This study investigated the relationship between age and pulse pressure (PP) in subjects with untreated essential hypertension based on their plasminogen activator inhibitor (PAI)-1 gene polymorphism. The results showed that subjects with the deletion/deletion genotype had a significantly steeper age-related increase in PP compared to those with the insertion/insertion or insertion/deletion genotypes, after adjusting for potential confounding factors. This suggests that the PAI-1 gene polymorphism modulates the effects of aging on PP increases in hypertensive individuals.
1. Journal of Internal Medicine 2005; 257: 93–99
Age-related increase of pulse pressure and plasminogen
activator inhibitor-1 I/D gene polymorphism in essential
hypertension
J.-J. MOURAD1, G. DU CAILAR2, E.-M. NAZAL1, M. E. SAFAR3 AND A. MIMRAN2
From the 1Department of Internal Medicine, Avicenne Hospital-AP-HP, Bobigny; 2Department of Internal Medicine, Lapeyronie Hospital,
Montpellier; and 3Diagnosis Center, Hotel Dieu Hospital, Paris, France
Abstract. Mourad J-J, du Cailar G, Nazal E-M, Safar between age and PP in subjects with never treated
ME, Mimran A (Avicenne Hospital-AP-HP, Bobigny; essential hypertension.
Lapeyronie Hospital, Montpellier; and Hotel Dieu Results. In the studied population, the genotype
Hospital, Paris, France). Age-related increase of deletion (D)/D at position )675 of the PAI-1
pulse pressure and plasminogen activator insertion (I)/D gene polymorphism was associated
inhibitor-1 I/D gene polymorphism in essential with a significant increase in the adjusted slope of
hypertension. J Intern Med 2005; 257: 93–99. the curve relating age to PP by comparison with the
two other genotypes. No comparable difference in
Background. Pulse pressure (PP), a marker of cyclic
age-related changes in systolic, diastolic or mean
strain on the arterial wall, is a significant predictor
blood pressure was found.
of cardiovascular (CV) risk, particularly regarding
Conclusion. In subjects with essential hypertension,
the incidence of coronary arterial stenosis. Genes
the PAI-1 I/D gene polymorphism modulates the
related to haemostatic and/or fibrinolytic factors are
age-mediated increase of PP, suggesting new
consistently influenced in vitro by mechanical
insights on the complex interactions between
strain.
genes, mechanical factors and CV risk.
Objective. The goal of the present study was to
determine, in the three genotypes of the plasminogen Keywords: ageing, hypertension, plasminogen acti-
activator inhibitor (PAI)-1 gene polymorphism, the vator inhibitor-1 gene polymorphism, pulse pressure.
gender-adjusted difference in the relationships
pathophysiological mechanisms. First, increased
Introduction
aortic SBP influences the level of end-systolic
Increased brachial systolic blood pressure (SBP) stress and promotes cardiac hypertrophy [1, 3,
and pulse pressure (PP) are independent predictors 5]. Secondly, reduced DBP impairs coronary per-
of cardiovascular (CV) risk, mainly for myocardial fusion and therefore favours myocardial ischaemia
infarction [1–6]. The predictive value of PP is [1–6].
simultaneously influenced by an increase of SBP Long-term follow up of large populations has
and a decrease of diastolic blood pressure (DBP). In shown that the rate of increase of SBP, PP and
large populations over 50 years of age, cross- arterial stiffness with age contributes greatly to CV
sectional and longitudinal studies have shown that risk [9, 10]. The demonstration of wide intersubject
CV mortality is not only positively correlated to variations in the slope of the curve relating SBP or
the level of SBP [7, 8], but also that, at any given PP to age suggests that a number of environmental
value of SBP, CV mortality is higher when DBP is or genetic factors or the combination of both might
lower [8]. In fact, the predictive value of SBP and substantially influence this relationship [9–14].
PP in CV mortality may result from two different Because a hereditary predisposition has previously
Ó 2005 Blackwell Publishing Ltd 93
2. 94 J . - J . M O U R A D et al.
been postulated for both PP and arterial stiffness mercury sphygmomanometer (mean of three meas-
[12], some investigations have suggested that gen- urements made with patients in the supine position)
etic factors play a role in the mechanism(s) of the and an automatic oscillometric device (see below)
age-related increase of SBP and PP [13, 14]. on the morning of the study; (iv) no clinical or
Perusse et al. [15] showed that particular geno- biological signs of secondary hypertension, as
types determined by single genes were associated previously defined [10]; and (v) no recent symp-
with a steeper increase in SBP and hence PP with toms of coronary artery disease, heart failure,
age amongst males and females in the general stroke, and/or peripheral arterial disease. Finally,
population. According to this observation, it seems 97 men and 75 women, all Caucasian, from 19 to
likely that the age-related increase of PP might be 75 years old (mean: 47 years) were investigated in
statistically influenced by genes potentially involved this study. Mean body weight and height and
in longevity and CV mortality. Till the recent years, clinical and biological characteristics are indicated
most of the results have been focused on gene in Table 1. Noninsulin diabetes was defined as a
polymorphisms related to the renin-angiotensin fasting plasma glucose >7 mmol L)1 and/or by the
system [13, 14, 16]. Few studies have been per- presence of antidiabetic agents (metformine and/or
formed on gene polymorphisms related to fibrino- biguanids).
lytic and/or haemostatic factors. However, All participants were examined in the morning
alterations of fibrinolysis play a role in the CV risk after fasting for at least 12 h, and all underwent
of myocardial infarction [17–19] and plasminogen the same procedure after providing written con-
activator inhibitor (PAI)-1 release as well as corres- sent. After 20 min of rest in the supine position,
ponding genes are significantly influenced in vitro by BP was measured automatically every 2 min with
cyclic strain [20, 21]. Furthermore, changes in PAI- the Dinamap device (Model 845, Critikon, Tampa,
1 plasma levels, sodium diet and renin-angiotensin FL, USA) [10]. The mean of five consecutive
system are highly interrelated in the pathophysio- measurements was calculated. This automatic
logical mechanisms of hypertension and athero- method was chosen for the statistical analysis to
sclerotic diseases [22]. avoid interobserver variations and to minimize
The objective of the present study was to investi- ‘white-coat’ phenomenon. At the end of this
gate whether the PAI-1 insertion (I)/deletion (D) procedure, blood was drawn for determination of
gene polymorphism might influence the relation- standard biochemical measurements and DNA
ships between age and PP in men and women with extraction [10].
never treated sustained essential hypertension.
Table 1 Population description (n ¼ 172): mean ± SD (1 SD)
Patients and methods and extremes (minimum–maximum)
The study was performed in patients which entered Mean (±SD) Extremes
the Department of Internal Medicine of Lapeyronie
Gender (men : women) 97 : 75
Hospital, Montpellier, France, for a CV check-up Age (years) 47.1 ± 11.6 19–75
ordered by their general practitioner or their cardi- Body height (cm) 167.7 ± 9.1 147–191
ologist because of the presence of one or several CV Body weight (kg) 75.5 ± 17.1 44–163
BMI (kg m)2) 26.7 ± 5.2 18.5–58.6
risk factors involving high BP, smoking, dyslipidae-
Tobacco consumption 65 : 35
mia, diabetes mellitus, and/or a family history of (no : yes) (%)
premature CV disease. From this overall population Plasma total cholesterol (g L)1) 2.26 ± 0.42 1.20–3.45
of patients, 172 consecutive subjects with never- Plasma creatinine (mmol L)1) 84.11 ± 15.32 54–123
SBP (mmHg) 163.0 ± 17.8 134–220
treated sustained essential hypertension were
DBP (mmHg) 96.9 ± 10.9 72–136
enrolled in the present investigation. MAP (mmHg) 118.9 ± 12.0 93–159
Hypertensive subjects were selected according to PP (mmHg) 66.1 ± 13.6 40–118
the following criteria: (i) a history of hypertension; HR (bpm) 67.5 ± 9.2 44–92
(ii) no antihypertensive or vasodilator treatments BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic
prior to the start of the study; (iii) SBP >140 mmHg blood pressure; MAP, mean arterial pressure; PP, pulse pressure;
and/or DBP >90 mmHg as measured with both HR, heart rate.
Ó 2005 Blackwell Publishing Ltd Journal of Internal Medicine 257: 93–99
3. PULSE PRESSURE AND PAI-1 GENE POLYMORPHISM 95
Genomic DNA was prepared by standard salting-
Results
out techniques. The set of primers used for ampli-
fication of relevant DNA regions, the probes and the Table 2 shows the mean values (adjusted to age and
methods used to detect the gene polymorphisms gender) of the different BP measurements according
have been described previously [13, 14]. In all cases, to the studied gene polymorphism. There was no
polymerase chain reaction amplification was used. difference between the three genotypes.
The deletion (4G)/insertion (5G) polymorphism at Table 3 shows the adjusted slopes of the age–BP
position )675 of the PAI-1 promotor was studied curves in the three genotypes. Regarding PP, there
according to standard methods published in detail were significant differences (P1 < 0.034) (indicated
elsewhere [23, 24]. as P1) between the three studied genotypes. For the
The statistical analyses was conducted using data DD genotype, PP represented 36% of the variance.
collected from patients without any missing values. This interslope difference was not observed for the
Allele and genotype frequencies were analysed using age–SBP, age–DBP and age–MAP relations.
the gene-counting method. The Hardy–Weinberg Figure 1 shows the age–PP relationships (adjusted
equilibrium, checked by chi-square test [13, 14] was on gender) for the DD + ID genotypes (lower panel)
considered to be maintained. Quantitative data were versus the II genotype (upper panel). The presence of
expressed as means ± SD (1 SD). Mean comparisons D allele was associated with a significant increase in
were made with adjustment to gender by anova the slope of the relationships between age and PP by
using the F-test to evaluate whether the mean comparison with the I allele.
values of BP and other variables differed according
to the different genotypes. The aim of the study was
Discussion
to show an intergroup difference in the slope of the
curve relating age and BP [either SBP or DBP or In this study, conducted in a population of never-
mean arterial pressure (MAP) or PP] according to treated subjects with sustained essential hyperten-
gender and, in a given gender, according to the sion, an increase of the slope of the age versus PP
different genotypes of a given gene. For this a test of curve was associated with the presence of the D/D
heterogeneity of slopes was used, which represents a variant of the PAI-1 gene polymorphism. This
natural extension of covariance analysis. This finding would have been even more meaningful if
method involves features from both anova and the result had been demonstrated using the alter-
multiple regression. In a first step, the relationship native approach of longitudinal analysis. However, a
between age and BP in each genotype was explored number of previous reports have already identified
by a multiple regression analysis in order to evaluate an adequate parallelism of the age–SBP and age–PP
the slope value. To do so, regression coefficients were relationship evaluated either from cross-sectional or
the least-squares estimates of the parameters. All longitudinal investigations [3, 7, 9, 10, 13, 14]. On
these coefficients were adjusted to the standard CV the contrary, long-term follow up would have taken
risk factors, including: body mass index (BMI), many years and raised ethical concerns in view of
plasma total and high-density lipoprotein (HDL) the availability of drug treatment (mainly blockade
cholesterol, triglycerides, diabetes (yes or no) and of the renin-angiotensin system) that might offset
smoking habits (yes or no). After the slopes of the
age–BP curves have been calculated for each geno-
type, a slope comparison test was applied to evaluate Table 2 Blood pressures (age- and gender-adjusted) mean values
and standard error according to genotypes
whether or not the age- and gender-adjusted
regression coefficients of the age–BP relationship II (n ¼ 45) ID (n ¼ 82) DD (n ¼ 45)
differed amongst the three groups. A F-test was used
SBP (mmHg) 161.5 ± 2.5 164.4 ± 1.9 161.1 ± 2.6
to determine the difference in the slopes. The result DBP (mmHg) 96.2 ± 1.7 96.4 ± 1.2 97.0 ± 1.7
of the test (indicated P1 in Table 3) was considered MAP (mmHg) 118.0 ± 1.8 119.0 ± 1.3 118.4 ± 1.8
as the principal outcome measure of the study. PP (mmHg) 65.3 ± 2.0 67.9 ± 1.5 64.1 ± 2.0
Analyses were performed with sas software version SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP,
6.7 (Cary, NC, USA) under Windows NT 98. A value mean arterial pressure; PP, pulse pressure; I, insertion; D, dele-
of P £ 0.05 was considered significant. tion.
Ó 2005 Blackwell Publishing Ltd Journal of Internal Medicine 257: 93–99
4. 96 J . - J . M O U R A D et al.
Table 3 Blood pressures regression on age (adjusted on gender) in plasminogen activator inhibitor (PAI)-1 I/D gene polymorphism
according to genotypes
II (n ¼ 45) ID (n ¼ 82) DD (n ¼ 45)
2 2
R (%) Slope ± SE P-value R (%) Slope ± SE P-value R2 (%) Slope ± SE P-value P1-value
SBP 2 0.244 ± 0.255 0.344 9 0.472 ± 0.147 0.002 14 0.791 ± 0.271 0.006 0.394
DBP 4 )0.025 ± 0.174 0.887 4 0.185 ± 0.092 0.048 3 )0.098 ± 0.172 0.572 0.257
MAP 4 0.061 ± 0.178 0.732 7 0.286 ± 0.102 0.006 2 0.199 ± 0.192 0.305 0.550
PP 2 0.269 ± 0.214 0.216 7 0.286 ± 0.105 0.008 36 0.889 ± 0.184 0.0001 0.034
BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; MAP, mean arterial pressure; PP, pulse pressure; HR,
heart rate; R2, partial adjusted explanation of blood pressure variance by age and gender; Slope, regression coefficient; SE, standard error of
regression coefficient; n, sample size.
P-value: probability of regression between blood pressure and age.
P1-value: probability of interslope comparison between genotypes (see text).
(a) and positively correlated in human populations, the
Pulse pressure (mmHg)
130
120
110 haemodynamic data should be interpreted with
100
90 caution [1–7]. However, in the present study, this
80
70 issue raised minor concern since the modulating
60
50 influence of the gene polymorphisms studied on the
40 age–BP relationship was observed exclusively for PP
30
20 and not for SBP, DBP or MAP.
15 25 35 45 55 65 75 85
Age (years)
In the present population, we observed that the
P(Univariate Reg.) = 0.22 r (Univariate Corr.) = 0.19 slope of the age versus PP curve was enhanced in
subjects with the D/D genotype of the PAI-1 gene
130 (b)
r (Univariate Corr.) = 0.34 polymorphism. Studies of this polymorphism have
Pulse pressure (mmHg)
120 shown a higher plasma PAI-1 activity in subjects
110
100
90
with the D than with the I allele in patients with
80
70
myocardial infarction, non insulin-dependent dia-
60
50
betes mellitus, insulin resistance and even in the
40 y = 0.38x + 49 general populations [17, 18, 27–33]. A relationship
30
20 between fibrinolysis and blood pressure was first
15 25 35 45 55 65 75 85
Age (years)
described in the northern Sweden MONICA project
P(Univariate Reg.) = 0.0001
[34], but when excluding groups of patients who
Fig. 1 Age and pulse pressure relationships between plasminogen would be expected to have high PAI-1 levels, the
activator inhibitor (PAI)-1 II (a) and DD+ID (b) genotypes. Reg, association was lost. On the contrary, it is widely
regression; Corr, correlation.
accepted that plasma PAI-1 levels are positively
(but inconstantly) associated with increased risk for
any deleterious consequences resulting from a atherosclerotic coronary disease [17–19, 35, 36]
variant genotype [22, 25]. and mostly with increased SBP and DBP levels [37,
In many reports, the hypothesis that genetic 38]. Since, in large populations, the slope of the
variability could lead to hypertension has been plasma PAI-1 level versus SBP or DBP is steeper for
tested on the comparison of DBP mean values of SBP than for DBP [37], a positive and significant
patients with different genotypes (see review in [12, statistical link is expected between PP and the
26]). In this investigation, we exclusively tested the plasma PAI-1 level. Our finding of a steeper slope of
hypothesis that genes might modulate the age- the age versus PP relationship for the D/D genotype
influence on PP and therefore play a role in CV risk (Fig. 1) agrees with this possibility and suggests
through an alteration in the course of this mechan- subtle links between increased PP, CV risk and
ical factor. Since PP represents the difference prothrombotic state [34–36]. Indeed, subjects with
between SBP and DBP, and PP and SBP are strongly the 4G allele might have the higher values of PAI-1
Ó 2005 Blackwell Publishing Ltd Journal of Internal Medicine 257: 93–99
5. PULSE PRESSURE AND PAI-1 GENE POLYMORPHISM 97
and of prothrombotic factors. They may have also a
Acknowledgements
steeper increase of PP with age and, accordingly,
the higher risk for myocardial infarction. Finally, This study was performed with the help of INSERM,
impaired fibrinolytic balance is a feature of many ´ ´ ´
Societe Francaise d’Hypertension Arterielle and
¸
patient populations including postmenopausal wo- GPH-CV (Groupe de Pharmacologie et d’Hemody- ´
men and individuals with hypertension, insulin namique Cardiovasculaire), Paris. Authors thank
resistance, type 2 diabetes and hypertriglyceridae- Mrs Anne Safar for her skilful secretarial help.
mia, all of them having an increased risk for CV
events [37, 39]. We suggest that PP may be the
References
main mechanical factor enabling to establish a
consistent link between these different populations. 1 Darne B, Girerd X, Safar M, Cambien F, Guize L. Pulsatile
versus steady component of blood pressure: a cross-sectional
Others and we have shown that the level of aortic
analysis and a prospective analysis on cardiovascular mor-
PP (but not of SBP or DBP) is significantly and tality. Hypertension 1989; 13: 392–400.
positively related with the number of arterial 2 Madhavan S, Ooi WL, Cohen H, Alderman MH. Relation of
coronary stenosis in subjects with coronary ischae- pulse pressure and blood pressure reduction to the incidence
of myocardial infarction. Hypertension 1994; 23: 395–401.
mic disease and preserved cardiac function [40–
3 Franklin SS, Khan SA, Wong ND, Larson MG, Levy D. Is pulse
42]. pressure useful in predicting risk for coronary heart disease?
In the present study, the age–PP relationship of The Framingham Heart Study. Circulation 1999; 100:
the D/D genotype could not be explained on the 354–60.
4 Millar JA, Lever AF, Burke V. Pulse pressure as a risk factor
basis of cardiac factors which are poorly related to
for cardiovascular events in the MRC mild hypertension trial.
fibrinolytic and/or haemostatic agents. The age–PP J Hypertens 1999; 17: 1065–72.
relationship was rather modulated by vascular 5 Benetos A, Safar M, Rudnichi A et al. Pulse pressure: a pre-
factors as arterial stiffness and wave reflections. dictor of long-term cardiovascular mortality in a French male
population. Hypertension 1997; 30: 1410–5.
This observation is consistent with previous inves-
6 Mitchell GF, Moye LM, Braunwald E et al. Sphygmomano-
tigations in the past where we showed subtle but metrically determined pulse pressure is a powerful inde-
consistent statistical links between plasma PAI-1, pendent predictor of recurrent events after myocardial
arterial stiffness and wave reflections, the two latter infarction in patients with impaired left ventricular function.
Circulation 1997; 96: 4254–60.
factors being the main determinants of PP. Indeed,
7 Benetos A, Zureik M, Morcet J et al. A decrease in diastolic
elevated plasma PAI-1 levels are significantly asso- blood pressure combined with an increase in systolic blood
ciated with reduced carotid distensibility and/or pressure is associated with high cardiovascular mortality.
altered PP amplification, both in subjects with J Am Coll Cardiol 2000; 35: 673–80.
8 Blacher J, Gasowski J, Staessen JA et al. Pulse pressure – not
hypertension and overweight and in patients with
mean pressure – determines cardiovascular risk in older
chronic renal disease [43–45]. hypertensive patients. Arch Intern Med 2000; 160: 1085–9.
This study had several limitations: the cohort is 9 Franklin SS, Gustin W IV, Wong ND et al. Hemodynamic
relatively small, and therefore required that the patterns of age-related changes in blood pressure. The Fra-
mingham Heart Study. Circulation 1997; 96: 308–15.
findings should be constantly adjusted to age and
10 Benetos A, Adamopoulos C, Bureau JM et al. Determinants of
sex. On the contrary, and for the same reason, gene– accelerated progression of arterial stiffness in normotensive
gene interactions were not evaluated, particularly subjects and in treated hypertensive subjects over a 6-year
with the renin-angiotensin system [14, 39, 46]. period. Circulation 2002; 105: 1202–7.
11 du Cailar G, Mimram A, Fesler P, Ribstein J, Blacher J, Safar
In conclusion, the results of present study have
ME. Dietary sodium and pulse pressure in normotensive and
shown a possible influence of PAI-1 polymorphism essential hypertensive subjects. J Hypertens 2004; 22: 697–
on the increase of PP with age in hypertensive 703.
subjects. Because PP is a consistent and independent 12 Avolio A. Genetic and environmental factors in the function
and structure of the arterial wall. Hypertension 1995; 26:
predictor of CV mortality [1–7], this result gives a
34–7.
new insight on the relationship between genes, 13 Lajemi M, Labat C, Gautier S et al. Angiotensin II type 1
mechanical factors and CV risk. receptor )153A/G and 1166A/C gene polymorphisms and
increase in aortic stiffness with age in hypertensive subjects.
J Hypertens 2001; 19: 407–13.
Conflict of interest statement 14 Safar ME, Lajemi M, Rudnichi A, Asmar R, Benetos A.
Angiotensin-converting enzyme D/I gene polymorphism and
No conflict of interest was declared.
Ó 2005 Blackwell Publishing Ltd Journal of Internal Medicine 257: 93–99
6. 98 J . - J . M O U R A D et al.
age-related changes in pulse pressure in subjects with 30 Margaglione M, Cappucci G, D’addedda M et al. PAI-1 plasma
hypertension. Arterioscler Thromb Vasc Biol 2004; 24: 782– levels in a general population without clinical evidence of
6. atherosclerosis. Relation to environmental and genetic
15 Perusse L, Moll PP, Sing C. Evidence that a single gene with determinants. Arterioscler Thromb Vasc Biol 1998; 18: 562–7.
gender and age-dependant effects influences systolic blood 31 Eriksson P, Kallin B, Van’T Hooft FM, Bavenholm P, Hamsten
pressure determination in a population-based sample. Am J A. Allele-specific increase in basal transcription of the
Hum Genet 1991; 49: 94–105. plasminogen-activator inhibitor 1 gene is associated with
16 Staessen JA, Wang JG, Brand E et al. Effects of three candidate myocardial infarction. Proc Natl Acad Sci U S A 1995; 92:
genes on prevalence and incidence of hypertension in a 1851–5.
Caucasian population. J Hypertens 2001; 19: 1349–58. 32 Ye S, Green FR, Scarabin PY et al. The 4G/5G genetic poly-
17 Hindorff LA, Schwartz SM, Siscovick DS, Psaty BM, Longst- morphism in the promoter of the plasminogen activator
reth WT Jr, Reiner AP. The association of PAI-1 promoter 4G/ inhibitor-1 (PAI-1) gene is associated with differences in
5G insertion/deletion polymorphism with myocardial infarc- plasma PAI-1 activity but not with risk of myocardial
tion and stroke in young women. J Cardiovasc Risk 2002; 9: infarction in the ECTIM study. Etude CasTemoins de I’nfarctus
131–7. du Mycocarde. Thromb Haemost 1995; 74: 837–41.
18 Ossei-Gerning N, Mansfield MW, Stickland MH, Wilson IJ, 33 Kohler HP, Grant PJ. Plasminogen-activator inhibitor type 1
Grant PJ. Plasminogen activator inhibitor-1 promoter 4G/5G and coronary artery disease. N Engl J Med 2000; 342: 1792–
genotype and plasma levels of myocardial infarction in pa- 801.
tients characterized by coronary angiography. Arterioscler 34 Eliasson M, Evrin PE, Lundblad D. Fibrinogen and fibrinolytic
Thromb Vasc Biol 1997; 17: 33–7. variables in relation to anthropometry, lipids and blood
19 Lijnen HR, Collen D. Impaired fibrinolysis and the risk for pressure. The Northern Sweden MONICA Study. J Clin Epi-
coronary heart disease. Circulation 1996; 94: 2052–4. demiol 1994; 47: 513–24.
20 Williams B. Mechanical influences on vascular smooth mus- 35 Hamsten A, De Faire U, Walldius G et al. Plasminogen acti-
cle cell function. J Hypertens 1998; 16: 1921–9. vator in plasma: risk factor for recurrent myocardial infarc-
21 Cheng JJ, Chao YJ, Wung BS, Wang DL. Cyclic strain-induced tion. Lancet 1987; 2: 3–9.
plasminogen activator inhibitor-1 (PAI-1) release from 36 Juhan-Vague I, Pyke SDM, Alessi MC, Jespersen J, Haverkate
endothelial cells involves reactive oxygen species. Biochem I, Thomson SG. Fibrinolytic factor and the risk of myocardial
Biophys Res Commun 1996; 225: 100–5. infarction and sudden death in patients with angina pectoris.
22 Brown NJ, Murphey LJ, Srikuma N, Koschachuhanan N, Circulation 1996; 94: 2057–63.
Williams GH, Vaughan DE. Interactive effect of PAI-1 4G/5G 37 Poli KA, Tofler GH, Larson MG et al. Association of blood
genotype and salt intake on PAI-1 antigen. Arterioscler pressure with fibrinolytic potential in the Framingham off-
Thromb Vasc Biol 2001; 21: 1071–7. spring population. Circulation 2000; 101: 264–9.
23 Miller RC, Schini V, Schoeffter P. Modulation by the endo- 38 Eliasson M, Jansson J-H, Nilsson P, Asplund K. Increased
thelium of agonist-induced contractions of vascular smooth levels of tissue plasminogen activator antigen in essential
muscle. In: Vanhoutte PM, ed. Relaxing and Contracting Fac- hypertension. A population-based study in Sweden. J Hyper-
tors, Biological and Clinical Research. Clifton, NJ, USA: Humana tens 1997; 15: 349–56.
Press Inc., 1988; 241–65. 39 Vaughan DE. Angiotensin and vascular fibrinolytic balance.
24 Henry M, Chomiki N, Scarabin PY et al. Five frequent poly- Am J Hypertens 2002; 15(1 Pt 2): 3S–8S.
morphisms of plasminogen activator inhibitor 1 gene: lack of 40 Danchin N, Benetos A, Lopez-Sublet M, Demicheli T, Safar M,
association between genotypes, PAI activity and triglyceride Mourad JJ. Aortic pulse pressure is related to the presence and
levels in a healthy population. Arterioscler Thromb Vasc Biol extent of coronary artery disease in men undergoing diag-
1997; 17: 851–8. nostic coronary angiography: a multicenter study. Am J
25 Morris BJ, Zee RYL, Schrader AP. Different frequencies of Hypertens 2004; 17: 129–33.
angiotensin-converting enzyme genotypes in older hyperten- 41 Nishijima T, Nakayama Y, Tsumura K et al. Pulsatility of
sive individuals. J Clin Invest 1994; 94: 1085–9. ascending aortic blood pressure waveform is associated with
26 Luft FC. Molecular genetics of human hypertension. an increased risk of coronary heart disease. Am J Hypertens
J Hypertens 1998; 16: 1871–8. 2001; 14: 469–73.
27 Dawson SJ, Wiman B, Hamsten A, Green F, Humphries S, 42 Nakayama Y, Tsumura K, Yamashita N, Yoshimaru K,
Henney AM. The two allele sequences of a common poly- Hayashi T. Pulsatility of ascending aortic pressure waveform
morphism in the promoter of the plasminogen activator is a powerful predictor of restenosis after percutaneous
inhibitor-1 (PAI-1) gene respond differently to interleukin-1 transluminal angioplasty. Circulation 2000; 101: 470–2.
in HepG2 cells. J Biol Chem 1993; 268: 10739–45. 43 Pannier B, Guerin AP, Marchais SJ, Metivier F, Safar ME,
28 Panahloo A, Mohamed-Ali V, Lane A, Green F, Humphries London GM. Postischemic vasodilation, endothelial activa-
SE, Yudkin JS. Determinants of plasminogen activator inhib- tion, and cardiovascular remodeling in end-stage renal dis-
itor 1 activity in treated NIDDM and its relation to a poly- ease. Kidney Int 2000; 57: 1091–9.
morphism in the plasminogen activator inhibitor 1 gene. 44 Achimastos A, Brahimi M, Raison J et al. Plasma insulin,
Diabetes 1995; 44: 37–42. plasminogen activator inhibitor, and ankle-brachial systolic
29 Mansfield MW, Stickland MH, Grant PJ. Environmental and blood pressure ratio in overweight hypertensive subjects.
genetic factors in relation to elevated circulating levels of J Hum Hypertens 1999; 13: 329–35.
plasminogen activator inhibitor-1 in Caucasian patients with 45 Nishiwaki Y, Takebayashi T, Omae K, Ishizuka C, Nomiyama
non insulin-dependent diabetes mellitus. Thromb Haemost T, Sakurai H. Relationship between the blood coagulation-
1995; 74: 842–7. fibrinolysis system and the subclinical indicators of arterio-
Ó 2005 Blackwell Publishing Ltd Journal of Internal Medicine 257: 93–99
7. PULSE PRESSURE AND PAI-1 GENE POLYMORPHISM 99
sclerosis in a healthy male population. J Epidemiol 2000; 10: ˆ
Correspondence: Prof. Michel Safar, Diagnosis Center, Hopital
34–41. ˆ
Hotel-Dieu, 1, place du Parvis Notre-Dame, 75181 Paris Cedex
46 Moore JH, Lamb JM, Brown NJ, Vaughan DE. Comparison of 04, France.
combinatorial partitioning and linear regression for the (fax: +33 1 42 34 86 32; e-mail: michel.safar@htd.ap-hop-
detection of epistatic effects of the ACE I/D and PAI-1 4G/5G paris.fr).
polymorphisms on plasma PAI-1 levels. Clin Genet 2002; 62:
74–9.
Ó 2005 Blackwell Publishing Ltd Journal of Internal Medicine 257: 93–99