Obesity (often complicated by type 2 diabetes), arterial hypertension, and hyperlipidemia show insulin resistance (1), metabolic abnormalities (triglycerides, insulin and blood glucose levels, and HDL cholesterol levels ), subclinical inflammation (3,4), functional and metabolic indexes of endothelial dysfunction (5–7), and sympathetic overactivity (8–12). Inflammation (C-reactive protein, white cells, and interleukins) is linked to metabolic abnormalities, sympathetic overactivity, and endothelial dysfunction (3,13,14).
Enlarged adipose tissue, infiltrated by macrophages, releases substances (adipokines) (15) that can explain most features of obesity (type 2 diabetes, arterial hypertension, and accelerated atherosclerosis) (16–18) including leptin, which is considered a risk factor for cardiovascular disease (19,20) and acts directly on the myocardium, stimulating cardiac hypertrophy (21,22).
Weight loss decreases the metabolic impact of obesity, reduces prevalence of type 2 diabetes and arterial hypertension (23), prevents their incidence (24,25), improves insulin resistance (26) and endothelial dysfunction (4,7), and reduces sympathetic overactivity (27), adipokines, and adhesion molecules (4,7), as well as white cells (28).
In this study, we analyzed interrelationships of metabolic abnormalities, sympathetic overactivity, inflammation, and endothelial dysfunction in obesity, including their changes induced by weight loss (laparoscopic adjustable gastric banding [LAGB]).
RESEARCH DESIGN AND METHODS—
LAGB is regularly performed at Ospedale San Raffaele and Ospedale San Paolo, Milano, Italy, in morbidly obese patients (World Health Organization criteria) (25,26). For this study, we considered 69 patients (15 men and 54 women, aged 42.3 ± 1.07 years, BMI 45.1 ± 0.78 kg/m2) undergoing LAGB during the period of 2002–2004. Inclusion/exclusion criteria, schedule for assessments, and diet after LAGB have already been published (25,26). Waist circumference was measured, and ultrasound scan (29) of visceral and subcutaneous adipose tissue was performed; all evaluations were repeated after 1 year.
QT intervals on electrocardiogram were measured according to Bazett’s formula, together with heart rate (corrected QT [QTc] = QT/√ RR) (30). Specimens for endothelin-1 (ET-1), E-selectin, and intracellular adhesion molecule (ICAM)-1 were collected in prechilled tubes (EDTA tubes with 200 μl aprotinin for ET-1), centrifuged at 4°C, and stored at −70°C. Blood glucose, insulin, leptin, triglyceride, HDL cholesterol levels, and adhesion molecules (ICAM-1, E-selectin, and ET-1) were measured as published before (7,22,25–28). The homeostasis model assessment (HOMA) index was calculated as insulin (in microunits per milliliter) × blood glucose (in micromoles per liter) × 22.5−1 (31).
Calculations and statistical analysis
Because normality had not been verified on all variables, data were log transformed. Changes of variables were then analyzed. Pairwise correlations between changes of variables were calculated. A stepwise regression analysis was carried out to estimate the independent contribution of variables (significant at linear regression plus age and sex) on heart rate and QTc. A P value <0.05 was considered statistically significant.
Weight loss (BMI at 1 year 37.0 ± 0.94 kg/m2) caused significant decreases (P < 0.01 to P < 0.001 vs. baseline) in visceral fat (88.5 ± 3.21 to 53.0 ± 3.08 mm), waist circumference (125.5 ± 1.74 to 109.5 ± 1.62 cm), metabolic variables (HOMA index 5.6 ± 0.48 to 2.3 ± 20.18), triglycerides (1.6 ± 0.14 to 1.2 ± 0.11 mmol/l), HDL cholesterol (1.21 ± 0.04 to 1.33 ± 0.04 mmol/l), sympathetic tone (systolic blood pressure [133.2 ± 1.85 to 126.5 ± 1.55 mmHg], diastolic blood pressure [84.1 ± 1.51 to 80.0 ± 1.61 mmHg], heart rate [73.9 ± 1.40 to 65.3 ± 1.39], QTc [418.5 ± 3.86 to 400.7 ± 3.49 ms], and leptin [39.0 ± 1.98 to 20.6 ± 1.34 ng/ml]), and inflammation/endothelial dysfunction (white cells [7,867 ± 336 to 7,012 ± 314 fl], ICAM-1 [325.8 ± 10.68 to 281.1 ± 6.79 ng/ml], E-selectin [66.7 ± 6.39 to 43.4 ± 2.99 ng/ml], and ET-1 [1.3 ± 0.07 to 1.1 ± 0.05 pg/ml]).
Changes of almost all metabolic variables, sympathetic tone, inflammation/endothelial dysfunction, and insulin resistance correlated with changes of BMI and visceral fat (data not shown). Changes of ICAM-1 significantly correlated with ET-1 (r = 0.328), E-selectin (r = 0.673), HOMA (r = 0.239), and leptin (r = 0.251); E-selectin correlated with ET-1 (r = 0.294), HOMA (r = 0.342), and leptin (r = 0.337) (P < 0.05 to P < 0.001).
Changes of sympathetic activation (heart rate and QTc) correlated with BMI and at least one measure of visceral fat, leptin, insulin, and insulin resistance. At stepwise regression analysis, an ultrasound scan of visceral adipose tissue and leptin were the only significant predictors of heart rate, while leptin and heart rate were the only significant predictors of QTc. Figure 1 shows the correlations for heart rate and QTc.
We confirmed that weight loss is accompanied by improvement of metabolic variables and indexes of inflammation/endothelial dysfunction and of sympathetic overactivity, thus pointing to a pivotal role of BMI (4,7,26–28); changes of sympathetic overactivity and adhesion molecules correlated with changes of BMI and visceral fat. Finally, changes of heart rate and QTc correlated with leptin and visceral fat, thus raising the hypothesis that heart rate and QTc are crucial aspects of this cluster.
The correlation between changes of BMI and visceral fat, and adhesion molecules, agrees with previous studies considering these molecules (3,4,7,32,33), tumor necrosis factor-α, P-selectin, and interleukin-6 (4). The correlations between adhesion molecules, and some metabolic variables not previously described, indicate a correlation between several aspects of the cluster and agree with previous findings of an association between C-reactive protein and metabolic variables (3,32,33).
Heart rate and QTc correlated with BMI, visceral fat, and insulin resistance (HOMA). These correlations have been reported in a few studies, supporting a causal role of BMI and visceral fat on these, as well as on other, indexes of sympathetic activity (11,12,33–39,40).
Leptin correlated with heart rate and QTc. Chronic leptin administration leads to hypertension and tachicardia in rats, and this effect is prevented by adrenergic β-blockade (41). Leptin increases sympathetic activity in the central nervous system (42) and has a direct action in the heart, as leptin has specific receptors in heart tissues (21,22).
Adhesion molecules did not correlate with heart rate and QTc; the possible correlation between sympathetic drive and immune function, in particular between adrenergic agonists/antagonists and adhesion molecules, has recently received a great deal of attention (43). For instance, β-adrenergic agonists and antagonists have different effects on white cells and ICAM-1 (43–48). Psychological stress, physical stress, hypertension, painful anestesia, adrenaline, and norepinephrine administration have specific effects on ICAM-1 levels, white cells, NK cells, E-selectin, P-selectin, L-selectin, and heart rate (49–56). An indirect association between adhesion molecules and sympathetic overactivity might be mediated by insulin resistance, leptin, or body weight. In our study, leptin correlated with ICAM-1 and E-selectin, and this agrees with experiments showing that leptin upregulates proinflammatory immune responses (57). At stepwise regression analysis, visceral fat and leptin were the only significant predictors of heart rate and QTc.
In conclusion, these data indicate a cluster of metabolic, immune, and sympathetic abnormalities. This cluster, favored by, if not essentially linked to obesity, is a more complex entity than originally supposed. Sympathetic activity and subinflammation coexist and are probably essential features of the cluster and can be corrected through stable weight loss by LAGB.
This research was supported by Grant FIRST 2002 (Universita‘ degli Studi di Milano), Ministero dell’Universita‘ e della Ricerca 2002 (grant 2002064582_003), and Ministero della Salute (grant 199/02).
A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C Section 1734 solely to indicate this fact.