Increased central pulse pressure and augmentation index in subjects with hypercholesterolemia

of 7

Please download to get full document.

View again

All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
7 pages
0 downs
Increased central pulse pressure and augmentation index in subjects with hypercholesterolemia
  Coronary Artery Disease Increased Central Pulse Pressureand Augmentation Index inSubjects With Hypercholesterolemia Ian B. Wilkinson, MA, BM,* Krishna Prasad, MB,† Ian R. Hall, BS C , MB,† Anne Thomas, BN,†Helen MacCallum, BN,* David J. Webb, MD,† Michael P. Frenneaux, MD,† John R. Cockcroft, BSc, MB†  Edinburgh and Cardiff, United Kingdom OBJECTIVES  The goal of this study was to investigate the relation between serum cholesterol, arterialstiffness and central blood pressure. BACKGROUND  Arterial stiffness and pulse pressure are important determinants of cardiovascular risk.However, the effect of hypercholesterolemia on arterial stiffness is controversial, and centralpulse pressure has not been previously investigated. METHODS  Pressure waveforms were recorded from the radial artery in 68 subjects with hypercholester-olemia and 68 controls, and corresponding central waveforms were generated using pulse wave analysis. Central pressure, augmentation index (AIx) (a measure of systemic stiffness)and aortic pulse wave velocity were determined. RESULTS  There was no significant difference in peripheral blood pressure between the two groups, butcentral pulse pressure was significantly higher in the group with hypercholesterolemia (37  11 mm Hg vs. 33  10 mm Hg [means  SD]; p  0.028). Augmentation index was alsosignificantly higher in the patients with hypercholesterolemia group (24.8  11.3% vs. 15.6  12.1%; p  0.001), as was the estimated aortic pulse wave velocity. In a multiple regressionmodel, age, short stature, peripheral mean arterial pressure, smoking and low-density lipoprotein cholesterol correlated positively with AIx, and there was an inverse correlation with heart rate and male gender. CONCLUSIONS  Patients with hypercholesterolemia have a higher central pulse pressure and stiffer blood vessels than matched controls, despite similar peripheral blood pressures. These hemody-namic changes may contribute to the increased risk of cardiovascular disease associated withhypercholesterolemia, and assessment may improve risk stratification. (J Am Coll Cardiol2002;39:1005–11) © 2002 by the American College of Cardiology Foundation  Arterial stiffening is an inevitable part of the aging processin nearly all societies (1,2), and it is associated with a rise inpulse pressure (3). The importance of pulse pressure as apredictor of cardiovascular mortality has been demonstratedby several recent population-based and interventional stud-ies (4–7) and has focused attention on arterial stiffness as akey determinant of cardiovascular risk. “Premature arterialstiffening” occurs in association with several importantcardiovascular risk factors including hypertension, diabetesmellitus and cigarette smoking (8), which are also associated with endothelial dysfunction (9). Whether arterial stiffeningrepresents a marker of occult atheroma, or endothelialdysfunction, or is more directly involved in the process of atherosclerosis remains unclear (10).Hypercholesterolemia is a major risk factor for cardiovas-cular disease, and it is also associated with endothelialdysfunction (11). However, there are conflicting data con-cerning the relationship between hypercholesterolemia and local   arterial stiffness (12). Moreover, the effect of hypercho-lesterolemia on  systemic   arterial stiffness has only beeninvestigated in one previous study (13), and central bloodpressures were not assessed. We hypothesized that hyper-cholesterolemia would be associated with increased  systemic  arterial stiffness and central pulse pressure, and the aim of this study was to test this hypothesis in a group of otherwisehealthy individuals with hypercholesterolemia and matchedcontrols using the technique of pulse wave analysis (PWA). We have previously demonstrated that PWA is a reproduc-ible, noninvasive method for assessing central blood pres-sure and augmentation index (AIx) (14)—a measure of thecontribution that wave reflection makes to the arterialpressure waveform. The amplitude and timing of thereflected wave ultimately depend on the stiffness of the small(preresistance) vessels and large arteries and thus, AIxprovides a measure of   systemic   arterial stiffness. From the *Clinical Pharmacology Unit, Department of Medical Sciences, Univer-sity of Edinburgh, Western General Hospital, Edinburgh, United Kingdom; and the†Department of Cardiology, University of Wales College of Medicine, University Hospital, Cardiff, United Kingdom. Supported, in part, by the High Blood PressureFoundation, by a local Research and Development Grant (Lothian Universities NHSHospital Trust) and by an unrestricted educational grant from Bayer Pharmaceuticals,PLC. Dr. I. B. Wilkinson, Dr. J. R. Cockcroft and Prof. Dr. J. Webb are supportedby a Biomedical Research Collaboration Grant from the Wellcome Trust (056223).Prof. D. J. Webb is currently in receipt of a Research Leave Fellowship from the Wellcome Trust (052633). Dr. Wilkinson is now at the Clinical Pharmacology Unit,University of Cambridge, Addenbrooke’s Hospital, Cambridge, United Kingdom.Manuscript received January 13, 2000; revised manuscript received December 7,2001, accepted December 21, 2001. Journal of the American College of Cardiology Vol. 39, No. 6, 2002© 2002 by the American College of Cardiology Foundation ISSN 0735-1097/02/$22.00Published by Elsevier Science Inc. PII S0735-1097(02)01723-0  METHODS Sixty-eight patients with hypercholesterolemia, defined as atotal serum cholesterol   6.5 mmol/l, were recruited fromthe cardiovascular risk clinics at the Western GeneralHospital in Edinburgh and the University Hospital of  Wales in Cardiff and from general practices local to bothhospitals. Concurrently, normocholesterolemic controls (to-tal serum cholesterol  6.5 mmol/l) were recruited from thesame sources and community databases of volunteers held atboth hospitals and selected such that, as a group, theirdistribution of age, gender and weight closely matched thepatient group. Approval for the study was obtained from therespective Local Research Ethics Committees, and in-formed consent was obtained from each participant. Pa-tients with elevated blood pressure (brachial artery bloodpressure  160/100 mm Hg), treated hypertension, diabetesmellitus (British Diabetic Association criteria) or a clinicalhistory of cardiovascular disease were excluded, as weresubjects receiving any medication. Cigarette smokers wereallowed to participate, having abstained for at least 4 h. Blood pressure measurement.  Systolic and diastolic bloodpressure were recorded in duplicate in the dominant armusing a validated oscillometric technique (HEM-705CP,Omron Corporation, Tokyo, Japan) (15). Peripheral meanarterial pressure (PMAP) was calculated by integration of the pressure waveform using the sphygmoCor version 6.1software (AtCor Medical, Sydney, Australia). Measurement of arterial stiffness and central blood pres-sure.  Pulse wave analysis was used to determine systemicarterial stiffness (sphygmoCor version 6.1 software) aspreviously described (14). A high-fidelity micromanometer(SPC-301, Millar Instruments, Houston, Texas) was usedto obtain accurate recordings of the peripheral pressure waveforms by flattening, but not occluding, the radial artery of the dominant arm using gentle pressure (16). Data werecollected directly into a microcomputer and, after 20 se-quential waveforms had been acquired, an averaged periph-eral waveform was generated. A corresponding averagedcentral pressure waveform was then generated using a validated transfer function (17–20), and, from this, augmen-tation, AIx, ascending aortic pressure and heart rate werethen determined using the integral software. Augmentation represents the difference between the sec-ond and first systolic peaks of the central pressure waveform,and AIx is defined as augmentation expressed as a percent-age of the pulse pressure (Fig. 1) and is a measure of   systemic  arterial stiffness. The tension time index (TTI), the areaunder the systolic portion of the pressure waveform perminute (the systolic pressure-time integral), an index of systolic stress and the diastolic pressure-time integral(DPTI) were also determined (21,22). The aortic pulse wave velocity was estimated by calculating the time betweenthe foot of the pressure wave and the inflection point, whichprovides a measure of the timing of the reflected wave, asdescribed previously (23,24). Study protocol.  After 5 min seated rest in a quiet room,blood pressure was measured. Augmentation, AIx andcentral arterial pressure were then determined using PWA. All measurements were made in duplicate and mean valuesused for analysis. Venous blood was then drawn from theantecubital fossa for measurement of plasma glucose, totalserum cholesterol, low-density lipoprotein (LDL), high-density lipoprotein (HDL) cholesterol and triglycerides.Finally, height and weight were recorded and body massindex calculated. Data analysis.  Data were analyzed using unpaired, two-tailed Student  t   tests and multiple regression analysis (SPSSsoftware package, version 10, SPSS Inc., Chicago, Illinois).Unless stated otherwise, all results are presented as means  SD. Significance was defined as p  0.05. RESULTS Sixty-eight subjects with hypercholesterolemia, aged 51   10 years (range: men 29 to 77 years, women 39 to 73 years),and 68 control subjects, aged 51  10 years (range: men 25to 75 years, women 32 to 74 years), were entered into thestudy. There was no significant difference in age, gender,height, weight or the number of smokers between the twogroups, but serum LDL cholesterol and triglycerides weresignificantly higher in the hypercholesterolemic group (Ta-ble 1). A summary of the peripheral and central hemodynamicsis presented in Table 2 and Figure 2. Peripheral bloodpressure did not differ between the groups, although the  Abbreviations and Acronyms  AIx    augmentation indexDPTI    diastolic pressure time integralFH    familial hypercholesterolemiaHDL    high-density lipoproteinLDL    low-density lipoproteinPMAP    peripheral mean arterial pressurePWA    pulse wave analysis TTI    tension time index Figure 1.  Representative ascending aortic waveforms from a young andolder subject (Murgo type “C” and “A” waves, respectively); note theprominence of the second systolic peak (*) in the older individual. Augmentation index is defined as the difference between the second andfirst systolic peaks (  P) expressed as a percentage of the pulse pressure(PP). 1006 Wilkinson  et al.  JACC Vol. 39, No. 6, 2002 Central Pressure and Hypercholesterolemia  March 20, 2002:1005–11  control group had a higher resting heart rate. Augmenta-tion, AIx and central pulse pressure were significantly higherin the subjects with hypercholesterolemia compared withcontrols, and there was a trend towards a higher centralsystolic pressure in the hypercholesterolemic group. Thedifference in AIx between the two groups persisted aftercorrection for height and heart rate (18.4  7.6 vs. 23.1  6.1, p    0.004). The TTI (2,296    380 vs. 2,320    438,p  0.7) and DPTI (3,743  559 vs. 3,613  455, p  0.2) were similar in both groups. There were fewer Murgo type“A” waves (23) in the control subjects compared with thesubjects with hypercholesterolemia (40 vs. 57) and moretype “B” (20 vs. 10) and “C” (8 vs. 1) waves (p  0.003 fortrend; chi-square).Data from all 136 subjects were used to construct twomultiple regression models with AIx and estimated aorticpulse wave velocity as the dependent variables. Age, gender,height, heart rate, PMAP, smoking history, LDL choles-terol, HDL cholesterol and triglycerides were entered intothe models as known or likely determinants of arterialstiffness. Age, short stature, PMAP, smoking and LDLcholesterol, but not HDL cholesterol or triglycerides, cor-related positively with AIx, and there was an inversecorrelation with heart rate and male gender (Table 3). Themodel explained  62% of the variability in AIx observed inthe study (p    0.001). Substituting augmentation for thedependent variable did not make a significant difference tothe correlations in the regression model, neither did theaddition of plasma glucose or changing the method of analysisto stepwise regression. Estimated aortic pulse wave velocity correlated positively with LDL cholesterol and smoking andinversely with male gender and height (Table 4). DISCUSSION  A number of studies have investigated the relationshipbetween hypercholesterolemia and arterial stiffness. In ani-mals, hypercholesterolemia, induced by a cholesterol-richdiet, results in an initial reduction in arterial stiffness,followed by a progressive increase over time (25,26), whichcan be reversed by lowering serum cholesterol (26,27). Inhumans, a variety of techniques have been used to assess vascular stiffness in vivo, with conflicting results. Lehmannet al. (28) demonstrated increased distensibility, as deter-mined by aortic pulse wave velocity, in a group of youngpatients with heterozygous familial hypercholesterolemia(FH) but, in contrast, reported reduced aortic distensibility in adults with FH (12). Others have also observed reducedradial artery compliance in adults with FH compared withcontrols (29). However, Toikka et al. (30) reported nodifference in aortic or carotid compliance between asymp-tomatic adults with FH and controls.In subjects without FH, most (30–33), but not all (13),studies have reported an inverse relationship between eitherLDL cholesterol or total cholesterol and aortic compliance.Differences in the vessels studied, patient selection andsmall sample sizes are likely to account for these discrepantfindings. Indeed, Kool et al. (34) reported reduced arterialdistensibility in some, but not all, arteries in individuals withhypercholesterolemia. However, all of these studies, withone exception (13), have focused on the stiffness of easily accessible arterial beds or even one small segment of anartery, and not on  systemic   arterial stiffness, which has amajor influence on the interaction between the arterial treeand the left ventricle (35). Indeed, the shape of the centralpressure waveform partly determines left ventricular work-load (35). Therefore, assessing systemic stiffness may be of more clinical importance than measuring compliance withinan isolated vascular bed. Moreover, the effect of hypercho-lesterolemia on central blood pressure has not been previ-ously investigated.In this study we recorded peripheral pressure waveformsusing applanation tonometry and generated corresponding  Table 2.  Hemodynamics HR (min  1 )PSBP(mm Hg)PDBP(mm Hg)PPP(mm Hg) Alx (%) Aug (mm Hg) T R  (ms)CSBP(mm Hg)CDBP(mm Hg)CPP(mm Hg)PPP:CPP(Ratio) HC 70  9 132  15 84  11 49  12 24.8  11.3 9.9  5.9 140  13 122  17 85  12 37  11 1.34  0.25Controls 74  12 131  15 84  11 47  13 15.6  12.1 5.8  5.3 145  11 118  17 85  11 33  10 1.45  0.24Significance(p Value)0.026 0.6 1.0 0.5   0.001   0.001 0.013 0.1 0.9 0.028 0.012  All values are expressed as means  SD. AIx  augmentation index; Aug  systolic augmentation; CDBP  central diastolic blood pressure; CPP  central pulse pressure; CSBP  central systolic blood pressure;HC  hypercholesterolemics; HR   heart rate; PDBP  peripheral diastolic blood pressure; PPP  peripheral pulse pressure; PSBP  peripheral systolic blood pressure.  Table 1.  Subject Characteristics Number (n)Men(n) Age(yrs)Height (m) Weight (kg)Smokers(n)Cholesterol TG(mmol/l)Glucose(mmol/l) Total(mmol/l)LDL(mmol/l)HDL(mmol/l) Hypercholesterolemics 68 25 51  10 1.70  0.08 77.5  14.4 24 7.1  0.5 4.7  0.7† 1.3  0.4 2.2  1.4* 5.1  1.0Controls 68 25 51  10 1.73  0.10 78.8  13.0 21 5.0  0.7 2.8  0.7 1.4  0.5 1.6  0.8 5.2  0.7  All values are expressed as means  SD. There were no significant differences between the groups for any parameter, unless stated. *p  0.05; †p  0.001 (unpaired Student t   test).HDL  high-density lipoprotein; LDL  low-density lipoprotein; TG  triglycerides. 1007 JACC Vol. 39, No. 6, 2002  Wilkinson  et al. March 20, 2002:1005–11  Central Pressure and Hypercholesterolemia  central arterial pressure waveforms using PWA. The mainnovel findings were an increased AIx and central pulsepressure in individuals with hypercholesterolemia compared with matched normocholesterolemic controls (Fig. 2). The AIx is a measure of the contribution made by the reflectedpressure wave to the ascending aortic pressure waveform(36) and, thus, provides a measure of   systemic   arterialstiffness (37). Therefore, the present data demonstrate thathypercholesterolemia is associated with increased  systemic  arterial stiffness; this conclusion is supported by the higherfrequency of Murgo type “A” waves among the subjects withhypercholesterolemia. Importantly, aortic pulse wave veloc-ity, estimated from the timing of the reflected wave, was alsoincreased in the subjects with hypercholesterolemia, indi-cating increased  aortic   stiffness and that the observedchanges in AIx were not solely due to differences in wavereflection. A number of factors, including gender, heart rate andheight are known to influence AIx, independently of changes in stiffness (22,38). However, none of these vari-ables, other than heart rate, differed significantly betweenthe two groups. Moreover, after correction for heart rate andheight, AIx remained significantly higher in the hypercho-lesterolemic group. Furthermore, in the multiple regressionmodel, which included known confounding variables, LDL,but not HDL, cholesterol was a significant independentdeterminant of AIx (Table 3). The model accounted for  65% of the observed variability of AIx, which comparesfavorably with other published multiple regression models(22,32). In addition, we also confirmed the positive relationof age, female gender, short stature, PMAP and low heartrate with AIx (22) and here, for the first time, demonstratedthat chronic smoking is associated with an increased AIx(Table 3). We and others have previously demonstrated thatindividuals with diabetes mellitus (39) and insulin resistance(40) have a higher AIx than matched controls and thatinsulin per se can influence arterial stiffness (41). However,there was no relation between AIx and serum glucose in thispresent study. The most likely explanation for this is that weactively excluded patients with diabetes mellitus and, there-fore, only studied individuals within a rather narrow rangeof serum glucose concentrations. The multiple regression model investigating the determi-nants of estimated aortic pulse wave velocity confirmed that  Table 3.  Results of the Multiple Regression Analysis With Augmentation Index as theDependent Variable Parameters UnitsRegressionCoefficient SE Beta Significance Gender male   8.98 2.13   0.34   0.001 Age years 0.45 0.08 0.38   0.001Heart rate min  1  0.40 0.07   0.33   0.001PMAP mm Hg 0.198 0.061 0.20 0.001LDL cholesterol mmol/l 2.18 0.71 0.20 0.003Height m   20.89 10.67   0.16 0.043Smoker 4.076 1.536 0.15 0.009HDL cholesterol mmol/l 3.41 2.03 0.12 0.1 Triglycerides mmol/l 1.34 0.78 0.12 0.1 Multiple regression analysis using all 136 subjects. R  2  value for the entire study group  0.651, p  0.001. The regression coefficient provides the slope of the regression line relating each parameter to augmentation index, and thecoefficient beta provides a measure of the relative strength of the association independent of the units of measurement.Parameters listed in descending value of beta.HDL    high-density lipoprotein; LDL    low-density lipoprotein; PMAP    peripheral mean arterial pressure; SE   standard error. Figure 2.  Box and whisker plots for augmentation index (AIx), augmentation and estimated aortic pulse wave velocity (T R  ) for the hypercholesterolemia(HC) and control subjects.  Solid horizontal lines  median values;  error bars  95% confidence intervals;  shaded area   interquartile range. n  163;*p  0.01; **p  0.001. 1008 Wilkinson  et al.  JACC Vol. 39, No. 6, 2002 Central Pressure and Hypercholesterolemia  March 20, 2002:1005–11  LDL cholesterol was independently associated with aorticstiffening. Height was inversely related to estimated pulse wave velocity, which is to be expected, as we did not correctfor path length but simply used the arrival of the reflected wave as a measure of aortic pulse wave velocity. Therefore,for a given pulse wave velocity, taller subjects with a longeraortae will have a prolonged transit time.Peripheral pulse pressure, a surrogate measure of arterialstiffness and important predictor of cardiovascular mortality (4,5,42), did not differ significantly between the two groups.However, we have demonstrated, for the first time, in-creased  central   pulse pressure in subjects with hypercholes-terolemia compared with controls and a reduction in thedegree of pressure amplification. Systolic pressure variesthroughout the arterial tree due to wave reflection anddifferences in vessel stiffness. Normally, there is an ampli-fication of pulse pressure from the central to peripheralarteries, but the degree of amplification depends on anumber of factors including heart rate (38) and posture (43). Therefore, peripheral pulse pressure does not always predictcentral pulse pressure, which is important because central,not brachial artery, pressure best defines left ventricular workload and, thus, left ventricular mass—an important andindependent predictor of cardiovascular mortality (44).Moreover, carotid pulse pressure correlates more closely  with carotid intima-media thickness than brachial pulsepressure (45) and predicts restenosis after angioplasty (46).However, whether measurement of central pulse pressure will improve risk stratification remains to be proven. Nev-ertheless, we have demonstrated that simply assessing bra-chial artery pressure does not reliably predict central hemo-dynamics, which may explain the lack of correlationbetween peripheral blood pressure and serum cholesterol ina recent study (47). Whether structural changes in the arterial wall alone areresponsible for the increase in arterial stiffness reported inthis and other studies is unclear. Although we did notinclude subjects with a history of vascular disease or mani-fest evidence of atheroma, we cannot exclude the possibility of occult atheroma in the hypercholesterolemic group beingresponsible for the observed arterial stiffening. However,mounting evidence suggests that there is a degree of functional regulation of stiffness by the vascular endothe-lium (48,49). Hypercholesterolemia is strongly associated with endothelial dysfunction and reduced nitric oxide bio-availability (11,49), which may lead to a degree of functionaland, therefore, potentially reversible arterial stiffening. In-deed, in animal models of hypercholesterolemia, cholesterollowering appears to reduce arterial stiffness (26,27). Inhumans, HMG-CoA reductase (statin) therapy has alsobeen associated with a reduction in aortic pulse wave velocity within six months (50), but not any improvement infemoral, carotid or brachial artery distensibility within eight weeks (34). Finally, although we assessed arterial stiffnessand central pressures noninvasively using PWA rather thandirectly, the transfer function involved has been validated.However, aortic pulse wave velocity was estimated from thetiming of the reflected pressure wave rather than measuredusing a foot-to-foot methodology.In summary, we have shown increased  systemic arterial  and  aortic   stiffness and, for the first time, central pulsepressure in asymptomatic patients with hypercholesterol-emia. Stiffness was independently correlated with LDL butnot HDL cholesterol. Large-scale studies are now requiredto investigate further the influence of hypercholesterolemiaon central arterial pressure and left ventricular load. Multi-center studies investigating the effect of various therapeuticinterventions on arterial stiffness, including cholesterol-lowering and folate supplementation in hypercholester-olemic subjects (Study of the Effectiveness of AdditionalReduction in Cholesterol and Homocysteine [SEARCH]),are currently ongoing (51), and these will serve not only toprovide data concerning the importance of systemic arterialstiffness as a predictor of outcome but will also determinethe effects of therapeutic intervention. Mechanistic studiesare also required to investigate the relationship betweencholesterol, arterial stiffness and endothelial function andmay serve to elucidate the role of the endothelium inregulation of arterial stiffness.  Table 4.  Results of the Multiple Regression Analysis With Estimated Aortic Pulse WaveVelocity (T R  ) as the Dependent Variable Parameters UnitsRegressionCoefficient SE Beta Significance Gender male 8.80 2.51 0.38 0.001LDL cholesterol mmol/l   1.63 0.84   0.17 0.046PMAP mm Hg   0.15 0.08   0.17 0.06Heart rate min  1  0.16 0.08   0.17 0.06Smoker   4.22 2.14   0.17 0.050 Age yrs   0.17 0.09   0.16 0.08Height m 13.45 12.60 0.12 0.037HDL cholesterol mmol/l 0.50 2.34 0.02 0.8 Triglycerides mmol/l   0.43 0.92   0.04 0.6 Multiple regression analysis using all 136 subjects. R  2  value for the entire study group  0.364, p  0.001. Parameters listed indescending value of beta.HDL  high-density lipoprotein; LDL  low-density lipoprotein; PMAP  peripheral mean arterial pressure. 1009 JACC Vol. 39, No. 6, 2002  Wilkinson  et al. March 20, 2002:1005–11  Central Pressure and Hypercholesterolemia
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks