Abstract
Background: Obesity increases the risk of cardiovascular disease and heart failure (HF). However, in patients with established HF, many studies observed greater survival with increasing adiposity. This counterintuitive observation has been termed the “obesity paradox.” In recent years, new HF therapies have emerged that improve prognosis in patients with HF. Some of these, such as sodium-glucose cotransporter 2 inhibitors (SGLT2i), cause weight loss and may therefore interfere with the obesity paradox. Summary: This article is a narrative review on the relationship between body weight and outcomes in patients with HF with special focus on new HF treatments. PubMed was searched for studies reporting the prognostic impact of obesity in HF from 2002 to February 22, 2022. More than 400 records were examined, with 150 being included in the present review. Literature provides evidence for an obesity paradox in a broad range of HF patients, including acute and chronic HF across the spectrum of left ventricular ejection fraction. It has been verified in HF patients treated with SGLT2i but not in those using sacubitril/valsartan. Cardiorespiratory fitness and the severity of HF seem to be important confounders of the obesity paradox in HF. While unintentional weight loss is associated with a poor prognosis in HF, weight loss associated with SGLT2i treatment appears safe. Key Messages: Consensus has yet to emerge as to whether the obesity paradox is a true phenomenon in HF. As cardiorespiratory fitness is strongly associated with prognosis and significantly modifies the relationship between adiposity and survival in HF, regular physical activity is recommended irrespective of the body weight. In HF patients with severe obesity, a modest weight reduction of 5–10% may be reasonable to improve HF symptoms and quality of life.
Introduction
Obesity is a global pandemic that affects more than 10% of the world’s adult population [1]. It is an established independent risk factor for the development of cardiovascular disease and heart failure (HF), and it adversely affects cardiovascular hemodynamics, structure, and function [2]. Recent data show that about 26 million people worldwide suffer from chronic HF [3]. Although new therapies have emerged in recent years, HF remains one of the leading causes of morbidity and mortality in developed countries [3]. While obesity increases the risk of developing HF, numerous studies have reported better survival rates with a high body mass index (BMI) in patients with established HF [4-10]. This counterintuitive phenomenon has been termed the “obesity paradox” [11, 12]. To date, the obesity paradox has been observed in a wide range of HF patients across the spectrum of left ventricular ejection fraction (LVEF) [4-10]. However, since the first description of the obesity paradox by Horwich et al. [11] in 2001, HF treatment and overall prognosis have improved considerably, with four new classes of HF drugs being approved to market since 2015 [2]. Some of these, such as sodium-glucose cotransporter 2 inhibitors (SGLT2i), cause a reduction in body weight and may therefore interfere with the obesity paradox in HF. In the present review, we summarize the current evidence on the obesity paradox in patients with HF with special focus on new HF therapies including sacubitril/valsartan, SGLT2i, vericiguat, and omecamtiv mecarbil.
Methods
We searched PubMed for original articles, reviews, meta-analyses, editorials, comments, and abstracts on the obesity paradox in HF using the search term “obesity paradox (title/abstract) AND heart failure (title).” Records published between January 1, 2002, and February 22, 2022, were considered. In addition, we examined references from the retrieved articles and explored a number of related websites. Overall, more than 400 records were studied. Evidence was synthesized in a narrative review.
Moreover, we performed an extensive literature search regarding the obesity paradox in light of new HF therapies. PubMed was searched for the term {(heart failure [Title]) AND (obesity [Title/Abstract] OR obese [Title/Abstract] OR overweight [Title/Abstract] OR body mass index [Title/Abstract] OR nutrition [Title/Abstract] OR weight [Title/Abstract])} AND (sodium glucose cotransporter 2 inhibitor [Title/Abstract] OR SGLT2i [Title/Abstract] OR SGLT2 inhibitor [Title/Abstract] OR empagliflozin [Title/Abstract] OR dapagliflozin [Title/Abstract] OR canagliflozin [Title/Abstract] OR ertugliflozin [Title/Abstract] OR sotagliflozin [Title/Abstract] OR ARNI [Title/Abstract] OR angiotensin receptor neprilysin inhibitor [Title/Abstract] OR sacubitril [Title/Abstract] OR LCZ696 [Title/Abstract] OR entresto [Title/Abstract] OR vericiguat [Title/Abstract] OR omecamtiv [Title/Abstract]) on February 22. All studies investigating the relationship between body weight and outcomes in patients with HF treated with either SGLT2i, sacubitril/valsartan, vericiguat or omecamtiv mecarbil on top of standard HF therapy were included in the analysis. No restrictions were applied with regard to the time course of HF (acute vs. chronic), LVEF, study size, or publication date.
Results
Obesity and HF
According to the World Health Organization (WHO), obesity is characterized by an excessive fat accumulation that may impair health [13]. In 2015, about 603 million people or 12% of the world’s adult population were obese as defined by a BMI ≥30 kg/m2 [1]. Since 1980, the prevalence of obesity has doubled in more than 70 countries and has continuously increased in most other countries [1]. In the general population, obesity is known as an independent risk factor for numerous chronic diseases including diabetes mellitus, musculoskeletal disorders, and cancer [5, 13-15]. Most importantly, obesity is strongly associated with cardiovascular diseases such as coronary artery disease, atrial fibrillation, hypertension, and HF [5]. Regarding HF, the Framingham Heart Study reported an increase of HF prevalence of 5% in men and 7% in women for every 1-unit increase in BMI after adjustment for other risk factors, and the risk of HF increased across the entire spectrum of BMI [16]. Other anthropometric parameters of excess adiposity such as waist circumference (WC), waist-to-hip ratio (WHR), and waist-to-height ratio have also been independently associated with HF risk [5, 17-19]. In patients with established HF, approximately 29–40% are overweight and 30–49% are obese [17]. Recent work has suggested that obesity is significantly more prevalent in the HF with preserved ejection fraction (HFpEF) population compared to those with reduced ejection fraction (HFrEF), with >80% HFpEF patients being classified as overweight or obese [20, 21].
Obesity contributes directly to the development of HF through effects on the myocardium and vasculature and indirectly through obesity-related comorbidities [5, 22]. Numerous studies have established obesity as a major risk factor for hypertension, cardiovascular disease, left ventricular hypertrophy, metabolic syndrome, dyslipidemia, and obstructive sleep apnea, which too are associated with the development of HF [23, 24]. Hemodynamic changes associated with obesity include higher blood volume and a higher blood pressure as a result of activation of the renin-angiotensin-aldosterone and sympathetic nervous systems. They lead to subsequent increases in cardiac output, cardiac work, lipotoxicity-induced cardiac myocyte injury, and myocardial lipid accumulation [2, 25, 26]. Obesity also directly affects the myocardium with myocardial fat accumulation and subsequent fibrosis that contribute to the development of diastolic dysfunction and HFpEF. In the setting of obesity, HFpEF shows a distinct pathophysiological phenotype that is characterized by greater concentric left ventricular remodeling, right ventricular dilatation, and right ventricular dysfunction [27].
Obesity and HF Outcomes
In 2001, Horwitch et al. [11] first analyzed the relationship between BMI and outcome in a cohort of 1,203 patients with advanced HFrEF. Although obese and overweight study patients had significantly higher rates of hypertension, diabetes, and hyperlipidemia, a higher BMI was not a risk factor for increased 1-year mortality, but was associated with a trend toward improved survival. Subsequent studies have supported the protective effect of a higher BMI on survival [4, 5, 8-10, 28-94]. A systematic review of 35 studies regarding the clinical significance of the nutritional status in patients with chronic HF reported that the risk of all-cause mortality for obese and overweight patients decreased by 24–59% and by 15–65%, respectively. When considering BMI as a continuous variable, morality decreased by 2–16% with increasing BMI [4]. Likewise, a study involving 108,927 patients with acute HF showed a 10% reduction in mortality for every 5-unit increase in BMI [50]. However, other studies point toward a U-shaped rather than a linear relationship between BMI and mortality, with the nadir in mortality rates in class I obesity (BMI 30–35 kg/m2) and the highest risk of death in patients who are normal or underweight [8, 10, 34, 75-80].
While most studies have focused on the relationship between BMI and survival, some studies have explored other outcomes including hospitalizations for HF [10, 32, 35, 47, 75, 95], the composite of death or hospitalization for HF [31, 69, 75, 96], sudden death [83], or HF symptoms [46, 97]. The results are inconclusive and contradictory, with either positive [10, 31, 46, 69, 75], neutral [35, 83, 95-97] or negative [32, 47] associations between adiposity and HF outcomes. On the one hand, a meta-analysis of six studies involving 22,807 patients with chronic HF confirmed the obesity paradox with regard to cardiovascular mortality and hospitalizations [10]. Similarly, a high BMI was associated with a lower rate of the composite outcome of all-cause mortality or HF hospitalizations in a study of 219 Japanese patients with acute HF [69]. On the other hand, studies including patients with acute [47] or chronic [32] HF noted that BMI was independently associated with an increased risk of HF hospitalizations despite the presence of the obesity paradox with respect to survival.
Some authors have criticized the use of BMI to classify obesity due to the inability of BMI to differentiate between fat, muscle, and skeletal weight [24]. However, an obesity paradox has also been proven for other anthropometric measures of adiposity including per cent body fat [9, 30, 87, 88], tricipital skinfold thickness [87, 89], brachial muscle circumference [53, 87], and visceral adiposity index [90]. A study of 209 HF patients reported that every 1% absolute increase in percent body fat was associated with >13% reduction in major clinical events [12]. With regard to WC, most studies support the presence of an obesity paradox in patients with HF [38, 39, 53, 72], although Puig et al. [72] indicate that the protective effect of abdominal adiposity may disappear for extreme values of WC (>120 cm). In a study of 570 Japanese patients with acute HF, brachial muscle circumference, but not WC, played a complementary role to BMI in predicting prognosis, with the best prognosis being observed in the high-BMI/high-arm circumference group [53]. Then again, a post hoc analysis of the TOPCAT trial reported a significantly higher risk of all-cause mortality in patients with HFpEF with abdominal obesity as defined by a WC of ≥102 cm in men and ≥88 cm in women [98]. Similarly, both a high WHR [99] and a high waist-to-height-ratio [67] have been associated with worse outcomes in patients with HF. In a multicenter study of 2,051 patients with HFrEF or HFpEF from 11 Asian countries, a high waist-to-height ratio was associated with a higher risk of 1-year HF hospitalization or death [67]. Moreover, a higher WHR was associated with a higher risk of death in female HF patients from the Scottish BIOSTAT-CHF validation study [99].
The Obesity Paradox across the Spectrum of HF
The obesity paradox has been established in chronic [4, 5, 28, 32-34, 37-44, 66, 68-74, 80, 81, 83] and acutely decompensated HF [30, 35, 36, 45-52, 54, 55, 58-61, 63, 64, 82, 85]. It has been observed in both patients with HFrEF [33-42, 44, 80, 85] and HFpEF [30-34, 77], although a recent analysis of 39,647 patients from the Get With The Guidelines-Heart Failure Registry reported absence of the obesity paradox in patients with HFpEF [85]. In the majority of studies, a higher BMI was associated with better survival irrespective from sex [38, 50, 57, 84] or race [85]. However, other reports noted an obesity paradox in men only [35, 86]. Several studies point out that the presence of an obesity paradox may depend on the etiology of HF [33, 37, 73]. In a multicenter observational study of 5,551 outpatients with chronic HF, the prognostic benefit of obesity was maintained only in nonischemic HF [33]. In contrast, obesity conferred improved prognosis irrespective from etiology in another cohort of 1,160 ambulatory patients with chronic HF [37].
Several studies have evaluated the obesity paradox with special focus on comorbidities [31, 41, 45, 59, 63, 70, 74, 82, 91, 92, 95, 100-102]. A study of 806 Chinese patients hospitalized with acute HF confirmed the obesity paradox in the presence of concomitant atrial fibrillation [91]. In contrast, the prognostic impact of BMI differed depending on the presence of hypertension, chronic obstructive pulmonary disease, and hyponatremia among 4,617 patients with acute HF enrolled in the ATTEND registry [63]. In addition, in an individual patient data analysis of 5,819 patients with chronic HF, the obesity paradox was restricted to patients older than 75 years or having at least one relevant comorbidity [95]. A number of studies have questioned the obesity paradox in patients with HF and diabetes mellitus [41, 59, 70, 74, 100, 101], whereas others noted an obesity paradox irrespective of concomitant diabetes [31, 45, 63, 82, 92]. A recent unpublished study indicates that insulin treatment and poor glycemic control may be important confounders of the obesity paradox in patients with chronic HFrEF and concomitant diabetes mellitus [94].
Several works indicate that the relationship between adiposity and survival in patients with HF may be modified by the degree of cardiorespiratory fitness [44, 81, 103-106]. In an analysis of 4,623 HFrEF patients from the Italian MECKI Score Research Group, the protective impact of BMI disappeared after adjustment for peak oxygen consumption (peakVO2) [106]. Lavie et al. [103] evaluated 2,066 patients with HFrEF who underwent symptom-limited cardiopulmonary exercise testing. They found that BMI was a significant predictor of improved survival in the low fitness group defined as peakVO2 <14 mL/kg*min, but not in the high fitness group. Results were confirmed in subsequent studies including an analysis from the Henry Ford Exercise Training Project, where the benefit of a high BMI was restricted to patients with an exercise capacity less than 4 metabolic equivalents [44, 81, 103-105]. Thus, in the unfit population, obese HF patients have better outcomes, whereas this benefit is lost when compared to groups with preserved fitness. The HF-ACTION trial, a multicenter trial of 2,331 medically stable outpatients with HFrEF randomized to either aerobic exercise training or usual care found nonsignificant reductions in all-cause mortality or hospitalizations and improvements in HF symptoms and quality of life across all BMI categories, suggesting that exercise training is safe and effective [107]. Consequently, regular physical activity receives a class I recommendation from both the American and European HF guidelines [2, 108].
Weight Change in HF
There is convincing evidence that unintentional weight loss is detrimental in patients with chronic HF [109-112]. Current HF guidelines define a >5% oedema-free body weight loss during the previous 12 months or less as the key feature of cardiac cachexia, which is associated with a poor prognosis [2]. According to a recent meta-analysis of eight studies including 226,506 patients with chronic HF, weight loss >5% is associated with a 74% higher mortality risk as compared to stable weight [113]. Then again, weight loss has been shown to reduce the incidence of HF [114]. In addition, intentional weight loss for the obese may improve hemodynamics and cardiac structure, including reductions in circulating blood volume, left ventricular stroke volume, cardiac output, and left ventricular work [115, 116]. Reducing body weight and increasing exercise in obese patients may further improve symptoms and functional capacity [117, 118]. In a small prospective trial including 100 obese older patients with clinically stable HFpEF, caloric restriction and aerobic exercise training had additive beneficial effects on peakVO2 [119].
To date, little is known about the prognostic impact of long-term weight gain in patients with stable HF. An observational study of 92 patients with acute HF reported that weight gain at 6-month follow-up was related to a reduced probability of death or rehospitalization after discharge [120]. In contrast, the ASCEND-HF trial found a worse prognosis in patients with acute HF who gained weight within 10–30 days after hospital discharge [121]. While weight gain shortly after hospitalization for HF is suspicious for recurrent congestion, a long-term moderate weight gain in patients with stable chronic HF may reflect a healthier metabolic state and thus a better prognosis. Recent unpublished data reveal that in patients with chronic stable HF, a gradual weight gain of ≈5% is associated with the best prognosis, whereas mortality steadily increases with increasing weight loss or excessive weight gain >5% [122].
Due to the paucity of randomized controlled trials, current HF guidelines do not provide any firm recommendations on weight management [2, 108, 123]. However, a scientific statement on nutrition, obesity, and cachexia in patients with HF from the HF Society of America advises for 5–10% weight loss in patients with HF and BMI ≥35 kg/m2 [124]. This recommendation is based on small trials with surrogate endpoints only, including incidence of atrial fibrillation, insulin resistance, or left ventricular hypertrophy. However, considering the presumably U-shaped relationship between measures of adiposity and mortality, recommendations for purposeful weight loss in severely obese patients seem reasonable.
Mechanisms for the Obesity Paradox in HF
The mechanisms behind the obesity paradox in HF are somewhat unclear and difficult to reconceal. Some authors suggest that normal weight or BMI does not always correspond to an appropriate nutritional status. Normal weight may result from weight loss throughout chronic HF and therefore reflect malnutrition or cardiac cachexia rather than a healthy state. Cardiac cachexia is known to have a poor prognosis, and obese patients may be less susceptible to cardiac cachexia [23, 125-127]. Moreover, the obesity paradox could be the result of the use of inadequate indicators of fat distribution or the failure to assess physical fitness, regardless of the category of obesity. As pointed out above, literature indicates an important impact of cardiorespiratory fitness over fatness for predicting prognosis [44, 81, 103-106]. Other aspects of the obesity paradox may include a greater metabolic reserve and reduced natriuretic peptide levels in obese patients [17, 23, 128]. Adipose tissue is also known to promote soluble tumor necrosis factor-alpha receptors that could be protective in patients with greater adiposity by neutralizing tumor necrosis factor [129]. Epicardial adipose tissue has been found to be low in patients with HF compared with the general population, and a recent study reported low epicardial adipose tissue in HF to be associated with higher HF mortality [17, 130, 131]. Due to a greater blood volume and higher blood pressure, obese patients tolerate more life-saving guideline-directed medical HF therapy which may improve prognosis [129]. Some authors, however, argue that the observance of an obesity paradox is simply due to research biases, such as lead-time bias [132] and collider bias [133, 134]. In addition, the obesity paradox may be confounded by a younger age, lower natriuretic peptide levels, and less severe HF in the obese population [10, 93, 135]. In a cohort of 1,970 outpatients with chronic systolic HF, the prognostic power of BMI disappeared after adjusting for indicators of disease severity and other confounders [93].
Influence of HF Medications on Body Weight
For almost two decades, β-blockers, renin-angiotensin system (RAS) blockers, and mineralocorticoid receptor antagonists have formed the mainstay of HF therapy. Interestingly, both β-blockers and RAS antagonists have been shown to either prevent weight loss or lead to weight gain [110, 111, 136, 137]. Among 2,289 patients with advanced HFrEF enrolled in the COPERNICUS trial, carvedilol prevented weight loss in patients with increased BMI and it promoted weight gain in those with reduced BMI at baseline (139). In addition, analyses from the SOLVD and Val-HeFT II trials found a ≈20% lower risk of ≥6% weight loss in patients treated with RAS blockers. It has been postulated that β-blockers impede the activation of the sympathetic nervous system, which seems to play an important role in the development of cardiac cachexia [138]. Moreover, improvements of hemodynamics and cardiac function induced by HF therapies may facilitate physical activity and thus lead to an increase in adipose tissue and muscle mass. Mineralocorticoid receptor antagonists have shown a neutral effect on body weight. However, a post hoc analysis from the EMPHASIS-HF trial detected a significant quantitative interaction between the effect of eplerenone and adiposity, whereby the benefit of this treatment was greater in patients with larger WC [139]. No data are available on the effects of either the soluble guanylate cyclase stimulator vericiguat or the myosin activator omecamtiv mecarbil on body weight. In a retrospective observational study of HFrEF patients treated with sacubitril/valsartan, body weight remained unchanged at 3-, 6-, and 12-month follow-up [140]. In contrast, SGLT2i caused a moderate weight loss of 1–3 kg in randomized trials and real-world studies including patients with diabetes or HF [141-143]. The weight loss is mainly due to increased glycosuria of approximately 50–100 g glucose per day which causes a daily caloric deficit [141-144]. In a post hoc analysis of the randomized DAPA-HF trial, dapagliflozin proved to reduce the composite endpoint of worsening HF or cardiovascular death in both HF patients with BMI ≥30 kg/m2 and patients with BMI <30 kg/m2 [143]. To date, no studies have evaluated the effects of SGLT2i or sacubitril/valsartan on body weight in patients with HFpEF.
The Obesity Paradox in Light of New HF Therapies
The literature search focusing on the obesity paradox in the context of new HF therapies yielded 106 results. Of these, three records were considered as relevant, including two original investigations [80, 96] and one editorial comment [144]. Examination of references from the retrieved articles engendered an additional conference abstract of relevance [145]. One original study and one editorial comment discussed the relationship between BMI and outcome in HF patients treated with SGLT2i [80, 144], whereas one original study and a conference abstract reported on the obesity paradox in HF patients receiving sacubitril/valsartan [96, 145]. All reports included HFrEF cohorts, whereas no data are available on the effects of new HF therapies on the obesity paradox in patients with HFpEF. In addition, no studies have been published that address the relationship between body weight and outcome in HF patients treated with either vericiguat or omecamtiv mecarbil.
Adamson et al. [80] examined the effect of the SGLT2i dapagliflozin according to baseline BMI in the DAPA-HF trial. DAPA-HF was a prospective randomized trial that compared the effects of dapagliflozin versus placebo on the composite outcome of worsening HF or cardiovascular death in 4,742 ambulatory patients with chronic symptomatic HFrEF and elevated natriuretic peptides. The authors reported a U-shaped relationship seen with higher risk with both low and very high BMI. The lowest rate of the primary outcome was found in patients of BMI around 30 kg/m2. A similar pattern was seen for the other outcomes of interest including cardiovascular death, worsening HF, and all-cause death. The U-shaped relationship between BMI and outcome was attenuated but not eliminated after adjustment for known prognostic variables including natriuretic peptides. Thus, the obesity paradox persisted in HFrEF patients treated with dapagliflozin. Of note, the treatment effect of dapagliflozin was not affected by baseline BMI. In DAPA-HF, the average weight reduction observed with dapagliflozin by 8 months was rather small (<1 kg). A BMI reduction of 2% in the overall population during follow-up was associated with an increased risk for the composite outcome. However, when dapagliflozin was analyzed separately, the benefits remained apparent despite the weight reduction. This supports that weight loss achieved with SGLT2i is safe [144]. In their editorial comment, Carbone et al. [144] discuss potential mechanisms through which SGLT2i may improve outcomes in HF. They point out that SGLT2i have been shown to improve cardiorespiratory fitness, which is a strong and modifiable risk factor for clinical outcomes in patients with HF [146, 147]. Of note, the study by Adamson et al. [80] does not provide any data on physical activity or cardiorespiratory fitness. As cardiorespiratory fitness seems to outweigh the impact of adiposity on prognosis [44, 81, 103-106], future studies are needed to investigate the complex interrelationship of SGLT2i treatment, body weight, fitness, and outcomes in patients with HF.
Kido et al. [96] investigated the association between BMI and clinical outcomes in a retrospective multicenter cohort study of 721 patients with HFrEF who were receiving sacubitril/valsartan. In unadjusted analyses, there was a trend to increased mortality in normal weight versus obese subjects (HR 1.49, 95% confidence interval: 0.99–2.24, p = 0.06). However, after adjusting for differences in baseline variables between BMI groups by use of propensity score weighting, no significant associations between BMI and all-cause death or hospitalizations for HF were apparent. Similarly, a post hoc analysis from the PARADIGM-HF trial found no evidence of an obesity paradox in 8,399 symptomatic patients with HFrEF who were undergoing treatment with sacubitril/valsartan [145]. Unfortunately, results are only available as a conference abstract at this time without publication of detailed analyses.
Conclusion
Although obesity adversely affects cardiac structure and function leading to an increased risk of cardiovascular disease and HF in the general population, a paradoxical survival benefit with class I obesity has been established in HF. The obesity paradox holds true for women and men with acute or chronic HF across the spectrum of LVEF. Obesity may confer some intrinsic benefits in HF such as a greater metabolic reserve and increased tolerance to guideline-directed HF treatments. However, literature indicates that younger age, cardiorespiratory fitness, and indicators of HF severity are important confounders of the obesity paradox. Consensus has yet to emerge as to whether the obesity paradox is a true phenomenon.
While unintentional weight loss is associated with a poor prognosis in HF, weight loss associated with SGLT2i treatment in patients with HFrEF appears to be safe. To date, prospective data on the effects of intentional weight loss in obese patients with HF are scarce. Interventions should therefore focus on the improvement of physical activity and cardiorespiratory fitness which are strongly associated with prognosis and recommended by international guidelines. In HF patients with severe obesity, a modest weight reduction of 5–10% may be reasonable to improve HF symptoms and quality of life.
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author Contributions
Hanna Fröhlich: literature review and drafting of the manuscript. Tobias Täger, Norbert Frey, and Lutz Frankenstein: review of the manuscript.