations and the severity of hepatic lesions found, all of which was independent of obesity and body composition [21]. A metaanalysis demonstrated OSA progressed in terms of liver enzymes and histological alterations and demonstrated OSA was independently associated with NAFLD developmen [19]. This evidence is in agreement that the chronic hypoxic state induced by the sleep disorder of OSA may be a key component as the second-hit in the pathogenesis of NAFLD. If left untreated, this hypoxia mechanism could lead to the progression from NAFLD to NASH.

Sleep Insuffi ciency, Circadian Dysrhythmia, and NAFLD
The direct effect of isolated sleep insuf iciency and NAFLD has only recently been explored and has remained largely inconclusive [27,28]. In a large study published in 2013 of 69,463 middle-age workers and their spouses, one group assessed sleep duration and quality using the Pittsburgh Sleep Quality Index (PSQI) and then determined the presence of fatty liver [26]. Results showed that after controlling for confounding factors such as age, alcohol intake, smoking, and sleep apnea, with an odds ratio for NAFLD decrease in sleep duration of 1.28 in men and 1.71 in women. This implies that in the middle-aged general population, a shorter sleep duration and poor sleep quality are signi icantly associated with an increased risk of NAFLD. Various follow-up studies have demonstrated con licting conclusions [27,28]. One study of Japanese subjects showed that short sleep duration of < 6 hours was associated with reduced risk of NAFLD in men [27]. Another study of found no signi icant association with short sleep of < 7 hours per day but found a signi icantly increased risk for NAFLD in those sleeping > 9 hours per day [28]. The con licting data on sleep duration and NAFLD, It is clear that work is needed to better understand the pathophysiologic mechanisms that may be involved regarding decreased sleep and NAFLD.
In addition to decrease in overall sleep time, alterations in sleep habits through circadian dysrhythmia can contribute to hepatic steatosis. One recent study evaluating the relationship between circadian misalignment and metabolic associated fatty liver disease (MAFLD) found prevalence of MAFLD to be 45% in the presence of circadian misalignment compared to 28% in the non-misaligned group [29]. The data also suggest that circadian misalignment is independently associated with MAFLD and ibrosis, while short sleep duration alone is not independently associated with this risk [29].

Microbiome Changes, Sleep, and NAFLD
The gut microbiome is an ecosystem within the human gut consisting of bacteria, fungi, and other microbes, which affect the host's physiology [30]. The microbiome is affected by the host's environmental factors such as diet, exercise, medications, circadian rhythm, and geographic location, and can also effect change in host physiology [30]. Environmental factors can lead to dysbiosis, a term used here to refer to a change in the composition of the microbiome that is associated with pathogenesis. The dysbiotic microbiome can alter activation of the immune system through direct interaction as well as indirectly through the production of metabolites that may trigger immune response locally or at other peripheral tissues through translocation. One such example would be through irst-pass of these bacterial products to the liver through the portal system (gut-liver axis) [30][31][32]. Metabolic products such as phenols and ammonia along with proin lammatory bacterial components such as peptidoglycans and lipopolysaccharides (LPS) can trigger an in lammatory cascade that can lead to ibrosis [32,33].
Several studies done in both humans and animals have demonstrated a link between gut dysbiosis and NAFLD [34][35][36][37]. One study examining non-obese patients with and without NAFLD, found a 20% increase in the phylum Bacteroidetes and a 24% decrease in the phylum Firmicutes among those with NAFLD [35]. Notably, lora capable of producing shortchain fatty acids (SCFA) and 7α-dehydroxylation of bile acids were severely reduced [35]. Proteobacteria, another mainly gram negative phylum, have also been found to be elevated in patients with steatosis and NASH [36]. Another study evaluated stool samples of patients with biopsy-proven NAFLD using 16S ribosomal RNA gene sequencing, and found signi icant increase in Bacteroides spp. in patients with NASH and F ≥ 2 ibrosis [37]. Rumminococcus spp. were found to be increased within patients with severe ibrosis [37]. Changes in bacterial composition are linked to.
An important homeostatic mechanism would be the regulation on the amount and quality of the sleep the host is getting. As mentioned previously patients with NAFLD can have close associations with OSA and chronic intermittent hypoxia, which lead to dysbiosis, and a propensity to increase obligate anaerobes [38]. A one study on murine models, showed that a group receiving chronic intermittent hypoxia for 6 weeks, had a higher abundance of Firmicutes and a smaller abundance of Bacteroidetes and Proteobacteria phyla compared to the normoxic control [39]. Dysbiosis of the gut can lead to an increase in bacterial products such as reactive oxygen species or lipopolysaccharide which can ultimately circulate and cause increased in lammation within the liver.

Lifestyle modifi cation
Lifestyle modi ication including diet and exercise is the irst line in managing patients with NAFLD. Primarily lifestyle modi ication is aimed at gradual weight loss by all main guidelines [2,40,41]. Various weight reduction targets have been proposed, including 3% -5%, 5% -7% and 7% -10% weight loss for improvement of steatosis, in lammation, and NAFLD/NASH remission along with decrease in ibrosis, respectively [42]. One study of patients with histologically proven NASH evaluated 52 weeks of lifestyle modi ication for weight loss [43]. At the end of the 52 weeks, 25% of patients achieved resolution of steatohepatitis, 47% had reductions in NAFLD activity scores, and 19% had regression of ibrosis [43]. Patients with the highest amount of improvement had ≥ 10% overall weight loss [43]. While there is no speci ic diet currently recommended for NAFLD, the Mediterranean diet has shown improvement in other metabolic and cardiovascular diseases. Recently the ATTICA prospective observational study of healthy volunteers without any chronic diseases or diagnosis of NAFLD evaluated the role of a Mediterranean diet with routine follow-up in cardiovascular disease and metabolic disorders [44]. The study demonstrated that a Mediterranean diet was protective from diabetes mellitus and cardiovascular disease in patients with NAFLD [44]. Given the signi icant potential for bene it and low risk of harm, lifestyle modi ication through activity and dietary change remains a major recommendation for management of NAFLD.

Melatonin
Melatonin is a neurohormone that regulates the sleepwake cycle. Its production in the pineal gland is inhibited by light perception-mediated input from the SCN [45]. In addition to pineal production, several peripheral tissues produce and bind melatonin [46,47]. In addition to its role in mediating sleep function, melatonin has several immunomodulatory effects. Melatonin increases proliferation of several immune cell lines, including lymphocytes (particularly T-helper 1 cells), natural killer cells, and neutrophils [48]. Additionally, it acts to suppress the systemic in lammatory response through inhibition of the NLRP3 in lammasome, decreased nuclear translocation of pro-in lammatory transcription factor NF-κB, anti-in lammatory transcription factor Nrf2, and decreased macrophage toll-like receptor 4 signaling when activated by bacterial LPS [49,50]. In addition, melatonin has an overall anti-oxidant effect through scavenging by its metabolites as well as through up-regulation of superoxide dismutase and down-regulation of NADPH oxidase [51,52]. One study of patients with NAFLD found that melatonin supplementation over three months decreased weight, waist size, blood pressure, transaminases, and C-reactive protein, compared to baseline, however this did not reach statistical signi icance when compared to a control group [53]. In animal models, melatonin has been demonstrated to reduce expression of Micro-RNA-34a-5p, and inhibitor of sirtuin 1 production [54]. Increased sirtuin 1 availability has been linked to decreased hepatic in lammation and ibrosis [54]. In combination, supplementation of melatonin has potential for bene it through improved sleep as well as decreased propagation of the in lammatory cascade.

Vitamin E
Vitamin E is known to be a free radical scavenger, with use in patients with NASH without diabetes. In the pioglitazone, vitamin E or Placebo for the Treatment of Nondiabetic Patients with Nonalcoholic Steatohepatitis (PIVENS) trial, usage of 800mg/day was found to be superior to placebo in patients without diabetes and cirrhosis for improvement of NASH histology 43% vs. 19% respectively [55]. Another study of longer term use of vitamin E (300 mg/day) over 3 years demonstrated improved NASH ibrosis [56]. Currently, 300-400 mg is being used rather than 800 mg to decrease adverse effects of vitamin E toxicity [57]. Increased doses of vitamin E have been observed with increased all-cause mortality, hemorrhagic stroke and risk for prostatic cancer in individuals with a family history of prostate cancer [58][59][60].

Thiazolidinediones
Thiazolidinediones are peroxisome proliferator activated receptor (PPAR)-γ ligands, which are used for type 2 diabetes mellitus (T2DM). Pioglitazone, a thiazolidinedione, has been shown in the past to signi icantly decrease steatosis and necroin lammation in patients with NASH and T2DM compared to placebos [61,62]. A meta-analysis showed improved steatosis, lobular in lammation, and ballooning on histology after usage of thiazolidinediones [63]. Thiazolidinediones have also been shown to improve ibrosis in NASH patients when compared to placebo, however it has not been seen in all studies to a signi icant extent [64][65][66]. A recent meta-analysis of 26 studies concluded that pioglitazone was associated with reductions in both AST and ALT and improved steatosis in both diabetic and non-diabetic NAFLD patients [67]. Pioglitazone is not without risk, it is well known to have a decently-sized side-effect pro ile ranging from weight gain, increased cardiovascular events, luid retention, and increased risk for bone fractures in women [40,67,68]. While there is no inalized recommendation on dosing of pioglitazone, some groups recommend a safe starting dose of 15 mg per day titrating up to 30 mg per day for histologic improvement [65,66,68]. Further research is still being done on this drug group, as it shows promise for improvement on the histologic level.

Fecal Microbiota transplantation
Gut microbiome dysbiosis and NAFLD have a close association, and direct manipulation through fecal microbiota transplantation as a potential therapy has been proposed and demonstrated in murine models [36,69]. Compared to mice that were fed only HFD, mice that underwent FMT and fed 8 weeks of HFD demonstrated correction of gut microbiota with increased abundance of Christensenella spp. and Lactobacillus spp., signi icant reduction in intrahepatic lipid accumulation and intrahepatic pro-in lammatory cytokines, and improvement in steatohepatitis [69]. Developing from murine models, a double-blinded randomized controlled trial involving 21 patients who underwent allogenic or autologous FMT demonstrated that allogenic FMT resulted in a relative decrease in small intestinal permeability [70]. The intestinal membrane dysfunction has been associated with NAFLD, and reduction in the permeability may re lect a therapeutic potential in NAFLD [71]. Current clinical investigation is underway to evaluate the ef icacy of FMT in NASH (NCT02469272).

Conclusion
Fatty liver disease and its complications are believed to be the result of altered metabolism as well as an associated cascade of systemic in lammation. Sleep-related disorders, including obstructive sleep apnea, insuf icient sleep duration, and circadian dysrhythmia, can all contribute to NAFLD pathogenesis. This can occur through direct effects on central and peripheral circadian regulators or through changes in the diurnal oscillation of gut microbial composition. Nonpharmacological therapies, including lifestyle modi ication through diet and exercise, have demonstrated ef icacy. Pharmacological therapies, including melatonin, vitamin E, thiazolidinediones, and fecal microbiota transplant, are all promising options.

Author contribution
Johnson DA, Parekh PJ, D'Souza SM and Yoo BS contributed construction of project; all authors wrote and edited the manuscript.