Enhanced Levels of Tissue-Type Plasminogen Activator in Borderline Hypertension
Abstract Despite effective antihypertensive therapy, essential hypertension is still associated with considerable residual risk of cardiovascular complications. The aim of the present study was to investigate the state of the endogenous fibrinolytic system in young subjects with borderline hypertension. Thirty-nine young (age, 24 to 34 years) male subjects with borderline hypertension (systolic BP [SBP] 140 to 160 mm Hg and/or diastolic BP [DBP] 85 to 95 mm Hg) and 17 normotensive control subjects (age, 22 to 31 years; SBP 110 to 130 and DBP 60 to 80 mm Hg) were recruited from a population screening. Plasma levels of tissue-type plasminogen activator (t-PA) antigen and activity and plasminogen activator inhibitor 1 (PAI-1) antigen were determined at rest and in response to a venous occlusion test. Borderline-hypertensive subjects had metabolic and anthropometric characteristics similar to normotensive individuals. In comparison with normotensive subjects, borderline-hypertensive subjects had higher plasma concentration of t-PA antigen both at rest and after venous occlusion but similar levels of t-PA activity or PAI-1 antigen. The increase in t-PA antigen and activity in response to venous occlusion was significantly greater in borderline-hypertensive subjects than in normotensive control subjects (P<.0001 and P=.003, respectively). In stepwise regression analyses, 24-hour mean arterial pressure emerged as the single most powerful predictor of t-PA antigen levels, but body mass index was the most important determinant of t-PA activity and PAI-1 antigen. However, PAI-1 was explained by both body mass index (partial r=.48, P<.001) and 24-hour mean arterial pressure (partial r=.29, P<.05). Thus, early hypertension may be associated with significant alterations in endogenous fibrinolysis.
Despite the introduction of effective blood pressure–lowering therapy, hypertension is still associated with increased cardiovascular morbidity and mortality, mainly from coronary events.1 Hence, it is likely that factors other than increased blood pressure per se contribute to the residual risk of cardiovascular complications in essential hypertension. Evidence is now accumulating for a central role of thrombogenic mechanisms in the pathogenesis of myocardial infarction and sudden cardiac death.2 In addition to enhanced coagulability, impaired endogenous fibrinolysis has long been thought to increase the risk of acute coronary thrombosis.3 Consequently, PAI-1 has been found to predict recurrent myocardial infarction.4 Recently, prospective data showed that the endogenous t-PA antigen concentration predicts future myocardial infarction,5 cardiac death,6 and thrombotic stroke.7
Systematic investigations of hemostatic functions in hypertension are sparse, and only a few studies on the fibrinolytic system in essential hypertension have been published. Studies have consistently reported increased PAI-1 activity and decreased t-PA activity in patients with established hypertension compared with normotensive control subjects.8 9 10 11 12 An increased level of t-PA antigen has been reported by some investigators11 12 13 but not by others.9 14
However, previous studies mainly investigated patients with established essential hypertension. It therefore cannot be ruled out that the hemostatic aberrations were early manifestations of preclinical atherosclerotic disease, the prevalence of which would be expected to be greater in established hypertensive subjects. Furthermore, human essential hypertension is frequently associated with a complex web of interrelated metabolic disturbances, including impaired glucose tolerance, fasting hyperinsulinemia, and blood lipid perturbations (eg, see References 15 through 1715 16 17 ). Although part of this effect is related to the coexistence of obesity in hypertension,18 there appears to be a primary link between high blood pressure and metabolic aberrations.19 Since these metabolic disturbances by themselves may have important effects on fibrinolytic functions,20 it has not been established whether impaired endogenous fibrinolysis is associated with elevated blood pressure as such.
The aim of the present study was to investigate the state of the endogenous fibrinolytic system in young subjects with a very mild and early borderline blood pressure elevation, in whom secondary adaptive changes of the cardiovascular system or atherosclerosis were likely to be limited. To avoid the confounding effects of obesity and/or metabolic aberrations on fibrinolysis, the study was based on a population-recruited sample of borderline hypertensive and normotensive control subjects with similar BMIs, devoid of major lipid metabolic derangements.
Study subjects were participants in a prospective long-term follow-up study of young men with borderline hypertension.21 The initial screening was performed in 1987 among 10 500 healthy young men living in the Göteborg area who were summoned to compulsory military enlistment. The screening procedure and blood pressure criteria used for selection were reported previously.21 By the time of the present investigation, the mean age of the subjects was 25.6 years (range, 22 to 34 years). The BH group had SBP equal to or greater than 145 mm Hg and/or DBP equal to or greater than 85 mm Hg at the screening, and SBP 140 to 160 mm Hg and/or DBP 85 to 95 mm Hg at two subsequent follow-ups at the Hypertension Center. Secondary hypertension was excluded by standard procedures. The NC group was recruited from the same population among subjects who had SBP 110 to 130 mm Hg and DBP 60 to 80 mm Hg in three of three consecutive blood pressure recordings. Forty of an initial 56 individuals with borderline hypertension and 17 of an initial 20 normotensive control subjects agreed to participate in the investigation. All subjects were apparently healthy, and none were on any regular medication.
One subject in the BH group had a fasting plasma glucose value of 17 mmol/L and therefore was excluded from analysis.
Informed consent was obtained from each subject before inclusion in the study. The protocol was approved by the Ethics Committee of the University of Göteborg. All methodological procedures were performed in accordance with the guidelines of our laboratory.
The present study was undertaken approximately 4 years after the baseline investigation. Studies were performed in the morning after an overnight fast (>10 hours).
Blood pressure was measured with a mercury sphygmomanometer in the supine position after 10 minutes of rest. On each occasion, an average of three readings was used. BMI was calculated from weight divided by height squared (kg/m2). WHR was determined to be the ratio of the waist circumference at the umbilical level to the hip circumference at the level of the major trochanter in the standing position. The sagittal diameter was defined as the midinspiratory height of the umbilicus in the supine position. Two baseline blood samples were collected at 8:30 and 10:30 am. Fibrinolytic capacity was measured by use of a standardized venous occlusion test according to Robertson et al.22 A sphygmomanometer cuff was applied to the upper arm and was inflated to a pressure midway between DBP and SBP for 15 minutes. Blood was withdrawn before and immediately after the occlusion.
24-Hour Ambulatory Blood Pressure Recordings
Ambulatory blood pressure was measured over a 24-hour period in the subjects’ natural environments by a portable Spacelabs SL-90202 recorder (Spacelabs). SBP, DBP, and mean arterial blood pressure were recorded every 20 minutes, and an average value was computed for each individual over the entire 24-hour period. Two subjects in the NC group and 4 subjects in the BH group were not willing to participate in the 24-hour recording.
Blood Sampling and Biochemical Assays
The first 3 to 4 mL of blood was always discarded. Blood samples were collected in tubes containing 1:10 (vol/vol) 0.13 mol/L sodium citrate (Venoject, Terumo Europe NV); 1:10 0.45 mol/L sodium citrate buffer, pH 4.3 (Stabilyte, Biopool AB); and 1:10 platelet-stabilizing buffer (Diatube H, Diagnostica Stago) for determination of t-PA antigen, t-PA activity, and PAI-1 antigen, respectively. The tubes were kept on ice, and plasma was isolated within 15 minutes by centrifugation at 4°C and 2000g for 20 minutes. Plasma was immediately frozen and stored at −70°C.
A solid-phase immunosorbent assay with trinitrobenzoylated poly-d-lysine as a stimulator was used for determination of t-PA activity in plasma (Novo Nordisk).23 For determination of t-PA and PAI-1 antigen, the reagent kits TintElize t-PA and TintElize PAI-1 (Biopool AB) were used. Intra-assay coefficients of variation in our laboratory are 4.3%, 3.9%, and 4.5% for t-PA activity, t-PA antigen, and PAI-1 antigen, respectively. Plasma insulin was assayed in duplicate by radioimmunoassay (Diagnostic Products Corporation). Serum cholesterol and triglyceride were determined by use of the enzymatic method (Boehringer Mannheim). HDL cholesterol was analyzed according to Seigler and Wu,24 and LDL cholesterol was computed according to the formula of Friedwald et al25 as follows:
None of the subjects had a triglyceride value >4.5 mmol/L, which would have made the use of this formula inappropriate.
Standard statistical methods were used, including univariate linear regression analyses. Unless otherwise stated, values are presented as mean±SEM. Between-group comparisons of normally distributed variables were performed by one-way ANOVA. Differences in abnormally distributed fibrinolytic variables between the groups were evaluated by the Mann-Whitney U test. Changes in response to venous occlusion were tested by the Wilcoxon rank sum test against the null hypothesis of no change. A value of P<.05 (two-tailed test) was considered significant.
Fibrinolytic variables were fit to univariate and multivariate regression models after logarithmic transformation. Significant correlations were checked by computation of Spearman’s rank correlation of crude values, which in all cases yielded similar results. Forward stepwise multiple regression was used to find patterns of significant predictors of t-PA and PAI-1 levels (F to enter=4). The following variables, all of which showed univariate correlation with the fibrinolytic parameters, were tested in the model: BMI, sagittal diameter, fasting insulin level, cholesterol level, triglyceride level, and 24-hour ambulatory blood pressure.
Table 1⇓ shows anthropometric and metabolic characteristics of the subjects. Borderline-hypertensive subjects had body weight, BMI, and regional fat distribution (sagittal diameter, WHR) similar to normotensive individuals. There were no differences in smoking habits or alcohol consumption between groups. SBP, mean blood pressure, and DBP were significantly higher in subjects in the BH group than in control subjects, although heart rates were similar. There were no significant differences among the groups regarding fasting insulin level; total, LDL, or HDL cholesterol levels; or triglyceride level.
Borderline-Hypertensive Subjects Versus Control Subjects
Borderline-hypertensive subjects had a higher plasma concentration of t-PA antigen both at rest and after venous occlusion than control subjects (Table 2⇓). There were no significant differences in t-PA activity or PAI-1 antigen levels between the two groups. As expected, t-PA and PAI-1 antigen levels declined and t-PA activity increased between 8 and 10:30 am, the changes being similar in the two groups. Venous occlusion caused significant increases in the plasma concentrations of t-PA activity and antigen as well as of PAI-1 antigen. The increase in t-PA antigen and activity in response to venous occlusion was significantly greater in borderline-hypertensive subjects than in normotensive control subjects.
To rule out the effect of obesity, we also did a matched-pair analysis in 15 normotensive and 15 borderline-hypertensive, nonobese subjects matched for BMI ≤28 kg/m2. Between-group comparison showed that borderline-hypertensive subjects had significantly higher t-PA antigen (P=.013) levels at rest and tended to have somewhat higher PAI-1 antigen levels (P=.070). t-PA activity levels were similar between groups. However, in response to venous occlusion, borderline-hypertensive subjects showed greater t-PA activity and greater release of t-PA antigen (P=.012 and P=.0009, respectively).
In univariate regression analyses across the whole group (n=56), PAI-1 and t-PA antigen were inversely correlated with t-PA activity (r=−.86 and r=.68, respectively, P<.0001 for both). PAI-1 and t-PA antigen were positively correlated (r=.68, P<.0001).
As shown in Table 3⇓, PAI-1 and t-PA antigen were directly correlated and t-PA activity was inversely correlated with BMI, sagittal diameter, fasting insulin, and triglyceride concentrations (P<.05 for each variable). In addition, PAI-1 and t-PA antigen were inversely correlated and t-PA activity was positively correlated with HDL cholesterol. t-PA and PAI-1 antigen correlated directly with 24-hour ambulatory SBP, mean arterial blood pressure, and DBP (r=.33 to r=.42, P<.05 for each variable), and somewhat weaker inverse correlations were observed for t-PA activity (r=−.26 to r=−.31). Auscultatory SBP and DBP recorded during the experiment correlated with t-PA antigen levels (r=.32, P<.05 for both) but were not significantly related to t-PA activity or PAI-1 antigen. t-PA activity and antigen and PAI-1 response to venous occlusion did not show significant correlation to any of the above variables.
Multiple Regression Analyses
In stepwise regression analyses, 24-hour mean arterial pressure emerged as the single most powerful predictor of t-PA antigen levels. The relation between t-PA antigen and 24-hour mean arterial pressure remained after introduction of BMI, fasting insulin, and triglyceride levels in the model. By contrast, in stepwise regression analyses of t-PA activity and PAI-1 antigen, BMI emerged as the primary determinant. In a bivariate regression model, the relation between t-PA activity and 24-hour mean arterial pressure was canceled out after introduction of BMI (partial r=−.50, P<.001). However, PAI-1 was explained by both BMI (partial r=.48, P<.001) and 24-hour mean arterial pressure (partial r=.29, P<.05).
The present study shows that young subjects with a mild, borderline elevation of blood pressure have higher plasma concentrations of t-PA antigen both at rest and after venous occlusion but similar levels of t-PA activity or PAI-1 antigen. The specific relation of t-PA and PAI-1 antigen to blood pressure became obvious only when average 24-hour blood pressure levels were introduced into the regression models. This was not unexpected since, given the great variability of casual blood pressure measurements, use of such measures may easily “scatter out” correlations between blood pressure and hemostatic parameters. Hence, it is likely that previous studies may have underestimated the association of fibrinolysis and hypertension as such by using less reliable indexes of average blood pressure levels.
The direct relation between global and regional adiposity and metabolic factors (mainly insulin and triglyceride) on the one hand and PAI levels on the other has been reported by several investigators (eg, Reference 2020 ). Unfortunately, in most studies on fibrinolytic function in essential hypertension, anthropometric and metabolic data have not been reported.10 11 12 13 However, in the study by Landin and coworkers,26 11 hypertensive men were carefully matched with normotensive subjects for anthropometric data. Despite this fact, hypertensive subjects not only displayed significantly elevated PAI activity compared with control subjects but also increased triglyceride, cholesterol, and insulin levels. t-PA was not measured in the study by Landin et al. In the study by Jansson and coworkers,9 the hypertensive group was selected on the criteria of elevated blood pressure (DBP 90 to 109 mm Hg) and high serum cholesterol levels (>6.0 mmol/L). In addition to increased PAI-1 activity and decreased t-PA activity, the hypertensive subjects also displayed increased triglyceride levels and a tendency toward higher BMI and WHR, and a greater number of hypertensive subjects were smokers compared with control subjects. The inverse relation between t-PA antigen and HDL cholesterol was demonstrated previously by Ridker and coworkers,27 although blood pressure was not reported in their study.
Studies of the capacity of the fibrinolytic system in hypertension are scarce. In line with the present results, Palermo and coworkers13 found an increased t-PA antigen release in response to mental and physical stress in hypertensive subjects compared with normotensive subjects; however, the increase in t-PA activity was attenuated compared with control subjects, which they ascribed to increased PAI activity in the hypertensive subjects. Jansson and coworkers9 found an attenuated t-PA activity response to venous occlusion in hypertensive subjects. However, as mentioned above, the hypertensive subjects in that study were also hypercholesterolemic, and cholesterol level was inversely correlated to t-PA capacity.
It is extremely difficult to conclude from the above-mentioned studies on established hypertensive subjects whether the observed increase in PAI activity in conjunction with decreased t-PA activity is due to the increased blood pressure per se or rather to early interrelated metabolic disturbances. This problem is further underscored by the fact that tissue PAI and t-PA activity in rats have been reported to be unaffected by experimentally induced hypertension.28
In the present study, the BH group had distribution of body weight, BMI, and regional fat distribution (as assessed by the sagittal diameter and WHR) similar to the control subjects and were devoid of blood lipid derangements often associated with established hypertension. Furthermore, none of the subjects was on any medication, and presence of atherosclerotic vascular disease would be highly unlikely in healthy subjects of this age group. Therefore, it is reasonable to assume that the derangement of the fibrinolytic system is a primary phenomenon associated with early blood pressure elevation rather than a secondary result of metabolic disturbances or atherosclerotic lesions. The validity of this interpretation is further strengthened by the observation that basal t-PA antigen levels and stimulated release during venous occlusion also were significantly greater when nonobese subjects were carefully matched for BMI.
The direct relation between PAI-1 antigen and degree of obesity is well established. Nevertheless, to the best of our knowledge, we are the first to describe an equally strong but inverse relation between BMI and t-PA activity. The reason that this has not been reported to date is probably that improper collection methods were used previously, with loss of t-PA activity during blood sampling resulting in t-PA activity near or even below the detection limit. The recent introduction of blood sampling tubes containing a citrate anticoagulant at low pH, which immediately reduces the blood pH, results in an effective inhibition of t-PA/PAI-1 complex formation and thus preserves t-PA activity.
Mean and median plasma t-PA activity levels (10 and 9.3 pmol/L, respectively) were lower than those we found previously (about 20 pmol/L) in our laboratory among young, healthy men in this age group. In the present study, however, the control subjects were selected from the same population as the borderline-hypertensive subjects. In our previous studies on hemostatic stress responses, the healthy volunteers had been recruited mainly among hospital employees and medical students. It may be that individuals with a low metabolic risk profile are overrepresented in the latter group. In fact, the NC group had a higher mean BMI (25.9) compared with the healthy men we studied previously (mean, 22 to 23). This assumption was further strengthened by the fact that there was a strong inverse relation between t-PA activity and BMI.
In conclusion, the main finding of the present study is that young subjects with a mild, borderline elevation of blood pressure have increased levels of t-PA antigen as well as an increased t-PA response to venous occlusion. The results suggest a direct relation between t-PA antigen and blood pressure levels, whereas t-PA activity is more dependent on obesity and metabolic factors. The increased t-PA antigen levels in this very early stage of hypertension are of particular interest because recent prospective studies have shown that t-PA antigen but not PAI activity is a risk factor for both myocardial infarction5 and thromboembolic stroke7 as well as for mortality in patients with stable coronary artery disease.6
Selected Abbreviations and Acronyms
|BH group||=||borderline-hypertensive group|
|BMI||=||body mass index|
|DBP||=||diastolic blood pressure|
|NC group||=||normotensive control group|
|PAI-1||=||plasminogen activator inhibitor 1|
|SBP||=||systolic blood pressure|
|t-PA||=||tissue-type plasminogen activator|
|WHR||=||waist-hip circumference ratio|
This study was supported by a grant and a postdoctoral fellowship from the Swedish Medical Research Council (Nos. 9046 and 7324, respectively) and by grants from the Swedish Heart-Lung Foundation, Bristol-Myers Squibb, the Swedish Hypertension Society, and the Göteborg Medical Society. The expert technical assistance of Hannele Korhonen and Annika Johansson is gratefully acknowledged.
- Received February 20, 1995.
- Revision received March 21, 1995.
- Accepted June 20, 1995.
Hamsten A, de Faire U, Walldius G, Dahlen G, Szamosi A, Landou C, Blombäck M, Wiman B. Plasminogen activator inhibitor in plasma: risk for recurrent myocardial infarction. Lancet. 1987:3-9.
Jansson J-H, Olofsson B-O, Nilsson TK. Predictive value of tissue plasminogen activator mass concentration on long-term mortality in patients with coronary artery disease: a 7-year follow-up. Circulation. 1993;88:2030-2034.
Palermo A, Bertalero P, Pizza N, Amelotti R, Libretti A. Decreased fibrinolytic response to adrenergic stimulation in hypertensive patients. J Hypertens. 1989;7(suppl 6):S162-S163.
Modan M, Halkin H, Almog S, Lusky A, Eshkol A, Shefi M, Shitrit A, Fuchs Z. Hyperinsulinemia: a link between hypertension obesity and glucose intolerance. J Clin Invest. 1985;75:809-817.
Reaven GM. Banting lecture 1988: role of insulin resistance in human disease. Diabetes. 1988;37:1595-1607.
Lucas CHP, Estigarribia JA, Darga LL, Reaven GM. Insulin and blood pressure in obesity. Hypertension. 1985;7:702-706.
Ferrannini E. Insulin and blood pressure: possible role of hemodynamics. Clin Exp Hypertens. 1992;A14:271-284.
Vague P, Juhan-Vague I. Insulin and the fibrinolytic system: a link between metabolism and thrombogenesis. In: Smith U, Bruun NE, Hedner T, Hökfelt B, eds. Hypertension as an Insulin-Resistant Disorder. Amsterdam, The Netherlands: Elsevier Science Publishing Co; 1991:321-334.
Petersen LC, Handest P, Brender J, Selmer J, Jørgensen M, Thorsen S. A sensitive solid-phase immunosorbent assay for tissue-type plasminogen activator activity in plasma using trinitrobenzoylated poly-d-lysine as a stimulator for plasminogen activation. Thromb Haemost. 1987;57:205-211.
Seigler L, Wu WT. Separation of serum high-density lipoprotein for cholesterol determination: ultracentrifugation vs precipitation with sodium phosphotungstate and magnesium chloride. Clin Chem. 1981;27:838-841.
Friedwald WT, Levy RI, Fredrickson DS. Determination of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18:499-502.
Ridker PM, Vaughan DE, Stampfer MJ, Sacks FM, Hennekens CH. A cross-sectional study of endogenous tissue plasminogen activator, total cholesterol, HDL cholesterol, and apolipoproteins A-I, A-II, and B-100. Arterioscler Thromb.. 1993;13:1587-1592.