Metabolic syndrome is defined by an increase in any three of the following parameters: waist circumference, blood pressure, blood glucose level, or triglyceride level, or high-density lipoprotein cholesterol concentration (6). The interplay among these features is difficult to sort out, as some are seen in one patient but not the next. However, in all the leading theories, insulin resistance is a core component. Metabolic syndrome is particularly important in the context of COVID-19, as it causes a three times higher chance of death, and a four to five times higher probability of invasive mechanical ventilation, acute respiratory distress syndrome, and admission to the intensive care unit. Obesity may also impair an immune response to the COVID-19 vaccine (3, 7). In addition, compared to COVID-19 patients with normal glucose or with diabetes, those with hyperglycemia at hospital admission had higher mortality and lower PaO2/FiO2 (8).

To address the significance of insulin resistance, we discuss a recent study by Yoshino et al., in which the authors concluded that weight loss, whether induced by bariatric surgery, or by strict dieting alone, was key to restoring insulin sensitivity (9). In this matched prospective cohort study, obese individuals with type 2 diabetes underwent either Roux-en-Y gastric bypass or were placed on a diet to achieve 16–24% weight loss. Weight loss in both groups increased hepatic insulin sensitivity as indicated by increased suppression of glucose production and increased glucose disposal, with no significant difference between groups (9). The report by Yoshino et al. suggests, at first glance, that the mechanism of weight loss does not impact the outcomes concerning insulin sensitivity. However, this study has considerable limitations, and the results must be interpreted with caution. As the authors mentioned, the subjects were not randomized, and the sample size may have been inadequate. We also noted that two-thirds of the participants were women, whose metabolic profiles differ from men with respect to lipid metabolism, and to their predisposition for central obesity. Most importantly, both groups, the surgical and the non-surgical group, were placed on diets to achieve weight loss, suggesting that the lack of between-group differences may be due to diet and weight loss combined. Several studies also suggest that various metabolic improvements, such as decreased visceral adipose tissue, increased insulin sensitivity, and increased levels of high-density lipoprotein cholesterol occur in the absence of weight loss (10). Given these confounders, it is difficult to determine whether other, more important, factors are at play.

Another confusing component of disrupted fuel homeostasis in metabolic syndrome is that there are two groups of outliers in the general population: “metabolically healthy” obese individuals and “metabolically obese” normal weight individuals. Metabolically obese normal weight individuals (24% of normal weight individuals) are primarily a result of genetics and lifestyle factors, whereas metabolically healthy obese individuals (32% of obese individuals) are impacted by a multitude of factors, such as inflammation of adipose tissue (11, 12). In both groups, poor metabolic “fitness” is strongly correlated with increased hepatic fat stores, which are likely a result of de novo lipid synthesis, increased uptake, and decreased disposal (13). These individuals with disrupted fuel homeostasis complicate the hypothesis that weight loss alone is the mechanism of restoring insulin sensitivity (14).

Additionally, considerable differences in metabolic risk factors in obese individuals (BMI 30–40 kg/m²) exist, depending primarily on ectopic fat and lifestyle factors, such as diet and exercise (4, 13). In a landmark paper by Reaven et al., the authors noted an association between insulin resistance and upregulation of the sympathetic nervous system, even suggesting a potential causal relationship (6). An increased adrenergic response may be the mechanism behind other components of metabolic syndrome, such as increased non-estrified long-chain fatty acids (FAs) in the plasma and arterial hypertension. Indeed, recent data have demonstrated a bidirectional relationship of leptin with the sympathetic nervous system and thus, energy storage and mobilization (15). Furthermore, lifestyle changes which decrease insulin resistance and sympathetic nervous system activity, including exercise and caloric control, appear to protect against the metabolic syndrome. It is not proven, however, that these lifestyle modifications are exclusively responsible for the mechanism to restore metabolic normalcy, and the genetic profile likely plays a role as well (6, 12).

The most common bariatric surgery procedures performed in the United States are currently sleeve gastrectomy (61%), and RYGB (17%), with LAGB and biliopancreatic diversion (2% together) falling out of favor (23). Novel gastrointestinal surgeries that target diabetes, such as duodenal-jejunal bypass and ileal interposition, provide helpful clues to the mechanism of restored insulin sensitivity following bariatric surgery (5). Briefly, RYGB excludes most of the stomach just distal to the gastroesophageal junction (GEJ). It bypasses the entire duodenum and part of the jejunum as the 30 cc stomach pouch attached to the esophagus is anastomosed to the distal jejunum. Sleeve gastrectomy removes a significant portion (60–80%) of the stomach along the greater curvature, and adjustable gastric banding is an inflatable/adjustable silicone band placed around the superior portion of the stomach, just distal to the GEJ. Biliopancreatic diversion removes the stomach just as in gastric sleeve, which is anastomosed to the distal portion of the ileum. The excluded duodenum, jejunum, and proximal ileum are anastomosed to the alimentary limb, allowing biliary and pancreatic secretions to mix with ingested nutrients (5).

The current standard of care in the United States is to allow the patient to choose among these procedures, and patients may find the following information helpful in the decision (Table 1) (23, 24). After 5 years, 35% of LAGB patients had failed to lose weight compared to 4% of RYGB patients, and metabolic parameters were inferior in LAGB patients after 2 years (25). In general, bypass surgeries have caused more durable weight loss and better insulin control than restrictive surgeries, or than lifestyle management (5, 23). Long-term improvements in dyslipidemia and hypertension have been less consistent across studies. Of note, RYGB has higher complication rates than sleeve gastrectomy, including a higher rate of reoperation, perioperative mortality, and readmission (23). However, overall perioperative mortality remains low with RYGB (around 0.2%), and adverse events occur in 1–9% of patients undergoing RYGB (23).

There are several findings that indicate that bariatric surgery has benefits independent of weight loss. There is a rapid resolution of type 2 diabetes within days following bariatric surgery that persists at one and five years (4, 5). Another important finding is a remarkably higher rate of type 2 diabetes remission at 2 years with malabsorptive RYGB (72%) compared to restrictive LAGB (17%) in patients who had lost approximately 30% of their body weight (26).

In addition to remission of diabetes, subjects in a non-randomized, prospective cohort study showed decreased glucose, insulin, and leptin levels within 3 months of surgery, and glucagon-like peptide 1 (GLP-1), glucose-sensitive insulinotropic peptide (GIP), peptide-YY, while pancreatic polypeptide concentrations increased within days (4, 5, 24), On the contrary, fasting did not affect high molecular weight adiponectin, ghrelin, leptin, or GLP-1, suggesting unique effects of surgery compared to diet (20). One explanation for the rapid change in disrupted fuel homeostasis may be that surgery excises or interrupts regions of the digestive system involved in some of the pathologic adaptations of fuel excess. Additional support for these unique benefits of surgery include the nearly identical skeletal muscle gene expression of diabetogenic hormones, such as stearoyl-CoA desaturase (SCD), pyruvate dehydrogenase kinase 4 (PDK-4), and peroxisome proliferator-activated receptor alpha (PPAR-α) among participants who underwent LAGB vs. RYGB despite significant variation in weight loss between the two groups (222% greater mean weight loss at 9 months in RYGB participants compared to LAGB participants) (24). However, lengthening of the intestinal bypass by RYGB does not affect GLP-1 secretion, which suggests that the GLP-1 response after RYGB may not require delivery of nutrients to more distal intestinal segments (27). It is not known whether the metabolic effects of RYGB persist long-term, but the duration of surgery's impact may provide clues as to the mechanism of restored fuel homeostasis. For example, return of a disruption in fuel homeostasis over time may suggest that surgery merely interrupts pathologic metabolic adaptations, whereas permanent improvement in metabolic health may indicate that surgery removes an element of the metabolic pathway essential to disrupted fuel homeostasis.

Given these important mediators, we postulate that there are intermediate steps between diet and surgery and increased insulin sensitivity (Figure 2). These interposed steps may include incretins, appetite-suppressing hormones, known metabolic risk factors, such as VAT, catecholamines, and unknown upper gastrointestinal effects that impact the regulation of blood glucose and lipogenesis. A possible mechanism for these hormonal changes may be related to the impressive reduction of oil-red-O staining in skeletal muscle post-operatively, which may occur in other organs, dramatically impacting the neuroendocrine signaling between adipose tissue and other organs (24). These findings also support the observation that ectopic fat, such as in skeletal muscle, is a key component of global metabolic dysfunction throughout the body (13).

Additional mechanisms may include surgical removal of the gastric fundus with sleeve gastrectomy, which is the site of ghrelin secretion, or hyperstimulation of appetite-suppressing hormone production by increased nutrient delivery to distal portions of the gastrointestinal tract, such as with RYGB. Antidiabetic effects are also observed in patients with decreased nutrient delivery to the upper gastrointestinal tract, possibly due to anti-incretin factors released from the proximal bowel (4, 26, 28).

Obesity affects the cardiovascular system at many different levels. In one study, 42% of obese participants had left ventricular diastolic dysfunction on tissue Doppler imaging (14, 24). Compared to non-obese patients with heart failure with preserved ejection fraction (HFpEF), obese patients with HFpEF had greater volume overload and right ventricular dysfunction, and decreased levels of brain natriuretic peptide and cardiac efficiency (4, 29). Obese patients also have an approximately 50% higher prevalence of atrial fibrillation than normal-weight patients, but the reversibility of arrhythmias following bariatric surgery is still unknown (4). These obesity-related abnormalities may be related to cardiac fibrosis, hypertrophy, and impaired microvascular coronary perfusion (in hearts of diabetic patients), and the cardiovascular impact of COVID-19 has the potential to further compound these problems (5).

Given the detrimental cardiac effects of obesity, pharmacologic agents have been tested as treatment options without success. We have previously proposed that obesity related cardiac dysfunction is likely a result of fuel excess, which results in production of reactive oxygen species (ROS). The overwhelmed heart may protect itself from further damage with insulin resistance in response to ROS. As a result, abruptly inducing insulin sensitivity overrides this protective response, resulting in cytotoxic damage (30). Worsening cardiac dysfunction with thiazolidinedione (TZD) use in a diabetic heart supports this proposition, as TZDs contribute to fuel overload by increasing glucose uptake and oxidation. Alternatively, metformin has protective effects on the heart by enhancing peripheral glucose oxidation and decreasing FAs (31).

In this context, it is important to emphasize that bariatric surgery-induced weight loss is a feasible solution to improve metabolic abnormalities, as well as cardiac function. Surgery decreased the incidence of major adverse cardiovascular events by more than 40% among a matched cohort of 2,600 obese patients with cardiac disease and by more than 55% among those with heart failure (32, 33). Surgery also decreases the use of antihypertensive drugs (33%), and diet induces a 0.43 mmHg decrease in systolic blood pressure per 1% decrease in body mass index (10, 20, 24). Interestingly, blood pressure is one of the first parameters to change with exercise before any meaningful weight loss occurs (10). Tissue Doppler diastolic velocity also increased by 1.9 cm/sec (95% CI 0.52–3.4) and 1.2 cm/sec (95% CI 0.32–2.1) at 3 and 9 months, respectively, along with improved septal mitral annular velocity (24). Lastly, left ventricular mass decreased linearly for 2 years post-operatively even as weight loss plateaued (14). Correction of metabolism by non-pharmacologic means is the most beneficial way to restore cardiac function.