Elevated Circulating Tissue Factor Procoagulant Activity, Factor VII, and Plasminogen Activator Inhibitor-1 in Childhood Obesity: Evidence of a Procoagulant State

Summary

Childhood obesity is rapidly increasing in prevalence. We compared circulating membrane-bound tissue factor (FIII, F3) procoagulant activity (TF-PCA) and plasma markers of coagulation, fibrinolysis and endothelial dysfunction in 21 obese (10.1±1.5 years, mean ± standard deviation) and 22 healthy weight children (9.9±1.6 years), classified by Body Mass Index (BMI). TF-PCA and factor VII coagulant activity (FVII:C), plasminogen activator inhibitor (PAI-1, SERPINE1) and soluble vascular cell adhesion molecule 1 (sVCAM1) were higher in obese children. BMI correlated positively with TF-PCA, FVII:C, and PAI-1. Childhood obesity is associated with a procoagulant state and endothelial dysfunction. Studies are needed to assess whether weight reduction reverses these abnormalities.

Introduction

The prevalence of obesity in children has tripled over the last three decades. Obesity in children is associated with childhood type 2 diabetes mellitus (T2DM) and increased cardiovascular risk. In adults both obesity and T2DM are associated with cardiovascular disease and alterations in blood that reflect an ongoing risk for thrombotic events and atherogenesis. Tissue factor (TF, FIII, F3) is the physiological initiating mechanism for blood coagulation and is both prothrombotic and proinflammatory (Tilley and Mackman 2006). Adipose tissue is a site of TF production in response to leptin and proinflammatory stimuli and, in mouse models of obesity, the gene encoding TF (F3) is upregulated (Faber, et al 2009). Studies in humans indicate that there is blood borne TF, which is functional, participates in thrombogenesis and is elevated in prothrombotic states (Key, et al 1998, Sambola, et al 2003, Vaidyula, et al 2006). We have reported increased circulating TF procoagulant activity (TF-PCA) in T2DM patients (Boden, et al 2007) and in healthy nondiabetic adults during hyperglycemia and hyperinsulinemia (Boden, et al 2007, Vaidyula, et al 2006). In adults, obesity is associated with a procoagulant state with elevated plasma factor VII coagulant activity (FVII:C), FVIII:C, fibrinogen and plasminogen activator inhibitor (PAI1, SERPINE1) (Faber, et al 2009, Mertens and Van Gaal 2002). Some of these proteins are also elevated in childhood obesity (Sudi, et al 2001). To our knowledge, circulating levels of membrane-bound TF have not been studied in childhood obesity. We tested the hypothesis that obese children have higher levels of circulating TF-PCA and of plasma markers of coagulation, fibrinolysis, and endothelial perturbation compared to healthy-weight children.

Materials and Methods

Participants

Forty-seven children (aged 8 – 12 years) were classified according to on Body Mass Index (BMI) percentile (CDC growth curves) as underweight (≤5th percentile; n=1), healthy weight (>5–84.9th percentile; n=22), overweight (≥85th-94.9th percentile; n=3) and obese (≥95th percentile; n=21) (Dean, et al 1996). We compared the findings in 21 obese children and 22 healthy weight children (Table I). Children with any medical conditions or taking any medications were excluded. The study was approved by the Institutional Review Board and informed consent and assent were obtained from all parents and children.

Table I. Demographic and clinical details, and markers of coagulation, endothelial function, fibrinolytic system, and microparticles in the healthy-weight and obese children of the healthy-weight and obese children.

Characteristics	Marker	Healthy-weight (n=22)	Obese (n=21)	p
Age (years)		9.9 ± 1.6	10.1 ± 1.5	0.624
Gender				0.280
	Male (%)	54.55	38.10
	Female (%)	45.45	61.90
Ethnicity				0.550
	African American (%)	68.18	80.95
	Hispanic (%)	18.18	4.76
	Mixed/African American (%)	4.55	4.76
	Mixed/Puerto Rican (%)	4.55	0
	Other (%)	0	4.76
	Puerto Rican (%)	0	4.76
	Puerto Rican/Caucasian (%)	1 4.55	0
Height (cm)		141.8 ± 13.56	150.9 ± 10.41	**0.018
Weight (kg)		34.9 ± 9.27	66.1 ± 17.67	*<0.0001
BMI		17.4 ± 1.39	28.7 ± 4.71	*<0.0001
BMI Percentile		52.7 ± 19.66	97.7 ± 1.19	*<0.0001
Coagulation
	TF-PCA (u/ml)	34.0 ± 13.5†	60.6 ± 32.5	*0.005
	TF-antigen (pg/ml)	44.7 ± 50.3	78.7 ± 88.5	0.136
	FVIIa (mu/ml)	119.4 ± 56.6	134.0 ± 99.6	0.933
	FVII:C (u/ml)	0.90 ± 0.20	1.03 ± 0.20	**0.032
	FVIII (u/ml)	1.56 ± 0.64	1.47 ± 0.49	0.971
	Fibrinogen (g/l)	3.07 ± 0.60	3.21 ± 0.45	0.193
	TAT (μg/l)	3.8.0 ± 8.6	2.1 ± 1.3	0.508
Fibrinolytic
	tPA (ng/ml)	5.1 ± 2.0	6.7 ± 3.4	0.070
	PAI-1 (ng/ml)	25.6 ± 15.0	37.3 ± 18.0	**0.034
Endothelial
	VWF:Ag (mu/ml)	987.9 ± 130.8	1057.4 ± 139.8	0.074
	sVCAM1 (ng/ml)	526.5 ± 184.8	709.2 ± 444.5	0.052
Microparticle (nM)		6.4 ± 4.5	9.2 ± 7.5	0.547

Values are mean ± standard deviation unless otherwise noted; BMI- Body mass index, * p<0.01, ** p<0.05, n- number of subjects, TF-PCA- tissue factor procoagulant activity, FVIIa-activated Factor VII, FVII:C- Factor VII coagulant activity, FVIII- Factor VIII, TAT- thrombin-antithrombin complex, VWF:Ag- von Willebrand factor antigen, sVCAM1- soluble vascular cell adhesion molecule 1, tPA- tissue plasminogen activator, PAI-1- plasminogen activator inhibitor-1;

^*p<0.01,

^**p<0.05,

^†n=21.

Blood collection and plasma preparation

Blood (9 ml) was drawn into 1 ml of 3.8% sodium citrate by venipuncture. Two aliquots (1 ml) were removed for TF-PCA assay; the remainder was centrifuged (20 min, 1590×g) to obtain plasma. A plasma aliquot was centrifuged (13,000×g, 2 min) to remove residual cells prior to microparticle assay.

Assays

Whole blood TF-PCA

TF-PCA was measured in whole blood cell lysates with a two-stage clotting assay using recombinant activated FVII (FVIIa; American Diagnostica Inc., Stamford, CT) and factor X (Haematologic Technologies Inc., Essex Junction, VT) (Key, et al 1998, Vaidyula, et al 2006). This assay measures cell-bound and microparticle-associated TF.

Plasma Measurements

TF (FIII), PAI-1, tissue plasminogen activator (tPA) and von Willebrand factor antigen (VWF:Ag) were determined by enzyme-linked immunosorbent assay (ELISA) (American Diagnostica Inc.). FVIII:C and FVII:C and fibrinogen were measured by standard clotting assays (Vaidyula, et al 2006); FVIIa activity was measured by a clotting assay (Diagnostica Stago, Parsippany, NJ) using soluble TF. Thrombin-antithrombin complexes (TAT) (Enzygnost; Dade Behring, Marburg, Germany) and sVCAM1 (Diaclone, Besancon Cedex, France) were measured by ELISAs. Procoagulant activity of microparticles in plasma was determined by a thrombin generation assay (Zymuphen; Hyphen Biomed, Neuville-sur-Oise, France).

Statistical analysis

Differences in demographic factors between the two groups were evaluated with Student t-test. The Satterthwaite approximation for p-values was applied when the equal variance assumption was suspect. Pearson chi-square tests were used to compare categorical variables, and the extension of the Fisher exact test was utilized when assumptions underlying the chi-square test were violated. Differences between groups with respect to biomarkers were evaluated with Wilcoxon rank-sums tests. Spearman correlation coefficients, using data from all 47 participants, were used to assess the relationships between BMI and the biomarkers and among biomarkers. A P value <0.05 was considered statistically significant.

Results

Circulating membrane-bound TF-PCA was significantly higher in obese children than in healthy weight children (Table I). Plasma FVII:C was significantly higher in the obese than in the healthy weight group, while plasma TF antigen, FVIIa, FVIII, fibrinogen and TAT were not different between the groups (Table I). PAI-1 was significantly higher in the obese children; tPA antigen levels were slightly but not significantly higher in obese children. Endothelial markers VCAM1 and VWF:Ag were higher in obese than in healthy weight children, but did not reach statistical significance. Using all 47 subjects, BMI was significantly and positively correlated with TF-PCA, FVII:C, and PAI-1 (Fig. 1) with suggestive positive correlations between BMI and sVCAM1 (r=0.27, p=0.07), fibrinogen (r=0.27, p=0.06) and VWF:Ag (r=0.26, p=0.08).

Figure 1.

Relationship between body mass index (BMI) and (A) circulating membrane-bound TF-PCA, (B) plasma FVII:C and (C) plasma PAI-1. The data shown includes all subjects studied: underweight, healthy weight, overweight and obese.

Plasma sVCAM1 correlated positively with TF-PCA (r=0.36, p=0.02) and FVII:C (r=0.35, p=0.02) (Supplemental Table 1), suggesting that TF-FVII driven thrombin generation contributes to endothelial perturbation and total TF-PCA. Plasma tPA (also endothelial-derived) correlated with FVIIa (r=0.42, p<0.01), and may reflect the same mechanism. PAI-1 levels correlated with circulating microparticles (r=0.30, p=0.04), which is interesting because PAI-1 promotes endothelial microparticle formation (Brodsky, et al 2002). Plasma FVIII (r=-0.42, p<0.01) and VWF:Ag (r=-0.37, p=0.01) correlated inversely with microparticles, possibly related to their removal from plasma via binding to microparticles.

Discussion

In this study, circulating TF-PCA levels were higher in obese than in healthy-weight children (Table I) and correlated with BMI (Fig. 1). Strikingly, the levels in obese children are comparable to those reported in adults with T2DM (Boden, et al 2007). Circulating TF plays a crucial role in coagulation and thrombosis (Key, et al 1998, Tilley and Mackman 2006). There is enhanced thrombus formation when blood with elevated circulating TF levels is perfused over a collagen-coated surface (Sambola, et al 2003). Our findings suggest that, even in young children, obesity is associated with a potential prothrombotic state.

Obesity, TF, and monocytes are closely linked. The monocyte, which is the macrophage precursor, is a major source of TF in circulation, and macrophage and adipocyte functions converge in obesity. Macrophages infiltrate adipose tissue and both cells produce proinflammatory cytokines; fat tissue from obese individuals has a gene expression profile suggestive of macrophages. Obesity upregulates F3 expression in genetically obese mice and in lean mice insulin induces F3 mRNA in several tissues, which is relevant because obesity is associated with hyperinsulinemia (Faber, et al 2009). Overall, these studies point to a cross-talk between obesity, inflammation and coagulation, with obesity promoting inflammation and coagulation (Zhang and Lawrence 2011). Moreover, TF triggers signalling through protease activated receptor-2 (PAR2) and recent studies (Badeanlou, et al 2011) implicate TF directly in the development of diet-induced obesity and insulin resistance. They showed that TF-VIIa-PAR2 signalling of adipocytes crucially regulates weight gain and, in adipose tissue, macrophages promote inflammation and insulin resistance. TF blockade has been proposed as a new therapeutic approach targeting obesity. Based on our findings, circulating TF may serve as a biomarker for these regulatory interactions in children.

In this study, plasma FVII:C was higher in obese than in healthy-weight children and correlated with BMI (Figure 1). Plasma FVII:C is elevated in adults with metabolic syndrome, diabetes mellitus and obesity and is associated with increased risk for acute cardiovascular events (Green, et al 2010, Mertens and Van Gaal 2002). The increased plasma FVII may contribute to the enhanced thrombin generation reported in the blood of obese children. In healthy young adults, plasma FVII levels have been positively associated with increased carotid intimal-media thickness (indicating subclinical atherosclerosis), and this association was modified by BMI (Green, et al 2010). Even in children, BMI was correlated with carotid intimal-media thickness (Beauloye, et al 2007) and in our studies, plasma FVII:C and TF-PCA correlated positively with BMI.

PAI-1 levels were higher in obese children and correlated with BMI. Elevated PAI-1 levels reflect endothelial dysfunction and are associated with increased cardiovascular risk (Mertens and Van Gaal 2002). PAI-1 release is stimulated by insulin and chronic inflammation and the Serpine1 gene is upregulated in genetically obese mice (Mertens and Van Gaal 2002). Together, the elevated TF-PCA, FVII and PAI-1 levels indicate that childhood obesity is a procoagulant state.

Endothelial surface markers are upregulated by cytokines and inflammation, and increased levels reflect endothelial dysfunction. Obese children had elevated sVCAM1, which correlated with TF-PCA. Our findings assume importance because elevated endothelial markers have predicted subsequent development of T2DM and cardiovascular events (Meigs, et al 2004).

In summary, we found that childhood obesity was associated with changes suggesting a procoagulant and proatherogenic state with concomitant endothelial dysfunction. Although children studied were mostly African American or Hispanic we are unaware of racial differences in these biomarkers in children; we believe the findings apply to other groups also. Larger studies are needed in children to define whether weight reduction can reverse these abnormalities and impact clinical consequences.