Obstructive sleep apnoea and cardiovascular complications: perception versus knowledge

Authors


Correspondence: Dr J Ren, University of Wyoming College of Health Sciences, Laramie, WY 82071, USA. Email: jren@uwyo.edu

Summary

  1. Epidemiological evidence has confirmed that obstructive sleep apnoea (OSA) significantly promotes cardiovascular risk, independent of age, sex, race and other common risk factors for cardiovascular diseases, such as smoking, drinking, obesity, diabetes mellitus, dyslipidaemia and hypertension.
  2. Patients with severe OSA exhibit a higher prevalence of coronary artery disease, heart failure and stroke. Despite the tight correlation between sleep apnoea and these comorbidities, the mechanisms behind increased cardiovascular risk in OSA remain elusive. Several theories have been postulated, including sympathetic activation, endothelial dysfunction, oxidative stress and inflammation.
  3. The association between OSA and cardiovascular diseases may be rather complicated and compounded by the presence of components of metabolic syndrome, such as obesity, hypertension, diabetes mellitus and dyslipidaemia. The present minireview updates current knowledge with regard to the cardiovascular sequelae of OSA and the mechanisms involved.
Abbreviations
AHI

Apnoea hypopnoea index

CIH

Chronic intermittent hypoxia

CPAP

Continuous positive airway pressure

FFA

Free fatty acids

H/R

Hypoxia–re-oxygenation

hs-CRP

High-sensitivity C-reactive protein

NAFLD

Non-alcoholic fatty liver disease

NEFA

Non-esterified fatty acids

OSA

Obstructive sleep apnoea

Introduction

Although the overall morbidity and mortality of cardiovascular disease have declined significantly in recent years, cardiovascular disease remains the number one cause of death and a major health threat. From 1997 to 2007 there was a 27.8% decline in the cardiovascular mortality rate in the US, although cardiovascular disease continues to impose a significant financial burden on health care.[1] Identifying risk factors and instituting early treatment for the risk factors is of paramount importance. One of the emerging risk factors for cardiovascular disease is obstructive sleep apnoea (OSA), a rather serious, potentially life-threatening condition. Activation of the sympathetic nervous system during respiratory events in sleep may potentiate vasoconstriction and trigger increases in blood pressure and heart rate.[2, 3] Obstructive sleep apnoea is also associated with several cardiorespiratory problems (e.g. loud snoring, loud gasps and daytime breathlessness). Both clinical and epidemiological evidence demonstrates a tight association between OSA and increased risk of cardiovascular disease, including endothelial dysfunction, hypertension, arrhythmias and heart failure.[4-6] Nonetheless, the presence of confounding factors in patients with OSA, such as obesity, makes proper delineation of sleep apnoea-induced cardiovascular risk somewhat difficult.

Obstructive sleep apnoea involves episodes of partial and/or total collapse of the upper airway alternating with normal breathing. It is characterized by repeated cessation of breathing in sleep, likely due to complete or partial pharyngeal obstruction. This may lead to cycles of desaturation and rapid re-oxygenation, referred to as intermittent hypoxia. Clinically, OSA is recognized and diagnosed by a combination of symptoms and laboratory results, including repetitive apnoeas and hypopnoeas accompanied by hypoxia, sleep arousals and haemodynamic changes.[7, 8] These hypopnoeic/apnoeic events usually lead to chronic intermittent hypoxia (CIH), increased sympathetic response, sleep fragmentation and increased afterload. Obstructive sleep apnoea can be classified as mild, moderate or severe. It has been reported that approximately one in five adults suffer from mild OSA, whereas one in 15 adults suffer from moderate OSA.[9] Mild OSA refers to an apnoea–hypopnoea index (AHI) of 5–15 (events/h), moderate OSA denotes an AHI of 15–30, whereas severe OSA suggests a AHI > 30.[10] Across the spectrum of sleep-disordered breathing, many researchers also consider snoring as a part of the ‘upper airway resistance syndrome’.[11] Usually, snoring may constitute the main preliminary symptom of OSA. The current most effective non-invasive management technique for OSA is application of continuous positive airway pressure (CPAP) to assist in the maintenance of proper airway patency.[12]

Cardiovascular Complications in OSA

Endothelial function

Endothelial dysfunction usually precedes visible clinical manifestations of cardiovascular disease, particularly hypertension. Endothelial dysfunction triggers the onset and progression of atherosclerotic injury in the vasculature. Hypoxia and hypercapnia, hallmarks of OSA, may serve as potent stimulators of vasoactive substances to initiate a cascade of events resulting in endothelial defect. Elevated adhesion molecules are implicated in the pathogenesis of atherosclerosis and other cardiovascular disorders. Adherence of monocytes to the vascular endothelium is mediated by either resident or circulating leucocytes. Monocyte adherence often results in the release of several inflammatory mediators, including tumour necrosis factor (TNF)-α and interleukin (IL)-1.[4, 5, 9] Dyugovskaya et al.[13] reported significantly increased adhesion molecules, including CD15 and CD11c, in monocytes and increased adherence of monocytes to human endothelial cells in patients with OSA. In addition, they reported reactive oxygen species (ROS) accumulation in monocyte and granulocyte subpopulations in patients with OSA.[13] Furthermore, nasal CPAP treatment was associated with downregulation of CD15 and CD11c monocyte expression and decreased ROS production in CD11c+ monocytes in patients with OSA.[13] Sleep apnoea was also found to be associated with the release of chemokines, such as IL-8 or monocyte chemoattractant protein-1, and adhesion molecules, including intercellular adhesion molecule 1 (ICAM-1) and selectins.[14] Heat shock protein (HSP) 70 and TNF-α levels were investigated in patients with OSA.[15] Basal levels of HSP70 in monocytes were significantly higher in patients with OSA and were positively correlated with both AHI and time below 90% arterial oxygen saturation. In contrast, heat stress-induced HSP70 was significantly decreased in patients with OSA and was inversely correlated with AHI and time below 90%. Furthermore, TNF-α levels were inversely correlated with the ability to synthesize de novo heat stress-induced HSP70.[15] Decreased levels of nitrite and nitrate have also been reported in hypertensive and normotensive subjects with OSA[16] and lower flow-mediated dilation was noted in patients with OSA who were otherwise free of any cardiovascular diseases.[17, 18] Continuous positive airway pressure treatment significantly improved flow-mediated dilation, even after only 4 weeks, in a small cohort of men (n = 28) with OSA.[17] Discontinuation of CPAP for 1 week compromised flow-mediated dilation,[17] suggesting a pivotal role of endothelial function in pathophysiological changes in patients with OSA.

Endothelial dysfunction is greater as the severity of OSA increases. This may be due, in part, to the fact that levels of essential adhesion molecules, such as vascular cell adhesion molecule-1 (VCAM-1) and ICAM-1, are correlated with the severity of OSA.[14, 19] Similarly, an association has been demonstrated between OSA and levels of soluble IL-6 receptors, IL-6 concentrations and TNF-α.[20, 21] In rodent models, CIH triggers an increase in TNF-α levels.[22] The mechanisms for endothelial dysfunction in patients with OSA may be associated with inflammation and oxidative stress. Figure 1 shows the proposed mechanisms involved in OSA-induced organ complications. It is of note that endothelial dysfunction in patients with OSA may easily facilitate cardiovascular anomalies, including atherosclerosis, myocardial ischaemia, stroke, arrhythmias and hypertension. A summary of the proposed mechanisms and the resulting complications from OSA-induced endothelial dysfunction is given in Fig. 2.

Figure 1.

Schematic diagram showing the proposed mechanisms involved in obstructive sleep apnoea (OSA)-induced organ complications. NAFLD, non-alcoholic fatty liver disease; LV, left ventricular; ROS, reactive oxygen species.

Figure 2.

Schematic diagram showing proposed mechanism involved in obstructive sleep apnoea (OSA)-induced endothelial dysfunction. ADMA, asymmetric dimethylarginine; hs-CRP, high-sensitivity C-reactive protein; IL, interleukin; NO, nitric oxide; ROS, reactive oxygen species; SMC, smooth muscle cells; TNF-α, tumour necrosis factor-α.

Hypertension

Obstructive sleep apnoea is associated with hypertension.[10, 23-25] Even moderate OSA may increase the propensity for hypertension.[26] Although OSA and hypertension often occur jointly in the presence of common cardiovascular comorbidities, such as obesity and metabolic syndrome, an independent association between the two has been consolidated. In the Wisconsin Sleep Cohort Study, a tight positive correlation has been found between the sleep-disordered breathing AHI and increases in blood pressure independent of other known comorbidities.[26] The authors reported that subjects with an AHI > 15 events/h exhibit a 3.2-fold increase in the odds of developing hypertension compared with individuals without OSA. Other large cohort studies have also documented a trend for a correlation between OSA and hypertension. For example, the Sleep Heart Health Study examined 2470 normotensive individuals and found that the odds ratio for hypertension according to AHI values was not significant after adjusting for body mass index (BMI).[27] In contrast, O'Connor et al.[27] reported a possible modest association between severe OSA and hypertension independent of BMI. Their findings indicate that subjects with an AHI ≥ 30 events/h exhibit a 1.5-fold increased risk of developing hypertension. The apparent discrepancies in the study outcomes for the association between OSA and hypertension may be related to differences in study design, sample population and baseline AHI.[28] Nevertheless, a trend for a positive association has been noted in these studies.

The prevalence of OSA is higher in hypertensive individuals compared with the estimated prevalence of the general population, who often go undiagnosed.[23] In a unique study by Drager et al.[29], subjects with established hypertension were screened for OSA. Overall, 56% of patients were diagnosed OSA based on sleep questionnaires and polysomnographic tests, representing a much greater incidence compared with the general population (2–9%). Moreover, levels of pro-inflammatory circulating markers, including high-sensitivity C-reactive protein (hs-CRP), IL-6, IL-18, TNF-α and asymmetric dimethylarginine, were higher in hypertensive individuals with than without OSA.[30, 31] Tumour necrosis factor-α is an inflammatory cytokine independently associated with OSA or hypertension.[32-34] Li et al. reported increased levels of TNF-α and neuropeptide Y in hypertensive subjects with OSA compared with hypertensive subjects without OSA, normotensive subjects with OSA, and controls.[35] These markers of inflammation in patients with OSA and hypertension give us a better understanding of the mechanisms underlying the health risks associated with OSA.

In addition to proinflammatory markers, an overt increase in sympathetic activity has been noted in hypertensive subjects.[2] This rise in sympathetic activity may be observed in the presence and absence of OSA.[36, 37] O'Driscoll et al.[38] investigated sympathetic activity in adolescents diagnosed with OSA and found a tight association between noradrenaline levels and AHI, as well as a significant association between adrenaline and AHI, supporting a likely role for sympathetic activation in the onset and development of hypertension in patients with OSA. The increase in sympathetic activity, which carries over into the waking hours, suggests a causative relationship between OSA and hypertension.

Treatment with CPAP has been reported to lower blood pressure in hypertensive OSA subjects. In a large multihospital cohort in Spain, newly diagnosed hypertensive patients with OSA exhibited decreases of 2.1 and 1.3 mmHg in systolic and diastolic blood pressure, respectively, following 3 months CPAP treatment.[39] In a similar study, 96 patients with resistant hypertension experienced a significant reduction in blood pressure following 3 months CPAP treatment.[40] Other studies also report similar results in terms of blood pressure reductions following CPAP treatment, despite unchanged TNF-α and IL-6 levels.[31] Further studies are warranted to identify the possible mechanisms and particular inflammatory markers involved in the blood pressure-lowering effect of CPAP treatment.

Cardiac dysfunction and arrhythmias

It is well known that sympathetic overactivation may promote ventricular dysfunction and arrhythmia. Subjects with severe OSA often have a much higher prevalence of atrial fibrillation, non-sustained ventricular tachycardia and complex ectopic excitation compared with individuals without OSA.[41] After correction for age, sex, BMI and prevalent coronary disease, subjects with severe OSA have an odds ratio of 4.02 for atrial fibrillation compared with controls. Obstructive sleep apnoea may serve as a useful predictor for atrial fibrillation independent of obesity, as found in a large cohort (3542 patients).[42] Not surprisingly, the degree of nocturnal oxygen desaturation as a result of OSA also independently predicts the incidence of atrial fibrillation.[43] In one study, the myocardial performance index was evaluated in patients with OSA.[43] In patients with severe OSA, the myocardial performance index was significantly higher compared with that in patients with mild OSA. Patients with both moderate and severe OSA exhibit left ventricular (LV) diastolic dysfunction, whereas patients with mild OSA often display normal LV function.[44] In addition, patients with severe OSA often develop moderate LV hypertrophy (LVH), as well as increases in the thickness of the interventricular septum (IVS) and LV posterior wall, LV mass (LVM) and LVM index compared with patients with moderate OSA.[44] Severe OSA is also associated with LVH, left atrial enlargement, right atrial enlargement and right ventricular hypertrophy.[44] A significant reduction in LVH was reported in patients with OSA following 6 months CPAP therapy in the absence of changes in the enlarged atria.[45] In obese patients with severe OSA, significantly lower LV ejection fraction and stroke volume were reported compared with values in overweight, age-matched controls.[45] After 6 months CPAP treatment, there was a significant reduction in the thickness of the IVS and a significant increase in stroke volume.

The cardiac arrhythmias associated with OSA may be the consequence of a reflex response to apnoea and hypoxia, and are much more common during sleep. Increased sympathetic activity can initiate myocardial irritability, leading to the onset of arrhythmia. These arrhythmias may occur in the absence of any anatomical defects in the cardiac conduction system.[9] It has been postulated that the severity of arrhythmias may be directly related to low nocturnal oxygen saturation and the duration of sleep with oxygen saturations < 90%.[46] These indices may be used to predict arrhythmias in patients with OSA. In addition, OSA has been used as a predictive marker for the incidence of postoperative arrhythmias in subjects undergoing coronary bypass surgery.[47] Although the precise mechanisms underlying the development of cardiac arrhythmias in the presence of OSA are unknown, increased sympathetic tone has been commonly considered as a likely culprit. Figure 1 summarizes several commonly accepted mechanisms underlying OSA-triggered cardiac arrhythmias. Further studies are needed to clearly delineate the risk of arrhythmia in patients with OSA and to develop optimal treatment regimens.

Myocardial ischaemia

Recent evidence indicates an increased risk of coronary artery disease in patients with OSA, independent of other cardiovascular comorbidities.[6, 48, 49] In the Sleep Heart Health Study, sleep-disordered breathing was associated with an all-cause mortality and, specifically, mortality associated with coronary artery diseases.[50] In particular, coronary artery calcification was present in 67% of patients diagnosed with OSA, compared with only 31% of patients without OSA.[48] Moreover, the degree of coronary artery calcification was significantly greater in patients with than without OSA. This association remained unchanged after adjusting for age and sex, and was reported in the absence of pre-existing coronary diseases. In rat models of chronic intermittent hypoxia, 1 week of hypoxic treatment is capable of compromising the function of the hypothalamic–pituitary–adrenal (HPA) axis.[51] This alteration may result in an increased sensitization to the HPA axis under acute stress and an increased release of adrenocorticotropic hormone. Increased sensitization of the HPA axis may be a central mechanism explaining the overall increase in cardiovascular disease risk.

Obstructive sleep apnoea has been shown to retard the recovery of ventricular function after acute myocardial infarction.[52] This phenomenon coincides with a much more frequent incidence of myocardial ischaemia, as well as haemodynamic and neurohormonal abnormalities at night, possibly due to the higher incidence of cardiac arrhythmias at night in patients with OSA.[5] Inflammation has been shown to be a potent mediator of myocardial ischaemia. Of the cellular mechanisms reported for myocardial ischaemic injury, hypoxia inducible factor (HIF)-1α and the endothelin system play significant roles in inflammation-associated myocardial injury in patients with OSA.[53] Similar to its beneficial roles in hypertension and arrhythmia, CPAP treatment for > 4 months reduced signs of early atherosclerosis, including carotid intima–media thickness, arterial thickness (evaluated by pulse wave velocity), hs-CRP and catecholamines, compared with controls.[54] Continuous positive airway pressure treatment for ≥ 4 h/night markedly reduced blood pressure, total cholesterol, TNF-α and insulin resistance in patients with severe OSA and metabolic syndrome.[55] In contrast, an independent trial failed to note any significant change in hs-CRP, IL-6, adiponectin and interferon-γ after 4 weeks CPAP treatment in patients with OSA.[56] More large-scale randomized trials are needed.

Oxidative stress also serves as a potential mechanism underlying coronary artery disease and myocardial ischaemia. The hypoxia–reperfusion cycles associated with OSA lead to an increase in the production of ROS. Changes in vascular reactivity also contribute to enhanced ROS production during OSA. Mice exposed to 14 days of CIH experienced enhanced vasoconstriction.[57] In mice with an intact sympathetic nervous system, noradrenaline produced changes in vascular resistance, whereas mice with a defective sympathetic nervous system (e.g. under sustained hypoxia) only responded to high doses of noradrenaline. Interestingly, responsiveness to acetylcholine remained unchanged.[57] This change in the vasoconstriction–vasodilatation balance may play a role in OSA-induced myocardial ischaemic injury.

Heart failure

Heart failure is usually the ultimate cardiac sequel of OSA. Given that independent associations have been established for OSA and cardiovascular diseases, it is not surprising that an independent association exists between OSA and heart failure.[58-61] In the Sleep Heart Health Study cohort, men with severe OSA were 58% more likely to develop heart failure than those without OSA.[59] In a cohort of post-myocardial infarction patients, patients with severe OSA had a significantly higher incidence of major adverse events (15.9%) compared with patients with mild and moderate OSA (3.3%).[62] Major adverse events included re-infarctions, unplanned target vessel revascularizations, hospitalization for heart failure and death. At the 18-month follow up, severe OSA was associated with a lower event-free survival rate compared with patients with mild and moderate OSA.[62]

The prevalence of OSA in patients with heart failure ranges from 15% to 50%.[61] In OSA, upper airway stability is altered and may be further compromised in patients with heart failure. Increased filling pressure in the supine position during sleep may contribute to increased airway collapsibility.[60] Obstructive sleep apnoea occurs more frequently in men with heart failure than in women.[61] In addition, patients with concurrent systolic and diastolic heart failure had a high prevalence of OSA, although a large majority of patients may go unrecognized and underreported.[60] The mortality rate has been evaluated in heart failure patients with either severe or moderate OSA.[63] In that study, patients with severe OSA were found to have a significantly higher mortality rate. Several mechanisms have been postulated to explain the association between heart failure and OSA, including increased sympathetic tone, oxidative stress and inflammation.[9, 61] Treatment with CPAP may improve LV ejection fraction,[64] although CPAP failed to improve LV ejection fraction following a pilot 3 month treatment in newly diagnosed OSA patients with stable systolic dysfunction.[65] The benefits of CPAP are inconclusive in patients with heart failure and further research is needed to better elucidate the interplay between OSA and heart failure.

Cellular and Molecular Mechanisms in OSA

Increased sympathetic activity

Patients with OSA are prone to episodes of intermittent hypoxia throughout sleep, triggering surges in blood pressure and sympathetic activation via carotid chemoreceptors. These surges and sympathetic activation (peripheral, adrenal and renal) may result in diurnal increases in blood pressure.[66] Long-term facilitation of sympathetic tone via carotid chemoreceptors has been reported in animals models of CIH.[67] In addition, the baroreceptor reflex may ‘reset’ to allow for increased blood pressure. Patients with OSA do not exhibit a nocturnal ‘dip’ in blood pressure. Rather, sympathetic activity will vary solely as a function of the cyclic apnoeic episodes.[68]

Increased noradrenaline levels are common manifestation of sympathetic overactivation and are found in hypertensive subjects with OSA. It has been reported that normotensive subjects with OSA may also exhibit increased noradrenaline levels.[69] An increase in noradrenaline levels was found in subjects with OSA during waking hours.[70] This increased sympathetic traffic has been reported in the presence of increased muscle nerve sympathetic activity compared with normotensive controls.[3] This increased sympathetic activity has been postulated to play a role in the link between OSA and cardiac arrhythmias and hypertension. In addition to this increased sympathetic activity, oxidative stress has also been proposed to play a pivotal role in OSA patients with hypertension.[36]

Oxidative Stress

Oxidative stress occurs as a result of an imbalance between the production of ROS and anti-oxidant capacity. This oxidative stress increases as a consequence of ischaemia–reperfusion cycles like those seen in ischemic disease.[5, 13] Obstructive sleep apnoea is characterized by intermittent hypoxaemia–re-oxygenation (H/R) cycles that closely resemble these ischaemia–reperfusion cycles.[13] Several cells sources have been proposed as the origin of the increased ROS, including mitochondria, leucocytes and/or endothelial cells, or the ROS may be the result of H/R cycling.[71] There are several redox-activated transcription factors pertinent to OSA, including HIF-1α, nuclear factor (NF-κB, activator protein-1, early growth response-1 and IL-6. Using an in vitro model of intermittent H/R, Ryan et al.[72] used HeLa cells transfected with reporter constructs and DNA binding assays to evaluate the master transcriptional regulators of the inflammatory and adaptive pathways (NF-κB and HIF-1, respectively). HeLa cells exposed to intermittent H/R exhibited selective activation of the proinflammatory transcription factor NF-κB, whereas the adaptive regulator HIF-1 was not activated. These investigators went on to examine 19 OSA patients and 17 matched normal subjects. Circulating levels of the proinflammatory cytokine TNF-α were found to be much higher in patients with OSA, but were normalized with CPAP therapy. Furthermore, circulating neutrophil levels were much higher in patients with OSA.[72] These findings received support from another independent study in which activated NF-κB was observed in mice exposed to CIH and in patients with OSA.[73] Not surprisingly, the NF-κB activity was significantly decreased when obstructive apnoeas and their resultant CIH were eliminated by CPAP therapy. Further scrutiny revealed the involvement of increased inducible nitric oxide synthase levels in CIH-induced increases in NF-κB activity.[73] Therefore, NF-κB may serve as an essential molecular messenger linking OSA and cardiovascular pathologies in patients with OSA. Nuclear factor-κB is known to upregulate a cascade of adhesion molecules involved in leucocyte recruitment and migration to the inflammatory site, including ICAM-1, E-selectin and VCAM-1.[74] This is in line with the pronounced endothelial dysfunction, platelet activation and adhesion molecule accumulation, as well as atherosclerotic lesions, in OSA.[74] Consequently, oxidative stress may be exacerbated during OSA, leading to inflammation and endothelial dysfunction.

Non-alcoholic fatty liver disease

Non-alcoholic fatty liver disease (NAFLD) is a spectrum of conditions characterized histologically by hepatic steatosis in individuals without significant alcohol consumption who are negative for markers of viral, congenital and autoimmune disease. The spectrum ranges from fat accumulation in hepatocytes without a concomitant inflammation or fibrosis (classified as simple hepatic steatosis) to hepatic steatosis with a necroinflammatory component (steatohepatitis) with or without fibrosis or cirrhosis.[75] Such hepatic lipid accumulation can result from an imbalance between lipid availability and lipid disposal, leading ultimately to lipoperoxidative stress and hepatic injury.[76] Non-alcoholic fatty liver disease was first described 30 years ago and represents the leading cause of liver disease in developed countries, with a prevalence of 20–35% in general populations.[77] It was originally postulated that NAFLD develops as a result of insulin resistance, although recent evidence favours a bidirectional mechanism between NAFLD and insulin resistance.[78]

Mounting evidence has been accumulated to indicate the role of OSA as a risk factor for hepatic injury.[79-81] Obstructive sleep apnoea has been associated with NAFLD. Animal research using CIH reported increased hepatic lipid biosynthesis in genetically obese ob/ob mice.[82] Savransky et al.[83]found overt lipid peroxidation, inflammation and fibrosis in livers from mice on a high-fat/high-cholesterol diet along with 6 months CIH. Oxidative stress is believed to play a role in OSA-induced NAFLD because hepatic protein expression and phosphorylation of the superoxide generating enzyme p47phox were found to be significantly elevated after 4 weeks exposure to CIH.[84]

Because it is difficult to examine the pure correlation between OSA and NAFLD independent of obesity and other confounding factors in human subjects, Norman et al.[85] evaluated the associations among OSA, serum aminotransferase levels, metabolic syndrome and markers of hypoxia and identified a tight association between hypoxia and serum aminotransferase levels. Oxyhaemoglobin desaturation may be associated with markers of NAFLD instead of metabolic syndrome. Mishra et al.[86] further evaluated patients undergoing bariatric surgery, with their data suggesting that patients with OSA and steatohepatitis have a significantly higher alanine aminotransferase : aspartate aminotransferase ratio and lower desaturation and nocturnal oxygen saturation. The lowest desaturation was independently associated with histological steatohepatitis in these individuals with OSA. In another study of bariatric surgery patients, NAFLD lesions, NAFLD activity score and fibrosis were significantly more severe in patients with the highest oxygen desaturation index.[87] This association remained after adjustment for age, obesity and insulin resistance. In patients already diagnosed with NAFLD, 46% were diagnosed with OSA, including lean subjects (BMI < 25 kg/m2).[88] Treatment with CPAP has been shown to have equivocal effects on hepatic enzymes and to provide beneficial effects against oxidative stress.[89] Shpirer et al.[90] investigated the effects of 2–3 years CPAP treatment in patients with deteriorating liver function. Patients with moderate–severe OSA exhibited a lower liver attenuation index compared with patients with mild OSA. Patients who were more compliant with CPAP treatment had significantly improved mean liver attenuation indices. It is therefore apparent that a link between OSA and NAFLD does exist, but the mechanisms are not entirely clear. Of all the mechanisms proposed for NAFLD, insulin resistance, altered hepatic metabolism and oxidative stress seem to play a role in the interaction between OSA and NAFLD. Table 1 provides a brief overview of the proposed mechanisms underlying the development of NAFLD in patients with OSA.

Table 1. Proposed mechanisms underlying the development of non-alcoholic fatty liver disease in patients with obstructive sleep apnoea
Risk factorMechanismReferences
  1. FFA, free fatty acids; NEFA, non-esterified fatty acid; SREBP, sterol regulatory element-binding proteins; SCD-1, stearoyl coenxyme A desaturase 1.

Hepatic insulin resistanceIncreased FFA from adipocytes in circulating blood is taken up by the liver, inhibiting hepatic glucose production and leading to increased glucose production [81]
Altered hepatic metabolism

Increased NEFA from peripheral fat in adipose tissue

Upregulation of SREBP-1c and SCD-1 increases monounsaturated fats and the biosynthesis of cholesterol esters and triglycerides

[84, 97-100]
Oxidative stressAltered cell membrane properties modify membrane receptor and enzyme activity, antigen expression, membrane permeability and intercellular interactions [72, 100]

Hepatic insulin resistance

Insulin resistance is instrumental in the onset and development of NAFLD.[80] The term ‘insulin resistance’ refers to a reduced insulin-mediated glucose disposal response in tissues such as the liver, muscle and adipose tissues. Insulin resistance may lead to an increased release of free fatty acids (FFA) from adipocytes, prompting elevated FFA levels in the circulation that can be taken up by the liver. With increased hepatic fat accumulation, the ability of insulin to inhibit hepatic glucose production becomes impaired.[91] Hepatic insulin resistance will lead to a rise in plasma glucose levels and a concomitant stimulation of insulin secretion, ultimately serving as an initiating factor for NAFLD. Barcelo et al.[92] identified a correlation between increased FFA levels and AHI, suggesting a role of sleep fragmentation in the release of FFA into the circulation.

In OSA, elevated circulating levels of adrenaline, noradrenaline and angiotensin II may suppress the inhibitory effect of insulin on glucose.[93] This exacerbates glucose intolerance and insulin resistance. Lin et al.[94] examined components of metabolic syndrome in non-obese patients diagnosed with OSA. Their data suggest a role for the AHI as an independent risk factor and predictor of insulin resistance. Nonetheless, one major argument against a role of OSA in triggering insulin resistance is that CPAP treatment seems to be somewhat ineffective against insulin resistance.[95]

Altered hepatic metabolism

Studies using stable (non-radioactive) isotopes have revealed metabolic pathways pivotal for fatty accumulation in the liver.[96] There are four major pathways identified that contribute to the fatty acid pool in the liver. First and foremost, peripheral fat stored in adipose tissues may be delivered to the liver via the plasma non-esterified fatty acid (NEFA) pool. The second pathway belongs to fatty acids newly made within the liver through de novo lipogenesis. Both the third and fourth pathways involve dietary fatty acids entering the liver through ‘spillover’ into the plasma NEFA pool (third pathway) or the intestinally derived chylomicron remnants (fourth pathway). It has been reported that the majority of NEFA in the liver originate from the first metabolic pathway.[96] Accumulation of NEFA from adipose tissue indicates a failure of suppression of adipose fatty acid flux in the fed and fasting state, a hallmark of NAFLD.[96] This altered hepatic metabolism may be further exacerbated by OSA. For example, using animal models has revealed an upregulation of sterol regulatory element binding protein 1c (SREBP-1c) and stearoyl coenxyme A desaturase 1 (SCD-1) associated with OSA[83, 97]; SREBP-1c is a hepatic transcription factor for lipid biosynthesis, whereas SCD-1 is an SREBP-1c-regulated enzyme that converts saturated fatty acids into monounsaturated fatty acids. The abundance of monounsaturated fatty acids promotes the biosynthesis of cholesterol esters and triglycerides. In humans, increased levels of cholesterol, low-density lipoprotein and triglycerides have all been reported in patients with OSA.[98]

Oxidative stress

Oxidative stress is a commonly accepted mechanism in OSA for the development of NAFLD and cardiovascular diseases.[66, 71, 99] In NAFLD, oxidative stress may directly modify the properties of cell membranes to impact membrane receptor integrity and enzyme activity, antigen expression, membrane permeability and intercellular interactions.[80] This will cause liver cell degeneration and necrosis. Liver stellate cells may be activated, leading to altered morphology, synthesis of extracellular matrix components, collagen deposition and liver fibrosis. Another important transcriptional factor turned on during OSA is NF-κB, which is known to promote inflammatory responses in the liver. As a result of NF-κB activation, TNF-α, FFA and oxidative stress may work in concert to facilitate overt apoptosis of hepatocytes, leading to inflammation, fibrosis, degeneration and necrosis of hepatocytes, all of which are characteristics of steatohepatitis and the progression from hepatic steatosis to steatohepatitis.[79, 87] The oxidative stress pathway, initiated by ROS-induced lipid peroxidation, is expected to represent a particularly important mechanism for the interaction between OSA and NAFLD.

Conclusions and Future Directions

Obstructive sleep apnoea represents a serious condition with grave ramifications. It has been associated with a myriad of cardiovascular complications, including endothelial dysfunction, hypertension, cardiac dysfunction, arrhythmias and heart failure. In addition, OSA may contribute to the metabolic dysfunction seen in metabolic syndrome and NAFLD. Although many complications have been identified regarding OSA, there are still many unanswered questions, including the exact mechanisms by which OSA can give rise to these cardiovascular and metabolic issues.

Although associations between OSA and cardiovascular disease are better established, studies investigating OSA and NAFLD are less prolific.[100] Despite the confounding factors, obesity in particular, making independent associations difficult, more human studies are needed to elucidate these associations. In addition, further investigations into the effects of CPAP are needed. Equivocal results regarding the effectiveness of CPAP as a therapeutic option for metabolic dysregulation call for more well-controlled studies. Although CPAP treatment has been shown to ameliorate the cardiovascular impairments of OSA, the exact mechanisms underlying this improvement need to be determined. As investigations continue, it is widely acknowledged that early identification of OSA is important to reduce its cardiovascular and metabolic ramifications.

Ancillary