| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Center for Womens Health, Department of Medicine, Columbia University College of Physicians and Surgeons, New York, New York 10032
Address all correspondence and requests for reprints to: Elsa-Grace V. Giardina, M.D., Center for Womens Health, Columbia University Medical Center, 630 West 168th Street, PH 3346, New York, New York 10032. E-mail: evg1{at}columbia.edu.
| Abstract |
|---|
|
|
|---|
Plasminogen activator inhibitor (PAI-1), tissue plasminogen activator, von Willebrand factor, fibrinogen, and fibrin D-dimer predict risk of coronary heart disease. Yet there is limited information describing the status of endogenous hormone concentrations and hemostatic and coagulation factors in premenopausal women.
Twenty premenopausal women, mean age 34 ± 7 yr, underwent testing over a cycle to measure endogenous hormones and hemostatic factors: estradiol and progesterone, FSH, LH; PAI-1, tissue plasminogen activator, von Willebrand factor, fibrin D-dimer, and fibrinogen as well as lipids: total cholesterol, high-density lipoprotein-cholesterol, low-density lipoprotein-cholesterol, and triglycerides.
There was cyclical variability in estradiol (P < 0.01) and progesterone (P < 0.001) during the follicular and luteal phases. Moreover, there was intra- and interindividual cyclical variation in hemostatic risk factors. Measures of PAI-1 (P < 0.01) and D-dimer (P < 0.05) differed during the follicular and luteal phases. As estradiol concentration increased, PAI-I decreased. There was a significant correlation between total cholesterol and PAI-1 (r = 0.56, P < 0.05), low-density lipoprotein-cholesterol and PAI-1 (r = 0.50, P < 0.05) as well as between total cholesterol and fibrinogen (r = 0.61, P < 0.05).
There is significant cyclical variability in estradiol, FSH, and progesterone as well as the hemostatic factors, PAI-1 and fibrin D-dimer. Characterization of emerging hemostatic risk factors enhances understanding of normal physiology, provides insight into the relation between estrogen and hemostatic factors, and raises the potential for predicting coronary heart disease even in relatively young women.
| Introduction |
|---|
|
|
|---|
Despite evidence for an association between atherothrombotic factors in postmenopausal women treated with ERT and HRT, there is limited information describing endogenous hemostatic and coagulation factors in premenopausal women. We considered that the known cyclical variability of estrogen and progesterone in premenopausal women might be associated with variability in coagulation and fibrinolytic factors, parallel to the effect seen after HRT in postmenopausal women. Accordingly, this study was designed to characterize the time course and expression of PAI-1, t-PA, vWF, fibrin D-dimer, and fibrinogen in premenopausal women. Characterization of hemostatic risk factors in the premenopausal woman could enhance our understanding of normal physiology, provide insight into the relation between estrogen and hemostatic factors, and raise the potential for predicting CHD.
| Subjects and Methods |
|---|
|
|
|---|
At enrollment and at the conclusion of the protocol, each subject underwent a history and physical examination as well as complete blood count, electrolytes, hepatic and renal function, total cholesterol, low-density lipoprotein (LDL)-cholesterol, high-density lipoprotein (HDL)-cholesterol, triglycerides, thyroid function, estradiol, FSH, progesterone, and LH. Urine for analysis was also collected. All subjects described normal menstrual cycles for more than 6 months before study entry. The cycles were divided into weekly phases based on history of the last menstrual cycle, duration of uterine bleeding, and morning body temperatures. Each week, samples were collected on 2 separate days for estradiol, FSH, progesterone, LH, PAI-1 antigen, t-PA antigen, vWF antigen, fibrin D-dimer, and fibrinogen. The twice-weekly measures respectively were averaged. The weeks corresponded to: wk 1 (d 7), early follicular phase, range 59 d; wk 2 (d 14), late follicular phase, range 1216 d; wk 3 (d 21), early luteal phase, range 1923 d; wk 4 (d 28), late luteal phase, range 2630 d.
Hemostatic variables
Blood samples were collected in the morning between 0800 and 0900 h after a 12-h fast to avoid the diurnal variation in coagulation and fibrinolytic variables. Blood samples were drawn at rest from a large antecubital vein into 0.1 M trisodium citrate tubes for analysis. The first 23 ml of blood was discarded, and samples were analyzed only if venous return was prompt throughout. Tubes were immediately centrifuged, and spun at 3000 x g for 20 min at 4 C, and centrifugation was performed within 1 h. Plasma was removed into aliquots and stored in 80 C for up to 4 months before analysis. Assays for PAI-1, t-PA, vWF, and human C-reactive protein (C-RP) were determined using the respective ELISA kits purchased from American Diagnostics Inc. (Greenwich, CT). PAI-1 antigen ELISA measured the total quantity of PAI-1 present (free or complexed with t-PA; in active or inactive form) (18). The t-PA antigen ELISA measured the total amount of t-PA present (free and complexed t-PA, single- and double-chain t-PA) (18). The detection limit for C-RP was 0.35 ng/ml; intra- and interassay coefficients of variation (CVs) were between 3 and 4.5% and 3 and 7%, respectively. Fibrin D-dimer was analyzed using the Asserachrom D-Di ELISA from Diagnostica Stago (Asnieres-Sur-Seine, France). Estradiol, progesterone, FSH, and LH concentrations were determined by their specific immunoassays from Diagnostic Products Corp. (Los Angeles, CA) according to the manufacturers instructions. Interassay CVs for estradiol at high, middle, and low control are 3.5, 4.6, and 9.6%, respectively; intraassay CVs ranged from 3 to 10%. Total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, and fibrinogen concentrations were determined enzymatically.
Statistical analysis
Data are expressed as mean ± 1 SD. The significance of changes in measured variables between the follicular and luteal phases was determined by the nonparametric Wilcoxon paired sample test. Weekly changes were analyzed using Friedmans test, the nonparametric analog of the repeated-measures ANOVA. Dunns multiple comparison test was used to test post hoc differences among weekly data. Spearmans rank correlation was used to determine the relation between variables at specific time points during the menstrual cycle as well as averaged across the menstrual cycle. P < 0.05 was considered significant for all analyses.
| Results |
|---|
|
|
|---|
Twenty-three premenopausal women enrolled; three did not complete the protocol because complete data points over a cycle were not collected due to noncompliance with strict timing of phlebotomies (2) or concurrent infection (1). Twenty subjects, free of overt disease as assessed by medical history, physical exam, and baseline laboratory data participated (Table 1
). The mean age of the subjects was 34 ± 7 yr (range 2344); weight was 150 ± 36 lb (range 108248); and body mass index (BMI) was 25 ± 6 (range 2040). The average duration of the menstrual cycle was 29 ± 5 d (range 2430).
|
Hormone concentrations
For estradiol there were differences between the follicular and luteal phases and mean change (
) = 30 ± 53 pg/ml (P < 0.01) (Table 2
) as well as wk 1 and 2,
= 39 ± 48 pg/ml (P < 0.05); wk 1 and 3,
= 51 ± 43 pg/ml (P < 0.01); and wk 1 and 4,
= 65 ± 65 pg/ml (P < 0.001) (Fig. 1
). For progesterone there were also differences between the follicular and luteal phases,
= 6.3 ± 3.2 ng/ml (P < 0.001) as well as wk 1 and 4,
= 7.7 ± 6.1 ng/ml (P < 0.001), and wk 2 and 4,
= 8.3 ± 6.0 ng/ml (P < 0.001). For FSH there were also significant differences between the follicular and luteal phases,
= 2.3 ± 6.7 mIU/ml (P < 0.05) as well as wk 1 and 4,
= 4.2 ± 10.4 mIU/ml (P < 0.01), and wk 2 and 4,
= 3.4 ± 6.2 mIU/ml (P < 0.001). The differences between follicular and luteal phases for LH did not reach significance.
|
|
There was intra- and interindividual variability over the cycle for each of the hemostatic variables (Table 2
). The difference between the follicular and luteal phases was significant for PAI-1 (
= 10 ± 14 ng/ml, P < 0.01 (Fig. 2
), and for D-dimer (
= 57 ± 93 ng/ml, P < 0.05); the changes for t-PA, vWF, and fibrinogen did not reach significance. There was an inverse relation between the concentration of estradiol and PAI-1 (Fig. 3
). There was no difference between measures of C-RP during the follicular (0.41 ± 0.18 mg/l) or luteal (0.38 ± 0.10 mg/l) phases.
|
|
There was a significant relation between the total cholesterol and PAI-1 (r = 0.56, P < 0.05) (Fig. 4
) and total cholesterol and fibrinogen (r = 0.61, P < 0.05). There was also a significant relation between LDL-cholesterol and PAI-1 (r = 0.50, P < 0.05). There was no significant relationship between PAI-1 and triglycerides.
|
There was no significant relation between BMI and hormones (estradiol, progesterone, FSH, LH) during the follicular and luteal phases. There was a significant correlation between BMI and t-PA during the follicular (r = 0.51, P < 0.05) and luteal phases (r = 0.64, P < 0.01). No change in BMI, body weight, blood pressure, or lipids occurred over the study.
| Discussion |
|---|
|
|
|---|
Despite substantial data in older subjects (8, 9, 10, 11, 12), information describing coagulation and fibrinolysis in premenopausal women is limited. The current study increased our understanding of the dynamic physiology of hemostatic factors by describing their variability during the menstrual cycle. In the current study, the time course of PAI-1, t-PA, vWF, fibrin D-dimer, and fibrinogen is described as well as intra- and interindividual variability. We report that in normal premenopausal women, cyclical changes in endogenous estradiol (P < 0.01), progesterone (P < 0.001), FSH (P < 0.05), PAI-1 (P < 0.01), and fibrin D-dimer (P < 0.05) were observed. Furthermore, an inverse relation between estradiol and PAI-1 was observed, such that as estradiol increased, PAI-1 decreased paralleling reports of ERT or HRT on PAI-1 in postmenopausal women (15, 16, 17) (Fig. 3
).
PAI-1 is the major regulator of the fibrinolytic system, and disturbances in the plasma fibrinolytic system have been implicated in the pathogenesis of CHD. Increased levels of PAI-1 result in less plasmin degradation of fibrin and, thus, would be expected to be prothrombotic. Both clinical and epidemiological data suggest that reduced endogenous fibrinolytic activity, characterized by increased PAI-1 antigen, t-PA antigen, and PAI-1 activity and reduced t-PA activity, is a major factor in both the development and severity of atherosclerosis and thrombosis (19, 20, 21, 22, 23). The finding that hemostatic factors, PAI-1, t-PA, and fibrinogen, are relatively lower in premenopausal women has been used to explain the protection against CHD experienced by younger women (15, 24). In the Framingham Offspring Study (24), subjects with high estrogen status (premenopausal women or postmenopausal women receiving HRT) had lower PAI-1 than subjects with low estrogen status (men or postmenopausal women without HRT). In postmenopausal women, the reduction in PAI-1, t-PA, and fibrinogen from hormones (15, 16, 23, 25) has been used to infer increased fibrinolytic potential; however, this finding as an underlying principle for cardiac protection, is open to question (25, 26, 27, 28).
Cyclical variation
The model estradiol curve expects a peak in phase II with a smaller peak in phase III. To accommodate outpatient subjects, weekly hormone and/or hemostatic factors were measured. Other time points, however, such as 36-h peak estradiol levels, compared with early follicular levels, and urinary LH surge time values in relation to the LH peak might have resulted in classical estradiol curves. Atypical cycling, intra- and interindividual variability, inadequate historical information, and the timing of collections in this outpatient study could account for the fact that the estradiol determinations did not reflect the idealized nadir and peaks of the follicular and luteal phases (Fig. 1
). Nonetheless, the finding of significant cyclical variability in PAI-1 and D-dimer reported here supplements the known diurnal variation in hemostatic and coagulation factors. The present findings highlight that an appreciation of endogenous cyclical timing and biologic variability is important for interpreting the nature of hemostatic and fibrinolytic factors in the mechanism of disease and for the design of studies. Future studies to determine whether the changes have biological significance are indicated. A fundamental question raised by the observed cyclical variation is whether the cardiac protection afforded to younger women results from diminished prolonged exposure to potentially prothrombotic variables like PAI-1.
Other variables also have dynamic characteristics during the menstrual cycle. Studies of blood flow and endothelial function report appreciable changes in brachial artery reactivity during the menstrual cycle (29, 30, 31, 32, 33), and cyclical low estradiol levels have been invoked to explain variant angina (34). Other examples of cyclical changes including higher HDL-cholesterol levels during the follicular phase (35) and variations in vascular endothelial growth factor associated with cyclical ovarian and uterine blood flow (36) have also been described.
PAI-1 and genetic polymorphism
We found interindividual variability in PAI-1 and plasma levels were substantially higher in four subjects (Fig. 2
). The observation is of potential clinical utility because an easily measurable concentration of PAI-1 function might be useful for selection and dosing of patients treated with anticoagulant and fibrinolytic factors as well as for predicting risk, particularly to prothrombotic agents such as ERT or HRT. Understanding the genetics of PAI-1 as well as the environmental factors that influence its concentration is potentially attractive for screening and treatment of vascular disease.
Whereas it is acknowledged that PAI-1 levels are regulated by genetic and environmental factors, the precise determinants of PAI-1 expression are not fully explained. Basic investigations have explored a specific insertion/deletion polymorphism in the promoter region of the PAI-1 gene whereby one allele sequence has four guanosines (4G) and the other has five (5G) (37). There is in vitro evidence reporting that the 4G allele has higher activity than the 5G allele, and subjects homozygous for the 4G allele have higher plasma PAI-1 concentrations (38, 39, 40). In a small study of young men with infarction, younger than 45 yr of age, the 4G/4G genotype, was found in 43% (38). Clinical studies have explored the influence of age and gender and the relation between PAI-1 polymorphism and outcome. An increase in the mutation has been correlated with postmenopausal women with coronary artery disease (41). Another study of middle-aged men, mean age 63 yr (42), found that the 4G/5G polymorphism in the promoter of the PAI-1 gene was not a major pathogenetic risk factor. Another found that the risk of myocardial infarction was decreased among women (<45 yr) carrying the 4G allele of the PAI-1 4G/5G polymorphism (43).
Limitations
First, the hormone levels and hemostatic factors associated with the phases of the menstrual cycle could have been over- or underestimated due to timing, unpredictable cycling, intra- or interindividual variability, or inadequate historical information. That the hormone and hemostatic determinations did not match the exact nadir and peak of the follicular and luteal phases, resulting in over- or underestimation of levels in some patients, is likely. In this outpatient setting, we collected and determined weekly hormone and hemostatic factors; however, other ways to collect the hormone data, such as 36-h peak estradiol levels or LH surge times might have resulted in model estradiol curves. Second, the study was designed to characterize the time course of prothrombotic factors and their relation to hormones and due to its size and short-term nature could not determine the clinical significance of these in relation to future clinical events.
Conclusions
Among premenopausal subjects there is significant variability over the menstrual cycle for estradiol, FSH, progesterone, PAI-1, and fibrin D-dimer. Cyclical variability, like diurnal variability, should be considered in studies of hemostatic and fibrinolytic factors in premenopausal subjects and those who may be at risk for CHD. These observations are of potential clinical utility because an easily measurable concentration of PAI-1 function might be useful for selection and dosing of patients as well as for predicting risk, particularly to prothrombotic agents such as ERT or HRT. Understanding normal variability of atherothrombotic factors is potentially attractive for screening and treatment of many forms of vascular disease. Future studies to determine whether the changes have biological significance are indicated.
| Acknowledgments |
|---|
| Footnotes |
|---|
Abbreviations: BMI, Body mass index; CHD, coronary heart disease; C-RP, C-reactive protein; CV, coefficients of variation; ERT, estrogen replacement therapy; HDL, high-density lipoprotein cholesterol; HRT, hormone replacement therapy; LDL, low-density lipoprotein cholesterol; PAI, plasminogen activator inhibitor; t-PA, tissue plasminogen activator; vWF, von Willebrand factor.
Received March 29, 2004.
Accepted September 7, 2004.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
G. G. Gosman, H. I. Katcher, and R. S. Legro Obesity and the role of gut and adipose hormones in female reproduction Hum. Reprod. Update, September 1, 2006; 12(5): 585 - 601. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |