801 Original Article Nomogram of the Relation of Brachial-Ankle Pulse Wave Velocity with Blood Pressure Akira YAMASHINA, Hirofumi TOMIYAMA, Tomio ARAI, Yutaka KOJI, Minoru YAMBE, Hiroaki MOTOBE, Zydem GLUNIZIA, Yoshio YAMAMOTO, and Saburoh HORI The present study was conducted to make a nomogram of the relation of brachial-ankle pulse wave velocity (bapwv) with blood pressure. In 11,375 subjects (age range: 30 to 79 years) in whom we measured bapwv, multivariate linear regression analysis demonstrated that age, systolic blood pressure, and gender were significant determinants of bapwv. Of them, we selected 5,302 subjects (2,630 men and 2,672 women, age 46 11 years) with either normal blood pressure or with high blood pressure without other atherosclerotic risk factors. Using data from these subjects, we tried to make an age-derived nomogram and a systolic blood pressure-derived nomogram for in both genders. The regression curve analysis demonstrated significant quadratic curves for both associations. The obtained nomograms were applied to 297 subjects with coronary heart disease. Their measured bapwv (1,569 344 cm/s) was significantly higher than the values calculated from the age-derived nomogram (1,502 241 cm/s) and the systolic blood pressure-derived nomogram (1,418 236 cm/s) ( p 0.01). Thus, a nomogram of the relation of bapwv with blood pressure was obtained. The comparison of the measured bapwv with the values calculated from the nomograms may be useful to not to underestimate the real risk for atherosclerotic cardiovascular diseases, including the severity of atherosclerotic vascular damage, reflected by measured bapwv in subjects with different blood pressure levels. (Hypertens Res 2003; 26: 801 806) Key Words: pulse wave velocity, blood pressure, age, nomogram, atherosclerosis Introduction Pulse wave velocity (PWV) reflects arterial stiffness, and it is a predictor of the prognosis of patients with cardiovascular diseases, and it serves, at least in part, as a marker of the severity of atherosclerotic vascular damage (1 3). Recently, brachial-ankle PWV (bapwv), a parameter more simple to obtain compared with the conventional carotid-femoral PWV, has been useful (4 6). bapwv can be obtained by only wrapping the four extremities with blood pressure cuffs, and it can be easily used to screen a large number of subjects (6). However, We demonstrated that blood pressure is a major determinant of PWV (6). In a subject with high blood pressure, it is difficult to estimate whether PWV reflects solely the level of blood pressure or the risk for cardiovascular diseases, including the severity of atherosclerotic vascular damage, in addition to blood pressure. Therefore, a nomogram of the relation of bapwv related with blood pressure may be useful to estimate more precisely the absolute risk for atherosclerotic cardiovascular disease as assessed based on the bapwv in a subject with high blood pressure. In the present cross-sectional study measured bapwv in two cohorts involving more than 10,000 subjects at the time of their annual health check up. Then, we made a nomogram of the relation of bapwv with blood pressure using the val- From the Second Department of Internal Medicine, Tokyo Medical University, Tokyo, Japan, Health Care Center, Kajima Corporation, Tokyo, Japan, and Preventive Medical Center, St. Luke s International Hospital, Tokyo, Japan. This study was partially supported by a grant in aid from Japanese Atherosclerosis Prevention Fund. Address for Reprints: Akira Yamashina, M.D., Second Department of Internal Medicine, Tokyo Medical University, 6 7 1, Nishi-shinjuku, Shinjukuku, Tokyo 160 0023, Japan. E-mail: akyam@tokyo-med.ac.jp Received June 9, 2003; Accepted in revised form July 9, 2003.
802 Hypertens Res Vol. 26, No. 10 (2003) ues from subjects with normal blood pressure and subjects with high blood pressure without other atherosclerotic risk factors. Furthermore, we compared the measured bapwv in patients with atherosclerotic cardiovascular disease with the value obtained using the nomogram to examine whether bapwv does reflect atherosclerotic vascular damage in subjects different blood pressure levels. Subjects Methods Table 1. Anthropometrics in All Participants (n 11,375) Gender Number of subjects 7,093 4,282 Age 48 11 48 12 SBP (mmhg) 127 150 119 180 DBP (mmhg) 81 11 73 11 MBP (mmhg) 98 13 91 14 bapwv (cm/s) 1,332 2620. 1,215 2610. BMI (kg/m 2 ) 24 30 21 30 BMI 25 (n) 2,095 442 TC (mg/dl) 198 310 205 350 TC 240 (n) 639 712 TG (mg/dl) 117 810 76 43 TG 150 (n) 1,325 243 FBS (mg/dl) 98 21 93 13 FBS 110 (n) 801 215 Smoke (n) 2,478 442 Treated hypertension (n) 361 60 Diabetes mellitus (n) 307 90 Dyslipidemia (n) 474 337 CHD (n) 81 37 Stroke (n) 20 5 SBP, systolic blood pressure; DBP, diastolic blood pressure; MBP, mean blood pressure; bapwv, brachial-ankle pulse wave velocity; BMI, body mass index; n, number of subjects. The present study was conducted in two cohorts. Some of the data obtained from these two cohorts have already been reported elsewhere (6, 7). The age of the subjects varied from 21 to 94, and to make the nomogram, data from subjects between 30 and 80 years of age were used, because the number of subjects under 30 and of those over 80 years of age was small. Both cohorts consisted of subjects having their annual health check up (including the measurement of bapwv) from October 2000 to the end of July 2002 at two institutions affiliated to Tokyo Medical University, where more than 4,000 subjects are screened annually. The medical history and symptoms of each subject were confirmed by the consulting doctors. In these cohorts, patients having diseases (except for atherosclerotic cardiovascular diseases) requiring medical treatment and subjects who had an abnormal ankle/brachial pressure index of less than 0.9 as determined by plethysmograph were excluded. Atherosclerotic cardiovascular diseases were described according to the criteria of the International Classification of Diseases (10th revision) regarding coronary heart disease and cerebrovascular disease. In addition, patients with a medical history of arteriosclerosis obliterans or aortic aneurysm were excluded, because these diseases are known to reduce PWV (8). Consecutive 11,375 subjects (age range: 30 to 79 years, 7,093 men and 4,282 women) were entered for this study. Informed consent was obtained from all the subjects. Then, we selected data from subjects without atherosclerotic risk factors except blood pressure level to make the nomogram. Among them, the group of subjects without atherosclerotic risk factors except blood pressure level was defined by the following criteria: fasting blood glucose 110 mg/dl, total cholesterol 240 mg/dl, triglycerides 150 mg/dl, body mass index 25 kg/m 2, and no history of smoking. To assess the applicability of the obtained nomogram, 297 patients with coronary heart disease (240 men and 57 women, age: 65 8 years old) were recruited, and the obtained nomogram was applied to them. All these patients were diagnosed as having coronary heart disease by coronary angiography performed at Tokyo Medical University Hospital, and their ankle/brachial pressure index was over 0.9 as determined by plethysmography. In subjects having their annual health check up, informed consent was obtained when they visited institutions for their check up, and, in patients with coronary heart disease, it was obtained when they admitted to Tokyo Medical University Hospital. The study protocol was approved by the ethical committee of Tokyo Medical University. Instruments Brachial-ankle PWV was measured using a plethysmograph (form PWV/ABI, Colin Co. Ltd., Komaki, Japan), which records PWV, blood pressure, electrocardiogram and heart sounds, simultaneously (4 6). The details of the measurement have been reported elsewhere (4). In all the studies, bapwv was obtained after at least 5 min of rest. Laboratory Measurements Plasma total cholesterol, triglycerides, blood sugar, and serum creatinine levels were measured enzymatically. All blood samples were obtained in a fasting state in the morning. Statistics Data are expressed as the mean SD. Statistical analysis was performed using the SPSS software package (SPSS, Chicago, USA). Differences between two groups were evaluated
Yamashina et al: Pulse Wave Velocity and Blood Pressure 803 Fig. 1. Association between age and brachial-ankle pulse wave velocity (bapwv) at each systolic blood pressure level in both genders. Table 2. Results of Multiple Regression Analysis of the Correlation of Brachial-Ankle Pulse Wave Velocity with Other Variables in All the Subjects (n 11,375) β t-value p-value Age 0.35 53.72 0.01 SBP 0.42 26.85 0.01 DBP ns MBP 0.16 7.49 0.01 BMI 0.09 14.43 0.01 TC ns TG 0.04 7.07 0.01 FBS 0.07 12.51 0.01 Smoking ns Gender 0.09 13.72 0.01 The abbreviations are the same as Table 1. by paired t-test. Multiple linear regression analysis was performed to evaluate the association between bapwv and other clinical variables. The regression curve analysis was performed to estimate the relationship between age and bapwv and between systolic blood pressure and bapwv. A p value of 0.05 was considered statistically significant. Results R 2 0.67 The clinical characteristics including medical history and conventional atherosclerotic risk factors in all participants (n 11,375) are summarized in Table 1. Systolic and diastolic blood pressures were used the values measured by form PWV/ABI in the right arm. Table 2 shows the results of multiple linear regression analysis including bapwv and other clinical variables. Age, systolic blood pressure, and gender were significant determinants of bapwv. To make the nomogram of the relation of bapwv with blood pressure, Table 3. Anthropometrics in All Participants (n 5,302) Gender Number of subjects 2,630 2,672 Age 47 11 45 11 SBP (mmhg) 125 150 116 160 DBP (mmhg) 79 10 72 10 MBP (mmhg) 96 12 89 13 bapwv (cm/s) 1,300 2390. 1,169 2370. BMI (kg/m 2 ) 22 20 20 20 TC (mg/dl) 191 240 194 260 TG (mg/dl) 95 53 66 32 FBS (mg/dl) 93 80 90 70 The abbreviations are the same as Table 1. subjects (n 5,302) without conventional atherosclerotic factors, except for blood pressure, were selected. Table 3 summarizes the anthropometrics of the 5,302 selected subjects (2,630 men and 2,672 women; age: 46 11 years; range: 30 to 79 years). The association between age and bapwv was assessed by the regression curve analysis every 10 mmhg of systolic blood pressure (from less than 120 mmhg to over 160 mmhg) in both genders. In all of the systolic blood pressure levels, the association between age and bapwv demonstrated a quadratic curve (Fig. 1 and Table 4). The association between systolic blood pressure and bapwv was assessed by regression curve analysis in subjects of both genders divided by age from 30 to 70 years of age. The association between age and bapwv demonstrated a quadratic curve (Fig. 2 and Table 5). The obtained nomograms were applied to 297 subjects with coronary heart disease (Table 6). In these patients as a whole, their measured bapwv was significantly higher than the values calculated from the age-derived and systolic blood
804 Hypertens Res Vol. 26, No. 10 (2003) Table 4. The Value in Quadratic Curves between Age and Brachial-Ankle Pulse Wave Velocity (bapwv) in Both Genders at Each Systolic Blood Pressure (SBP) Level SBP (mmhg) n a b c r 2 120 1,061 0.1478 8.6 1,238 0.17 120 129 730 0.1812 9.8 1,326 0.28 130 139 434 0.2785 17.3 1,543 0.39 140 149 216 0.3256 22.3 1,740 0.35 150 159 80 0.1581 3.3 1,306 0.39 160 109 0.4575 34.5 2,190 0.39 120 1,823 0.1356 4.8 1,014 0.33 120 129 415 0.1537 6.4 1,181 0.41 130 139 217 0.3026 18.6 1,500 0.46 140 149 103 0.1769 6.9 1,336 0.33 150 159 57 0.4733 34.8 2,001 0.55 160 57 0.6580 60.9 2,984 0.43 Equation of a quadratic curve: bapwv a age 2 b age c. Fig. 2. Association between systolic blood pressure and brachial-ankle pulse wave (bapwv) velocity in subjects of both genders divided by age. pressure-derived nomograms. This result was confirmed also in both male and female subjects and in elderly (65 years or older) subjects with coronary heart disease. However, in subjects with a systolic blood pressure of 140 mmhg or more (n 58), their measured bapwv was higher than the value calculated from the systolic blood pressure-derived nomogram, but not significantly higher than that calculated from the age-derived nomogram. Discussion Blood pressure is a determinant of PWV (6). Franklin et al. demonstrated that, while systolic blood pressure and pulse pressure increase according to age, the increase of diastolic blood pressure and that of mean blood pressure are attenuated in people over 60 years of age (9). High systolic blood pressure and high pulse pressure are known as risk factors for cardiovascular diseases (9, 10). In our previous report, we demonstrated that age, gender and systolic blood pressure were major determinants of bapwv (6). Therefore, we tried to make an age-derived nomogram according to systolic blood pressure and a systolic blood pressure-derived nomogram according to age for subjects of both genders. Previously, we demonstrated that the association between age and bapwv is a quadratic curve at blood pressure levels within the normal range (blood pressure less than 140/90 mmhg) (6). The present study confirmed the same association between age and bapwv existed at blood pressure levels above the upper limit of normal (systolic blood pressure over 140 mmhg). Blacher et al. reported that the carotid femoral PWV was higher in hypertensive patients with atherosclerotic cardiovascular disease than in those without ath-
Yamashina et al: Pulse Wave Velocity and Blood Pressure 805 Table 5. The Value in Quadratic Curves between Systolic Blood Pressure and Brachial-Ankle Pulse Wave Velocity (bapwv) in Subjects of Both Genders Divided by Age Age (years old) n a b c r 2 30 39 883 0.0353 15.55 178 0.30 40 49 633 0.0106 11.05 42 0.43 50 59 757 0.0274 17.25 393 0.46 60 69 271 0.0666 27.67 955 0.37 70 79 84 0.0277 21.22 624 0.59 30 39 1,027 0.0234 2.16 514 0.36 40 49 707 0.0026 8.67 165 0.52 50 59 596 0.0328 17.09 328 0.55 60 69 251 0.0195 15.24 169 0.53 70 79 90 0.0618 28.92 1,060 0.49 Equation of a quadratic curve: bapwv a systolic blood pressure 2 b systolic blood pressure c. Table 6. Measured Brachial-Ankle Pulse Wave Velocity (bapwv) and the Values Calculated from the Nomograms in Patients with Coronary Heart Disease All (n 297) (n 240) (n 57) SBP 140 (n 58) Age 65 (n 166) Measured bapwv (cm/s) 1,569 344 1,548 342 1,656 340 1,851 283 1,715 343 Age-dr bapwv (cm/s) 1,502 241 1,582 229 1,586 275 1,833 189 1,623 236 SBP-dr bapwv (cm/s) 1,418 236 1,406 239 1,466 220 1,640 279 1,514 241 All, all patients with coronary heart disease; SBP, systolic blood pressure; Age-dr, age derived; SBP-dr, SBP derived. p 0.05, p 0.01 vs. measured bapwv. erosclerotic cardiovascular disease (11). To confirm the usefulness of the present nomograms regarding the assessment of the severity of atherosclerotic vascular damage in subjects with different levels of blood pressure, we applied these nomograms to patients diagnosed as having coronary heart disease by coronary angiography. In all the patients with coronary heart disease in this study, the measured bapwv was higher than the values calculated from both nomograms. Therefore, bapwv might reflect atherosclerotic vascular damage regardless of blood pressure levels. In obtaining the age-derived nomogram, the systolic blood pressure of subjects whose values were the bases for the equation for subjects with a systolic blood pressure of 160 mmhg or more, varied widely from 160 to 220 mmhg, and thus this equation may have some limitations regarding accuracy. It was therefore considered that this wide range of systolic blood pressure might have affected the accuracy of the estimated value obtained by applying the age-derived nomogram to subjects with a systolic blood pressure of 140 mmhg or more, resulting in no significant difference between the measured bapwv and the value calculated from the age-derived nomogram in subjects with coronary heart disease whose systolic blood pressure was 140 mmhg or more. Asmar et al. reported a simple linear equation to obtain the carotid-femoral PWV derived from systolic blood pressure and age (12). However, the association of bapwv with either systolic blood pressure or age was a quadratic equation rather than a linear equation. Therefore, we believe that the accuracy of a simple linear equation to obtain bapwv derived from systolic blood pressure and age may not be sufficiently high to apply it to clinical practice. On the other hand, researchers have attempted to correct PWV for diastolic blood pressure, and thereby be able to evaluate the severity of angiopathy separately from blood pressure (13). This method, however, does not allow a definite evaluation for it does not consider the risks inherent to a high blood pressure itself. A high blood pressure is a major risk for atherosclerosis (9, 10), and the actual value of bapwv is an important parameter to be considered in the management of atheroslclerotic cardiovascular diseases. In patients with high blood pressure, it is difficult to discriminate whether the measured bapwv reflects only an elevated blood pressure or further risk for atherosclerotic cardiovascular diseases, including the severity of atherosclerotic vascular damage, and it might be possible to underestimate the real risk for atherosclerotic cardiovascular diseases in some of such patients. The use of the present systolic blood pressure-derived nomogram may prevent this pressure-related underestimation. In patients with high blood pressure, we believe that the sys-
806 Hypertens Res Vol. 26, No. 10 (2003) tolic blood pressure-derived nomogram can be used to decide the management strategy that is, whether to provide the intensive treatment of high blood pressure or to further evaluate the patients for the presence of atherosclerotic vascular damage through examinations, such as carotid artery ultrasonography, flow-mediated vasodilatation test of the brachial artery and brain magnetic resonance imaging, among others (14). To assess the association between blood pressure and PWV, the evaluation of the relationship between changes in PWV and changes in blood pressure caused by vasoactive drugs may be another possible approach. However, such drugs influence PWV by changing the arterial tonus independent of the change in blood pressure (15). In addition, it is difficult to increase the number of subjects of the study when using this method. Therefore, such method does not seem to be suitable to make a nomogram. On the other hand, the method used in the present study to make the nomogram of PWV based on data from subjects with a normal blood pressure and subjects with a high blood pressure without other atherosclerotic risk factors is easy to increase the number of subjects for the study. However, even a high blood pressure in the absence of other atherosclerotic risk factors may affect the vascular structure and function. Such possibilities should be considered as a limitation for the use of the present nomograms in clinical practice. In conclusion, we made the nomograms of the relation of bapwv with blood pressure level. The comparison of the measured bapwv with the values calculated from the nomograms may be useful not to underestimate the real risk for atherosclerotic cardiovascular diseases, including the severity of atherosclerotic vascular damage, reflected by the measured bapwv in subjects with different blood pressure levels. Especially, the systolic blood pressure-derived nomogram may be applicable for subjects with a high blood pressure. References 1. Laurent S, Boutouyrie P, Asmar R, et al: Aortic stiffness is an independent predictor of all-cause and cardiovascular mortality in hypertensive patients. Hypertension 2001; 37: 1236 1241. 2. Cruickshank K, Riste L, Anderson SG, Wright JS, Dunn G, Gosling RG: Aortic pulse-wave velocity and its relationship to mortality in diabetes and glucose intolerance: an integrated index of vascular function? Circulation 2002; 106: 2085 2090. 3. van Popele NM, Grobbee DE, Bots ML, et al: Association between arterial stiffness and atherosclerosis: the Rotterdam Study. Stroke 2001; 32: 454 460. 4. Yamashina A, Tomiyama H, Takeda K, et al: Validity, reproducibility, and clinical significance of noninvasive brachial-ankle pulse wave velocity measurement. Hypertens Res 2002; 25: 359 364. 5. Suzuki E, Kashiwagi A, Nishio Y, et al: Increased arterial wall stiffness limits flow volume in the lower extremities in type 2 diabetic patients. Diabetes Care 2001; 24: 2107 2114. 6. Tomiyama H, Yamashina A, Arai T, et al: Influences of age and gender on results of noninvasive brachial-ankle pulse wave velocity measurement a survey of 12517 subjects. Atherosclerosis 2003; 166: 303 309. 7. Yamashina A, Tomiyama H, Arai T, et al: Brachial-ankle pulse wave velocity as a marker of atherosclerotic cardiovascular risk. Hypertens Res 2003; 26: 615 622. 8. Asmar R: Arterial Stiffness and Pulse Wave Velocity. Amsterdam, Elsevier, 1999, pp 117 121. 9. Franklin SS, Gustin W 4th, Wong ND, et al: Hemodynamic patterns of age-related changes in blood pressure: the Framingham Heart Study. Circulation 1997; 96: 308 315. 10. Franklin SS, Khan SA, Wong ND, Larson MG, Levy D: Is pulse pressure useful in predicting risk for coronary heart Disease?: the Framingham heart study. Circulation 1999; 100: 354 360. 11. Blacher J, Asmar R, Djane S, London GM, Safar ME: Aortic pulse wave velocity as a marker of cardiovascular risk in hypertensive patients. Hypertension 1999; 33: 1111 1117. 12. Asmar R, Benetos A, Topouchian J, et al: Assessment of arterial distensibility by automatic pulse wave velocity measurement: validation and clinical application studies. Hypertension 1995; 26: 485 490. 13. Hasegawa M: Pulse wave velocity in human. Jikei Med J 1970; 85: 742 760. 14. Pearson TA: New tools for coronary risk assessment: what are their advantages and limitations? Circulation 2002; 105: 886 892. 15. Ortiz PA, Garvin JL: Intrarenal transport and vasoactive substances in hypertension. Hypertension 2001; 38: 621 624.