SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Objective: In order to improve our understanding of high-density lipoprotein cholesterol (HDL-C) cardiovascular (CV) impact in obesity, the association of HDL-C plasma level with circulating early endothelial progenitor cell (early-EPC) number and endothelium-dependent vasodilatation (EDV) in obese women with normal or high low-density lipoprotein cholesterol (LDL-C) plasma levels was evaluated.

Design and Methods: One hundred thirteen obese female subjects and a control group of 78 healthy female subjects were recruited. Circulating early-EPC were assessed by single- and two-color flow cytometric analyses with a fluorescence activated cell sorting (FACScan) flow cytometer. EDV was evaluated as response to ischemia by strain gauge plethysmography.

Results: Both early-EPC number and EDV were significantly decreased in obese women compared with the control group. Obese women with low HDL-C showed a further decrease of early-EPC and EDV in the presence of both high or normal LDL-C plasmatic levels. In the normal HDL-C level subgroup, hypercholesterolemic and nonhypercholesterolemic subjects showed no difference in early-EPC number, whereas slight EDV impairment was present in hypercholesterolemic subjects.

Conclusion: In obese women, low HDL-C is associated to decreased early-EPC number and impaired EDV, suggesting the need to assess whether evaluation of early-EPC and EDV may increase HDL-C prognostic value in the stratification of CV risk.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Obesity is now a major public health problem worldwide, both in industrialized and emerging countries. Obesity significantly increases the risk of morbidity and mortality from hypertension, dyslipidemia, diabetes mellitus, heart failure, stroke, and coronary artery disease [1], but it is documented to be an independent cardiovascular (CV) risk factor even when not associated with other CV risk factors [2]. Cardiovascular disease (CVD) is the leading cause of death in western countries, and coronary heart disease (CHD) is a major cause of morbidity and mortality in women, responsible for more deaths than all forms of cancer combined [3].

Insulin resistance is widely held responsible for the relationship between obesity, especially abdominal obesity, and increased CV risk [4, 5], but the overall mechanisms are not completely understood and are likely to be very complex. One of them may reside in the close association of obesity with an atherogenic-dyslipidemic state characterized by high triglycerides, increased apolipoprotein B, increased proportion of small dense low-density lipoprotein (sdLDL), and low levels of high-density lipoprotein (HDL) cholesterol [6].

High-density lipoprotein cholesterol (HDL-C), in contrast with other lipoproteins, exerts many physiological functions, including removal of free cholesterol from the macrophage in the arterial wall [7], and lowering of oxidized lipid species in low-density lipoprotein cholesterol (LDL-C) particles. All these activities positively influence the CV system preventing atherosclerosis, acute coronary syndromes, and restenosis after coronary angioplasty [8]. In addition, HDL-C has been recently suggested to promote proliferation [9], endothelial Nitric Oxide Synthase (eNOS) protein expression and colony formation [10] in endothelial progenitor cell (EPC) in vitro, but no information is available on the relationship between HDL-C and EPC in vivo, excluding our recent report documenting that circulating EPC number is affected according to the HDL-C status in hypercholesterolemic lean subjects [11].

To improve our understanding of HDL-C CV impact in obesity, this study was aimed to evaluate the association of HDL-C plasma level with circulating early-EPC number and endothelium-dependent vasodilatation (EDV) in obese women.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Subjects

One hundred thirteen obese female subjects were consecutively recruited from the outpatient service of the Department of Experimental Medicine of Rome Sapienza University in about one year. This study was approved by the Bioethical Committee of Policlinico Umberto I Sapienza University of Rome. Before inclusion in the study, all candidates were screened using an extensive medical history and routine clinical screening. All tests were conducted between day 1 and 7 of the menstrual cycle. Waist circumference (WC) was calculated as media of 3 measures by each of 3 independent examiners. The cutoff for abdominal obesity was set at ≥88 cm according to the ATP III Guidelines.

Inclusion criteria: (body mass index (BMI) > 30). Each subject was assigned to the opportune subgroup according first to LDL-C (LDL-C plasma level more or less than 130 mg/dL) and then to HDL-C plasma level (HDL-C more or less than 50 mg/dL).

Exclusion criteria: triglycerides >150 mg/dL, previous history of CV events, presence of hypertension or diabetes mellitus, any medical treatment including estrogen therapy, smoking, menopause, and menstrual cycle alterations. A control, age-matched, group of 78 healthy female subjects was enrolled as well. Characteristics of the subjects enrolled in the study are detailed in Table 1.

Table 1. Demographic characteristics
 ControlObese
  1. BMI, body mass index; CHD N, family history of coronary heart disease; EPC, circulating endothelial progenitor cell number; FBFr, forearm blood flow ratio; WC, waist circumference. Values expressed as mean ± SEM. **P < 0.01, ***P < 0.001.

N78113
Age, y35.1 ± 3.234.4 ± 4.35
BMI, kg/m220.9 ± 2.935.3 ± 4.6**
WC, cm74 ± 2.4110.5 ± 1.5***
EPC42.8±10.813.2 ± 7.3***
FBFr17.6 ± 0.628.1 ± 0.42***
Systolic BP, mm Hg122 ± 4.3125 ± 3
Diastolic BP, mm Hg76 ± 3.175 ± 3.4
Fasting glucose, mmol/L4.1 ± 0.84.4 ± 0.8
Fasting insulin, mmol/L5.0 ± 0.45.5 ± 0.4
CHD N, %9 (23.7)24 (21.3)

Evaluation of endothelial function

The plethysmographic study began at 09.00 h, after the subjects fasted for at least 12 h. The subjects were kept in a supine position in a quiet, dark, air-conditioned room (constant temperature of 22°C). After 30 min in the supine position, the basal forearm blood flow (FBF) was measured in both arms simultaneously by strain gauge plethysmography, as already reported [4, 12]. Having assessed the basal FBF, EDV was evaluated as response to ischemia, and endothelium-independent vasodilatation as response to nitrates. In brief, cuff was inflated at 80 mm Hg exceeding the systolic blood pressure for 5 min around the nondominant arm, and maximum FBF (postischemic vasodilator response) was estimated as mean of two measurements. FBF peak response to ischemia was expressed as FBF ratio (FBFr) computed as maximum FBF in the experimental arm divided by maximum FBF in the contralateral arm. After administration of 0.3 mg nitroglycerin, maximum FBF was measured as nitroglycerin-induced hyperemia. Nitroglycerin-induced FBFr was calculated as nitroglycerin-induced FBF divided by the baseline value of FBF. FBF evaluation was performed by the same examiner (without any knowledge of the experimental condition under scrutiny); coefficient of variation was 3.8% ± 0.4% for all the tests.

Assessment of early-EPC number

As already reported [11], peripheral mononucleate cells were isolated by 20 mL of peripheral blood samples collected from the antecubital vein into a Vacutainer containing 3.8% buffered sodium heparin. Mononuclear cells were then isolated by density-gradient centrifugation of Histopaque 1077 (Sigma-Aldrich, Italy) at 1600 rpm for 30′ and 4°C. The mononuclear cells were washed twice with phosphate buffer solution and centrifuged before incubation with 1 mL blocking buffer for 30 min at 4°C. Cells were then incubated for 30 min at 4°C in a dark room with monoclonal antibodies against PE-conjugated CD133 (Miltenyi Biotec Macs) and FITC-conjugated CD34 (Miltenyi Biotec Macs, Italy) antibody and fixed with 2% paraformaldehyde. Single- and two-color flow cytometric analyses were performed using a FACScan flow cytometer EPICS® XL-MCL™. Each analysis included 60.000 events per sample. The assay for early-EPC in each sample was performed in duplicate and mean level was reported. Dual-staining cells positive for both CD133 and CD34 were judged and counted as early-EPC.

Statistical analysis

Statistical analysis was performed using Prism for windows version 4 (GraphPad software). Data are expressed as mean ± SEM. Unpaired t tests or, when appropriate, nonparametric tests were used for comparisons of the data after testing for normal distribution. Group comparisons were made with unpaired and two-tailed t tests or two way ANOVA tests when appropriate. To detect determinants of FBFr, univariate linear regression analyses were performed. Multivariate models, including variables showing significant correlation in univariate analysis, were carried out. Statistical significance was accepted at the 95% confidence level (P < 0.05). Odds ratio (OR) was performed and cumulative OR were calculated according to normality test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Obese women had significantly lower number of circulating EPC and impaired endothelium-dependent vasodilation, evaluated as FBF ratio (FBFr), compared with control group (Table 1). Abdominal obesity was tightly associated with both EPC decrease and endothelial dysfunction as documented by the OR of, respectively, 31 (CI 12-81, P < 0.0001) and 22 (CI 5.2-95, P < 0.0001). Univariate analysis of the data documented that, in obese women, EDV was related to EPC circulating number (Pearson r = 0.86, P < 0.0001) (Figure 1a), and to HDL-C (Pearson r = 0.62, P < 0.001) (Figure 1b). In addition, EDV was inversely related to WC (Pearson r = −0.60, P < 0.001) (Figure 1c) and LDL-C (Pearson r = −0.56, P < 0.001) (Figure 1d). EDV was associated as well with LDL-C level greater than 130 mg/dL and EPC number, as shown by the OR of, respectively, 4.3 (CI 1.2-21, P < 0.05) and 3.9 (CI 1.4-11, P < 0.005).

image

Figure 1. Correlation among EDV and other variables. EDV showed a direct correlation with circulating EPC number (Pearson r = 0.86, P < 0.0001) (A) and HDL-C (Pearson r = 0.62, P < 0.001) (B), an inverse correlation was assessed among EDV and waist circumference (Pearson r = −0.60, P < 0.001) (C) as well as LDL-C plasma levels (Pearson r = −0.56, P < 0.001) (D).

Download figure to PowerPoint

To evaluate the impact of HDL-C on early-EPC number and EDV, obese women were opportunely separated in two population according to HDL-C cut off of 50 mg/dL. Early-EPC number and EDV resulted significatively impaired in lower HDL-C subgroup (Figure 2a and 2b). Then, data were analyzed according to LDL-C plasma level to investigate whether the HDL-C impact on early-EPC number and EDV was dependent upon it. Circulating early-EPC number and EDV were significantly decreased in the hypercholesterolemic (LDL-C > 130 mg/dL) subgroup compared with normocholesterolemic (LDL-C < 130 mg/dL) subjects (Table 2). Data were then evaluated according to HDL-C plasmatic levels in both LDL-C < 130 mg/dL and ≥130 mg/dL subgroups. When we focused on early-EPC, HDL-C lower than 50 mg/dL resulted in significantly (P < 0.001) decreased early-EPC number, in the presence of both normal cholesterol plasma levels and hypercholesterolemia (Figure 3a). Interestingly, in the normal HDL-C level subgroup, no difference in early-EPC number was present between hypercholesterolemic and nonhypercholesterolemic subjects. High association of low circulating early-EPC number with low HDL-C was shown by OR 8 (CI 1.4-74, P < 0.05) and 10 (CI 2.0-50, P < 0.01), respectively, for LDL-C level greater or less than 130 mg/dL. In addition, univariate analysis of the data demonstrated a marked direct correlation between circulating early-EPC number and HDL-C plasma levels (Pearson r = 0.53, P < 0.001).

image

Figure 2. Obese women were evaluated according to HDL-C < 50 mg/dL and ≥50 mg/dL. Both EPC number (A) and EDV (B) resulted affected according to the HDL-C status. Each histogram represents the mean ± S.E.M. “*” indicates significant difference (P < 0.01), ***P < 0.0001 vs. control group.

Download figure to PowerPoint

image

Figure 3. Obese women, subdivided into LDL-C ≥ 130 mg/dL and LDL-C < 130 mg/dL subgroups, were evaluated according to HDL-C < 50 mg/dL and ≥50 mg/dL. Both EPC number (A) and EDV (B) resulted affected according to the HDL-C status. Each histogram represents the mean ± S.E.M. “*” indicates significant difference (P < 0.01), ***P < 0.0001 vs. control group.

Download figure to PowerPoint

Table 2. Comparison between hypercholesterolemic and normocholesterolemic subjects
 Obese
 LDL-C < 130 mg/dLLDL-C ≥ 130 mg/dL
  1. BMI, body mass index; CHD N, family history of coronary heart disease; EPC, circulating endothelial progenitor cell number; FBFr, forearm blood flow ratio; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; WC, waist circumference. Values expressed as mean ± SEM.

  2. *P < 0.05, ***P < 0.001.

N4865
Age, y34.5 ± 0.834.3 ± 0.6
BMI, kg/m234 ± 0.935 ± 0.84
WC, cm109 ± 1.5112 ± 1.4
EPC15 ± 811.8 ± 6.1*
FBFr9.2 ± 0.67.4 ± 0.47*
Systolic BP, mm Hg124 ± 2.2126 ± 3.8
Diastolic BP, mm Hg76 ± 4.278 ± 3.1
Fasting glucose, mmol/L4.3 ± 1.24.4 ± 0.6
Fasting insulin, mmol/L5.4 ± 0.35.6 ± 0.4
Total cholesterol, mg/dL164 ± 8210 ± 12***
HDL-C, mg/dL40 ± 1.639 ± 1.4
LDL-C, mg/dL99 ± 2.4180 ± 3.9***
Triglycerides, mg/dL90.5 ± 10.392.3 ± 8.3
CHD N, %11 (23.4)13 (27.1)

Analyzing EDV in terms of HDL-C status (Figure 3b), we documented that low HDL-C was associated with decreased EDV in both hypercholesterolemic (P < 0.001) and nonhypercholesterolemic subjects (P < 0.001). At variance, normal HDL-C was associated with only slightly reduced EDV in hypercholesterolemic subjects (P < 0.05). Univariate data analysis indicated a good correlation between HDL-C and EDV (Pearson r = 0.63, P < 0.0001).

To further study the association among endothelial dysfunction, early-EPC number, and additional CV risk factors affecting our study population, multivariable analysis was performed for obese subjects, entering in the model independent variables significantly related in univariate analysis. To this purpose, EDV was entered as a dependent variable, and subsequently, age, WC, LDL-C, HDL-C, and early-EPC number were entered into the model as independent variables (Table 3). The model obtained entering age, WC, and LDL-C explained about the 52% of the variance in EDV in our population. Inclusion of HDL-C increased the variance in EDV explained by the model to 61%. Further inclusion of early-EPC number increased the variance in EDV explained by the model to 78%. In addition, adjustment for age, WC, HDL-C, and LDL-C had no effect on the association between early-EPC number and EDV documented by univariate analysis (Table 3).

Table 3. Multivariate regression models for parameters associated with endothelial dysfunction
 FBFr Beta (95% CI)P valuear2P valueb
  1. Multivariate analysis including variables showing significant correlation in univariate analysis.

  2. BMI, body mass index; EPC, circulating endothelial progenitor cell number; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; WC, waist circumference.

  3. a

    Level of significance for the association between FBFr (as evaluation of EDV) and the separate components of the model.

  4. b

    Level of significance of the model.

Model 1  0.34<0.0001
Age    
Model 2  0.51<0.0001
Age−0.44 to −0.19<0.0001  
WC−0.22 to −0.11<0.0001  
Model 3  0.52<0.0001
Age−0.43 to −0.18<0.0001  
WC−0.03 to 0.005<0.0001  
LDL-C−0.21 to −0.09ns  
Model 4  0.61<0.0001
Age−0.35 to −0.12<0.0001  
WC−0.17 to −0.06<0.0001  
LDL-C−0.02 to 0.01ns  
HDL-C0.09-0.23<0.0001  
Model 5  0.78<0.0001
Age−0.14 to 0.06ns  
WC−0.08 to 0.00ns  
LDL-C−0.012 to 0.01ns  
HDL-C0.05-0.15<0.001  
EPC0.17-0.27<0.0001  

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Our data indicate that both circulating early-EPC number and EDV are significantly decreased in obese women compared with the control group. In addition, low HDL-C plasmatic levels are associated to further decrease of early-EPC and EDV in the presence of both high or normal LDL-C plasmatic levels.

Marked direct correlation between HDL-C plasma levels and circulating early-EPC number and good correlation between HDL-C and EDV were documented in univariate analysis. In addition, direct correlation was also demonstrated between early-EPC number and EDV. The adjustment for age, WC, and LDL-C had no effect on the association, indicating that HDL-C and early-EPC number are independent predictors of EDV.

The relationship between obesity and CVD is indeed complex. Some investigators suggested that the connection is indirect and dependent on the increased prevalence of diabetes, hypertension, and dyslipidemia, on the grounds of the widely recognized association between metabolic syndrome and increased risk of CVD [5]. Others have supported independent association between obesity and CVD risk [2], relying on the strong evidence suggesting that BMI is closely associated with CV mortality [13]. However, the use of BMI as a CV risk factor has been intensely debated [14]. BMI is not a good index of visceral fat [15], which is believed to be the prominent component of obesity associated with increased CV risk. This widely held view is supported by several studies and the recent meta-regression analysis reporting that 1 cm increase in WC enhances relative risk of CVD event by 2% [16].

Abdominal obesity is associated with low HDL-C levels due to impaired lipoprotein lipase activity and enhanced cholesteryl ester transfer protein-mediated lipid exchange. In addition, TG-rich HDL-C constitutes a better substrate for hepatic lipase, leading to further lowering of HDL-C levels. Atherogenic dyslipidemia clinically presents as elevated serum triglycerides levels, increased levels of sdLDL particles, and decreased levels of HDL-C [17]. Indeed, evidence suggests that, as BMI increases, dyslipidemia progressively develops and sdLDL concentration rises [18]. These changes are postulated to increase CVD risk by threefold to sixfold [19]. On this ground, we were interested in evaluating the role of HDL-C in the mechanisms that lead to the increased CV risk in obesity.

In our previous report, we documented that HDL-C is a strong determinant of EPC number. These cells were recently proposed as a marker of vascular bed health. Their number, as well as proliferative potential, may change under pathological conditions, including CV risk factors such as coronary artery disease [20] and diabetes mellitus. In survival analyses of longitudinal studies, low EPC count has been shown to independently predict CV events in patients with CVD, metabolic syndrome [21], or chronic renal failure [22]. In addition, it has been reported that some therapeutic procedures can increase EPC number. Treatment with statins [23], erythropoietin [24], antagonists of the angiotensin II receptor [25], as well as constant moderate physical activity have been documented to increase circulating EPC [26].

In this study, the effect of HDL-C on early-EPC number was evidenced when analyzing our subjects according to LDL-C level. Indeed, no early-EPC number change was appreciable between normocholesterolemic and hypercholesterolemic subjects with normal HDL-C, suggesting a protective role for HDL-C, and supporting recently reported in vitro evidence [10]. In addition, low HDL-C level resulted in a decrease of early-EPC number in normocholesterolemic obese women of the same order of magnitude of that found in hypercholesterolemic ones.

Contrasting data have been reported on the EPC number change in obesity. Our data showed decreased number of circulating early-EPC in obese women, and concur with the investigations reporting decreased EPC count, colony-forming capacity, proliferative activity, and weight loss-induced increased number of circulating EPC in obesity [27]. Why other studies found increased EPC number in obese patients [28] is not clear. This variance is likely due to methodological issues and subtype of EPC investigated. At now, in fact, a full agreement on the gold standard markers to identifying these cells is lacking [29].

We then evaluated the impact of HDL-C on endothelial dysfunction. The assessment of endothelial function as EDV is widely acknowledged to be a sensitive prognostic marker in the assessment of CV risk [30], and it has been related to EPC [31]. Low HDL-C was related, in the group of obese hypercholesterolemic women, to marked reduction in EDV, whereas normal HDL-C was related to less impaired EDV, again indicating a protective role of HDL-C, even when associated to major CV risk factors such as obesity and high LDL-C, and supporting the evidence that low HDL-C is an independent CV risk factor. The OR analysis indicated that low HDL-C has a negative impact on EDV in obese women, suggesting a mechanism responsible for the inverse relationship evidenced by clinical and epidemiological studies between HDL-C and the risk of CHD events.

The limitation of this study is represented by the inclusion of a population of female subjects only. The great difference in gender-related EPC number and variation of HDL-C is such that male and female subjects are to be considered two independent populations. However, although limited to female subjects, the data reported in this paper document, for the first time, the correlation between HDL-C and EDV and early-EPC number in obese subjects.

In conclusion, our data indicate that low HDL-C is strictly associated to low early-EPC number and impaired endothelial function, which are widely recognized as independent CV risk factors [30, 31] in obese subjects. The nature of this association and the underlying mechanisms as well as the impact of early-EPC and EDV evaluation on HDL-C prognostic value in the assessment of cardiovascular risk need to be further investigated.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

This work was partially supported by Ricerche Universitarie e Ricerche Ateneo Federato di Scienze Politiche Pubbliche Sanitarie MIUR grants.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References