Surgical Endoscopy

, Volume 25, Issue 8, pp 2650–2659

Early effects of gastric bypass on endothelial function, inflammation, and cardiovascular risk in obese patients

Authors

    • Bariatric and Metabolic Institute, M61, Cleveland Clinic
  • Helen M. Heneghan
    • Bariatric and Metabolic Institute, M61, Cleveland Clinic
  • Shai Eldar
    • Bariatric and Metabolic Institute, M61, Cleveland Clinic
  • Patrick Gatmaitan
    • Bariatric and Metabolic Institute, M61, Cleveland Clinic
  • Hazel Huang
    • Department of PathobiologyLerner Research Institute, Cleveland Clinic
  • Sangeeta Kashyap
    • Endocrinology and Metabolism Institute
  • Heather L. Gornik
    • Heart and Vascular Institute
  • John P. Kirwan
    • Department of PathobiologyLerner Research Institute, Cleveland Clinic
  • Philip R. Schauer
    • Bariatric and Metabolic Institute, M61, Cleveland Clinic
Article

DOI: 10.1007/s00464-011-1620-6

Cite this article as:
Brethauer, S.A., Heneghan, H.M., Eldar, S. et al. Surg Endosc (2011) 25: 2650. doi:10.1007/s00464-011-1620-6

Abstract

Background

Obesity is associated with a chronic low-grade inflammatory state, insulin resistance, and endothelial dysfunction, all of which contribute to increased risk of cardiovascular disease. We hypothesized that gastric bypass would produce rapid improvements in endothelial function, reduce inflammation, and lead to a decrease in cardiovascular risk.

Methods

We performed a prospective study of morbidly obese patients who underwent laparoscopic Roux-en-Y gastric bypass (RYGB). Clinical data, biochemical markers of inflammation, and parameters indicative of cardiovascular risk were collected preoperatively and at 3 and 6 months postoperatively. Metabolic and inflammatory mediators that were quantified included C-reactive protein, fibrinogen, PAI-1, IL-6, IL-10, IL-1Ra, adiponectin, leptin, triglycerides, total cholesterol, HDL, LDL, glucose, insulin, and HbA1c. Brachial artery reactivity testing (BART) was performed to assess peripheral arterial endothelial function, and Framingham cardiovascular risk score (FRS) was calculated on all study participants pre- and postoperatively.

Results

Fifteen patients (11 female) were enrolled (age = 49.2 ± 10.4 years; BMI = 48.1 ± 5.3 kg/m2). Six months post RYGB, mean BMI decreased to 35.4 ± 4.5, corresponding to 51.7% excess weight loss (P < 0.001). Mean waist circumference decreased significantly from 132 cm at baseline to 110 cm at 3 months (P = 0.003) and 107 cm at 6 months (P < 0.001). Six months after RYGB, weight loss led to significant improvements in clinical parameters indicative of cardiovascular disease or risk, including brachial artery diameter, endothelial independent vasodilation, and FRS. Favorable improvements in the proinflammatory markers CRP (P = 0.01) and leptin (P = 0.005), the anti-inflammatory mediator adiponectin (P = 0.002), and insulin sensitivity (HOMA-IR, P = 0.007) were evident at 3 months. At 6 months, improvements in CRP, leptin, and fasting insulin were maintained and fibrinogen levels also decreased (P = 0.047). Adiponectin continued to increase at 6 months (P = 0.004).

Conclusion

Gastric bypass is associated with early reversal of endothelial dysfunction, a more favorable inflammatory milieu, and, most importantly, a reduction in cardiovascular risk.

Keywords

Gastric bypassInflammationCardiovascularBARTEndothelial

Obesity is now a global epidemic of major public health concern; at present approximately 35% of US adults are obese (body mass index [BMI] ≥ 30 kg/m2), with an annual increase in the proportion categorized as morbid (BMI > 40) or superobese (BMI > 50) [1]. Obesity contributes significantly to global morbidity, mortality, and socioeconomic burden; the US Centers for Disease Control estimate that there are almost 300,000 premature deaths per year in the US due to this disease. This accounts for more deaths than lung, breast, colon, and prostate cancers combined [2]. A plethora of comorbidities are associated with obesity, including dyslipidemia, hypertension, insulin resistance, diabetes mellitus, coronary artery disease, venous thromboembolism, obstructive sleep apnea, and various malignancies. Notably, there is a positive correlation between increasing severity of obesity and propensity for developing these comorbid processes [3, 4]. Furthermore, all-cause mortality rates and years of life lost due to obesity are significantly greater among the morbidly obese [5, 6]. Understanding the link between obesity and this diverse array of comorbid processes has been the focus of clinicians and scientists in recent years, as efforts intensify to unravel the mechanisms linking increased BMI with these diseases. In particular, the pathophysiology underlying obesity-induced atherosclerotic disease and diabetes mellitus has been scrutinized, given their considerable morbidity and the prevalence of these conditions among obese individuals [79].

The adipocyte is no longer considered a passive storage cell. An abundance of evidence exists to demonstrate the role of adipose tissue as an active endocrine and paracrine organ that secretes numerous hormones, peptides, and other molecules that affect metabolism, vascular function, and glucose homeostasis [10, 11]. However, the precise cellular mechanisms explaining the pathophysiology of adipose tissue and its association with the comorbid conditions of obesity are not entirely understood. It is postulated that a chronic state of low-grade inflammation, created by proinflammatory cytokines and peptides secreted by adipose tissue, mediates the link between obesity and both cardiovascular disease and insulin resistance [12, 13]. C-reactive protein (CRP), an acute-phase reactant produced mainly in the liver in response to IL-6, is also a reliable marker of current inflammatory status [13, 14]. Given that IL-6 is secreted by adipose tissue and levels are increased in obesity, it is not surprising then that circulating CRP levels have also been shown to correlate with BMI, the degree of insulin sensitivity, endothelial dysfunction, and cardiovascular risk [13, 1517]. Other markers of inflammation [e.g., fibrinogen, plasminogen activation inhibitory-1 (PAI-1)] have also been found to be predictive of cardiovascular disease in healthy and symptomatic men and women [1821]. Presently, there are limited studies reporting on anti-inflammatory mediators such as IL-10 and adiponectin in obese patients and conflicting data exist regarding their association with clinicopathological parameters [22, 23].

There are many parallels between the pathophysiology of type 2 diabetes and obesity; adult onset diabetes also appears to be associated with a chronic low-grade inflammatory state in which there are increased serum levels of cytokines secreted from various cells, including adipocytes [24, 25]. Several studies have shown that weight loss after bariatric procedures leads to a decrease in levels of inflammatory biomarkers, including CRP, with concomitant improvement in insulin resistance and endothelial function [13, 2628]. However, diverging data exist regarding alterations in serum levels of Il-6 after weight loss; Monzillo et al. [29] describe significant decreases in IL-6 levels in morbidly obese patients following weight loss. These findings have not been consistently reproducible [13, 30].

There are few well-designed prospective studies, however, that evaluate the global effect of massive weight loss after laparoscopic gastric bypass on markers of inflammation, diabetes, endothelial function, and cardiovascular disease status [30]. The purpose of this pilot study was to evaluate changes in inflammatory mediators in response to laparoscopic Roux-en-Y gastric bypass (RYGB) in severely obese patients and to correlate postoperative weight loss to changes in inflammatory markers, diabetes, and cardiovascular disease risk parameters.

Methods

Study cohort

This prospective longitudinal study was approved by the Cleveland Clinic Institutional Review Board and written informed consent was obtained from all participants. Our study cohort consisted of 15 morbidly obese patients who underwent laparoscopic RYGB, and all met the criteria for bariatric surgery as outlined by the National Institutes of Health Consensus Development Panel report of 1991 [32]. Eligibility for participation in this study was limited to individuals between the ages of 18 and 75 years with a BMI of ≥40 kg/m2, or a BMI ≥35 kg/m2 with comorbid conditions such as diabetes mellitus, obstructive sleep apnea, and hypertension. Type 2 diabetes (T2DM) was defined as a fasting blood glucose ≥126 mg/dL or HbA1c ≥ 6.5% or the need for pharmacotherapy to treat T2DM at the time of initial evaluation in our program. Hypertension was defined as the need for antihypertensive medication or an untreated blood pressure >140/90 mm Hg on repeated measurements. Other comorbidities were considered present preoperatively if patients met standard diagnostic criteria or were receiving therapy for a particular condition. All patients underwent an extensive preoperative evaluation, including history and physical examination, nutritional and psychiatric assessments, consultation with internal medicine and anesthesiologists, in addition to mandatory attendance at a bariatric informational workshop. Failure to comply with the required preoperative workup disqualified a patient from participation in the study. Additional consultations with medical subspecialists (endocrinology, cardiology, pulmonary medicine) were obtained as clinically indicated. Individuals with known autoimmune disease, cancer, thrombotic disorders, and valvular heart disease were excluded, as were those who were unable or unwilling to cooperate with postoperative follow-up.

Study protocol

Consenting participants were evaluated at three time points for the purpose of this study: preoperatively and at 3 and 6 months postoperatively. Each evaluation consisted of a review of systems and physical exam, anthropometric measurements, cardiovascular risk factor assessment, and blood sampling for measurement of biochemical, metabolic, and inflammatory biomarkers. Vascular reactivity was measured as a surrogate marker for endothelial function by means of a brachial artery reactivity test at baseline and at the final (6 month) assessment [33, 34]. The brachial reactivity testing protocol included assessment of both the vasodilator response to reactive hyperemia (endothelium-dependent vasodilation) and the vasodilator response to sublingual nitroglycerin (endothelium-independent vasodilation) using standard methodology [35].

Analytic determinations

Fasting venous blood samples (40 ml at each time point) were collected from each participant for measurement of biochemical (hepatic enzymes AST and ALT), metabolic (glucose, HbA1C, lipid profile, insulin, leptin, adiponectin), and inflammatory biomarkers (C-reactive protein, fibrinogen, albumin, IL-6, PAI-1, IL-10, IL-1Ra). Blood samples were collected in serum, EDTA, Citrate, and SCAT-1 tubes as appropriate for specific analyte measurements and processed immediately according to protocols established in our institution’s Clinical Research Center, and serum/plasma aliquots were stored at −70°C until analyzed. Quantitation of inflammatory markers and adipokines (adiponectin, insulin, and leptin) was by means of a high-sensitivity human cytokine multiplex kit and human gut hormone LINCOplex kit (LINCOplex; LINCO Research, St. Charles, MO) using a Luminex instrument (Luminex 100; Bio-Rad Laboratories, Inc., Hercules, CA), which facilitated the simultaneous evaluation of multiple immune and endocrine mediators. To correct for interassay variability, all preoperative, 3-, and 6-month measurements for each individual subject were run on the same plate for each kit. Serum CRP concentration was determined in duplicate with a high-sensitivity sandwich enzyme-linked immunosorbent assay kit (ALPCO Diagnostics, Salem, NH) using a microtiter plate reader (ELx808; Bio-Tek Instruments, Inc., Winooski, VT). Routine biochemistry, lipid profiling (total cholesterol, HDL and LDL cholesterol, and triglycerides), glucose, and HbA1C were measured in the certified core clinical laboratory. Insulin resistance was estimated from fasting plasma glucose and insulin levels using the previously validated homeostasis model assessment of insulin resistance [HOMA-IR; fasting serum insulin (μU/ml) × fasting plasma glucose (mg/dl)/405] [36].

Cardiovascular risk assessment

Cardiovascular risk was determined at baseline and at both postoperative time points using the gender-specific Framingham Coronary Heart Disease risk score developed by Wilson et al. [37]. This algorithm determines an individual’s 10-year risk of coronary heart disease based on categorical values, including age, total cholesterol, HDL cholesterol, LDL cholesterol, smoking status, blood pressure, and diabetes history.

Brachial artery reactivity testing (BART)

Impairment of endothelial function is a recognized early component of atherosclerotic cardiovascular disease and can be characterized by noninvasive means by vasodilation of a conduit artery, typically the brachial artery, using high-resolution ultrasound (brachial artery reactivity testing or BART) [35, 38]. Assessment of endothelial function using brachial artery reactivity testing includes assessment of both endothelium-dependent vasodilation (flow-mediated vasodilation or FMD) and endothelium-independent vasodilation (vasodilator response to nitroglycerin). FMD of the brachial artery is measured by comparing the diameter of the brachial artery at rest to that after reactive hyperemia leads to nitric oxide-mediated vasodilation [35]. Brachial artery FMD correlates well with measures of coronary artery endothelial function and has been shown to be an independent predictor of cardiovascular risk and future adverse cardiovascular events [35, 39, 40]. Brachial artery reactivity testing flow-mediated vasodilation is a frequently used noninvasive ultrasonographic assessment of FMD and is indicative of endothelium-dependent response to shear stress; thus, it is considered a surrogate marker of endothelial function. Blood pressure cuff occlusion of the brachial artery and subsequent shear stress produced by hyperemia upon cuff release provides a stimulus for release of nitric oxide from the endothelium that leads to vasodilation. A normal response of the brachial artery in a healthy volunteer to reactive hyperemia is 10–15% vasodilation; impaired relaxation suggests subclinical atherosclerotic disease as has been demonstrated in patients with coronary risk factors and established coronary heart disease [41]. In this study, BART was performed on all patients in accordance with guidelines issued by the International Brachial Artery Reactivity Task Force [35], at baseline and repeated 6 months postoperatively. In preparation for the test, patients were required to fast and abstain from caffeine and smoking for 12 h prior to the test. BART testing was performed under standardized ambient conditions in a quiet room. The test was initiated by placing a blood pressure cuff around the forearm and positioning an ultrasound probe over the brachial artery for baseline images to be obtained using high-resolution B mode ultrasound (Seimens Sequoia, 8 l/8 or 5 MHz linear array transducer). The blood pressure cuff was then inflated to 200 mmHg (or 40 mmHg above systolic blood pressure) and maintained for 5 min. Hyperemic blood flow was measured 10 s after cuff release (at an optimal Doppler angle of 60°). Vessel diameter was measured 1 min after cuff deflation with a series of ECG-gated images (55–65 s). After a rest period of 15 min, repeat baseline images were obtained. The patient was then given 0.4 mg sublingual nitroglycerin (to assess endothelium-independent vasodilation) and a final series of ECG-gated brachial artery images were obtained 3 min after nitrate administration to measure endothelium-independent vasodilation. All brachial artery diameter measurements (baseline, post FMD, repeat baseline, post nitroglycerin) were analyzed by a single investigator (HG) using specialized edge-detection software (Brachial Analyzer, Vascular Research Tools, version 5.0.4, Medical Imaging Applications, LLC, Coralville, IA).

Surgical procedure

Laparoscopic Roux-en-Y gastric bypass (RYGB) was performed on all patients in this study. With the patient supine and under general anesthesia, the abdomen was entered with an optical trocar and pneumoperitoneum established. In addition, three 12-mm and one (or two, as required) 5-mm ports were inserted in standard fashion. A small 15-ml proximal gastric pouch was created with several firings of a linear stapler. The jejunum was divided with a linear stapler 50 cm distal to the ligament of Treitz and a 150-cm Roux limb measured. The jejunojejunostomy was created with a linear stapler and the Roux limb brought up in the antecolic, antegastric position to the gastric pouch. A 15-mm linear stapled gastrojejunostomy was created and then oversewn. Prior to closure a 15-Fr drain was placed posterior to the gastrojejunal anastomosis in the left upper quadrant. The drain remained in place for 1 week. Postoperatively, patients routinely underwent an upper GI water-soluble-contrast study on postoperative day 1 and commenced on a clear liquid diet after confirmation of an intact anastomosis without evidence of leak or obstruction. The mean length of stay was 3 days.

Statistical analysis

Data were analyzed using the software package SPSS 17.0 for Windows. Descriptive statistics were computed for all variables. Distribution of the data was checked for normality using the Kolmogorov–Smirnov test; parametric data are presented as mean ± standard deviation and analyzed using Student’s two-sample t test for any two-sample comparisons and ANOVA, followed by Tukey HSD post hoc test, where appropriate. All nonparametric data were analyzed using Mann–Whitney tests or analysis of variance based on ranks. Differences between proportions and categorical variables were determined using the χ2 test. All tests were two-tailed and results with a P < 0.05 were considered statistically significant.

Results

Patient demographics

Fifteen patients were enrolled in this study and all subjects adhered to the protocol and completed follow-up. The main demographic and clinical characteristics of the patients are described in Table 1. In this cohort, the mean preoperative BMI was 48.9 ± 5.8 kg/m2 and 46.7% of the group (n = 7) were “superobese” (BMI > 50 kg/m2). As expected, there was a high prevalence of obesity-associated comorbidities among these patients at the time of study enrollment; hyperlipidemia and hypertension were the most commonly encountered medical conditions (80 and 73.3% respectively), while type 2 diabetes mellitus was present in 20% (n = 3) and obstructive sleep apnea in 46% (n = 7). Overall, every patient (n = 15) had at least one comorbidity confirmed or newly diagnosed during their preoperative assessment.
Table 1

Demographic and metabolic characteristics of study group [mean (SD)]

 

Gastric bypass patients (n = 15)

Male/female

36%/64%

Age (years)

49.2 (10.4)

BMI (kg/m2)

48.7 (5.8)

Comorbidities

 Diabetes mellitus type II

20%

 Hypertension

73.3%

 Hyperlipidemia

80%

 Obstructive sleep apnea

46.7%

Waist circumference (cm)

132.98 (12.5)

Hip circumference (cm)

144.85 (13.1)

Blood pressure, baseline (mm HG)

140/81 (12.5/10.7)

Surgical outcomes: weight loss and resolution of comorbidities

At 6 months postoperatively, the mean BMI had decreased significantly from 48.9 ± 5.8 to 35.4 ± 4.5 kg/m2, corresponding to 51.7% excess weight loss on average (P < 0.001). The mean waist circumference decreased significantly from 132 to 110 cm at 3 months (P = 0.003) and to 107 cm at 6 months (P < 0.001). Comorbidity improvement or remission at 6 months, which was determined by review of clinical findings, medication usage, and standard biochemical markers, was documented in 86.7% (n = 13; 7 patients had remission and 6 had improvement of their comorbidities). There were no complications in this group during the 6-month follow-up period and there were no mortalities.

Inflammatory markers and insulin resistance

At 3 months, there was a significant decrease in levels of the proinflammatory markers CRP (P < 0.001), PAI-1 (P < 0.001), and leptin (P = 0.005). Circulating insulin levels had also decreased significantly by 3 months post gastric bypass (P = 0.004). At 6 months, levels of the hormones leptin and insulin remained significantly decreased (P = 0.005 and P = 0.006, respectively), as did levels of the proinflammatory markers CRP and PAI-1 (P < 0.001 and P < 0.001, respectively); fibrinogen levels had also decreased significantly (P = 0.0028). IL-6 levels did not change significantly at any time point postoperatively. The alterations in metabolic parameters of obese individuals after RYGB are summarized in Table 2.
Table 2

Alterations in metabolic parameters of obese individuals after RYGB, mean (SD)

 

Baseline

3 Months post RYGB

6 Months post RYGB

P Value (ANOVA)*

BMI

48.9 (5.8)

38.3 (4.6)

35.4 (4.5)

<0.001

%EWL

38.5 (6.2)

51.7 (8.8)

<0.001

Blood pressure

 Systolic

140 (12)

133 (27)

130 (22)

0.428

 Diastolic

81 (10)

81 (10)

78 (10)

0.665

Fasting glucose

106.7 (43.1)

88.4 (28.1)

84.2 (18.8)

0.150

HbA1C

6.2 (1.5)

5.5 (0.9)

5.4 (0.8)

0.152

Lipid profile

 Total cholesterol

200.2 (47.7)

168.5 (38.6)

163.0 (45.0)

0.062

 LDL

124.5 (38.3)

98.5 (27.1)

88.3 (27.6)

0.012

 Triglycerides

146.1 (64.5)

114.0 (48.5)

105.5 (47.4)

0.124

 HDL

46.4 (17.6)

47.2 (14.3)

53.7 (22.8)

0.533

Hepatic enzymes

 AST

31.9 (14.0)

23.4 (12.4)

24.8 (22.7)

0.356

 ALT

32.1 (17.5)

21.6 (16.7)

19.5 (14.4)

0.102

Albumin

4.3 (0.3)

4.2 (0.3)

4.3 (0.4)

0.776

Insulin

19.9 (12.7)

8.5 (5.0)

7.5 (4.6)

0.006

Leptin

63.8 (14.2)

50.9 (13.3)

20.2 (8.5)

0.005

Adiponectin

6.4 (4.2)

11.5 (6.0)

12.4 (6.1)

0.002

HOMA-IR

5.7 (4.9)

2.0 (2.1)

1.6 (1.0)

0.002

Waist circumference

133 (14.4)

111 (10.7)

107 (10.8)

<0.001

Hip circumference

145 (12.4)

126 (12.8)

121 (10.9)

<0.001

The italicized P-values represent those that are statistically significant

* Difference between baseline and 6 month post-operative level

At 3 months postoperatively, the anti-inflammatory mediator adiponectin was significantly increased (P = 0.0087) compared to baseline levels, and this rise in adiponectin levels persisted at 6 months follow-up (P = 0.002) (Fig. 1). Levels of the anti-inflammatory and antiatherogenic marker IL-1Ra decreased significantly 6 months after gastric bypass (P = 0.032), and there was no significant change in levels of the other anti-inflammatory cytokine IL-10 (P = 0.302, Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs00464-011-1620-6/MediaObjects/464_2011_1620_Fig1_HTML.gif
Fig. 1

Changes in CRP (P < 0.001 at 3 and 6 months) and adiponectin levels (P = 0.002 at 3 months, P = 0.004 at 6 months) after gastric bypass. * Statistically significant (P < 0.05) compared to baseline level

https://static-content.springer.com/image/art%3A10.1007%2Fs00464-011-1620-6/MediaObjects/464_2011_1620_Fig2_HTML.gif
Fig. 2

Percent change in pro- and anti-inflammatory markers, and in markers of cardiovascular risk, at 3 and 6 months after gastric bypass

Using the HOMA-IR as an index of insulin resistance, we observed a significant decrease from the average baseline measurement, to the 3- and 6-month postoperative values (mean HOMA-IR = 5.70 ± 4.9, 1.99 ± 2.1, and 1.57 ± 1.1 at baseline, 3 months, and 6 months, respectively, P = 0.002, Fig. 3).
https://static-content.springer.com/image/art%3A10.1007%2Fs00464-011-1620-6/MediaObjects/464_2011_1620_Fig3_HTML.gif
Fig. 3

HOMA-IR for each subject before and after gastric bypass (mean HOMA-IR = 5.70 ± 4.9, 1.99 ± 2.1, and 1.57 ± 1.1 at baseline, 3 months, and 6 months, respectively, P = 0.002)

Cardiovascular risk: BART and FRS

Six months after surgery, the percent change in endothelium-dependent vasodilation following flow-mediated vasodilation was improved compared to the presurgical percent change, though not significantly so (2.59 ± 2.03% at baseline vs. 4.44 ± 3.38% 6 months postoperatively, P = 0.21). However, endothelium-independent vasodilation (after nitroglycerin administration) was observed to be significantly improved (20.51 ± 6.71% at baseline vs. 23.62 ± 6.37% 6 months postoperatively, P = 0.047). The baseline brachial artery diameter was observed to have decreased significantly postoperatively (3.39 ± 0.37 vs. 3.26 ± 0.38 mm, P = 0.001, Table 3). Relative risk for 10-year coronary heart disease, based on gender-specific Framingham Risk Score, was significantly improved at 6 months (1.69 ± 0.20 vs. 1.12 ± 0.20, P = 0.005, Fig. 4).
Table 3

Alterations in vascular characteristics before and after weight loss

Brachial artery characteristics

Baseline

6 Months post RYGB

P Value (t-test)*

Diameter (mm)

3.39 (0.37)

3.26 (0.38)

0.001

FMD (% change)

 Endothelium dependent

2.59 (2.03)

4.44 (3.38)

0.210

 Endothelium independent

20.51 (6.71)

23.62 (6.37)

0.047

The italicized P-values represent those that are statistically significant

* Difference between levels at baseline and 6 months post-RYGB

https://static-content.springer.com/image/art%3A10.1007%2Fs00464-011-1620-6/MediaObjects/464_2011_1620_Fig4_HTML.gif
Fig. 4

Reduction in Framingham Risk Score 6 months after gastric bypass surgery (33.7% relative risk reduction, P = 0.005). * Statistically significant (P < 0.05) compared to baseline level

Discussion

This prospective cohort study illustrates the positive effects of substantial weight loss on inflammatory and metabolic biomarkers, endothelial function, and cardiovascular risk. The observation that acute-phase reactants and proinflammatory mediators (CRP, leptin, insulin, and fibrinogen) decreased significantly 6 months after RYGB, in conjunction with significant improvements in cardiovascular risk scores and markers of endothelial function, supports the hypothesis that inflammatory cytokines at least in part mediate the link between obesity and cardiovascular comorbidities. These improvements occurred after patients lost more than 50% excess weight, despite the fact that the patients remained morbidly obese at the 6-month follow-up (mean BMI = 35.4 kg/m2). Our results are consistent with those of Williams et al. [31] and Gokce et al. [42]; both groups reported significant improvements in endothelial function in obese people after surgically induced weight loss, in association with moderate improvements in insulin sensitivity, glucose homeostasis, and systemic inflammation.

Decrease in proinflammatory state associated with obesity after RYGB

A growing body of evidence supports the role of chronic low-grade inflammation as a link between obesity and its major comorbidities such as cardiovascular disease and diabetes mellitus. We present further evidence in support of this hypothesis by demonstrating that substantial weight loss after gastric bypass is associated with significant early reductions in the proinflammatory mediators CRP, fibrinogen, PAI-1, and leptin, and a concurrent increase in anti-inflammatory marker adiponectin. These bioactive marker alterations occurred in concert with significant changes in patients’ body habitus, including greatly reduced BMI and hip and waist circumference. These findings imply that a balance in cytokine production was achieved with rapid loss of visceral fat. It is well documented that visceral adipose tissue is a distinct and biologically active fat depot compared to subcutaneous fat [43, 44]. Williams et al. [31] proposed that a greater depletion of bioactive visceral fat relative to subcutaneous depot is a key factor in the prompt reduction in inflammatory and metabolic markers.

Similar to Williams and other authors, we did not observe a significant reduction in IL-6 levels following surgically induced weight loss [13, 30]. One possible explanation for this cytokine remaining elevated in obese patients 6 months post bariatric surgery is that these individuals remained obese despite significant weight loss. Perhaps pathways involving some inflammatory mediators remain activated even in moderate obesity and require further fat loss to achieve normalcy. Our follow-up period of 6 months was designed to demonstrate the early effects of surgical weight loss and may have missed a later, biologically relevant alteration in serum levels of these cytokines. Observing serial levels of these mediators over time, as weight loss continues, will be important to further elucidate the role IL-6 in obesity. Furthermore, technical difficulties in measuring IL-6 levels, such as its short half-life (70 min [45]), may affect the accuracy of levels quantified in real time and also imply that levels at a specific timepoint may not be reflective of a chronic inflammatory process. A more accurate determination of the activity of pathways involving IL-6 would be to measure its expression at messenger RNA level or the biologically active isoforms or the levels of soluble receptors for this cytokine. Unlike IL-6, CRP has a long half-life and therefore has greater stability of levels in plasma; hence, it has proven to be a very useful marker of inflammation in clinical and epidemiological studies [46, 47].

Alterations in anti-inflammatory markers after RYGB

Parallel to reduced levels of proinflammatory cytokines following weight loss, one may have expected a simultaneous increase in anti-inflammatory marker levels if it is assumed that a state of equilibrium in inflammatory pathways is being reached. We analyzed circulating levels of the anti-inflammatory markers IL-10, IL-1Ra, and adiponectin, before and after RYGB, to determine the effects of weight loss on the “good” side of the inflammation cascade. Adiponectin levels were significantly increased in our bariatric patients 6 months postoperatively, consistent with previous reports by Linscheid et al. [48] and Trakhtenbroit et al. [49]. No difference was observed for IL-10 levels and interestingly IL-1Ra levels decreased significantly at 6 months postoperatively, contrary to what may have been expected. This decrease in IL-1Ra levels may be explained by the simultaneous postoperative decrease in leptin levels in this cohort. Leptin is known to promote expression and secretion of the anti-inflammatory cytokine IL-1Ra, through its stimulatory effect on monocytes. The observed decrease in leptin levels after gastric bypass may be largely responsible for decreased stimulation of IL-1Ra release from monocytes in adipose tissue, its primary source.

Improved insulin sensitivity post RYGB

Perhaps the most striking clinical finding consistently reported after RYGB, even more significant than the dramatic weight loss that ensues, is the rapid restoration of euglycemia in patients who had impaired glucose tolerance secondary to their obese habitus. Obesity has long been associated with insulin resistance (IR) and only recently have certain mechanisms that mediate the link between excess adipose tissue and IR been identified, such as a state of low-grade inflammation and the effects of hormones and gut peptides. Our results support the fact that weight loss leads to improved insulin sensitivity as evidenced by the significant postoperative reductions observed in circulating fasting insulin and HOMA-IR scores. A plausible mechanism for this dramatic improvement in insulin sensitivity may be the interplay between altered gut peptide and secretion of adipokines (including adiponectin and leptin) due to exclusion of the foregut and the reduced inflammatory environment observed following substantial loss of visceral adipose tissue. Other groups have reported similar findings; Trakhtenbroit et al. [49] document that the effects of surgically induced (RYGB) weight loss on adipokines, inflammatory markers, and insulin levels persists even at 24 months.

Changes in brachial artery reactivity reflect improved endothelial function post RYGB

Obesity is associated with maladaptive enlargement of conduit arteries such as the brachial artery. Vascular remodeling occurs as a consequence of atherosclerosis resulting in luminal expansion at plaque locations, primarily as a protective response to maintain distal perfusion. Indeed, this response may be exaggerated in the obese state secondary to chronic excess inflammatory activation [31]. We observed a significant improvement in endothelium-independent vasodilation, 6 months post-RYGB. Additionally, there was a smaller, though non-significant, improvement in endothelium-dependent vasodilation. It is likely that the small sample size of our cohort precluded adequate power to achieve statistical significance for this latter end point, and this finding should be investigated further in larger studies.

We observed a statistically significant decrease in baseline resting brachial artery diameter (0.13 mm on average) after substantial weight loss; a finding also reported by Hamburg et al. [31] in their cohort of 24 patients who achieved a greater than 10% weight loss after 1 year, with an associated reduction in baseline brachial artery diameter of 0.19 mm and a corresponding highly significant decrease in CRP levels. Notably, there was no decrease in brachial artery diameter in patients who had moderate (or less) weight loss achieved either surgically or through nonoperative interventions. These results and the findings from the present study suggest that significant weight loss can reverse pathologic remodeling of diseased vasculature at least in part through a reduction in inflammation.

Decreased cardiovascular risk post RYGB

A primary goal of weight reduction for obese individuals with comorbidities such as diabetes, hypertension, hyperlipidemia, and coronary artery disease is to improve their cardiovascular risk profile given that this commonly represents their greatest threat to life. The Framingham Risk Score (FRS), which calculates the 10-year risk of developing “hard” cardiovascular disease outcomes such as myocardial infarction and coronary death, is designed to be independent of weight and is therefore an ideal model to estimate the impact of bariatric surgery on modifying cardiovascular risk. However, quantification of risk alteration from surgically induced weight loss has been infrequently reported [50]. By determining individual patients’ pre- and postoperative FRS, we report a 33.7% relative risk reduction only 6 months after gastric bypass (Fig. 4). In our cohort of severely obese patients, all of whom had at least one obesity-related comorbidity, the baseline FRSs were relatively low (mean = 1.69%, range = 1–3.4%) compared to two similar cohort studies in patients with BMI > 35 who also underwent gastric bypass (5.4–6.7% baseline FRS [50, 51]). This may be a consequence of the intensive preoperative program our patients must complete prior to bariatric surgical intervention. Patients are frequently required to lose a moderate amount of weight, stop smoking, and have all of their cardiovascular risk factors optimized prior to surgery. This means that a single preoperative calculation of FRS in our surgical patient population is typically based on a relatively young, normotensive, lipid-controlled nonsmoker and therefore may not reflect true cardiovascular risk of a demographically similar group of patients who are not pursuing bariatric surgery or receiving this intensive medical attention. Furthermore, the FRS is heavily weighted for age as the most powerful risk factor, and so it may underestimate the actual cardiovascular risk in a young bariatric cohort such as ours where the average age was 49 years. Nonetheless, a significant reduction in cardiovascular risk was still achieved with rapid weight loss in our apparently low-risk cohort.

Conclusion

This study is the first to expand the in vivo analysis of the effects of weight loss on inflammatory mediators and adipokines and to incorporate clinical measures such as cardiovascular risk scores and endothelial function and arterial remodeling. Our results indicate that weight loss following gastric bypass is associated with an early decrease in proinflammatory mediators, an increase in adipokines with anti-inflammatory properties, markedly enhanced insulin sensitivity, improved vascular endothelial function, and improved cardiovascular risk at 6 months. These global systemic effects highlight the major benefits of surgically induced weight loss in reducing comorbidities of severe obesity. This provides further evidence to support bariatric surgery as a primary therapy for reducing cardiovascular morbidity and mortality associated with obesity.

Acknowledgment

This study was supported by a SAGES research grant.

Disclosures

Dr. Brethauer is a speaker, consultant, and scientific advisory board member for Ethicon Endo-Surgery, a speaker for Covidien, and receives research support from Bard/Davol. Dr. Schauer is a consultant and scientific advisory board member of and has received research support from Ethicon Endo-Surgery; is on the board of directors of Remedy MD; is on the scientific advisory board of and has received an educational grant from Stryker Endoscopy; is on the scientific advisory board of and is a consultant for Bard/Davol; is a consultant for and has received an educational grant from Gore; has received an educational grant from Baxter; is on the scientific advisory board of Barosense, Surgiquest, Cardinal/Snowden Pencer; has received an educational grant from Covidien; an educational grant from Allergan; and is on the board of directors of Surgical Excellence LLC. Hazel Huang and Drs. Heneghan, Eldar, Gatmaitan, Kashyap, Kirwan, and Gornik have no conflicts of interest or financial ties to disclose.

Copyright information

© Springer Science+Business Media, LLC 2011