Background/Aims: Obesity leads to increased risk of thromboembolic events in adults, but few studies have addressed the relationship between obesity and thrombogenic risk during childhood. The aim of this study was to evaluate the prothrombotic state of obese children in comparison with healthy children. Methods: Thrombin generation, fibrinogen, and D-dimer levels, along with metabolic parameters, were measured in 72 prepubertal children, of which 47 were obese and 25 eutrophic. Results: A significant increase in thrombin generation, fibrinogen, and dyslipidemia was found among obese patients. Conclusion: A prothrombotic state develops in childhood obesity during the prepubertal phase.

The prevalence of obesity has been rising across all age groups. Between 1980 and 2013, the number of people overweight or obese has increased globally by 27.5% in adults and 47.1% in children [1]. The prevalence of obesity among children and adolescents has increased substantially in developing countries, for boys (from 8.1% to 12.9%) and girls (from 8.4% to 13.4%) [1].

Excess body weight and adipose tissue accumulation leads to pathological processes such as insulin resistance, systemic arterial hypertension, and dyslipidemia, associated with the development of atherosclerosis. Obesity is also a significant risk factor for the development of venous thromboembolism [2], and the combination of metabolic and hemostatic alterations promotes the development of coronary heart disease in adults. Clinical and epidemiological studies have clearly established a link between obesity and cardiovascular diseases, thrombosis, insulin resistance, and type 2 diabetes [3].

Risk factors for venous thromboembolism in obese patients include inflammation, reduced fibrinolysis, increased thrombin generation, and platelet hyperactivity [4, 5]. The generation of the enzyme thrombin, whose precursor is prothrombin, is central to fibrin clot formation. Fibrin cleavage during fibrinolysis results in the formation of D-dimers [6‒8]. Besides acting as the fibrin precursor, fibrinogen is the major determinant of plasma viscosity and erythrocyte aggregation, functioning as a key element in the coagulation cascade. In obese adults and children, a significant increase in fibrinogen concentration may occur [9, 10].

The thrombin generation (TG) test has been increasingly recognized as a versatile diagnostic tool to evaluate thrombosis and hemostasis [11]. The assay provides a curve of thrombin concentration over time characterized by the following parameters: latency time (lag time), endogenous thrombin potential (ETP), and peak (Fig. 1). Shorter lag times or increased ETP and peak heights indicate higher thrombin generation [12‒15].

Fig. 1.

Representative graph of the thrombin generation curve illustrating the variables analyzed. Adapted from Tripodi et al. [13]. ETP, endogenous thrombin potential.

Fig. 1.

Representative graph of the thrombin generation curve illustrating the variables analyzed. Adapted from Tripodi et al. [13]. ETP, endogenous thrombin potential.

Close modal

Some modifications have been introduced in TG tests to make them better reproduce in vivo coagulation, such as the use of thrombomodulin (TM). This protein activates protein C, one of the most powerful natural anticoagulants, which provides a coagulation control mechanism by reducing thrombin formation [15].

Few studies have evaluated the prothrombotic state of obese children. Our study aimed to assess the occurrence of such condition in children with obesity [16, 17].

The study involved a convenience sample of 50 obese and prepubertal patients, aged 7–10 years, who attended an outpatient obesity clinic at a tertiary hospital between October 2012 and December 2015. Three patients left the study at the beginning of follow-up. A total of 47 children, aged 7–10 years, completed the study. Twenty-five healthy children of similar age and gender distribution were invited to participate as a control group. This sample size has a statistical power of 90%. There is no family history of dyslipidemia or hypercoagulability states.

Inclusion criteria were children aged 7–10 years, prepubertal, and classified as obese according to the WHO criteria, that is, with a BMI Z-score above +2 SD [2]. Exclusion criteria were pubertal patients using corticoids, oral contraceptives, or anticoagulant therapy, syndromic patients with a history of coagulopathies, neoplasms, or type 2 diabetes.

Anthropometric data were obtained for all patients. Blood samples were also collected to perform laboratory testing. To assess prothrombotic state, thrombin generation, fibrinogen, and D-dimer were measured. For metabolic evaluation, glucose, insulin, and total cholesterol and fractions were assessed. Samples were collected and prepared according to the following criteria and procedures:

  • Blood samples were collected in the morning, after a 12-h fast.

  • Blood was drawn by peripheral venipuncture from a large vein in the upper limb, applying a tourniquet only during puncturing. The first tube filled with blood was used for other tests while subsequent draws were used for hemostatic tests.

  • For hemostatic tests, a total volume of 10.8 mL of whole blood was collected, distributed by 3 test tubes of 3.6 mL buffered with 3.2% sodium citrate.

  • Platelet-poor plasma was separated by centrifugation at 2,500 g for 15 min at room temperature, aliquoted in polystyrene tubes, and stored at −80°C. Samples were thawed immediately before measurements at 37°C.

  • Samples were separated and sent to the Laboratory of Hormones and Molecular Genetics at the University of São Paulo (LIM42, USP) for analysis.

D-dimer was assessed by classical ELISA using a commercial kit (Horiba ABX, Stago). Fibrinogen levels were measured according to the modified Clauss method using Fibriquik® reagent (Biomérieux).

Thrombin generation was measured on a microplate fluorometer (Fluoroskan Ascent; Thermo Scientific), using a software that converts fluorescence intensity into thrombin activity by comparison with an internal reference. The TG curve parameters analyzed were lag time, ETP, and peak, with or without TM (Fig. 1). Insulin was assessed by immunofluorometric assay using a commercial kit (AutoDELFIA; PerkinElmer). C-reactive protein (CRP) was measured by nephelometry. Blood glucose was assessed by the enzymatic colorimetric method GOD/PAP. Total cholesterol, high-density lipoprotein (HDL), and low-density lipoprotein (LDL) cholesterol were measured by enzymatic colorimetric methods, namely, CHOD/PAP (total cholesterol) and homogeneous assays (HDL and LDL).

The homeostatic model assessment of insulin resistance (HOMA-IR) was calculated using the formula: fasting insulin (µU/L) × fasting glucose (mg/dL)/22.5. The frequency distribution of categorical variables was expressed as the number of observations and percentage. Continuous variables were expressed as median and interquartile range considering that most of them did not have a Gaussian distribution.

The Mann-Whitney U test was used to compare the medians of each variable between the obesity group and the control group. A p value of 0.05 was considered significant. The Spearman’s correlation coefficient (rho) was used to evaluate the correlation between BMI and hemostatic variables.

The obesity group (OB) consisted of 47 children, of whom 29 were boys (61.7%), with a mean age of 9 years and a mean BMI Z-score of +3.3. The control group (C) consisted of 25 healthy children, including 13 boys (52%), with a mean age of 8 years and a mean BMI Z-score of +0.2 (Table 1).

Table 1.

Comparison of clinical and hemostatic variables between obese and healthy children

 Comparison of clinical and hemostatic variables between obese and healthy children
 Comparison of clinical and hemostatic variables between obese and healthy children

Among the OB group, 5 children (10.6%) showed increased fibrinogen levels and 9 (19%) presented higher than normal values of D-dimer. All participants in the control group had normal fibrinogen and D-dimer. There is no definition of normal values for the test of generation of thrombin for this age group.

The metabolic evaluation of OB subjects revealed alterations in total cholesterol (>170 mg/dL) in 33 children (70%), HDL concentrations below 40 mg/dL in 14 children (30%), and high levels of LDL (>100 mg/dL; 31 patients = 66%), triglycerides (>100 mg/dL; 24 children = 51%), CRP (>5 mg/dL; 9 patients = 19%), and HOMA-IR (>2.5; 21 children = 44.6%). The control group had normal levels of total cholesterol, HDL, LDL, and triglycerides (Tables 2, 2, 3).

Table 2.

Comparison of metabolic variables between obese and controls

 Comparison of metabolic variables between obese and controls
 Comparison of metabolic variables between obese and controls
Table 3.

Comparison of clinical, metabolic variable and hemostatic variables between obese and healthy children

 Comparison of clinical, metabolic variable and hemostatic variables between obese and healthy children
 Comparison of clinical, metabolic variable and hemostatic variables between obese and healthy children

When comparing OB and C groups, we observed that fibrinogen (p = 0.0003) and triglyceride levels (p = 0.007) were significantly increased in patients with obesity, while HDL levels were significantly reduced (p = 0.02) (Tables 1, 1, 2, 3). The thrombin generation measurements showed that lag time, with and without TM, was shorter in the OB group (p = 0.0001), indicating early thrombin formation among obese subjects. Both ETP and the peak of the TG curve were higher in OB subjects in the presence of TM (p = 0.01 and p = 0.008, respectively) (Table 1).

In addition, there was a positive correlation between BMI and ETP, peak, fibrinogen, and D-dimer and a negative correlation between BMI and lag time. There was no significant correlation between insulin and hemostatic variables (Table 4).

Table 4.

Correlation between body mass index, insulin, and hemostatic variables

 Correlation between body mass index, insulin, and hemostatic variables
 Correlation between body mass index, insulin, and hemostatic variables

In this study, we assessed metabolic and hemostatic parameters in prepubertal children with obesity and healthy controls. We observed higher levels of fibrinogen and D-dimer in 10% and 19% of obese patients, respectively. In addition, dyslipidemia was present in most obese children evaluated. Another important finding was an increase in the levels of the inflammatory marker CRP and in the values of HOMA index, affecting about 20% and 40% of obese children, respectively. These observations are consistent with previous literature and reinforce the hypothesis that childhood obesity is associated with a prothrombotic state and increased cardiovascular risk [10].

Compared to healthy children, obese patients showed significantly higher levels of fibrinogen and triglycerides, along with a reduction in HDL. All these data also agree with previous findings [9, 10].

Compared to tests based on coagulation time such as prothrombin time or activated partial thromboplastin time, the TG test is a more reliable indicator of the overall function of the hemostatic system. Among all TG variables analyzed (lag time, ETP, and peak), ETP is the most robust because it derives from the final amount of fluorogenic product released by free thrombin. Peak is more sensitive to variations in the mechanism of thrombin generation but is not considered robust because it results from the initial product-time curve [11‒14].

In 2006, Cimenti et al. [16] demonstrated for the first time that ETP is increased in children and adolescents with severe obesity, compared to normal-weight children, confirming the hypothesis that obese children present a hypercoagulable state. Later, in 2010 and 2011, Fritsch et al. [18] and Siklar et al. [17] also observed hypercoagulability in obese children and adolescents, demonstrated by greater thrombin generation. Most patients assessed in these studies, however, were pubertal.

Our analysis of thrombin generation showed a statistically significant reduction in lag time, with and without TM, in the group of obese patients, supporting early thrombin production in these children. Two other TG parameters, ETP and peak, also were elevated in obese patients, supporting the hypothesis that prepubertal children with obesity have greater thrombin generation than healthy children.

We observed a significant and positive correlation between BMI and ETP, peak, fibrinogen, and D-dimer, a negative correlation between BMI and lag time, and no correlation between insulin and hemostatic parameters. This suggests that increases in BMI correlate with worsening in prothrombotic states but not with insulin resistance. Our hypothesis is that the increase in adipose tissue is an independent factor to lead to a prothrombotic state in prepubertal obese children.

Our results show that childhood obesity is associated with a prothrombotic state characterized by a significantly increased fibrinogen and thrombin generation, as well as metabolic alterations (higher total cholesterol and triglycerides and lower HDL). Our findings are consistent with previous literature showing the occurrence of alterations in the coagulation system of obese children [14‒16]. To the best of our knowledge, this is the first study to demonstrate a significant change in thrombin generation in prepubertal children with obesity.

Our study shows that children with obesity present significantly increased thrombin generation, in addition to higher fibrinogen levels, even before the onset of puberty. These findings demonstrate for the first time that obese children develop a prothrombotic state early in life, during the prepubertal phase.

The authors would like to thank the patients and their families. They acknowledge Tania Rubia Flores da Rocha for her substantive contributions to this study.

Approval was obtained from the CAPPesq Ethics Committee, written informed consent was obtained, and all data were anonymized. The CAPPesq reference number is 0138/10. This study was conducted in accordance with the Declaration of Helsinki.

There are no conflicts of interest that could be perceived as prejudicing the impartiality of the research reported.

The authors did not have external funding for this research.

Cominato L. and Damiani D. partook in the conception of the project, the acquisition of data, data analysis, and writing and approving the manuscript. Carneiro J.D.A. partook in the conception of the project, data analysis, and writing and approving the manuscript. Franco R.R., Ybarra M., Frascino A.V., and Steinmetz L. partook in data acquisition, data analysis, and approving the manuscript. Ferraro A.A. partook in data analysis and writing and approving the manuscript.

All data generated or analyzed during this study are included in this article. Further enquiries can be directed to the corresponding author.

1.
Ng
M
,
Fleming
T
,
Robinson
M
,
Thomson
B
,
Graetz
N
,
Margono
C
,
.
Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: a systematic analysis for the global burden of disease study 2013
.
Lancet
.
2014
;
384
(
9945
):
766
81
.
2.
Obesity: preventing and managing the global epidemic. Report of a WHO consultation
.
World Health Organ Tech Rep Ser
.
2000
;
894
:
i
xii
3.
Yusuf
S
,
Hawken
S
,
Ounpuu
S
,
Bautista
L
,
Franzosi
MG
,
Commerford
P
,
.
Obesity and the risk of myocardial infarction in 27,000 participants from 52 countries: a case-control study
.
Lancet
.
2005
;
366
(
9497
):
1640
9
.
4.
Samad
F
,
Ruf
W
.
Inflammation, obesity, and thrombosis
.
Blood
.
2013
;
122
(
20
):
3415
22
.
5.
Stein
PD
,
Goldman
J
.
Obesity and thromboembolic disease
.
Clin Chest Med
.
2009
;
30
(
3
):
489
viii
.
6.
Stein
PD
,
Beemath
A
,
Olson
RE
.
Obesity as a risk factor in venous thromboembolism
.
Am J Med
.
2005
;
118
(
9
):
978
80
.
7.
RF F. Fisiologia da coagulação, anticoagulação e fibrinólise revistas usp2001.
8.
Orkin
SH
,
Fisher
DE
,
Ginsburg
D
,
Look
AT
,
Lux
SE
,
Nathan
DG
.
Nathan and Oski’s hematology and oncology of infancy and childhood
;
2015
.
9.
Hafez
M
,
El-Masry
S
,
Musa
N
,
Fathy
M
,
Hassan
M
,
Hassan
N
,
.
Relationship between visceral obesity and plasma fibrinogen in obese children
.
J Pediatr Endocrinol Metab
.
2016
;
29
(
3
):
289
96
.
10.
Balagopal
PB
,
de Ferranti
SD
,
Cook
S
,
Daniels
SR
,
Gidding
SS
,
Hayman
LL
,
.
Nontraditional risk factors and biomarkers for cardiovascular disease: mechanistic, research, and clinical considerations for youth: a scientific statement from the American heart association
.
Circulation
.
2011
;
123
(
23
):
2749
69
.
11.
Ninivaggi
M
,
Apitz-Castro
R
,
Dargaud
Y
,
de Laat
B
,
Hemker
HC
,
Lindhout
T
.
Whole-blood thrombin generation monitored with a calibrated automated thrombogram-based assay
.
Clin Chem
.
2012
;
58
(
8
):
1252
9
.
12.
Al Dieri
R
,
de Laat
B
,
Hemker
HC
.
Thrombin generation: what have we learned
.
Blood Rev
.
2012
;
26
(
5
):
197
203
.
13.
Brummel-Ziedins
K
,
Vossen
CY
,
Rosendaal
FR
,
Umezaki
K
,
Mann
KG
.
The plasma hemostatic proteome: thrombin generation in healthy individuals
.
J Thromb Haemost
.
2005
;
3
(
7
):
1472
81
.
14.
Brummel-Ziedins
KE
,
Everse
SJ
,
Mann
KG
,
Orfeo
T
.
Modeling thrombin generation: plasma composition based approach
.
J Thromb Thrombolysis
.
2014
;
37
(
1
):
32
44
.
15.
Tripodi
A
.
Thrombin generation assay and its application in the clinical laboratory
.
Clin Chem
.
2016
;
62
(
5
):
699
707
.
16.
Cimenti
C
,
Mangge
H
,
Haidl
H
,
Zach
D
,
Muntean
W
.
Thrombin generation in severely obese children
.
J Thromb Haemost
.
2006
;
4
:
1834
6
.
17.
Siklar
Z
,
Öçal
G
,
Berberoğlu
M
,
Hacihamdioğlu
B
,
Savas Erdeve
S
,
Eğin
Y
,
.
Evaluation of hypercoagulability in obese children with thrombin generation test and microparticle release: effect of metabolic parameters
.
Clin Appl Thromb Hemost
.
2011
;
17
(
6
):
585
9
.
18.
Fritsch
P
,
Kleber
M
,
Rosenkranz
A
,
Fritsch
M
,
Muntean
W
,
Mangge
H
,
.
Haemostatic alterations in overweight children: associations between metabolic syndrome, thrombin generation, and fibrinogen levels
.
Atherosclerosis
.
2010
;
212
(
2
):
650
5
.