| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Other Original Articles |
Department of Medicine and Therapeutics (C.J.G.K., J.M.C.C.), Obstetrics and Gynecology (H.L.), Division of Biochemistry and Molecular Biology (G.W.G.), Pathological Biochemistry (N.S.), and Medicine (J.R.P., A.R., G.D.O.L), University of Glasgow, Glasgow G31 2ER, United Kingdom
Address all correspondence and requests for reprints to: Dr. Naveed Sattar, Consultant/Senior Lecturer in Endocrinology and Metabolism, Department of Pathological Biochemistry, Glasgow Royal Infirmary University NHS Trust, Glasgow G31 2ER, United Kingdom. E-mail: . nsattar{at}clinmed.gla.ac.uk
Abstract
There is increasing evidence that elevated plasma levels of hemostatic factors [fibrinogen, factor VII, von Willebrand factor, fibrin D-dimer, and tissue plasminogen activator (t-PA) antigen] are independently linked to risk for coronary heart disease (CHD). Women with polycystic ovary syndrome (PCOS) are insulin-resistant and have increased risk for CHD and type 2 diabetes, but there are few data on hemostatic markers in women with PCOS. Seventeen women with PCOS (defined on the basis of elevated testosterone and oligomenorrhea) and 15 healthy women matched as a group for body mass index (BMI) were recruited. Insulin sensitivity was assessed using the hyperinsulinemic euglycemic clamp technique. Factor VIIc was determined by a clotting assay; fibrinogen was determined by nephelometry; and t-PA, D-dimer, and von Willebrand factor antigens were measured by ELISA techniques. Of these hemostatic markers, only t-PA concentration was significantly (P = 0.013) elevated in women with PCOS relative to controls. t-PA correlated with BMI in both PCOS and controls (r = 0.428, P < 0.1; and r = 0.686, P < 0.01) and inversely with the insulin sensitivity index (r = -0.590, P < 0.05; and r = -0.620, P < 0.05, respectively). After further adjustment for BMI and insulin sensitivity, there remained a significant difference in t-PA between cases and controls (P = 0.017). Together, age and insulin sensitivity explained 39% of the variance in t-PA in women with PCOS (P < 0.05). Total testosterone did not correlate significantly with t-PA in either group. We conclude that women with PCOS have significantly increased t-PA concentrations relative to women with normal menstrual rhythm and normal androgens. We suggest that elevated t-PA and dysfibrinolysis may be a factor in the increased cardiovascular morbidity seen in PCOS.
POLYCYSTIC OVARIAN SYNDROME (PCOS) affects between 5 and 10% of premenopausal women (1). It is recognized to have a metabolic component consisting of hyperinsulinemic insulin resistance. Women with PCOS exhibit a decrease in insulin sensitivity of between 30 and 40%, a deficit similar to that seen in subjects with type 2 diabetes mellitus (2). PCOS also carries with it an increased lifetime risk of diabetes, hypertension, and coronary heart disease CHD (3, 4, 5). For example, postmenopausal women with a past history of PCOS have a 3-fold increased prevalence of hypertension, and whereas evidence for an association between PCOS and cardiovascular disease is less robust, cross-sectional studies show an increased prevalence and severity of coronary artery disease in PCOS (6, 7, 8).
There is increasing evidence that elevated plasma levels of hemostatic factors, [fibrinogen, factor VII, von Willebrand factor (vWF), fibrin D-dimer, and tissue plasminogen activator (t-PA) antigen] are independently linked to risk for CHD (9, 10). There is also increasing evidence that t-PA antigen and plasminogen activator inhibitor-1 (PAI-1) are positively associated with insulin resistance (11, 12, 13, 14). Hemostatic factors have been studied previously in PCOS, most extensively PAI-1 (15, 16, 17). The results are however conflicting, and there are few data on t-PA antigen or other factors. Moreover, most studies have included only surrogate indices of insulin resistance such as fasting insulin rather than a direct measure. We therefore designed a study to test whether concentrations of hemostatic factors (fibrinogen, factor VII, fibrin D-dimer, vWF, and t-PA antigen) are elevated in women with PCOS relative to healthy controls, and if so, whether these differences are accounted for by differences in insulin resistance.
Subjects and Methods
Subjects
The study was approved by the West Glasgow University Hospitals National Health Service Trust Ethics Committee, and all subjects gave written informed consent. PCOS was defined as androgen excess (total testosterone >3.6 nmol/liter, or a free androgen index 
9) with ovulatory dysfunction (less than six menstrual cycles per year), once specific disorders such as adult onset congenital adrenal hyperplasia, hyperprolactinemia, and androgen secreting neoplasia had been excluded. The patients were recruited from the endocrine clinics of the North Glasgow University Trusts and were on no medication at the time of study. Control patients, recruited from the Glasgow area, were volunteers who had regular menstrual cycles and normal androgens and were on no medication. For each woman with PCOS, we attempted to recruit a control within two body mass index (BMI) units (kilograms per square meter). This was achieved for all but two patients. All subjects were studied in the early follicular phase if regularly cycling, or at least 8 wk after their last menstrual period if oligomenorrhoeic. All patients and controls were Caucasian, and none had clinical evidence (history or examination) of recent or ongoing infection.
Whole-body insulin sensitivity
Subjects attended after an overnight fast during which water intake was permitted. Baseline blood pressure and heart rate were recorded after 20 min supine rest. Indwelling cannulae were then inserted into the left antecubital vein for administration of glucose and insulin and retrogradely into the right dorsal hand vein for blood sampling. Insulin sensitivity was assessed using a modified version of the hyperinsulinemic euglycemic clamp described by DeFronzo et al. (18). In brief, a primed, constant rate infusion of soluble insulin (1.5 mU/kg·min) was administered for 180 min, and a variable rate infusion of 20% glucose was administered to maintain euglycemia (5.2 mmol/liter). Insulin was prepared in 45 ml of 0.9% NaCl; 5 ml of the patient’s own blood was added to prevent adsorption of insulin to plastic surfaces. The infusion was administered using a Braun perfusor pump (Braun, Melsungen, Germany). Glucose (20%) was infused from 4–180 min using an IVAC IV infusion system (IVAC, Basingstoke, UK); the glucose infusion rate was adjusted manually according to glucose concentrations (Beckman Coulter, Inc., Fullerton, CA) and measured at 5-min intervals on 2 ml arterialized blood samples collected from a dorsal right hand vein surrounded by a heated box (60 C; Department of Physiology, University of Nottingham, Nottingham, UK). Under steady-state conditions, insulin sensitivity was calculated from the glucose infusion rate and the blood glucose concentration (15). Samples were also collected for the measurement of serum insulin concentrations at 60-, 120-, 150-, and 180-min intervals. Insulin sensitivity index (M/I) was calculated using the steady-state insulin concentration. Blood pressure and heart rate were recorded every 30 min using a semiautomatic sphygmomanometer (Dinamap, Bracknell).
Clinical measures
Ovarian morphology was assessed (in each case by the same operator) using either transabdominal or transvaginal ultrasound. Anthropometric measurements were made by the same trained observer using standard techniques (World Health Organization). Body weight was measured using analog scales (Seca, Hamburg, Germany) to within 500 g in light clothing; height was measured barefoot using a stadiometer to within 0.5 cm. BMI was calculated as follows: weight (kg)/height (m)2.
Biochemistry
Routine biochemical analysis on all subjects was carried out using an Olympus AU5200 autoanalyser. Baseline hormonal profiles (testosterone, progesterone, PRL, 17-hydroxy progesterone, and FSH/LH) were measured using a Bayer Immuno 1 autoanalyser, and the SHBG using the DPC Immulite 2000. Free androgen index was calculated by total testosterone/SHBG x 100.
Hemostatic factors
Factor VIIc was determined by a one-stage clotting assay in an MDA-180 coagulometer (Organon Teknika, Cambridge, UK). t-PA, fibrin D-dimer (Biopool, Stockholm, Sweden), and vWF (DAKO Corp., Copenhagen, Denmark) antigens were measured by ELISA techniques. Fibrinogen was measured using the Clauss assay on the MDA-180 (Organon Teknika, Cambridge, UK). The intra-assay coefficients of variation for the hemostatic factors were all less than 5%.
Statistics
Results are shown as mean and standard errors or as median and range for skewed variables. Data from the two groups were compared using standard statistical tests (unpaired t tests) with adjustment for age where appropriate using regression analyses. Simple linear correlation was performed on the data from the PCOS cohort and control groups separately, with t-PA taken as the response variable. Standard multivariate regression analyses were used to adjust for the combined effects of age, BMI, and M/I on the difference in t-PA between the cases and controls.
Results
Patient characteristics are shown in Table 1
. There was no significant difference in BMI between groups, although the women with PCOS had a significantly lower mean age by more than 6 yr. The proportion of smokers in each group was similar. The women with PCOS had significantly elevated mean testosterone and lower mean SHBG concentrations. Sixteen of the 17 women with PCOS but none of the controls had evidence of polycystic ovaries on ultrasound. M/I was lower in women with PCOS, although this did not reach statistical significance (P = 0.069).
|
). t-PA correlated with BMI in both PCOS and controls (r = 0.43, P < 0.1; and r = 0.69, P < 0.01, respectively) and inversely with M/I (r = -0.59, P < 0.05; and r = -0.62, P < 0.05, respectively; Fig. 1). The slope of association between t-PA and M/I was greater in the PCOS group, but the difference was not statistically significant (P > 0.1). In the women with PCOS, the relationship between t-PA and insulin sensitivity came close to significance (P = 0.075) even after adjustment for BMI. There remained a significant difference in t-PA between cases and controls on further adjustment for BMI and insulin sensitivity (Table 3). Together, age and insulin sensitivity explained 39% of the variance in t-PA concentration in the women with PCOS (P < 0.05). Total testosterone did not correlate with t-PA in either group (data not shown).
|
|
|
Disturbed fibrinolysis, as measured by elevated plasma levels of t-PA antigen, is increasingly recognized as an important cardiovascular risk factor. High levels of t-PA antigen (which largely measure t-PA/PAI-1 complexes) have been shown to correlate both with the severity of coronary artery disease at angiography and independently with CHD event rate (9, 10, 19).
This is the first study to show elevated levels of t-PA antigen in women with PCOS relative to healthy women with normal menstrual rhythm matched for BMI. By contrast, we did not see any differences in any of the other hemostatic parameters examined in this study. In keeping with data from other populations (11, 12, 13, 14), we noted that t-PA concentrations in both PCOS and controls correlated directly with the degree of obesity, and inversely with insulin sensitivity. Indeed, the relationship between t-PA and insulin sensitivity was close (P = 0.075) to being independent of BMI in women with PCOS. These observations suggest that, like increased PAI-1, elevated t-PA antigen may be another feature of the metabolic syndrome. However, PCOS seems to be independently associated with increased levels of t-PA, because the group difference remained significant after correction for BMI and insulin sensitivity. This suggests that other unmeasured factors are important in promoting higher t-PA concentrations in women with PCOS, for example genetic determinants.
Increased plasma t-PA antigen is thought to represent largely inactive circulating t-PA/PAI-1 complexes, which, in turn, may reflect endothelial disturbance (t-PA release) and/or elevated PAI-1 levels. Although not measured in this study, there is some evidence that PAI-1 activity is raised in women with PCOS in relation to fasting insulin. For example, Sampson et al. (17) showed elevated PAI-1 activity in women with polycystic ovaries on ultrasound and menstrual disturbance, relative to controls with normal menstrual rhythm and normal ovaries. Similarly, Atiomo et al. (16) did observe a trend toward higher (P = 0.07) PAI-1 activity in women with PCOS relative to a control group of healthy women with similar mean BMI. However, unlike t-PA in a recent meta-analysis, PAI-1 levels did not independently predict CHD event rate (10), perhaps because PAI-1 exhibits greater biological variability within individuals.
Further indirect support for an alteration in the t-PA/PAI-1 axis in women with PCOS, in association with greater insulin resistance, comes from observations that treatment with metformin or thiazolidinediones reduces PAI-1 and/or t-PA antigen concentrations in women with PCOS (20, 21). Significantly, troglitazone-induced reductions in t-PA and PAI-1 concentrations were not related simply to a decrease in weight, because BMI does not fall with this treatment.
Vascular endothelial cells are a source of t-PA, and therefore high circulating t-PA levels may also reflect endothelial dysfunction. Our observation of high t-PA in women with PCOS may thus reflect generalized endothelial dysfunction in women with PCOS. Currently, there is inconsistent evidence for the presence of endothelial dysfunction (as measured by vascular reactivity) in women with PCOS (22, 23, 24). It should be noted, however, that vWF levels (another marker of endothelial function) were not statistically different between cases and controls in this study.
Finally, the lack of an association between t-PA, other hemostatic markers, and total testosterone is not surprising because we have previously noted that total testosterone and lipid and lipoprotein subfraction concentrations do not correlate in women with PCOS (25). These combined observations accord with a recent prospective case-referent study, which, by employing measurement of urinary sex hormone metabolites, showed that high androgen output is not related to an increased CHD risk in women (26).
In conclusion, we have shown for the first time that women with PCOS exhibit significantly increased t-PA antigen concentrations relative to BMI-matched, healthy women with a normal menstrual rhythm. We therefore suggest that elevated t-PA and dysfibrinolysis may be a factor in the increased cardiovascular morbidity seen in PCOS.
Acknowledgments
We acknowledge support from the Scottish Hospitals Endowment Research Trust.
Footnotes
Abbreviations: BMI, Body mass index; CHD, coronary heart disease; M/I, insulin sensitivity index; PAI-1, plasminogen activator inhibitor-1; PCOS, polycystic ovary syndrome; t-PA, tissue plasminogen activator; vWF, von Willebrand factor.
Received November 5, 2001.
Accepted March 17, 2002.
References
This article has been cited by other articles:
 |
|
 |
|
  N. Sattar Review: PCOS, insulin resistance and long-term risks for diabetes and vascular disease The British Journal of Diabetes & Vascular Disease, January 1, 2009; 9(1): 15 - 18. [Abstract] [PDF]
|
|
|
|
 |
|
 S. G. Wannamethee, N. Sattar, A. Rumley, P. H. Whincup, L. Lennon, and G. D.O. Lowe Tissue Plasminogen Activator, von Willebrand Factor, and Risk of Type 2 Diabetes in Older Men Diabetes Care, May 1, 2008; 31(5): 995 - 1000. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Sattar and S. M. Nelson Polycystic Ovarian Syndrome, Biomarkers, and Metformin: Research, Risk, and Reality J. Clin. Endocrinol. Metab., January 1, 2008; 93(1): 34 - 36. [Full Text] [PDF]
|
|
|
|
 |
|
 R. S. Legro Polycystic Ovary Syndrome and Cardiovascular Disease: A Premature Association? Endocr. Rev., June 1, 2003; 24(3): 302 - 312. [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 |
