The Effects of Bariatric Surgery on Cardiovascular Outcomes and Cardiovascular Mortality: A Systematic Review and Meta-Analysis

Obesity is a major public health problem that is associated with serious comorbidities and premature mortality. Cardiovascular disease (CVD) is the major cause of morbidity and mortality associated with obesity. Lifestyle modifications, pharmacological therapy, and weight reduction surgery are the major interventions to date available for obesity management. Bariatric surgery has been increasingly utilized as a therapeutic option for obesity. In this meta-analysis, we aim to assess the effects of bariatric surgery on CVD outcomes and cardiovascular mortality. This study was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist. PubMed, Embase, Cochrane Library, Google Scholar, and Web of Science were searched until 03/01/2022. Our search included three types of bariatric surgery: Roux-en-Y gastric bypass (RYGB), sleeve gastrectomy, and gastric banding (GB). All were searched in conjunction with “coronary artery disease,” “ischemic heart disease,” “myocardial infarction,” “cerebrovascular accident,” “stroke,” “atrial fibrillation,” “heart failure,” “arrhythmias,” and “mortality.” We included 49 studies meeting the study criteria. Bariatric surgery showed a beneficial effect on coronary artery disease (CAD) (hazard ratio (HR) of 0.68 {95% confidence interval (CI): 0.52-0.91}, p = 0.008), myocardial infarction (MI) (HR of 0.53 {95% CI: 0.44-0.64}, p < 0.01) heart failure (HF) (HR of 0.45 {95% CI: 0.37-0.55}, p < 0.01), cerebrovascular accident (CVA) (HR of 0.68 {95% CI: 0.59-0.78}, p < 0.01), and cardiovascular mortality (HR of 0.48 {95% CI: 0.40-0.57}, p < 0.01). The effect on atrial fibrillation (AF) did not reach statistical significance: HR of 0.81 (95% CI: 0.65-1.01), p = 0.07. Our study, that is, an updated meta-analysis, including the three types of procedure, confirms beneficial effects on the major CVD outcomes, including coronary artery disease, myocardial infarction, cerebrovascular accident, and heart failure, and on CVD mortality. This study provides updated insights into the long-term CV effects of bariatric surgery, an increasingly common intervention for obesity.


Introduction And Background
Obesity is a multifactorial disorder associated with serious complications including diabetes, dyslipidemia, cancer, and cardiovascular disease (CVD) [1,2]. Its prevalence has been uptrending over the last few decades, and it has become a modern-day epidemic [3]. Per the 2013 American Heart Association (AHA)/American College of Cardiology (ACC) guidelines, overweight is defined as a body mass index (BMI) of 25 to <30 kg/m 2 and obesity as a BMI of 30 kg/m 2 [4]. According to the 2017-2018 National Health and Nutrition Examination Survey (NHANES), at least two in five adults (42.4% prevalence) have obesity. This is an increase from the 1999-2000 data with a much lower prevalence of 30.5% [3]. The etiologies leading to obesity could be biological, psychosocial, socioeconomic, and environmental factors [2]. Although unhealthy dietary habits play a major role, racial differences [5] and socioeconomic factors play a major role in the high prevalence of obesity and its complications among minority populations [6]. A higher BMI was strongly associated with higher comorbid cardiovascular risk factors [1]. Of the BMI-related deaths, 41% were notably due to cardiovascular diseases [7].
Obesity is a major contributor to cardiovascular risk factors including hypertension, hyperlipidemia, coronary artery disease (CAD), heart failure (HF), stroke, sleep apnea, and arrhythmias [8]. Its pathogenesis is linked to proinflammatory factors and vessel wall remodeling, among others. Obesity accelerates atherosclerosis by promoting lipid deposition and atherothrombosis formation. It further activates the cytokines and interleukins causing endothelial dysfunction and vascular remodeling [2]. This translates into cardiovascular disease (CVD) events including CAD, myocardial infarction (MI), and stroke. Excess visceral adiposity leads to the activation of renin-angiotensin-aldosterone system, cytokine gene expression, and increased systemic circulation of proatherogenic factors [2,9]. This in turn leads to myocardial fat accumulation, increased stroke volume, cardiac wall remodeling, and fibrosis manifesting as heart failure [2,10]. Similar mechanisms lead to left atrial enlargement and fibrosis contributing to arrhythmogenesis [11].

Statistical Analysis
The meta-analysis was performed with Cochrane's Review Manager (RevMan) version 5.4. Adjusted hazard ratios were considered for the final analysis as the event rates were available for fewer studies. Hazard ratios (HR) were log transformed, and the confidence interval (CI) was used to measure standard error (SE). Genetic inverse variance and random effects model were used to obtain pooled HR and hence study the association between bariatric surgery and cardiovascular outcomes. Heterogeneity was assessed by Cochran's Q statistic and quantified by I 2 index. I 2 values of <50%, 50%-75%, and >75% were considered to have low, moderate, and high heterogeneity, respectively. Publication bias was assessed using funnel plot analysis. A funnel plot was obtained for outcomes involving >10 studies.

Results
Out of the 3982 articles, 49 studies were included for data abstraction. All the included studies were cohort studies, both prospective and retrospective. Some of the studies excluded are the following: (i) studies involving malabsorptive surgery such as biliopancreatic diversion, (ii) studies that looked at outcomes in cohorts having preexisting MI and atrial fibrillation (since this would corroborate our outcome data, they were excluded), (iii) studies that had a high comorbid CVD at baseline, and (iv) studies assessing only allcause mortality. The event rates and the hazard ratios for all the included studies are shown in Table 1  Baseline study characteristics are shown in Table 2. The studies reported a mean age ranging from 32 to 62 and a mean BMI ranging from 37 to 50. All the studies were nonrandomized. Thirty-two of the studies were retrospective cohort studies, and the rest were either prospective or population-based studies.    Sources: [13,17,20,[22][23][24][25][26][27][28][29][30][31][32][33][34][35]40] Like previously mentioned, Bouchard et al. [13] reported a combined incidence and hence was excluded. Johnson et al. [26], Sjöström et al. [23,25], Michaels et al. [31], and Ardissino et al. [20] provided only the incidence data and were not included in the analysis. Naslund et al. [62] studied the outcomes in patients with preexisting MI and was excluded.

Effect on Atrial Fibrillation
Eight studies reported atrial fibrillation outcomes. Seven had adjusted HR data and were included in the analysis ( Figure 5). These studies amounted to a sample size of 18309 in the intervention group and 32933 in the control group. The effect on atrial fibrillation was not significant with a pooled HR of 0.81 (95% CI: 0.65-1.01) (p = 0.07). Lynch et al. [47] provided relative risk data only and hence was excluded. Sources: [18,19,21,35,39,[45][46][47]

Effect on Cardiovascular Mortality
Twenty-six studies reported cardiovascular disease-specific mortality. Fifteen studies had adjusted HR data and were included in the analysis (Figure 7). There were 157750 in the surgery group and 643770 in the control groups. The effect on cardiovascular disease (CVD) mortality was significant with a pooled HR of 0.48 (95% CI: 0.40-0.57) (p < 0.01). The studies showed high heterogeneity with an I 2 = 71%.

Publication bias
Publication bias was assessed for MI, HF, CVA, and CVD. The studies included had a moderate-to-high amount of heterogeneity. This is likely from many smaller studies included leading to effect size variation. This is suggestive of likely publication bias in favor of positive studies. But the funnel plots (shown in Figures 8-11) show the studies being symmetrically scattered around the midline. This is in concordance with the inverted funnel appearance reassuring that there is no publication bias [63].

Discussion
In this updated meta-analysis, we analyzed six major long-term cardiovascular outcomes post-bariatric surgery. Five outcomes including CAD, MI, HF, CVA, and CVD mortality showed a significant risk reduction, whereas atrial fibrillation showed a non-significant risk reduction.

Bariatric Surgery and Atherosclerotic Disease
Obesity poses a high risk for atheroma formation [2]. Bariatric surgery provides a beneficial effect by altering molecular mechanisms involving inflammation. Bariatric surgery decreases the levels of oxidative stress and inflammatory markers [64]. It reduces circulating levels of adhesion molecules and improves endotheliumdependent vasodilatory response [65]. Objectively, several studies have shown that surgery reduces carotid intimal wall thickness in concordance with weight loss [66]. These processes in turn contribute to the risk reduction of atherosclerotic diseases such as CAD, MI, and CVA.
Although CAD and MI are atherosclerotic processes, they differ in their pathophysiology and clinical manifestations. CAD is defined as the presence of atherosclerotic plaque within the epicardial coronary arteries. Over time, risk factors potentiate plaque growth. During periods of myocardial oxygen demand, there is endothelial dysfunction causing plaque rupture. This in turn leads to atherothrombosis, vessel occlusion, and myocardial infarction [67]. Of significance, there was a 29.3% cumulative decrease in MIrelated inpatient deaths and 3.6% cumulative increase in CAD-related inpatient deaths from 2001 to 2014 [68]. It is important to differentiate MI and CAD, as bariatric surgery is protective against both MI and CAD. Hence, we have studied the effects separately.  [69].

Bariatric Surgery and Heart Failure
Bariatric surgery counteracts the effects of obesity on the heart, as described previously. Although there are no randomized controlled trials to show this effect on heart failure, few observational studies have been conducted. The mechanism by which this occurs could be multifactorial. Bariatric surgery reduces heart failure risk factors including hypertension, hyperlipidemia, and diabetes [51]. It also directly acts on the myocardium causing changes in the left ventricle (LV) wall and ejection fraction (EF) percentage. Vest et al. showed that bariatric surgery improved left ventricular systolic dysfunction and resulted in a statistically significant improvement in left ventricle ejection fraction (LVEF) [71]. Another study showed a 43% reduction in left ventricular mass with subsequent reduction in left atrial and right ventricular wall diameter and epicardial fat [72]. A meta-analysis done by Cuspidi et al. showed significant changes in LV thickness, improvement in LV diastolic function, and a decrease in left atrial diameter [73]. Cuspidi et al. also showed no significant improvement of EF percentage [73]. The pooled HR for HF in our study was 0.45 (95% CI: 0.37-0.55) from 15 studies. This is consistent with a prior similar meta-analysis [70,74].

Bariatric Surgery and Cardiovascular Mortality
Scandinavian countries have the most comprehensive obesity registries with a long-term follow-up [14,25,35,43,57]. The data from these have provided significant insight into the long-term outcomes after bariatric surgeries. Carlsson et al. followed 2007 patients over a mean of 24 years and found 457 deaths, of which 167 were from cardiovascular causes, the most common cardiovascular cause of death being myocardial infarction, heart failure, and sudden death [14]. Kauppila [75].
Given the significant cardiovascular benefits offered by bariatric surgery, the referral from primary care physicians has been lower. This could be attributed to knowledge gaps, hesitancy, or concerns regarding postoperative care. A recent Canadian survey showed that more than 50% of physician respondents did not feel equipped to counsel the patients on surgical options. And only 11.6% of the obese patients were being counselled [76]. In a Swedish survey, interestingly, 84% of respondents stated that the patients themselves initiated bariatric surgery referral [77]. Physician's knowledge showed a positive correlation toward referral and management of postoperative issues [77]. This brings into perspective that education and awareness would lead to better patient sampling, thereby cumulatively improving cardiovascular outcomes.

Limitations
Firstly, the studies included are all nonrandomized cohort studies, which could involve selection and publication biases. Henceforth, longer randomized controlled trials are required. Secondly, most of the outcomes had high heterogeneity, which could be owed to the many smaller studies that were included. Thirdly, some studies had non-generalizable populations such as type 1 diabetes or type 2 diabetes specifically. However, we omitted populations that had cardiovascular diseases at baseline. Fourthly, only English studies were included owing to the ease of interpretation and analysis. Lastly, we failed to study the HR specific to each bariatric surgery, likely due to the scarcity of data for a pooled analysis.

Conclusions
Although the management of obesity requires a multimodal approach, recognizing the necessity for bariatric surgery early in the disease course is important. Both the physician and the patients should be aware of the treatment strategies to make a well-informed decision. Our study is an updated meta-analysis highlighting the consistency with the prior data. We included additional studies to provide more comprehensive data on six major cardiovascular outcomes. In conclusion, bariatric surgery showed a statistically significant risk reduction withCAD, MI, HF, CVA, and cardiovascular disease-specific mortality and a non-significant risk reduction of atrial fibrillation. However, these data are inclusive of RYGB, SG, and laparoscopic banding. Further research needs to be conducted to determine if these individual procedures have better overall outcomes than one another.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.