Impact of obesity on cardiac metabolism, fibrosis, and function

Abstract

Obesity is a global pandemic with a huge burden on the healthcare system. Obesity is not only linked to the development of risk factors for atherosclerotic vascular disease but also has a strong association with ventricular hypertrophy, heart failure, atrial fibrillation, and stroke. Recent experimental and clinical studies have demonstrated that obesity is associated with cardiac dysfunction, adipokine dysregulation, and activation of the pro-fibrotic signaling pathways leading to cardiac fibrosis, which is a key structural change responsible for atrial fibrillation. Importantly, these also have been shown to be reversible with weight reduction strategies. This review discusses the alterations in cardiac metabolism and function due to obesity. In addition, it addresses the complex and not yet fully understood mechanisms underlying cardiac fibrosis, with a focus on atrial substrate predisposing to atrial fibrillation in obesity.

Introduction

Obesity is a global pandemic with a huge burden on the healthcare systems [1], [2]. The prevalence of obesity remains high in developed countries with more than two-thirds of adults being classified as overweight or obese [1], [2], [3]. Obesity is not only linked to the development of metabolic risk factors for atherosclerotic vascular disease, it also has strong and independent associations with left ventricular hypertrophy, heart failure, atrial fibrillation, and stroke. Indeed, many studies over the last decades have shown the impact of obesity on cardiac remodeling with structural and functional abnormalities. However, the mechanisms contributing to these changes remain incompletely understood with a complex interplay of hemodynamic, neurohumoral, and metabolic factors as well as inflammation and oxidative stress, contributing to cellular apoptosis, hypertrophy, and interstitial fibrosis. Specifically, concomitant conditions seen frequently in obese individuals such as hypertension, sleep apnea, and diabetes mellitus also have direct cardiac remodeling effects. This review is focused on the cardiac functional, morphological, and metabolic abnormalities due to obesity alone, with specific attention on the underlying signaling pathways underlying cardiac fibrosis seen with increased adiposity.

Defining the cardiomyopathy due to obesity has proven to be a challenge due to the confounding co-morbid conditions such as hypertension, diabetes, and coronary artery disease that are often present in obese individuals. Conflicting data have been reported in the literature with different imaging modalities and indices used for comparisons. Nevertheless, a causal association has been demonstrated in a large longitudinal population-based study whereby a graded increase in the risk of heart failure was observed across all categories of body mass index. There was increase in heart failure risk by 5% in men and 7% in women for each increment of one in the body mass index after adjustments for known risk factors such as age, alcohol consumption, serum cholesterol, diabetes, hypertension, left ventricular hypertrophy, and myocardial infarction [4].

Left ventricular systolic function, as assessed with standard echocardiographic left ventricular ejection fraction, has been shown to be either normal or supranormal in obese subjects. Even in the severely obese, overt systolic dysfunction is uncommon in the absence of concomitant heart disease [5], [6]. Despite the variable findings, more recent studies using novel echocardiographic techniques of tissue velocity and deformation imaging have demonstrated the presence of subclinical systolic contractile abnormalities in obese subjects without coronary or structural heart disease. The obese individuals in these studies demonstrate decreased spectral pulsed-wave systolic velocity as well as reduced regional and global strain, although left ventricular (LV) ejection fraction remained in the normal range [7], [8], [9], [10]. These tissue Doppler abnormalities, being load-independent, may indicate an inherent abnormality in contractility with obesity not appreciated with load-dependent parameters like LV ejection fraction.

Likewise, variable findings have been reported in different experimental models of obesity including mildly reduced or preserved LV systolic dysfunction [11], [12]. Mildly reduced LV systolic function has been demonstrated in obese Zucker diabetic fatty rats, transgenic and chronic high-fat-fed obese insulin-resistant mice and transgenic mice with cardiac steatosis [11], [13], [14], [15], [16]. The LV systolic function was, however, preserved in leptin-deficient homozygous obese mice despite a slight reduction seen with dobutamine challenge [17]. The significance of these findings in transgenic models of cardiac steatosis and their applicability to humans is not clear. Furthermore, LV systolic function was unchanged or mildly reduced in rats with diet-induced obesity [12], [18]. In sheep with high-caloric-diet-induced obesity, LV systolic function as assessed by LV ejection fraction was unaffected [19].

Although earlier reports on the effect of obesity on LV diastolic dysfunction have been variable, numerous recent studies have reported the presence of mild diastolic dysfunction in obese individuals [7], [8], [20]. These involved a variety of echocardiographic measures such as prolonged left ventricular relaxation time, increased E/e ratio, and lower E/A ratio, suggestive of diastolic filling abnormalities and elevated filling pressures [7], [8]. These have been in keeping with the elevated LV end diastolic and pulmonary wedge pressures reported in obesity [21], [22]. The prevalence of diastolic dysfunction appears to increase with increasing severity of obesity [23].

The diastolic dysfunction in obesity has been shown to correlate with left ventricular hypertrophy. However, studies in the severely obese have demonstrated diastolic dysfunction even after adjusting for left ventricular hypertrophy [8], [20]. One such study suggested that diastolic dysfunction might be related to cardiotoxicity secondary to elevated free fatty acids [20]. Similarly, diastolic abnormalities have also been shown in leptin-deficient homozygous obese mice, diet-induced obese mice, and diet-induced obese rabbits [12], [17], [24], [25]. Further, persistent diastolic dysfunction has been demonstrated in a transgenic murine model of cardiac steatosis even in the absence of systemic metabolic derangement or weight gain [11]. Elevated right and left atrial and pulmonary artery pressures have also been observed during invasive measurements in an ovine model of diet-induced obesity [19]. As with humans, elevated atrial pressures in presence of normal left ventricular pressure suggested diastolic dysfunction in this obese ovine model.

Recent data from a meta-analysis of 22 echocardiographic studies confirmed a consistent relationship between obesity and left ventricular hypertrophy [26]. Specifically, the overall prevalence of left ventricular hypertrophy was 56% in obesity, with a significant relationship between body mass index and left ventricular hypertrophy. There was an odds ratio of 4.19 for developing left ventricular hypertrophy with obesity [26]. Likewise, several smaller studies on younger healthy obese patients without co-morbid conditions have also demonstrated a positive relationship between obesity and left ventricular hypertrophy [7], [8]. Furthermore, a higher prevalence of eccentric rather than concentric left ventricular hypertrophy (66% vs. 34%) has been observed with obesity in an echocardiographic study [26]. However, a recent cardiac magnetic resonance study in patients without identifiable cardiac risk factors suggested a predominant concentric hypertrophic pattern in obese men and both concentric and eccentric hypertrophy in obese women [27]. The cardiac remodeling process due to obesity is likely to be progressive with more ventricular hypertrophy seen with longer duration and larger amount of weight gain (Fig. 1) [19].

Left atrial enlargement has been associated with obesity in both children [28] and adults [8], [29]. The echocardiographic measures of the left atrium from different planes suggested an oval-shaped left atrium in obese individuals. The anterior–posterior (parasternal long-axis view) and longitudinal diameters (apical four-chamber view) were increased without a significant change in the transverse apical four-chamber measure [30]. In addition, an echocardiographic imaging study has demonstrated a lower atrial peak systolic strain rate indicative of atrial dysfunction in obese individuals [28]. The Framingham observational cohort study has further highlighted the strong association between left atrial enlargement in obese individuals and incident atrial fibrillation [29]. In keeping with clinical observations, cardiac magnetic resonance imaging has also demonstrated increased bi-atrial and pericardial fat volume in an ovine model of diet-induced obesity (Fig. 1) [19]. Indeed, increased left atrial epicardial adiposity and pericardial fat have been shown to have a strong association with atrial fibrillation incidence, burden, and ablation outcome [31], [32].

Cardiac fibrosis and metabolic abnormalities secondary to obesity may partly account for the various functional and morphological changes described above.

Obesity is known to be associated with increased myocardial oxygen consumption, increased fatty acid metabolism, and reduced myocardial efficiency [33]. The rate of fatty acid uptake appears to be greater than its oxidation, contributing to myocardial lipid accumulation, as demonstrated with non-invasive proton magnetic resonance spectroscopy in obese individuals [34]. Importantly, lipotoxicity from the accumulated lipid has been linked to cardiomyopathy in obese animals [12]. There is also evidence of defects in mitochondrial oxidative capacity, mitochondrial uncoupling, and increased mitochondrial reactive oxygen species from transgenic obese mice studies [35]. These further contribute to abnormal myocardial substrate utilization and reduced cardiac efficiency that may result in contractile abnormalities in obese hearts.

Both pre-clinical and clinical studies have demonstrated an association between cardiac fibrosis and obesity. Although direct histological evidence of cardiac fibrosis from obese subjects without associated co-morbidities is unavailable, alterations in extracellular matrix regulatory proteins such as the metalloproteinases and their tissue inhibitors have been demonstrated [36]. Furthermore, serum markers for cardiac collagen synthesis such as pro-collagen type III have been shown to correlate with left ventricular diastolic dysfunction in obese individuals [37]. Moreover, histological evidence of increased interstitial fibrosis has been confirmed in a variety of diet-induced and transgenic obesity models [11], [38], [39] with increased cardiac collagen levels and altered cardiac expression of extracellular matrix turnover proteins [11]. Taken together, the alterations in extracellular matrix regulation may account for the left ventricular remodeling seen in obese subjects to result in both diastolic and systolic dysfunction as described earlier.

More recently, atrial interstitial fibrosis also has been demonstrated with short-term weight gain [19] in an ovine model (Fig. 2). The abnormal structural milieu is likely to be a key contributor to abnormal atrial conduction leading to increased arrhythmia. This is consistent with epidemiological studies demonstrating a robust relation between obesity and atrial fibrillation [29]. In addition, epicardial fat has also been implicated as a contributor to atrial fibrillation [31], [32]. There is pre-clinical evidence to suggest that epicardial fat may promote fatty infiltration and myocardial fibrosis through paracrine effects [40]. A recently validated non-invasive method for quantitation of atrial pericardial adipose tissue will facilitate further clinical studies [41].

The diverse pathophysiological abnormalities due to increased adiposity include chronic stretch with volume and pressure overload, metabolic derangements, neurohumoral changes, increased inflammation, and dysregulation of adipokines. All of these may contribute to the complex signal transduction pathways underlying cardiac fibrosis, although their pathogenic roles and interactions in the obese hearts remain incompletely understood. Here, we provide a concise review of key signal transduction pathways involved in cardiac fibrosis in obese hearts. A simplified schema illustration of the mechanisms associated with atrial fibrosis in obesity is shown in Fig. 3.

A state of heightened inflammation has been demonstrated in various obese animal studies. Accumulation of macrophages or inflammatory infiltration of the myocardium has been reported in obese rodent ventricles [42] and ovine atria, respectively, (Fig. 2) [19]. An increased expression of various pro-inflammatory cytokines and pro-fibrotic growth factors has also been shown in obese hearts [19]. These findings are not surprising given that obesity has long been associated with chronic systemic inflammation with increased leukocyte count and various pro-inflammatory cytokines including C-reactive protein, interleukin 6, and tumor necrosis factor-α [43].

Activation of the renin–angiotensin–aldosterone system has been shown in various animal models of obesity with elevated myocardial angiotensin II, angiotensin-converting enzyme, and angiotensin II type I receptor expression [12], [39]. In obese individuals without other significant co-morbidities, treatment with an aldosterone antagonist resulted in reduced pro-collagen levels and improved cardiac function [44]. This association between the renin–angiotensin–aldosterone system and cardiac remodeling was further affirmed in animal studies. These studies have demonstrated that treatment with an angiotensin receptor blockage agent (losartan) or angiotensin-converting enzyme inhibitor (perindopril) attenuated angiotensin expression, transforming growth factor β1, and mitogen-activated protein kinase cascade, resulting in reduced cardiac fibrosis [38], [45].

TGF-β superfamily overactivity has been described in human obesity [46]. TGF-β has three isoforms, the most important being TGF-β1. TGF-β1 is expressed in human adipose tissues and significantly correlates with body mass index [47]. The increase in TGF-β1 expression has also been shown in the left ventricles of rabbits with diet-induced obesity, together with increased cardiac interstitial collagen content [48]. TGF-β1 overexpression has also been shown in atrial tissues of obese sheep [19]. Further, overexpression of constitutional TGF-β1 was associated with selective atrial fibrosis, conduction heterogeneity, and increased atrial fibrillation vulnerability in transgenic mice [49].

The TGF-β signaling pathway is a crucial regulator in cardiac fibrogenesis acting primarily through the Smad protein pathway [50]. Increased atrial TGF-β1 expression has been shown in human AF with increased pro-fibrotic phospho-Smad-2 and -4 proteins [51], [52]. Phospho-Smad-2 is also elevated in the atria of transgenic mice with TGF-β1 overexpression [51]. In-vitro experiments in Smad-3^−/− fibroblasts and intervention study in rat atria organo-culture have shown that Smad-3 and activin A are essential for TGF-β1-mediated collagen deposition [40], [53]. TGF-β also acts downstream to Angiotensin II and acts in a paracrine–autocrine fashion [54]. Angiotensin induces the synthesis of TGF-β1, which in turn stimulates fibroblast to produce angiotensin II. TGF-β1 also activates a variety of non-canonical signaling pathways including: ras/MEK/ERK; p38 [55]; and c-Jun N-terminal kinase (JNK) [56]. In addition, TGF-β induces fibroblasts to differentiate into α-smooth muscle actin (α-SMA)-expressing myofibroblasts with connective tissue growth factor (CTGF) acting as a co-factor [57]. When applied to fibroblasts, TGF-β directly induces extracellular matrix gene expression and promotes collagen deposition by simultaneously suppressing matrix metalloproteinase gene expression and inducing tissue inhibitors of matrix metalloproteinase gene expression [58].

Enhanced endothelin-1 system activity has been demonstrated in obese individuals [59]. There is also pre-clinical evidence of elevated serum and myocardial endothelin-1 in diet-induced obesity that is mediated by leptin [60]. In an obese ovine model, increased atrial endothelin-1 and endothelin-A/B receptor expression have also been demonstrated with increased atrial fibrosis and atrial fibrillation [19]. The pathogenic role of endothelin-1 can be inferred from clinical studies demonstrating higher endothelin-1 expression in myocytes and fibroblasts from the left atrial appendage of patients with atrial fibrillation than those in sinus rhythm and a positive association between endothelin-1 and left atrial size or atrial fibrosis [61]. Similarly, plasma endothelin-1 levels are associated with recurrence of atrial fibrillation following catheter ablation [62].

Endothelin-1 induces extracellular matrix production and, in association with TGF-β1, promotes differentiation of fibroblasts to myofibroblasts [58], [63]. It also stimulates aldosterone release via the endothelin-B receptor [64]. In mouse cardiomyocytes, endothelin-1 has been shown to induce extracellular matrix accumulation by mediating CTGF expression [65]. Previous studies in hypertensive rats have shown that the pro-fibrotic effect of endothelin-1 is likely to be mediated through the endothelin-B receptor [66]. Further, endothelin-1 is essential in angiotensin-II-induced cardiac fibrosis, with one study demonstrating reduced expression of TGF-β1, CTGF, and collagen I/III in mice with endothelin-1 deficiency as compared to their wild-type counterparts [67].

The progression from lean to obese state is accompanied by a hypoxic state of the expanded adipose tissue, resulting in dysregulation of gene expression and adipokines. Although relative hyperoxia also has been reported in obese adipose tissue [68], the balance of evidence indicates adipose tissue hypoxia with obesity that could contribute to a chronic inflammatory and pro-fibrotic state [69], [70], [71]. Recent studies have suggested that hypoxia-inducible factor-1α (HIF-1α) induces fibrosis in the hypoxic adipose tissue. This has been confirmed in intervention studies demonstrating that the fibrotic response in obesity pads can be prevented by a chemical HIF-1α inhibitor or by overexpression of a dominant-negative HIF-1α mutant [72].

Obesity is also associated with increased leptin [73] and reduced adiponectin levels [74]. Recent experimental evidence also suggests pro-fibrotic effects of leptin in the myocardium [75], [76] that may play a crucial role in atrial fibrogenesis and development of atrial fibrillation [77]. The pro-fibrotic mechanisms of leptin are likely to be mediated by the Janus-activated kinase and mitogen-activated protein kinase pathways with impact on myocardial matrix metabolism [78], [79]. Leptin also promotes cardiomyocyte hypertrophy, as is evident from in-vitro leptin treatment of human ventricular myocytes and from in-vivo obese mouse hearts [78], [80]. The anti-fibrotic role of adiponectin has been shown in rodents where it protected against angiotensin-II-induced cardiac fibrosis [81]. Further, lower adiponectin levels have been demonstrated in obese individuals that also independently correlated with larger left atrial size [82]. Further studies are warranted to further delineate its role in atrial fibrosis and consequent atrial fibrillation.

Recently, it has been hypothesized that mediators of lipid metabolism and inflammation may be supplied by the epicardial adipose tissue surrounding the heart, which contains a wide variety of bioactive cytokines [83]. The contiguity of epicardial adipose tissue to the myocardium is highly conducive to paracrine interactions involving adipokines, free fatty acids, and pro-inflammatory cytokines that may contribute to lipotoxic cardiomyopathy [84]. Venteclef et al. have provided insights to the link between epicardial fat and atrial fibrosis. They performed elegant experiments in a rat organo-culture model and demonstrated that secretome from epicardial adipose tissue can induce myocardial fibrosis with likely involvement of activin A [40].

The negative impact of obesity on cardiac metabolism, fibrosis, and function is likely to be reversible with weight reduction. Experimental studies have demonstrated attenuation in left ventricular hypertrophy, fibrosis, and diastolic dysfunction following weight reduction by caloric restriction [39], [85]. Clinical studies have also shown a reduction with serum markers of extracellular matrix regulatory proteins [36] and significant improvements in echocardiographic parameters of left ventricular remodeling with weight reduction [86]. Regression of microvascular fibrosis in the adipose tissue [87] and improvement in adipokine profile has also been shown with weight reduction following bariatric surgery [88]. More recently, a randomized clinical trial in patients has shown improvement in left ventricular hypertrophy, reduction in left atrial size, together with reduced atrial fibrillation burden and severity following weight reduction achieved by caloric restriction [89]. Purposeful weight loss in conjunction with aggressive risk factor management has also been shown to improve long-term sinus rhythm maintenance following catheter ablation for atrial fibrillation in overweight individuals [90]. Taken together, a weight reduction strategy must be actively pursued to reduce the cardiovascular impact of obesity and its burden on the healthcare system. Whether the method of weight reduction has differential impact on cardiac metabolism and remodeling remains unknown.