Postmortem studies have demonstrated that morbidly obese subjects, surprisingly, have less coronary atherosclerosis than obese subjects. However, the reasons for this apparent protection from atherosclerosis are not yet clear. Thromboxane A2, a marker of platelet activation, is greater in obese subjects than in lean subjects, and this might be a clue to their increased cardiovascular risk. However, data on thromboxane A2 in morbidly obese subjects are lacking; therefore, we hypothesized that lower levels of thromboxane A2 in morbidly obese subjects might play a role in their lower atherothrombotic burden. We measured the serum levels of thromboxane B2 (TxB2), a stable metabolite of thromboxane A2, high-sensitivity C-reactive protein (hs-CRP) and leptin in 17 lean subjects (body mass index [BMI] 22.9 ± 1.6 kg/m 2 ), 25 obese subjects (BMI 32.6 ± 2.4 kg/m 2 ), and 23 morbidly obese subjects (BMI 48.6 ± 7.1 kg/m 2 ), without insulin resistance, diabetes, or overt cardiovascular disease. The serum TxB2 levels were lower in the lean subjects than in the obese subjects (p = 0.046) and in the morbidly obese subjects than in the lean and obese subjects (p = 0.015 and p <0.001, respectively). In contrast, the hs-CRP and leptin levels were greater in the obese than in the lean subjects (hs-CRP, p <0.001; leptin, p <0.001) and in the morbidly obese subjects than in the lean subjects (p <0.001 for both). Leptin was also higher in the morbidly obese subjects than in the obese subjects (p <0.001). TxB2 negatively correlated with leptin and BMI. hs-CRP correlated with leptin, and both also correlated with waist circumference, BMI, and homeostasis model assessment of insulin-resistance. In conclusion, insulin-sensitive morbidly obese subjects had lower levels of TxB2 than the obese subjects and lean subjects, suggesting that reduced platelet activation could play a role in the paradoxical protection of morbidly obese subjects from atherosclerosis, despite the greater levels of leptin.
Although obesity is clearly a risk factor for coronary artery disease, the correlation of the degree of obesity and atherosclerosis is complex and still controversial. Autopsy studies involving large numbers of obese patients dying from heart failure have consistently demonstrated a surprisingly low rate of coronary atherosclerosis in morbidly obese subjects. Platelets play a crucial role in the pathogenesis of acute coronary syndromes, and thromboxane A2 is a key mediator of platelet activation and aggregation. Increased platelet activation has been considered a potential mechanism of increased cardiovascular risk in obese patients. Previous studies have shown that insulin resistance contributes to increased platelet activation in obesity, independently of underlying inflammation. The renewed attention to the “obesity paradox” has emphasized the need to better understand the mechanisms linking obesity to cardiovascular disease. To our knowledge, no studies have specifically investigated whether fat tissue per se affects platelet function, avoiding the confounding presence of insulin resistance. Therefore, the aim of our study was to compare the thromboxane A2 levels across various body mass index (BMI) categories in insulin-sensitive subjects.
From September 2008 to May 2009, we screened 46 morbidly obese subjects who were undergoing an evaluation for bariatric surgery and 40 age- and gender-matched obese subjects at the Metabolic Disease Clinic of our institution. We also included 17 age- and gender-matched lean subjects as a control group. All participants were in good general health, with normal medical history and physical examination findings and were not taking any medication. The exclusion criteria included previous cardiovascular disease. In particular, myocardial infarction or angina pectoris were excluded on the basis of the medical history, a history of a coronary revascularization procedure, or the presence of left bundle branch block, Q or QS waves on the electrocardiogram. Additional exclusion criteria were: thyroid dysfunction, any kind of treatment for obesity or a loss of weight >5 kg in the previous 3 months, chronic lung disease, malignancies in the previous 5 years, any inflammatory or infectious disease, and renal failure (creatinine clearance <90 ml/min). When present, hypertension and dyslipidemia were a new diagnosis and the patients were not receiving medical treatment. Women were excluded if they were taking hormonal contraception or replacement therapy. Fifteen women were postmenopausal. To avoid confounding by diabetes and insulin resistance, we performed a standard 75-g oral glucose tolerance test that was used to calculate the homeostasis model assessment of insulin-resistance (HOMA-IR) index, defined by the formula: (fasting plasma glucose × fasting plasma insulin)/22.5. After the oral glucose tolerance test, 10 obese and 15 morbidly obese subjects were excluded because of diabetes, and 5 obese and 8 morbidly obese subjects were excluded because of insulin resistance, defined according to a HOMA-IR index cutoff of >2.5. A total of 65 subjects were finally enrolled and were divided by their BMI into 3 groups: 17 lean subjects (BMI 22.9 ± 1.6 kg/m 2 ; age 48.9 ± 4.8 years; 7 women), 25 obese subjects (BMI 32.6 ± 2.4 kg/m 2 ; age 50.8 ± 6.5 years; 14 women), and 23 morbidly obese subjects (BMI 48.6 ± 7.1 kg/m 2 ; age 49.6 ± 10.1 years; 8 women).
Information on the participant demographics and medical history, including the cardiovascular risk factor profile (age, gender, dyslipidemia according to the National Cholesterol Education Program screening criteria, hypertension according to the Joint National Committee 7 criteria, and cigarette smoking) were carefully recorded for each participant.
All patients gave written informed consent, and the ethics committee of the Catholic university approved the study protocol. Anthropometric measurements were taken according to standardized procedures, in the morning, before breakfast, with the patient wearing indoor clothes without shoes. Height was measured using a clinic stadiometer, weight was measured using a Tanita bioimpedance balance (Tanita, Middlesex, United Kingdom). Waist circumference was defined as the minimal abdominal circumference between the xiphoid process and the iliac crest. The blood pressure was measured 3 times by an automatic blood pressure measuring device (Omron, Hoofddorp, The Netherlands), and the median measurement was used. Peripheral blood samples were drawn after an overnight fast from an antecubital vein and kept for 1 hour at 37°C before centrifugation. The serum and plasma were prepared by centrifugation of the whole blood samples at 2,000 g for 15 minutes and then were divided into aliquots and stored at −80°C until assayed. Insulin sensitivity was evaluated from the 75-g 2-hour oral glucose tolerance test values. Plasma samples were taken at 0, 30, 60, 90, and 120 minutes after glucose loading. The plasma glucose was measured using the glucose oxidase method (Beckman, Fullerton, California). Insulin was assessed using radioimmunoassay kits (Abbott Diagnostic, Milan, Italy). Serum triglycerides were measured using the enzymatic colorimetric method. The total, low-density lipoprotein, and high-density lipoprotein cholesterol levels were measured using an automated enzymatic assay. Thromboxane B2 (TxB2), a stable metabolite of thromboxane A2, was measured in serum samples using an EIA commercial kit (Cayman Chemical, Ann Arbor, Michigan).
Serum high-sensitivity C-reactive protein (hs-CRP) levels were assessed using high-sensitivity immunonephelometry (Siemens Health Care Diagnostic BN, Deerfield, Delaware) with a detection limit of 0.1 mg/L. The plasma leptin and adiponectin levels were measured using a radioimmunoassay for human leptin (Phoenix Pharmaceuticals, Phoenix, Arizona) and human adiponectin (LINCO, St. Charles, Missouri). Because the distributions of TxB2, hs-CRP, leptin, adiponectin, and HOMA-IR were not normal according to the Kolmogorov-Smirnov method, the differences between groups were analyzed using Kruskal-Wallis as a nonparametric test. The remaining continuous variables were compared using analysis of variance. Corrections for multiple comparisons were performed using Dunn’s or Dunnett’s test, as appropriate. The categorical data were compared using Fisher’s exact test. The nonparametric Spearman correlation coefficient was used to assess the association between different measurements. Significance was accepted at p = 0.05 (2 sided). Data are reported as the median and range or the mean ± SD, as appropriate. Because of the exploratory nature of our study, it was impossible to reliably calculate the sample size in the absence of previous data on the serum TxB2 levels and platelet aggregation tests across the various BMI categories. Statistical analyses were performed using the Statistical Package for Social Sciences, version 17.0 (SPSS, Chicago, Illinois).
The baseline characteristics of the study population and main findings are listed in Table 1 . The weight, waist circumference, and BMI were significantly different among the 3 groups (p <0.0001 for all comparisons). The diastolic blood pressure was greater in the obese subjects and morbidly obese subjects than in the lean subjects (p = 0.008 and p = 0.001, respectively), but the systolic blood pressure (p = 0.015) and fasting plasma insulin (p = 0.013) were greater only for the morbidly obese subjects compared to the lean subjects. Although all participants were insulin sensitive, the HOMA-IR index was significantly greater in the obese subjects than in the lean subjects (p = 0.019) and in the morbidly obese subjects than in the lean subjects (p <0.001). Serum TxB2 was significantly lower in the lean subjects than in the obese subjects (p = 0.046), the morbidly obese subjects than in the lean subjects (p = 0.015), and the morbidly obese subjects than in the obese subjects (p <0.001; Figure 1 ). In contrast, hs-CRP increased according to the BMI and was significantly greater in the obese subjects than in the lean subjects (p <0.001) and in the morbidly obese subjects than in the lean subjects (p <0.001; Figure 2 ). Similarly, the leptin levels were higher in the obese subjects than in the lean subjects (p <0.001) and in the morbidly obese subjects compared to the lean subjects (p <0.001). However, they were also higher in the morbidly obese subjects than in the obese subjects (p <0.001; Figure 2 ). No significant differences were found in the adiponectin levels among the 3 groups (lean subjects versus obese subjects, p = 0.362; lean subjects versus morbidly obese subjects, p = 0.308; and obese subjects versus morbidly obese subjects, p >0.99). The correlations are listed in Table 2 . Specifically, TxB2 correlated negatively with leptin, weight, and BMI. hs-CRP correlated significantly with leptin, weight, waist circumference, BMI, and the HOMA-IR index. Leptin showed a significant correlation with weight, waist circumference, BMI, and the HOMA-IR index.
|Variable||Lean (n = 17)||Obese (n = 25)||Morbidly Obese (n = 23)||p Value|
|Age (years)||48.9 ± 4.8||50.8 ± 6.5||49.6 ± 10.1||>0.99 ⁎|
|Gender||0.53 † ; 0.74 ‡ ; 0.16 §|
|Postmenopausal||4 (57%)||6 (43%)||5 (62%)||>0.99 ⁎|
|Weight (kg)||65.3 ± 10.2||96.5 ± 10.5||134.2 ± 25.2||<0.0001 ⁎|
|Waist circumference (cm)||78.5 ± 10.7||105.9 ± 8.4||145.2 ± 19.7||<0.0001 ⁎|
|Body mass index (kg/m 2 )||22.9 ± 1.6||32.6 ± 2.4||48.6 ± 7.1||<0.0001 ⁎|
|Smoker||4 (23%)||6 (24%)||5 (22%)||>0.99 ⁎|
|Hypertension||7 (41%)||12 (48%)||14 (60%)||0.75 † ; 0.53 ‡ ; 0.78 §|
|Total cholesterol (mg/dl)||186.6 ± 41.4||.208 ± 37.2||202.8 ± 42.8||0.33 † ; 0.674 ‡ ; >0.99 §|
|Low-density lipoprotein cholesterol (mg/dl)||133.6 ± 26.2||133.6 ± 33.6||135.9 ± 31||>0.99 ⁎|
|High-density lipoprotein cholesterol (mg/dl)||47.1 ± 10.4||51.8 ± 9.5||47.4 ± 12||0.505 † ; >0.99 ‡ ; 0.488 §|
|Triglycerides (mg/dl)||111.7 ± 60.8||119.4 ± 62.9||150.9 ± 67.9||>0.99 † ; 0.346 ‡ ; 0.294 §|
|Systolic blood pressure (mm Hg)||129.6 ± 11.8||138.7 ± 15||142.8 ± 14.5||0.135 † ; 0.015 ‡ ; 0.95 §|
|Diastolic blood pressure (mm Hg)||81.2 ± 7.9||88.4 ± 7||90 ± 7||0.008 † ; 0.001 ‡ ; >0.99 §|
|Fasting plasma glucose (mmol/L)||4.5 (3.9–7.2)||4.6 (3.9–6.6)||5 (4.3–6.6)||>0.99 † ; 0.914 ‡ ; 0.350 §|
|Fasting plasma insulin (mU/L)||4.1 (2.6–6.2)||5.2 (2.3–7.8)||5.5 (3.5–9)||0.082 † ; 0.013 ‡ ; >0.99 §|
|Homeostasis model assessment-insulin resistance index||0.8 (0.5–1.2)||1.2 (0.4–1.5)||1.3 (0.7–1.7)||0.019 † ; <0.001 ‡ ; 0.112 §|
|Thromboxane B2 (pg/ml)||24,105 (1,974–67,209)||38,732 (1,705–104,728)||7,972 (161–44,893)||0.046 † ; 0.015 ‡ ; <0.001 §|
|High-sensitivity C-reactive protein (mg/L)||0.8 (0.1–5.4)||4.5 (0.4–13.9)||6.4 (0.9–35.5)||<0.001 † ; <0.001 ‡ ; 0.32 §|
|Leptin (ng/ml)||5.4 (1.9–25.7)||13 (2–55.5)||85 (36.3–117)||<0.001 † ; <0.001 ‡ ; <0.001 §|
|Adiponectin (μg/ml)||10.4 ± 6.7||13.6 ± 7.3||14 ± 7.2||0.36 † ; 0.30 ‡ ; >0.99 §|