Abstract
Objective. To investigate determinants of increase in aortic pulse wave velocity (PWV) in elderly subjects free from overt cardiovascular disease and not treated for arterial hypertension at baseline. Methods. The present study included 90 lecture attendees (“Continuing Adult Education”) who were examined at baseline and after a median follow-up of 9.5 years, including the PWV measurement using SphygmoCor. We used multiple linear regression analyses to assess predictors of PWV change. As independent covariates, we considered parameters with known effect on arterial stiffness and use of antihypertensive and lipid lowering medication. Results. At baseline, mean age was 66.9 ± 5.1 years, and 37.8% of subjects had arterial hypertension, respectively. The PWV increased from 9.4 to 10.2 m/s; p = 0.035. While accounting for covariates, PWV was significantly and independently associated with four factors: baseline heart rate (β = 0.074, χ2 = 7.40; p = 0.0079), mean arterial pressure (β = 0.070, χ2 = 11.39; p = 0.0011), fasting glucose (β = 0.790, χ2 = 11.30; p = 0.0012), and use of antihypertensive medication (β = − 1.416, χ2 = 7.95; p = 0.0060). We did not observe correlation between PWV increase and lipid or renal parameters and lipid lowering medication. Conclusions. In elderly subjects without manifest cardiovascular disease, mechanical load and glucose concentration play a major role in the aortic stiffening. Use of antihypertensive treatment was associated with smaller PWV increase.
Introduction
Aortic pulse wave velocity (PWV) is considered a surrogate measurement of aortic stiffness. Aortic PWV is a powerful predictor of cardiovascular outcome in the general population (Citation1), in well-functioning elderly population (Citation2), as well as in patients with hypertension (Citation3), diabetes or end-stage renal disease (Citation4). Moreover, in elderly subjects aortic stiffness was strongly and independently associated with coronary atherosclerosis (Citation5).
The European population is getting older, therefore understanding which factors worsen arterial stiffness in elderly is important. There is currently little information regarding the determinants of aortic stiffening in elderly. The Reference Values Project showed that PWV increased both with increasing age and blood pressure levels (Citation6). However, in this study, the diabetics and patients treated for hypertension and dyslipidaemia were not included. Another study in postmenopausal women showed that aortic PWV was significantly associated with age, mean arterial pressure (MAP), pulse pressure, heart rate and pack-years of cigarette smoking (Citation7). Dart et al. observed in hypertensive subjects older than 65 years a relationships between arterial stiffness and age, gender, MAP, height and heart rate (Citation8). However, these studies were of cross-sectional design. Benetos et al. (Citation9) observed that middle-aged hypertensives had higher annual rate of progression in PWV than normotensives. Only treated hypertensives with well-controlled blood pressure levels at the time of both visits had a PWV progression similar to that of normotensives. In treated hypertensives, heart rate and creatinine were positively associated with an accelerated progression of arterial stiffening (Citation9).
The aim of present longitudinal study was to evaluate the determinants of PWV progression over a 9-year period in elderly subjects free from overt cardiovascular disease and not treated for arterial hypertension at baseline. Furthermore, we examined effect of blood pressure and lipid lowering medication on PWV progression. The subjects were recruited from attendees of the Continuing Adult Education programme of Faculty of Medicine in Pilsen.
Methods
Subjects and study design
The Ethics Committee of University Hospital in Pilsen approved the protocol and all participants gave their written informed consent. From September 1998 to November 1999, we recruited 160 subjects attending Continuing Adult Education programme of Faculty of Medicine, free from overt cardiovascular disease as defined by history of angina pectoris, documented myocardial infarction, stroke or use of antihypertensive medication. Eighteen subjects (11.2%) were deceased by the time of the follow-up visit. From February 2008 to July 2008, we re-examined 116 (72.5 %) subjects who agreed to participate in follow-up visit. Of those, we excluded 24 subjects because of insufficient quality of the arterial measurement and two subjects who had diagnosed an ischaemic heart disease after baseline examination.
Baseline as well as follow-up protocol included an administration of a standardized questionnaire to obtain information on each subject's medical history, smoking and drinking habits, and use of medications. At each visit, we measured blood pressure with an Omron 705CP device using an appropriately sized cuff. Blood pressure was determined by the average of two consecutive readings. We defined pulse pressure as the difference between systolic and diastolic pressure. MAP was diastolic pressure plus one third of pulse pressure. Furthermore, blood and urine samples were obtained for biochemical analyses. Height and weight were measured for all participants. Body mass index (BMI) was calculated as body weight (kg)/height2 (m2). Arterial hypertension was defined as systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg or use of antihypertensive treatment. Estimated glomerular filtration rate (eGFR) was calculated according to Modification of Diet in renal disease (MDRD) formula (Citation10).
Arterial measurement
To ensure steady state, the arterial measurements were obtained in a quiet examination room, after the subjects had rested for 15 min in the supine position and had refrained from smoking, heavy exercise, and drinking alcohol or caffeinated beverages for at least 2 h prior to the examination.
We recorded the arterial waveform for 8 s in the dominant arm by applanation tonometry. We used a high-fidelity SPC-301 micromanometer (Millar Instruments, Inc., Houston, TX) interfaced with a laptop computer running the SphygmoCor software (AtCor Medical Pty. Ltd., West Ryde, New South Wales, Australia). We discarded recordings when the systolic or diastolic variability of consecutive waveforms exceeded 5% or when the amplitude of the pulse wave signal was less than 80 mV. We calibrated the pulse wave by measuring blood pressure at the contra-lateral arm immediately before the recordings. We computed the aortic PWV from recordings of the arterial pressure wave at the carotid and femoral arteries (Citation11). The intra-observer and inter-observer variability was 0.07 ± 1.17 m/s and − 0.30 ± 1.25 m/s for aortic PWV (Citation11). We measured the distance between the site of the carotid recordings and the suprasternal notch and between the suprasternal notch and the site of the femoral recordings. We subtracted these two distances to obtain travel distance. Aortic PWV was calculated as the ratio of the travel distance in meters to the transit time in seconds. Heart rate used for present analysis was measured during PWV measurement and calculated as an arithmetic mean of the heart rate values obtained at the carotid and femoral artery.
Statistical methods
For database management and statistical analyses, we used the SAS software, version 9.2 (SAS Institute Inc., Cary, NC, USA). Data are presented as mean± SD or proportions. A Student t-paired test and a Fisher test were used to compare differences between the first and second visit.
The predictors of an increase in PWV were evaluated using univariate and multivariate regression analyses. As covariates, we considered observer, gender, age, MAP, heart rate, BMI, total or low-density cholesterol or triglycerides, fasting serum glucose, serum creatinine or eGFR, smoking and alcohol intake. The parameters found significant in univariate analysis then enter multivariate analysis. Furthermore, we analysed effect of use of antihypertensive or lipid lowering medication on PWV progression with age by adding these parameters into a regression model.
Results
Characteristics of participants
The median follow-up was 9.5 years (range 8.5–9.7 years). Baseline and follow-up characteristics of the study cohort are given in . The mean age at baseline was 66.9 years (range, 47–81); the majority (80%) of subjects were women. In comparison with baseline values, the body height, eGFR, serum cholesterol and triglycerides were lower, while BMI and serum creatinine were higher at follow-up. At follow-up, subjects more frequently used antihypertensive treatment (0 vs 52.2%; p < 0.0001) and lipid lowering medication (1.1% vs 36.7%; p < 0.0001). From 47 patients on antihypertensive treatment at follow-up, 66.0% used drugs interfering with renin–angiotensin system, 36.2% used beta-blockers, 25.5% diuretics, 23.4% calcium channel blockers and 17.0% other classes of blood pressure lowering medication. The proportion of diabetic patients was marginally higher (5.6 vs 15.6%; p = 0.051) at follow-up compared with baseline examination. With exception of pulse pressure (57.1 vs 59.7; p = 0.28), all haemodynamic measurements increased during the study period (P ≤ 0.023, ). For PWV, this increase amounted to approximately 10% ().
Table I. Basic characteristics of the study cohort.
Table II. Laboratory and haemodynamic characteristics of the study cohort.
In a further step, we divided subjects according to the use of antihypertensive medication. In 43 subjects who never took blood pressure lowering drugs, both systolic and blood pressure increased during the 9 years of follow-up (125/74 vs 141/82 mmHg; p < 0.0009), as well as PWV (8.7 vs 10.2 m/s; p = 0.0091, ). On the other hand, in subjects treated with antihypertensive medications, neither blood pressure (144/82 vs 145/84 mmHg; p < 0.28) nor PWV (10.0 vs 10.4 m/s; p = 0.47) changed. Heart rate was lower during follow-up examination compared with baseline in both groups, but the difference did not reach statistical significance (p ≥ 0.067, ).
Determinants of aortic stiffening
gives results of multiple linear analysis of increase in PWV during follow-up. While accounting for baseline PWV, MAP, heart rate, fasting glucose, total serum cholesterol, and use of antihypertensive and lipid lowering medication , a 1 − SD of baseline heart rate (8.5 beats/min), MAP (12.4 mmHg) and fasting glucose (0.93 mmol/l) were associated with PWV increase amounting to 0.63 m/s (95% CI 0.18–1.08; p = 0.0079), 0.86 m/s (95% CI 0.36–1.37; p = 0.0011) and 0.74 m/s (95% CI 0.31–1.17; p = 0.0012), respectively. Moreover, use of antihypertensive medication was associated with lower PWV increase (β = − 1.42 ± 0.50, p = 0.0060).
Table III. Predictors of increase in aortic stiffness.
Sensitivity analyses, after excluding those 14 patients who had diabetes mellitus at follow-up, confirmed these results. Serum glucose was associated with increased PWV (β = 0.99 ± 0.37, χ2 = 7.03; p = 0.0099). There was no significant association between serum glycated haemoglobin and PWV either in all subjects (β = 0.06 ± 0.43, χ2 = 0.02; p = 0.89) or after exclusion of the diabetic patients (β = − 0.54 ± 0.69, χ2 = 0.62; p = 0.43).
None of lipid parameters [total cholesterol, low-density lipoprotein (LDL)- or high-density lipoprotein (HDL)-cholesterol, and triglycerides] as well as use of lipid lowering medication were associated with a progression in aortic PWV (0.13 < p < 0.88). Separate analysis of subjects who did not take lipid lowering medication (n = 57) confirmed results from total study population. The predictors of PWV increase were as follows: heart rate (β = 0.11 ± 0.03, χ2 = 10.2; p = 0.0024), MAP (β = 0.06 ± 0.03, χ2 = 6.0; p = 0.018) and serum glucose (β = 0.78 ± 0.28, χ2 = 7.8; p = 0.0071).
Subjects using antihypertensive medication had higher increase of serum creatinine from baseline to follow-up (Δ7.8 μmol/l; p = 0.0092) than untreated subjects (Δ3.7 μmol/l; p = 0.16). However, in both subgroups as well as in whole study population, serum creatinine or eGFR were not significant predictors of arterial stiffness either in univariate (p ≥ 0.48) or multivariate (p ≥ 0.063) analyses.
Discussion
We investigated predictors of arterial stiffening in elderly subjects free from overt cardiovascular disease and not treated with antihypertensive drugs at baseline. To our knowledge, our current study is the first to examine determinants of aortic stiffness in the elderly in a longitudinal manner. In our settings, an increase of aortic PWV was determined by MAP, heart rate and fasting glucose and was modified by use of an antihypertensive medication started after baseline examination.
Our prospective observations are in agreement with known physiological concepts (Citation12). Mechanical load represented by MAP and heart rate represent both sustained and pulsatile strain on arterial wall. This leads to an elastin defragmentation, with parallel increase in collagen within the media. In our study, MAP and heart rate were main determinants of aortic stiffening. It seems that driving force behind baseline PWV and its increase during follow-up is represented by blood pressure and blood pressure changes. In our study, more than half of participants started using an antihypertensive medication during follow-up. In untreated subjects, both blood pressure and PWV increased during follow-up period. On the other hand, blood pressure of treated patients remained similar and PWV was not raised. Consequently, at the study follow-up, both groups had similar PWV for a given blood pressure. This is in agreement with French study in middle-aged population (Citation9). Indeed, treated hypertensives with well-controlled blood pressure levels at the time of both visits had a PWV progression similar to that of normotensives (Citation9). Moreover, in the Reference Values Project (Citation6), untreated sexagenarians with grade I hypertension had PWV increased by 1.8 m/s per decade. In our study, after initiation of antihypertensive treatment, PWV increase over a decade was four times smaller.
Several studies proposed advanced glycation end products (AGEs) to be associated with increased arterial stiffness (Citation13,Citation14). AGEs accumulate in arteries with age (Citation15). In addition to introducing the cross-linking of collagen in arterial walls, AGEs contribute to endothelial dysfunction by several mechanisms. In our population of highly selected elderly subjects free from manifest cardiovascular disease, the increase in aortic PWV was associated with serum glucose, and this association remained significant after adjustment for classical cardiovascular risk factors. Classifying subjects into tertiles according serum glucose (data not shown) showed significantly higher PWV progression in subjects in second (serum glucose range: 5.2–5.4 mmol/l) and third tertile (serum glucose ≥ 5.5 mmol/l; p = 0.0095 for trend across tertiles). However, we did not find any association between aortic PWV increment and glycated haemoglobin. This might be explained by narrow variance of glycated haemoglobin in non-diabetic subjects.
In cross-sectional analysis of 868 hypertensive subjects aged 65–84 years with a low prevalence of known symptomatic atheromatous disease, systemic arterial compliance and aortic distensibility were not related to the level of total or HDL-cholesterol or presence of diabetes mellitus (Citation8). In multiple regression analyses, Dart et al. (Citation8) found significant independent association between arterial stiffness parameters and MAP, gender, age, height, and heart rate. Blood glucose was not included in the regression models. In the Reference Values Project, dyslipidaemia and smoking did not significantly influence PWV (Citation6). Our study is in line with these observations. The absence of an association between increase in aortic stiffness and cholesterol in our study is perhaps not surprising, given that two major determinants of arterial stiffness, namely MAP and heart rate, might reduce the potential role for other known cardiovascular risk factors. Other explanation might be that the deleterious effect of serum cholesterol on arterial stiffness was prevented by statins. Indeed, several reports demonstrated improvement of arterial distensibility after the use of statins (Citation16,Citation17). In our study, statin use was significantly higher in the follow-up compared with baseline period (36.7 vs 1.1 %); however, it was not associated with smaller PWV increase.
The present study must be interpreted within the context of its limitations and strengths. First, the number of subjects in our analysis was smaller than in other studies (Citation7,Citation8). On the other hand, these studies were of cross-sectional design, in contrast to our prospective study. Second, we excluded 24 subjects from analysis because of insufficient arterial measurement at one of the sessions because our intention was to use only high-quality measurements. Therefore, we implemented strict quality criteria. Third, in our population women to men ratio was 5:1. Therefore, our results probably could not be extrapolated into males. Fourth, in our current study, as in all other studies, blood pressure and heart rate were the overriding determinants of arterial stiffness. Our sample size might have been insufficient to identify other arterial stiffness correlates. Fifth, we did not measure AGEs and therefore we can only speculate about their function in an observed increase of aortic stiffness.
In conclusion, in elderly subjects without manifest cardiovascular disease, mechanical load, as demonstrated by positive association with MAP and heart rate, and glycaemia are main factors leading to increase of aortic stiffness. We might speculate that maintenance of glycaemia in lower values might favourably modulate aortic stiffness. We therefore advocate for tight blood pressure and glycaemic control to prevent increase in aortic stiffness in the elderly. Moreover, our findings suggest that antihypertensive treatment started at age over 60 years might have a beneficial effect on aortic stiffness.
Acknowledgements
The authors acknowledge the expert technical assistance of Alena Maříková.
This study would not have been possible without the voluntary collaboration of the participants.
Declaration of interest: This research was supported by the Charles University Research Fund (project number P36).
References
- Willum-Hansen T, Staessen JA, Torp-Pedersen C, Rasmussen S, Thijs L, Ibsen H, et al. Prognostic value of aortic pulse wave velocity as index of arterial stiffness in the general population. Circulation. 2006;113:664–670.
- Sutton-Tyrrell K, Najjar SS, Boudreau RM, Venkitachalam L, Kupelian V, Simonsick EM, et al. Elevated aortic pulse wave velocity, a marker of arterial stiffness, predicts cardiovascular events in well-functioning older adults. Circulation. 2005;111:3384–3390.
- Boutouyrie P, Tropeano AI, Asmar R, Gautier I, Benetos A, Lacolley P, et al. Aortic stiffness is an independent predictor of primary coronary events in hypertensive patients: A longitudinal study. Hypertension. 2002;39:10–15.
- Blacher J, Safar ME, Guerin AP, Pannier B, Marchais SJ, London GM. Aortic pulse wave velocity index and mortality in end-stage renal disease. Kidney Int. 2003;63:1852–1860.
- van Popele NM, Mattace-Raso FU, Vliegenthart R, Grobbee DE, Asmar R, van der Kuip DA, et al. Aortic stiffness is associated with atherosclerosis of the coronary arteries in older adults: The Rotterdam Study. J Hypertens. 2006;24:2371–2376.
- Reference Values for Arterial Stiffness’ Collaboration. Determinants of pulse wave velocity in healthy people and in the presence of cardiovascular risk factors: “Establishing normal and reference values”. Eur Heart J. 2010;31:2338–2350.
- Lebrun CEI, van der Schouw YT, Bak AAA, de Jong FH, Pols HAP, Grobbee DE, et al. Arterial stiffness in postmenopausal women: Determinants of pulse wave velocity. J Hypertens. 2002;20:2165–2172.
- Dart AM, Gatzka CD, Cameron JD, Kingwell BA, Liang YL, Berry KL, et al. Large artery stiffness is not related to plasma cholesterol in older subjects with hypertension. Arterioscler Thromb Vasc Biol. 2004;24:962–968.
- Benetos A, Adamopoulos C, Bureau JM, Temmar M, Labat C, Bean K, et al. Determinants of accelerated progression of arterial stiffness in normotensive subjects and in treated hypertensive subjects over a 6-year period. Circulation. 2002;105:1202–1207.
- Levey A, Bosch J, Lewis J, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: A new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med. 1999;130:461–470.
- Wilkinson IB, Fuchs SA, Jansen IM, Spratt JC, Murray GD, Cockcroft JR, et al. Reproducibility of pulse wave velocity and augmentation index measured by pulse wave analysis. J Hypertens. 1998;16:2079–2084.
- Safar ME, Levy BI, Struijker-Boudier H. Current perspectives on arterial stiffness and pulse pressure in hypertension and cardiovascular diseases. Circulation. 2003;107:2864–2869.
- Semba RD, Najjar SS, Sun K, Lakatta EG, Ferrucci L. Serum carboxymethyl-lysine, an advanced glycation end product, is associated with increased aortic pulse wave velocity in adults. Am J Hypertens. 2009;22:74–79.
- McNulty M, Mahmud A, Feely J. Advanced glycation end-products and arterial stiffness in hypertension. Am J Hypertens. 2007;20:242–247.
- Schleicher ED, Wagner E, Nerlich AG. Increased accumulation of the glycoxidation product Nϵ-(carboxymethyl)lysine in human tissues in diabetes and aging. J Clin Invest. 1997;99:457–468.
- Orr JS, Dengo AL, Rivero JM, Davy KP. Arterial destiffening with atorvastatin in overweight and obese middle-aged and older adults. Hypertension. 2009;54:763–768.
- Yokoyama H, Kawasaki M, Ito Y, Minatoguchi S, Fujiwara H. Effects of fluvastatin on the carotid arterial media as assessed by integrated backscatter ultrasound compared with pulse-wave velocity. J Am Coll Cardiol. 2005;46:2031–2037.