- Email: [email protected]
Electrocardiographic Left Ventricular Hypertrophy and Arterial Stiffness: The Ohasama Study
AJH
ORIGINAL CONTRIBUTIONS
2006; 19:1199 –1205
Blood Vessels
Electrocardiographic Left Ventricular Hypertrophy and Arterial Stiffness: The Ohasama Study Daisuke Watabe, Junichiro Hashimoto, Rieko Hatanaka, Tomohiro Hanazawa, Hiromi Ohba, Takayoshi Ohkubo, Masahiro Kikuya, Kazuhito Totsune, and Yutaka Imai Background: Whether arterial stiffness per se contributes to left ventricular hypertrophy (LVH) independently of blood pressure (BP) remains unknown. We examined the relationship between pulse wave velocity (PWV) and LVH in a large population. Methods: The PWV was measured between the brachial and ankle regions (baPWV) of 798 individuals. We diagnosed LVH using electrocardiographic criteria: Cornell voltage-duration product ⬎2440 mm ⫻ msec or Sokolow-Lyon voltage ⬎38 mm. The participants were initially separated into those with and without LVH [LVH(⫹) and LVH(⫺) groups, respectively]. To determine theoretical baPWV, we first constructed a nomogram for the LVH(⫺) group, calculated the PWV index (measured baPWV ⫺ theoretical baPWV) for each individual and then compared the two groups. We also examined the factors associated with LVH(⫹) using multivariate analyses.
Mean arterial pressure (MAP) (mm Hg) ⫹ 0.05 ⫻ Heart rate (beats/min) ⫺ 11.74 (R2 ⫽ 0.56). The PWV index was greater in the LVH(⫹) than in the LVH(⫺) group (P ⫽ .025). The baPWV was independently related to LVH(⫹) along with MAP, medication for hypertension, and for diabetes; a 1 SD (4.3 m/sec) increase in baPWV was associated with a 26% increase in the risk of LVH(⫹) (P ⫽ .022). When LVH(⫹) risk factors were defined as hypertension, diabetes, and high baPWV (ⱖ14.6 m/sec), the prevalence of LVH(⫹) linearly increased with the number of concomitant LVH(⫹) risk factors (P ⬍ .001). Conclusions: Arterial stiffness is independently related to electrocardiographically determined LVH in the general population. Am J Hypertens 2006;19:1199 –1205 © 2006 American Journal of Hypertension, Ltd.
Results: Linear regression analysis revealed that the theoretical baPWV (m/sec) ⫽ 0.20 ⫻ age (years) ⫹ 0.13 ⫻
Key Words: Arterial stiffness, pulse wave velocity, left ventricular hypertrophy, electrocardiography.
eft ventricular hypertrophy (LVH) is an independent predictor of cardiovascular risk1,2 and the most common cardiac abnormality associated with hypertension. The development of LVH is generally ascribed to a chronic increase in cardiac afterload that arises due to elevated blood pressure (BP).3 On the other hand, arterial stiffness might also play an important role in the development of LVH. Several markers of vascular functional properties, including aortic impedance,4 pulse wave velocity (PWV),5 and carotid distensibility6 are significantly correlated with LV mass in hypertensive patients. These
L
studies might mean that considerable arterial stiffening in response to hypertension increases end-systolic stress on the left ventricle, resulting in the development of LVH. However, whether arterial stiffness is similarly related to LVH in the general population as well as in hypertensive patients remains unknown. Moreover, whether arterial stiffening per se contributes to LVH independently of BP remains to be determined. Understanding the respective roles of high BP and arterial stiffness as determinants of LVH will uncover the mechanisms underlying the development of LVH.
Received November 6, 2005. First decision April 27, 2006. Accepted May 4, 2006. From the Department of Clinical Pharmacology and Therapeutics (DW, RH, TH, HO, KT, YI) and Department of Planning for Drug Development and Clinical Evaluation (JH, TO, MK), Tohoku University Graduate School of Pharmaceutical Science and Medicine, and Tohoku University 21th Century COE Program “Comprehensive Research and
Education Center for Planning of Drug Development and Clinical Evaluation” (JH, TO, KT, YI), Sendai, Japan. Address correspondence and reprint requests to Dr. Junichiro Hashimoto, Department of Planning for Drug Development and Clinical Evaluation, Tohoku University Graduate School of Pharmaceutical Science and Medicine, 6-3 Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan; e-mail: [email protected]
© 2006 by the American Journal of Hypertension, Ltd. Published by Elsevier Inc.
0895-7061/06/$32.00 doi:10.1016/j.amjhyper.2006.05.001
1200
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
Quantitative information about the elastic properties of large arteries is usually obtained by measuring PWV,5,7,8 which can predict cardiovascular risk.9,10 In addition to conventional carotid–femoral PWV (cfPWV) measurements, the brachial–ankle PWV (baPWV) can also evaluate arterial stiffness.11 Other studies have shown that this technique can provide useful information on arterial stiffness and that it is appropriate for screening a general population.11,12 The present cross-sectional study evaluates the relationship between baPWV and electrocardiographically determined LVH in the general population.
Methods Study Participants The present study was based on a health survey performed in the Japanese city of Ohasama, the geographic and demographic characteristics of which have been described elsewhere.12,13 Of 1739 individuals who participated in the health checkup, 808 completed baPWV measurements, an electrocardiogram, and both anthropometric and biochemical examinations. The baPWV measurement includes information about peripheral components of the arterial tree, and thus might be affected by arterial occlusive disease. From this viewpoint, individuals with suspected arterial occlusive disease, such as those having an ankle– brachial index (ABI) of ⬍0.9, were excluded (n ⫽ 10), therefore the study included 798 individuals. The Department of Health of the Ohasama Town Government approved the present study and informed consent was obtained from all participants. During the checkup, brachial systolic BP and diastolic BP were measured twice while seated after a 2-min rest, using an automated cuff oscillometric device (Omron HEM907; Omron Life Science, Kyoto, Japan). The average of the two readings was defined as the BP value. The pulse pressure (PP) and mean arterial pressure (MAP) were calculated according to the formula: PP ⫽ Systolic BP ⫺ Diastolic BP, and MAP ⫽ Diastolic BP ⫹ PP/3, respectively. Biochemical data were obtained from venous blood samples collected on the same days as the PWV measurements. Information concerning lifestyle, habits, and therapeutic status was collected using a questionnaire. Although a possible determinant of arterial stiffness,8 no data about the duration of hypertension were available to the present study. Hypertension was defined by the following criteria: systolic BP ⱖ140 mm Hg or diastolic BP ⱖ90 mm Hg or taking medication for hypertension. Diabetes was defined as fasting blood glucose ⱖ126 mg/dL, postprandial glucose ⱖ200 mg/dL, hemoglobin A1c (HbA1c) ⱖ6.5%, or taking medication for diabetes. Electrocardiography We electrocardiographically diagnosed LVH from a 12-lead electrocardiogram (ECG) in all participants.
AJH–December 2006 –VOL. 19, NO. 12
According to the criteria of the Losartan Intervention For Endpoint Reduction in Hypertension (LIFE) study,14 we defined LVH as follows: Cornell voltage– duration product [(RaVL ⫹ SV3) ⫻ QRS duration, in men; (RaVL ⫹ SV3 ⫹ 6 mm) ⫻ QRS duration, in women] ⬎2440 mm ⫻ msec,15–17 or Sokolow-Lyon voltage (SV1 ⫹ RV5/6) ⬎38 mm.18 The R waves in leads aVL, V5, and V6 and S waves in leads V1 and V3 were measured to the nearest 0.5 mm (0.05 mV). The rationale for using the criteria in the LIFE study has been described.19 The specificity of both of the Cornell and Sokolow-Lyon criteria (⬎35 mm) is more than 95% in normal adults but the sensitivity is only 30% to 40%.20 However, combining the two criteria (Cornell voltage– duration product ⬎2440 mm ⫻ msec or Sokolow-Lyon voltage ⬎38 mm) increased the sensitivity by more than 10% without a loss of specificity.15–18,21 Furthermore, the ECG criteria used in the LIFE study identified hypertensive patients with a ⬎70% prevalence of anatomic LVH by echocardiography.22 Measurement of PWV The baPWV was measured in supine individuals after at least 5 min of rest using an automatic device (form PWV/ ABI; Colin Co., Ltd., Komaki, Japan), as described.11,12 Briefly, pressure waveforms of the brachial and tibial arteries were simultaneously recorded by placing occlusion cuffs connected to a plethysmographic sensor around both the brachia and ankles. The time delay (T) of the two waveforms was measured between the feet. The path lengths from the suprasternal notch to the brachium (Lb) and from the suprasternal notch to the ankle (La) were automatically calculated according to individual height. The baPWV was calculated using the following equation: baPWV ⫽ 共La ⫺ Lb兲/T 共m/sec兲. Values of baPWV were measured for an average of 10 sec and we used right brachial-to-right ankle baPWV. The validity and reproducibility of this method have been reported; the intraobserver repeatability coefficient is 0.87 and the interobserver repeatability coefficient is 0.98.11 The rationale for the use of baPWV instead of cfPWV was based on the finding that this parameter closely correlates with aortic PWV determined by an invasive method.11 Calculation of PWV Index We calculated the PWV index for each individual according to Blacher et al.23 Briefly, the PWV index was obtained by subtracting the theoretical baPWV (t-baPWV) from the measured baPWV. The PWV index represents inherent arterial stiffness and was adjusted for the influence of some relevant factors. To determine t-baPWV, we constructed a nomogram of baPWV for the subjects without electrocardiographic LVH [LVH(⫺)]. We constructed a multivariate linear regression equation to estimate tbaPWV using age, BP, heart rate (HR), and gender as explanatory variables, because these are generally considered major influences on PWV.12,23,24
AJH–December 2006 –VOL. 19, NO. 12
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
1201
Table 1. Characteristics of 798 participants with and without LVH LVH
Age (y) Gender (male/female) SBP (mm Hg) DBP (mm Hg) MAP (mm Hg) PP (mm Hg) HR (beats/min) baPWV (m/sec) ABI BMI (kg/m2) Total cholesterol (mg/dL) HDL cholesterol (mg/dL) Triglycerides (mg/dL) Serum creatinine (mg/dL) HbA1c (%) Medication for hypertension (%) Medication for diabetes (%) Medication for hyperlipidemia (%) Smoking (%) Alcohol consumption (%)
(ⴙ) n ⴝ 168
(ⴚ) n ⴝ 630
P
65.1 ⫾ 10.8 62/106 143.5 ⫾ 20.0 78.0 ⫾ 13.1 99.8 ⫾ 14.2 65.5 ⫾ 14.6 73.7 ⫾ 13.4 18.6 ⫾ 5.3 1.11 ⫾ 0.07 24.1 ⫾ 3.2 203.3 ⫾ 29.5 62.5 ⫾ 15.6 142.5 ⫾ 91.7 0.67 ⫾ 0.14 5.14 ⫾ 0.76 31.5 8.9 8.3 14.3 38.7
61.9 ⫾ 10.7 205/425 134.3 ⫾ 18.1 74.6 ⫾ 10.6 94.5 ⫾ 11.8 59.7 ⫾ 14.2 73.4 ⫾ 11.7 16.6 ⫾ 3.9 1.10 ⫾ 0.08 24.2 ⫾ 3.1 203.9 ⫾ 33.3 60.7 ⫾ 16.0 135.9 ⫾ 73.8 0.66 ⫾ 0.14 5.04 ⫾ 0.70 17.6 3.3 6.5 12.7 34.6
.001 .287 ⬍.001 .002 ⬍.001 ⬍.001 .814 ⬍.001 .191 .740 .826 .193 .327 .574 .106 ⬍.001 .002 .407 .587 .325
Abbreviations as in Table 1.
In calculating the PWV index, the t-baPWV was regarded as the assumed value that represents a population and subtracted from the measured baPWV value of each subject without and with electrocardiographic LVH [LVH(⫺) and LVH(⫹), respectively]. Statistical Analysis The participants were divided into LVH(⫹) and LVH(⫺) groups, and their characteristics were compared using the Student t test for continuous variables and the 2 test for categorical variables. Second, we compared the PWV index between LVH(⫹) and LVH(⫺) groups using Student t test to evaluate whether inherent arterial stiffness is enhanced in LVH(⫹) individuals. Third, multivariate stepwise logistic regression analysis clarified the factors related to LVH(⫹). The adjusted odds ratio (OR) and 95% confidence intervals (CI) of the risk of LVH(⫹) were determined using possibly independent variables including age, gender, MAP, HR, baPWV, ABI, body mass index (BMI), total cholesterol, HDL-cholesterol, triglycerides, serum creatinine, HbA1c, medication for hypertension, diabetes and hyperlipidemia, smoking status, and alcohol consumption. To evaluate whether baPWV would be a better independent predictor of LVH(⫹) than brachial PP, multivariate stepwise logistic regression analysis was applied using a model that included brachial PP instead of baPWV and using a model that included both baPWV and brachial PP instead of MAP. Furthermore, to evaluate the relationship between PWV index and LVH(⫹), a similar analysis used a model that included the PWV index. In this
regard, age, MAP, and HR were eliminated from this model, because the PWV index was adjusted for these confounding factors. Finally, we evaluated the prevalence of LVH(⫹) in relation to concomitant LVH(⫹) risk factors. The participants were classified into four groups according to the number of concomitant LVH(⫹) risk factors to compare the prevalence of LVH(⫹). We established an optimal cutoff value of baPWV using receiver operating characteristic (ROC) analysis to discriminate individuals in terms of LVH(⫹). All statistical analyses were performed using SPSS software version 11.0 (SPSS Inc., Chicago IL). Statistical significance was accepted at a value of P ⬍ .05.
Results Characteristics of Study Population The study population consisted of 267 men and 531 women with a mean age of 62.6 ⫾ 10.8 years (range, 34 to 88 years). Of these, 164 (20.6%) were taking medication for hypertension. Table 1 lists the characteristics of the LVH(⫹) and LVH(⫺) individuals. Relationship Between Arterial Stiffness and LVH(ⴙ) The baPWV in the LVH(⫺) group was significantly and independently correlated with age, MAP, and HR (P ⬍ .001 for all, multiple regression R2 ⫽ 0.56), but not with gender. Thus, the t-baPWV was calculated from the following
1202
AJH–December 2006 –VOL. 19, NO. 12
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
When using a model that included the PWV index, this value was also an independent determinant of LVH(⫹), together with medication for hypertension and for diabetes (data not shown). Prevalence of LVH(ⴙ) and Multiple Risk Factors
FIG. 1. Comparison of pulse wave velocity (PWV) indices between LVH(⫹) (n ⫽ 168) and LVH(⫺) (n ⫽ 630) groups. Values are presented as means ⫾ SE.
equation: t-baPWV (m/sec) ⫽ 0.20 ⫻ Age (years) ⫹ 0.13 ⫻ MAP (mm Hg) ⫹ 0.05 ⫻ HR (beats/min) ⫺ 11.74. Fig. 1 shows that the PWV index (baPWV ⫺ t-baPWV) was significantly greater in the LVH(⫹) than in the LVH (⫺) group. Multivariate logistic regression analysis revealed that the factors independently associated with LVH(⫹) were MAP, baPWV, as well as medication for hypertension and for diabetes (Table 2). An increase in baPWV of 1SD (4.3 m/sec) was associated with a 26% increase in the risk of LVH(⫹). In contrast, when brachial PP replaced baPWV, brachial PP did not associate with LVH(⫹) (Table 2). When both baPWV and brachial PP were included in the model but MAP was not, they were independent predictors of LVH(⫹) (Table 2).
Based on the results of the logistic regression analysis (Table 2), we defined the putative risk factors for LVH(⫹) as hypertension, diabetes, and high baPWV. According to the ROC analysis, the optimal cutoff value of baPWV to detect LVH(⫹) was 14.6 m/sec, with 82.7% of sensitivity and 34.1% specificity (data not shown). Therefore, individuals with baPWV ⱖ14.6 m/sec were regarded as having a high baPWV. Fig. 2 shows that when classified into four groups by the number of concomitant LVH(⫹) risk factors, the prevalence of LVH(⫹) increased linearly with increasing numbers of concomitant risk factors.
Discussion The key finding of the present study is that arterial stiffness measured through baPWV was significantly associated in the general population with electrocardiographically determined LVH [LVH(⫹)] independently of peripheral brachial BP. As with previous studies using different methods,4 – 6 this result supports the notion that arterial distensibility plays an important role in the development
Table 2. Determinants of LVH(⫹) based on multiple stepwise logistic regression analysis Adjusted odds ratio (95% CI) Model including baPWV* MAP (mm Hg) (per 12.5 mm Hg)§ baPWV (m/sec) (per 4.3 m/sec)§ Medication for hypertension Medication for diabetes Model including PP† Age (y) (per 10.8 y)§ MAP (mm Hg) (per 12.5 mm Hg)§ Brachial PP (mm Hg) (per 14.4 mm Hg)§ Medication for hypertension Medication for diabetes Model including both baPWV and brachial PP instead of MAP‡ Brachial PP (mm Hg) (per 14.4 mm Hg)§ baPWV (m/sec) (per 4.3 m/sec)§ Medication for hypertension Medication for diabetes
P
1.36 1.26 1.77 2.64
(1.11–1.67) (1.03–1.53) (1.19–2.64) (1.28–5.44)
.003 .022 .005 .008
1.26 1.53 1.15 1.69 2.64
(1.04–1.53) (1.28–1.82) (0.94–1.41) (1.13–2.54) (1.28–5.41)
.021 ⬍.001 .183 .012 .008
1.26 1.31 1.73 2.40
(1.02–1.50) (1.08–1.59) (1.16–2.59) (1.17–4.90)
.031 .006 .007 .017
ABI ⫽ ankle– brachial pressure index; baPWV ⫽ brachial–ankle pulse wave velocity; BMI ⫽ body mass index; CI ⫽ confidence interval; DBP ⫽ diastolic blood pressure; HbA1c ⫽ hemoglobin A1c; HDL ⫽ high-density lipoprotein; HR ⫽ heart rate; LVH ⫽ left ventricular hypertrophy; MAP ⫽ mean arterial pressure; PP ⫽ pulse pressure; SBP ⫽ systolic blood pressure. * Factors excluded from the model: age, gender, HR, BMI, ABI, total cholesterol, HDL ⫽ cholesterol, triglycerides, serum creatinine level, HbA1c, medication for hyperlipidemia, smoking, and alcohol consumption. † Factors excluded from the model: gender, HR, BMI, ABI, total cholesterol, HDL ⫽ cholesterol, triglycerides, serum creatinine level, HbA1c, medication for hyperlipidemia, smoking, and alcohol consumption. ‡ Factors excluded from the model: age, gender, HR, BMI, ABI, total cholesterol, HDL cholesterol, triglycerides, serum creatinine level, HbA1c, medication for hyperlipidemia, smoking, and alcohol consumption. § 1 SD change.
AJH–December 2006 –VOL. 19, NO. 12
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
FIG. 2. Prevalence of LVH(⫹) among four groups according to number of concomitant risk factors. Risk factors for LVH(⫹) included hypertension, diabetes, and high brachial–ankle pulse wave velocity (baPWV) (baPWV ⱖ14.6 m/sec).
of LVH. To our knowledge, however, this study is the first to demonstrate that PWV and LVH are independently related in the general population. Few previous studies have demonstrated a direct relationship between PWV and LVH. Bouthier et al5 have shown that in hypertensive patients the LV mass-to-volume ratio is positively correlated with cfPWV, but the relationship between the two parameters was not independent of BP. In contrast, our findings indicate that arterial stiffening as assessed by baPWV is directly involved in the pathogenesis of LVH. The association between increased PWV and LVH(⫹) identified in the present study can be explained by the following mechanism. Stiffening of the arteries increases the PWV and might be responsible for earlier return of the reflected wave from the periphery toward the ascending aorta. This means that a reflected wave would return during the systolic, rather than the diastolic phase, thus augmenting the systolic part of the incident pressure wave and further contributing to the increase in systolic BP and PP in the ascending aorta.24,25 Therefore, increased arterial stiffness might contribute to LVH through an increase in central PP, independently of MAP. In fact, our recent study showed that late systolic pressure augmentation is associated with LVH in hypertensive patients.26 Other reports have shown that in normotensive patients, the shapes of central pressure waveforms are significantly related to LVH,27,28 which also supports our notion. To interpret our findings, mechanical loads imposed on the heart should be considered to comprise steady and pulsatile components.24 Increased MAP, which reflects steady pressure stress imposed on the heart, is the key determinant of LVH(⫹), whereas increased arterial stiffness as a contributory factor to the pulsatile pressure stress, also appears to be an important determinant of LVH(⫹). In addition, baPWV and brachial PP were both independently associated with LVH(⫹) (Table 1). This result
1203
indicates that baPWV did not substantially overlap with brachial PP, whereas brachial PP as well as PWV is used in clinical practice as an approximate measure of arterial stiffness.29 Therefore, it seems that PWV is a key predictor of LVH(⫹) in that PWV includes different information than brachial PP. Here, we used the concept of the PWV index according to the study of Blacher et al.23 To assess arterial wall distensibility using PWV measurement, whether changes in arterial stiffness are due to inherent changes in the arterial wall are difficult to reliably determine, because arterial stiffness passively responds to physiologic factors. However, calculation of the PWV index could help resolve this issue. The present study showed that baPWV correlates with age, BP, and HR. Using the PWV nomogram obtained from the age, MAP, and HR of the LVH(⫺) group, calculation of the PWV index should generate a value that is independent of these parameters. Furthermore, the PWV index was significantly greater in the LVH(⫹) group than in the LVH(⫺) group. This suggests that inherent arterial stiffness is enhanced in individuals with LVH compared with those without LVH, independently of age, BP, and HR. Moreover, the multivariate analysis showed that the PWV index was an independent determinant of LVH(⫹). These findings might emphasize a deleterious effect of inherent arterial stiffness on the pathogenesis of LVH, independently of age, BP, and HR. The present study also demonstrated that prevalence of LVH(⫹) increased linearly with increasing the number of concomitant LVH(⫹) risk factors, including hypertension, diabetes, and high baPWV. This result does not necessarily support the fact that the physiologic composite effect is theoretically linear, because the composition of the set of even the same number of risk factors differed in individual patients. However, the Strong Heart Study30 associated hypertension and diabetes with independent and additive increases in the LVH prevalence. The present study supports the notion that a high baPWV is associated with an additive increase in LVH prevalence, together with hypertension and diabetes. This finding indicates that an increase in arterial stiffening might lead to LVH. Therefore, an improvement of arterial stiffening should be one of the major objectives in the prevention and treatment of LVH. A recent report indicated that statins have beneficial effects on a patient vasculopathy, independent of the BP level.31 From this viewpoint, our findings warrant further studies to clarify whether a reduction in arterial stiffness could substantially contribute to the regression of LVH independent of an effect of BP. We used baPWV as a surrogate for cfPWV to determine the potential influence of arterial stiffness on the development of LVH. However, baPWV does not exactly agree with cfPWV, because baPWV depends on the wall properties of both central elastic and peripheral muscular arteries. Taking this into account, our results may indicate that small artery stiffness, as well as large artery stiffness,
1204
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
AJH–December 2006 –VOL. 19, NO. 12
2.
Casale PN, Devereux RB, Milner M, Zullo G, Harshfield G, Pickering T, Laragh J: Value of echocardiographic measurement of left ventricular mass in predicting cardiovascular morbid events in hypertensive men. Ann Intern Med 1986;105:173–178. Frohlich ED, Tarazi RC, Dustan HP: Clinical–physiological correlations in the development of hypertensive heart disease. Circulation 1971;44:446 – 455. Merillon JP, Motte G, Masquet C, Azancot I, Guiomard A, Gourgon R: Relationship between physical properties of the arterial system and left ventricular performance in the course of aging and arterial hypertension. Eur Heart J 1998;3(Suppl A):95–102. Bouthier JD, De Luca N, Safar ME, Simon AC: Cardiac hypertrophy and arterial distensibility in essential hypertension. Am Heart J 1985;109:1345–1352. Boutouyrie P, Laurent S, Girerd X, Benetos A, Lacolley P, Abergel E, Safar M: Common carotid artery stiffness and patterns of left ventricular hypertrophy in hypertensive patients. Hypertension 1995;25:651– 659. Asmar R, Benetos A, Topouchian J, Laurent P, Pannier B, Brisac A-M, Target R, Levy BI: Assessment of arterial distensibility by automatic pulse wave velocity measurement. Validation and clinical application studies. Hypertension 1995;26:485– 490. Hashimoto J, Chonan K, Aoki Y, Nishimura T, Ohkubo T, Hozawa A, Suzuki M, Matsubara M, Araki T, Imai Y: Pulse wave velocity and the second derivative of the finger photoplethysmogram in treated hypertensive patients: their relationship and associating factors. J Hypertens 2002;20:2415–2422. Lehmann ED: Pulse wave velocity as a marker of vascular disease. Lancet 1996;348:741. Blacher J, Asmar R, Djane S, London GM, Safer ME: Aortic pulse wave velocity as a marker of cardiovascular risk in hypertensive patients. Hypertension 1990;33:1111–1117. Yamashina A, Tomiyama H, Takeda K, Tsuda H, Arai T, Hirose K, Koji Y, Hori S, Yamamoto Y: Validity, reproducibility, and clinical significance of noninvasive brachial-ankle pulse wave velocity measurement. Hypertens Res 2002;25:359 –364. Hashimoto J, Watabe D, Kimura A, Takahashi H, Ohkubo T, Totsune K, Imai Y: Determinants of the second derivative of the finger photoplethysmogram and brachial–ankle pulse-wave velocity: the Ohasama study. Am J Hypertens 2005;18:477– 485. Imai Y, Nagai K, Sakuma M, Sakuma H, Nakatsuka H, Satoh H, Minami N, Munakata M, Hashimoto J, Yamagishi T, Watanabe N, Yabe T, Nishiyama A, Abe K: Ambulatory blood pressure of adults in Ohasama, Japan. Hypertension 1993;22:900 –912. Dahlof B, Devereux RB, de Faire U, Fyhrquist F, Hedner T, Ibsen H, Julius S, Kjeldsen S, Kristianson K, Lederballe-Pedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H: The Losartan Intervention For Endpoint Reduction (LIFE) in Hypertension Study. Rationale, design, and methods. Am J Hypertens 1997; 10:705 Dahl FB 713. Molloy TJ, Okin PM, Devereux RB, Kligfield P: Electrocardiographic detection of left ventricular hypertrophy by the simple QRS voltage– duration product. J Am Coll Cardiol 1992;20:1180 –1186. Okin PM, Roman MJ, Devereux RB, Kligfield P: Gender differences and the electrocardiogram in left ventricular hypertrophy. Hypertension 1995;25:242–249. Okin PM, Roman MJ, Devereux RB, Kligfield P: Electrocardiographic identification of increased left ventricular mass by simple voltage duration products. J Am Coll Cardiol 1995;25:417– 423. Sokolow M, Lyon TO: The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and link leads. Am Heart J 1949;37:161–186. Dahl FB, Devereux RB, Julius S, Kjeldsen SE, Beevers G, de Faire U, Fyhrquist F, Hedner T, Ibsen H, Kristianson K, LederballePedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H: Characteristics of 9194 patients with left ventricular hypertrophy: The LIFE study. Hypertension 1998;32:989 –997.
is associated with LVH(⫹). This is likely because the timing and amplitude of wave reflection, and hence late systolic pressure augmentation as a LV load, depend not only on central PWV but also on peripheral PWV, considering that peripheral artery–arteriole junctions are major reflecting sites.32 Based on these explanations, using the baPWV might be advantageous. The present study has some limitations. First, LVH was diagnosed using ECG instead of echocardiography. The ECG is simpler than echocardiography for screening a general population. However, ECG is undoubtedly less sensitive than echocardiography in diagnosing LVH.33 The findings of this study would have been more generalizable by clarifying the relationship of PWV to echocardiography-derived LVH as well as to ECG-derived LVH, because the two methods could provide somewhat different information about LVH. Second, we constructed a t-baPWV based on the LVH(⫺) status on ECG rather than on normalcy. Some of the LVH(⫺) subjects actually had high BP, had received treatment for hypertension, or might have had anatomical LVH. Therefore, the currently used method for calculating t-baPWV may not be applicable to a different population. Establishment of a more precise t-baPWV (namely, a normal reference value of baPWV for each individual) requires a further study aimed at the subjects who have no history of hypertension or evidence of LVH by echocardiography. Third, we did not measure central BP. Although baPWV was related to LVH(⫹), independent of peripheral brachial BP, it remains to be determined whether baPWV would still be a LVH predictor that is independent of central aortic BP. Further studies are necessary to clarify such an important relationship among arterial stiffness, central BP, and LVH. Finally, the possibility cannot be completely ruled out that the reported relationship between arterial stiffness and LVH(⫹) is not necessarily independent of BP, but is only as good as the adjustment for BP. Sampling variability and the limitations of modeling could influence the effectiveness of the adjustment. Prospective studies are still required to resolve this key issue. In conclusion, the present study showed that arterial stiffness might be related to electrocardiographically determined LVH, independent of peripheral brachial BP, in the general population. Thus, stiffening in large arteries appears to significantly contribute to the pathogenesis of LVH, together with elevated BP. Further longitudinal studies are necessary to clarify causal relationship between arterial stiffening and LVH development.
3.
4.
5.
6.
7.
8.
9. 10.
11.
12.
13.
14.
15.
16.
17.
18.
References 19. 1.
Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP: Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990;322:1561–1566.
AJH–December 2006 –VOL. 19, NO. 12
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
20. Casale PN, Devereux RB, Alonso DR, Campo E, Kligfield P: Improved sex-specific criteria of left ventricular hypertrophy for clinical and computer electrocardiogram interpretation. Circulation 1987;75:565–572. 21. Norman JE Jr, Levy D: Improved electrocardiographic left ventricular hypertrophy: results of a correlated data base approach. J Am Coll Cardiol 1995;26:1022–1029. 22. Devereux RB, Bella J, Boman K, Gerdts E, Nieminen MS, Rokkedal J, Papademetriou V, Wachtell K, Wright J, Paranicas M, Okin PM, Roman MJ, Smith G, Dahl FB: Echocardiographic left ventricular geometry in hypertensive patients with electrocardiographic left ventricular hypertrophy: the LIFE study. Blood Press 2001;10:74 – 82. 23. 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. 24. Nichols WW, O’Rourke MF: Vascular impedance, in Nichols WW, O’Rourke MF, (eds): McDonald Blood Flow in Arteries: Theoretical, Experimental and Clinical Principles, 4th ed. London, Edward Arnold, 1998, pp 54 –97, 243–283, 347–395. 25. Kelly RP, O’Rourke MF: Evaluation of arterial wave forms in hypertension and normotension, in Laragh JH, Brenner BM (eds): Hypertension: Pathophysiology, Diagnosis, and Management, New York, Laven Press, 1995, pp 343–364. 26. Hashimoto J, Watabe D, Hatanaka R, Hanasawa T, Metoki H, Asayama K, Ohkubo T, Totsune K, Imai Y: Enhanced radial late-
27.
28.
29.
30.
31.
32.
33.
1205
systolic pressure augmentation in hypertensive patients with left ventricular hypertrophy. Am J Hypertens 2006;19:27–32. Saba PS, Roman MJ, Pini R, Spitzer M, Ganau A, Devereux RB: Relation of arterial pressure waveform to left ventricular and carotid anatomy in normotensive subjects. J Am Coll Cardiol 1993;22: 1873–1880. Marchais SJ, Guerin AP, Pannier BM, Levy BI, Safar ME, London GM: Wave reflections and cardiac hypertrophy in chronic uremia. Influence of body size. Hypertension 1993;22:876 – 883. Stergiopulos N, Meister JJ, Westerhof N: Simple and accurate way for estimating total and segmental arterial compliance: the pulse pressure method. Ann Biomed Eng 1994;22:392–397. Bella JN, Devereux RB, Roman MJ, Palmieri V, Liu JE, Paranicas M, Welty TK, Lee ET, Fabsitz RR, Howard BV: Separate and joint effects of systemic hypertension and diabetes mellitus on left ventricular structure and function in American Indians (the strong heart study). Am J Cardiol 2001;87:1260 –1265. Raison J, Rudnichi A, Safar ME: Effects of atorvastatin on aortic pulse wave velocity in patients with hypertension and hypercholesterolaemia: a preliminary study. J Hum Hypertens 2002;16:705–710. Nichols WW, O’Rourke MF: McDonald Blood Flow in Arteries: Theoretical, Experimental and Clinical Principles, 5th ed. London, Arnold, 2005. Reichek N, Devereux RB: Left ventricular hypertrophy: relationship of anatomic, echocardiographic and electrocardiographic findings. Circulation 1981;63:1391–1398.
ORIGINAL CONTRIBUTIONS
2006; 19:1199 –1205
Blood Vessels
Electrocardiographic Left Ventricular Hypertrophy and Arterial Stiffness: The Ohasama Study Daisuke Watabe, Junichiro Hashimoto, Rieko Hatanaka, Tomohiro Hanazawa, Hiromi Ohba, Takayoshi Ohkubo, Masahiro Kikuya, Kazuhito Totsune, and Yutaka Imai Background: Whether arterial stiffness per se contributes to left ventricular hypertrophy (LVH) independently of blood pressure (BP) remains unknown. We examined the relationship between pulse wave velocity (PWV) and LVH in a large population. Methods: The PWV was measured between the brachial and ankle regions (baPWV) of 798 individuals. We diagnosed LVH using electrocardiographic criteria: Cornell voltage-duration product ⬎2440 mm ⫻ msec or Sokolow-Lyon voltage ⬎38 mm. The participants were initially separated into those with and without LVH [LVH(⫹) and LVH(⫺) groups, respectively]. To determine theoretical baPWV, we first constructed a nomogram for the LVH(⫺) group, calculated the PWV index (measured baPWV ⫺ theoretical baPWV) for each individual and then compared the two groups. We also examined the factors associated with LVH(⫹) using multivariate analyses.
Mean arterial pressure (MAP) (mm Hg) ⫹ 0.05 ⫻ Heart rate (beats/min) ⫺ 11.74 (R2 ⫽ 0.56). The PWV index was greater in the LVH(⫹) than in the LVH(⫺) group (P ⫽ .025). The baPWV was independently related to LVH(⫹) along with MAP, medication for hypertension, and for diabetes; a 1 SD (4.3 m/sec) increase in baPWV was associated with a 26% increase in the risk of LVH(⫹) (P ⫽ .022). When LVH(⫹) risk factors were defined as hypertension, diabetes, and high baPWV (ⱖ14.6 m/sec), the prevalence of LVH(⫹) linearly increased with the number of concomitant LVH(⫹) risk factors (P ⬍ .001). Conclusions: Arterial stiffness is independently related to electrocardiographically determined LVH in the general population. Am J Hypertens 2006;19:1199 –1205 © 2006 American Journal of Hypertension, Ltd.
Results: Linear regression analysis revealed that the theoretical baPWV (m/sec) ⫽ 0.20 ⫻ age (years) ⫹ 0.13 ⫻
Key Words: Arterial stiffness, pulse wave velocity, left ventricular hypertrophy, electrocardiography.
eft ventricular hypertrophy (LVH) is an independent predictor of cardiovascular risk1,2 and the most common cardiac abnormality associated with hypertension. The development of LVH is generally ascribed to a chronic increase in cardiac afterload that arises due to elevated blood pressure (BP).3 On the other hand, arterial stiffness might also play an important role in the development of LVH. Several markers of vascular functional properties, including aortic impedance,4 pulse wave velocity (PWV),5 and carotid distensibility6 are significantly correlated with LV mass in hypertensive patients. These
L
studies might mean that considerable arterial stiffening in response to hypertension increases end-systolic stress on the left ventricle, resulting in the development of LVH. However, whether arterial stiffness is similarly related to LVH in the general population as well as in hypertensive patients remains unknown. Moreover, whether arterial stiffening per se contributes to LVH independently of BP remains to be determined. Understanding the respective roles of high BP and arterial stiffness as determinants of LVH will uncover the mechanisms underlying the development of LVH.
Received November 6, 2005. First decision April 27, 2006. Accepted May 4, 2006. From the Department of Clinical Pharmacology and Therapeutics (DW, RH, TH, HO, KT, YI) and Department of Planning for Drug Development and Clinical Evaluation (JH, TO, MK), Tohoku University Graduate School of Pharmaceutical Science and Medicine, and Tohoku University 21th Century COE Program “Comprehensive Research and
Education Center for Planning of Drug Development and Clinical Evaluation” (JH, TO, KT, YI), Sendai, Japan. Address correspondence and reprint requests to Dr. Junichiro Hashimoto, Department of Planning for Drug Development and Clinical Evaluation, Tohoku University Graduate School of Pharmaceutical Science and Medicine, 6-3 Aoba, Aramaki, Aoba-ku, Sendai 980-8578, Japan; e-mail: [email protected]
© 2006 by the American Journal of Hypertension, Ltd. Published by Elsevier Inc.
0895-7061/06/$32.00 doi:10.1016/j.amjhyper.2006.05.001
1200
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
Quantitative information about the elastic properties of large arteries is usually obtained by measuring PWV,5,7,8 which can predict cardiovascular risk.9,10 In addition to conventional carotid–femoral PWV (cfPWV) measurements, the brachial–ankle PWV (baPWV) can also evaluate arterial stiffness.11 Other studies have shown that this technique can provide useful information on arterial stiffness and that it is appropriate for screening a general population.11,12 The present cross-sectional study evaluates the relationship between baPWV and electrocardiographically determined LVH in the general population.
Methods Study Participants The present study was based on a health survey performed in the Japanese city of Ohasama, the geographic and demographic characteristics of which have been described elsewhere.12,13 Of 1739 individuals who participated in the health checkup, 808 completed baPWV measurements, an electrocardiogram, and both anthropometric and biochemical examinations. The baPWV measurement includes information about peripheral components of the arterial tree, and thus might be affected by arterial occlusive disease. From this viewpoint, individuals with suspected arterial occlusive disease, such as those having an ankle– brachial index (ABI) of ⬍0.9, were excluded (n ⫽ 10), therefore the study included 798 individuals. The Department of Health of the Ohasama Town Government approved the present study and informed consent was obtained from all participants. During the checkup, brachial systolic BP and diastolic BP were measured twice while seated after a 2-min rest, using an automated cuff oscillometric device (Omron HEM907; Omron Life Science, Kyoto, Japan). The average of the two readings was defined as the BP value. The pulse pressure (PP) and mean arterial pressure (MAP) were calculated according to the formula: PP ⫽ Systolic BP ⫺ Diastolic BP, and MAP ⫽ Diastolic BP ⫹ PP/3, respectively. Biochemical data were obtained from venous blood samples collected on the same days as the PWV measurements. Information concerning lifestyle, habits, and therapeutic status was collected using a questionnaire. Although a possible determinant of arterial stiffness,8 no data about the duration of hypertension were available to the present study. Hypertension was defined by the following criteria: systolic BP ⱖ140 mm Hg or diastolic BP ⱖ90 mm Hg or taking medication for hypertension. Diabetes was defined as fasting blood glucose ⱖ126 mg/dL, postprandial glucose ⱖ200 mg/dL, hemoglobin A1c (HbA1c) ⱖ6.5%, or taking medication for diabetes. Electrocardiography We electrocardiographically diagnosed LVH from a 12-lead electrocardiogram (ECG) in all participants.
AJH–December 2006 –VOL. 19, NO. 12
According to the criteria of the Losartan Intervention For Endpoint Reduction in Hypertension (LIFE) study,14 we defined LVH as follows: Cornell voltage– duration product [(RaVL ⫹ SV3) ⫻ QRS duration, in men; (RaVL ⫹ SV3 ⫹ 6 mm) ⫻ QRS duration, in women] ⬎2440 mm ⫻ msec,15–17 or Sokolow-Lyon voltage (SV1 ⫹ RV5/6) ⬎38 mm.18 The R waves in leads aVL, V5, and V6 and S waves in leads V1 and V3 were measured to the nearest 0.5 mm (0.05 mV). The rationale for using the criteria in the LIFE study has been described.19 The specificity of both of the Cornell and Sokolow-Lyon criteria (⬎35 mm) is more than 95% in normal adults but the sensitivity is only 30% to 40%.20 However, combining the two criteria (Cornell voltage– duration product ⬎2440 mm ⫻ msec or Sokolow-Lyon voltage ⬎38 mm) increased the sensitivity by more than 10% without a loss of specificity.15–18,21 Furthermore, the ECG criteria used in the LIFE study identified hypertensive patients with a ⬎70% prevalence of anatomic LVH by echocardiography.22 Measurement of PWV The baPWV was measured in supine individuals after at least 5 min of rest using an automatic device (form PWV/ ABI; Colin Co., Ltd., Komaki, Japan), as described.11,12 Briefly, pressure waveforms of the brachial and tibial arteries were simultaneously recorded by placing occlusion cuffs connected to a plethysmographic sensor around both the brachia and ankles. The time delay (T) of the two waveforms was measured between the feet. The path lengths from the suprasternal notch to the brachium (Lb) and from the suprasternal notch to the ankle (La) were automatically calculated according to individual height. The baPWV was calculated using the following equation: baPWV ⫽ 共La ⫺ Lb兲/T 共m/sec兲. Values of baPWV were measured for an average of 10 sec and we used right brachial-to-right ankle baPWV. The validity and reproducibility of this method have been reported; the intraobserver repeatability coefficient is 0.87 and the interobserver repeatability coefficient is 0.98.11 The rationale for the use of baPWV instead of cfPWV was based on the finding that this parameter closely correlates with aortic PWV determined by an invasive method.11 Calculation of PWV Index We calculated the PWV index for each individual according to Blacher et al.23 Briefly, the PWV index was obtained by subtracting the theoretical baPWV (t-baPWV) from the measured baPWV. The PWV index represents inherent arterial stiffness and was adjusted for the influence of some relevant factors. To determine t-baPWV, we constructed a nomogram of baPWV for the subjects without electrocardiographic LVH [LVH(⫺)]. We constructed a multivariate linear regression equation to estimate tbaPWV using age, BP, heart rate (HR), and gender as explanatory variables, because these are generally considered major influences on PWV.12,23,24
AJH–December 2006 –VOL. 19, NO. 12
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
1201
Table 1. Characteristics of 798 participants with and without LVH LVH
Age (y) Gender (male/female) SBP (mm Hg) DBP (mm Hg) MAP (mm Hg) PP (mm Hg) HR (beats/min) baPWV (m/sec) ABI BMI (kg/m2) Total cholesterol (mg/dL) HDL cholesterol (mg/dL) Triglycerides (mg/dL) Serum creatinine (mg/dL) HbA1c (%) Medication for hypertension (%) Medication for diabetes (%) Medication for hyperlipidemia (%) Smoking (%) Alcohol consumption (%)
(ⴙ) n ⴝ 168
(ⴚ) n ⴝ 630
P
65.1 ⫾ 10.8 62/106 143.5 ⫾ 20.0 78.0 ⫾ 13.1 99.8 ⫾ 14.2 65.5 ⫾ 14.6 73.7 ⫾ 13.4 18.6 ⫾ 5.3 1.11 ⫾ 0.07 24.1 ⫾ 3.2 203.3 ⫾ 29.5 62.5 ⫾ 15.6 142.5 ⫾ 91.7 0.67 ⫾ 0.14 5.14 ⫾ 0.76 31.5 8.9 8.3 14.3 38.7
61.9 ⫾ 10.7 205/425 134.3 ⫾ 18.1 74.6 ⫾ 10.6 94.5 ⫾ 11.8 59.7 ⫾ 14.2 73.4 ⫾ 11.7 16.6 ⫾ 3.9 1.10 ⫾ 0.08 24.2 ⫾ 3.1 203.9 ⫾ 33.3 60.7 ⫾ 16.0 135.9 ⫾ 73.8 0.66 ⫾ 0.14 5.04 ⫾ 0.70 17.6 3.3 6.5 12.7 34.6
.001 .287 ⬍.001 .002 ⬍.001 ⬍.001 .814 ⬍.001 .191 .740 .826 .193 .327 .574 .106 ⬍.001 .002 .407 .587 .325
Abbreviations as in Table 1.
In calculating the PWV index, the t-baPWV was regarded as the assumed value that represents a population and subtracted from the measured baPWV value of each subject without and with electrocardiographic LVH [LVH(⫺) and LVH(⫹), respectively]. Statistical Analysis The participants were divided into LVH(⫹) and LVH(⫺) groups, and their characteristics were compared using the Student t test for continuous variables and the 2 test for categorical variables. Second, we compared the PWV index between LVH(⫹) and LVH(⫺) groups using Student t test to evaluate whether inherent arterial stiffness is enhanced in LVH(⫹) individuals. Third, multivariate stepwise logistic regression analysis clarified the factors related to LVH(⫹). The adjusted odds ratio (OR) and 95% confidence intervals (CI) of the risk of LVH(⫹) were determined using possibly independent variables including age, gender, MAP, HR, baPWV, ABI, body mass index (BMI), total cholesterol, HDL-cholesterol, triglycerides, serum creatinine, HbA1c, medication for hypertension, diabetes and hyperlipidemia, smoking status, and alcohol consumption. To evaluate whether baPWV would be a better independent predictor of LVH(⫹) than brachial PP, multivariate stepwise logistic regression analysis was applied using a model that included brachial PP instead of baPWV and using a model that included both baPWV and brachial PP instead of MAP. Furthermore, to evaluate the relationship between PWV index and LVH(⫹), a similar analysis used a model that included the PWV index. In this
regard, age, MAP, and HR were eliminated from this model, because the PWV index was adjusted for these confounding factors. Finally, we evaluated the prevalence of LVH(⫹) in relation to concomitant LVH(⫹) risk factors. The participants were classified into four groups according to the number of concomitant LVH(⫹) risk factors to compare the prevalence of LVH(⫹). We established an optimal cutoff value of baPWV using receiver operating characteristic (ROC) analysis to discriminate individuals in terms of LVH(⫹). All statistical analyses were performed using SPSS software version 11.0 (SPSS Inc., Chicago IL). Statistical significance was accepted at a value of P ⬍ .05.
Results Characteristics of Study Population The study population consisted of 267 men and 531 women with a mean age of 62.6 ⫾ 10.8 years (range, 34 to 88 years). Of these, 164 (20.6%) were taking medication for hypertension. Table 1 lists the characteristics of the LVH(⫹) and LVH(⫺) individuals. Relationship Between Arterial Stiffness and LVH(ⴙ) The baPWV in the LVH(⫺) group was significantly and independently correlated with age, MAP, and HR (P ⬍ .001 for all, multiple regression R2 ⫽ 0.56), but not with gender. Thus, the t-baPWV was calculated from the following
1202
AJH–December 2006 –VOL. 19, NO. 12
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
When using a model that included the PWV index, this value was also an independent determinant of LVH(⫹), together with medication for hypertension and for diabetes (data not shown). Prevalence of LVH(ⴙ) and Multiple Risk Factors
FIG. 1. Comparison of pulse wave velocity (PWV) indices between LVH(⫹) (n ⫽ 168) and LVH(⫺) (n ⫽ 630) groups. Values are presented as means ⫾ SE.
equation: t-baPWV (m/sec) ⫽ 0.20 ⫻ Age (years) ⫹ 0.13 ⫻ MAP (mm Hg) ⫹ 0.05 ⫻ HR (beats/min) ⫺ 11.74. Fig. 1 shows that the PWV index (baPWV ⫺ t-baPWV) was significantly greater in the LVH(⫹) than in the LVH (⫺) group. Multivariate logistic regression analysis revealed that the factors independently associated with LVH(⫹) were MAP, baPWV, as well as medication for hypertension and for diabetes (Table 2). An increase in baPWV of 1SD (4.3 m/sec) was associated with a 26% increase in the risk of LVH(⫹). In contrast, when brachial PP replaced baPWV, brachial PP did not associate with LVH(⫹) (Table 2). When both baPWV and brachial PP were included in the model but MAP was not, they were independent predictors of LVH(⫹) (Table 2).
Based on the results of the logistic regression analysis (Table 2), we defined the putative risk factors for LVH(⫹) as hypertension, diabetes, and high baPWV. According to the ROC analysis, the optimal cutoff value of baPWV to detect LVH(⫹) was 14.6 m/sec, with 82.7% of sensitivity and 34.1% specificity (data not shown). Therefore, individuals with baPWV ⱖ14.6 m/sec were regarded as having a high baPWV. Fig. 2 shows that when classified into four groups by the number of concomitant LVH(⫹) risk factors, the prevalence of LVH(⫹) increased linearly with increasing numbers of concomitant risk factors.
Discussion The key finding of the present study is that arterial stiffness measured through baPWV was significantly associated in the general population with electrocardiographically determined LVH [LVH(⫹)] independently of peripheral brachial BP. As with previous studies using different methods,4 – 6 this result supports the notion that arterial distensibility plays an important role in the development
Table 2. Determinants of LVH(⫹) based on multiple stepwise logistic regression analysis Adjusted odds ratio (95% CI) Model including baPWV* MAP (mm Hg) (per 12.5 mm Hg)§ baPWV (m/sec) (per 4.3 m/sec)§ Medication for hypertension Medication for diabetes Model including PP† Age (y) (per 10.8 y)§ MAP (mm Hg) (per 12.5 mm Hg)§ Brachial PP (mm Hg) (per 14.4 mm Hg)§ Medication for hypertension Medication for diabetes Model including both baPWV and brachial PP instead of MAP‡ Brachial PP (mm Hg) (per 14.4 mm Hg)§ baPWV (m/sec) (per 4.3 m/sec)§ Medication for hypertension Medication for diabetes
P
1.36 1.26 1.77 2.64
(1.11–1.67) (1.03–1.53) (1.19–2.64) (1.28–5.44)
.003 .022 .005 .008
1.26 1.53 1.15 1.69 2.64
(1.04–1.53) (1.28–1.82) (0.94–1.41) (1.13–2.54) (1.28–5.41)
.021 ⬍.001 .183 .012 .008
1.26 1.31 1.73 2.40
(1.02–1.50) (1.08–1.59) (1.16–2.59) (1.17–4.90)
.031 .006 .007 .017
ABI ⫽ ankle– brachial pressure index; baPWV ⫽ brachial–ankle pulse wave velocity; BMI ⫽ body mass index; CI ⫽ confidence interval; DBP ⫽ diastolic blood pressure; HbA1c ⫽ hemoglobin A1c; HDL ⫽ high-density lipoprotein; HR ⫽ heart rate; LVH ⫽ left ventricular hypertrophy; MAP ⫽ mean arterial pressure; PP ⫽ pulse pressure; SBP ⫽ systolic blood pressure. * Factors excluded from the model: age, gender, HR, BMI, ABI, total cholesterol, HDL ⫽ cholesterol, triglycerides, serum creatinine level, HbA1c, medication for hyperlipidemia, smoking, and alcohol consumption. † Factors excluded from the model: gender, HR, BMI, ABI, total cholesterol, HDL ⫽ cholesterol, triglycerides, serum creatinine level, HbA1c, medication for hyperlipidemia, smoking, and alcohol consumption. ‡ Factors excluded from the model: age, gender, HR, BMI, ABI, total cholesterol, HDL cholesterol, triglycerides, serum creatinine level, HbA1c, medication for hyperlipidemia, smoking, and alcohol consumption. § 1 SD change.
AJH–December 2006 –VOL. 19, NO. 12
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
FIG. 2. Prevalence of LVH(⫹) among four groups according to number of concomitant risk factors. Risk factors for LVH(⫹) included hypertension, diabetes, and high brachial–ankle pulse wave velocity (baPWV) (baPWV ⱖ14.6 m/sec).
of LVH. To our knowledge, however, this study is the first to demonstrate that PWV and LVH are independently related in the general population. Few previous studies have demonstrated a direct relationship between PWV and LVH. Bouthier et al5 have shown that in hypertensive patients the LV mass-to-volume ratio is positively correlated with cfPWV, but the relationship between the two parameters was not independent of BP. In contrast, our findings indicate that arterial stiffening as assessed by baPWV is directly involved in the pathogenesis of LVH. The association between increased PWV and LVH(⫹) identified in the present study can be explained by the following mechanism. Stiffening of the arteries increases the PWV and might be responsible for earlier return of the reflected wave from the periphery toward the ascending aorta. This means that a reflected wave would return during the systolic, rather than the diastolic phase, thus augmenting the systolic part of the incident pressure wave and further contributing to the increase in systolic BP and PP in the ascending aorta.24,25 Therefore, increased arterial stiffness might contribute to LVH through an increase in central PP, independently of MAP. In fact, our recent study showed that late systolic pressure augmentation is associated with LVH in hypertensive patients.26 Other reports have shown that in normotensive patients, the shapes of central pressure waveforms are significantly related to LVH,27,28 which also supports our notion. To interpret our findings, mechanical loads imposed on the heart should be considered to comprise steady and pulsatile components.24 Increased MAP, which reflects steady pressure stress imposed on the heart, is the key determinant of LVH(⫹), whereas increased arterial stiffness as a contributory factor to the pulsatile pressure stress, also appears to be an important determinant of LVH(⫹). In addition, baPWV and brachial PP were both independently associated with LVH(⫹) (Table 1). This result
1203
indicates that baPWV did not substantially overlap with brachial PP, whereas brachial PP as well as PWV is used in clinical practice as an approximate measure of arterial stiffness.29 Therefore, it seems that PWV is a key predictor of LVH(⫹) in that PWV includes different information than brachial PP. Here, we used the concept of the PWV index according to the study of Blacher et al.23 To assess arterial wall distensibility using PWV measurement, whether changes in arterial stiffness are due to inherent changes in the arterial wall are difficult to reliably determine, because arterial stiffness passively responds to physiologic factors. However, calculation of the PWV index could help resolve this issue. The present study showed that baPWV correlates with age, BP, and HR. Using the PWV nomogram obtained from the age, MAP, and HR of the LVH(⫺) group, calculation of the PWV index should generate a value that is independent of these parameters. Furthermore, the PWV index was significantly greater in the LVH(⫹) group than in the LVH(⫺) group. This suggests that inherent arterial stiffness is enhanced in individuals with LVH compared with those without LVH, independently of age, BP, and HR. Moreover, the multivariate analysis showed that the PWV index was an independent determinant of LVH(⫹). These findings might emphasize a deleterious effect of inherent arterial stiffness on the pathogenesis of LVH, independently of age, BP, and HR. The present study also demonstrated that prevalence of LVH(⫹) increased linearly with increasing the number of concomitant LVH(⫹) risk factors, including hypertension, diabetes, and high baPWV. This result does not necessarily support the fact that the physiologic composite effect is theoretically linear, because the composition of the set of even the same number of risk factors differed in individual patients. However, the Strong Heart Study30 associated hypertension and diabetes with independent and additive increases in the LVH prevalence. The present study supports the notion that a high baPWV is associated with an additive increase in LVH prevalence, together with hypertension and diabetes. This finding indicates that an increase in arterial stiffening might lead to LVH. Therefore, an improvement of arterial stiffening should be one of the major objectives in the prevention and treatment of LVH. A recent report indicated that statins have beneficial effects on a patient vasculopathy, independent of the BP level.31 From this viewpoint, our findings warrant further studies to clarify whether a reduction in arterial stiffness could substantially contribute to the regression of LVH independent of an effect of BP. We used baPWV as a surrogate for cfPWV to determine the potential influence of arterial stiffness on the development of LVH. However, baPWV does not exactly agree with cfPWV, because baPWV depends on the wall properties of both central elastic and peripheral muscular arteries. Taking this into account, our results may indicate that small artery stiffness, as well as large artery stiffness,
1204
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
AJH–December 2006 –VOL. 19, NO. 12
2.
Casale PN, Devereux RB, Milner M, Zullo G, Harshfield G, Pickering T, Laragh J: Value of echocardiographic measurement of left ventricular mass in predicting cardiovascular morbid events in hypertensive men. Ann Intern Med 1986;105:173–178. Frohlich ED, Tarazi RC, Dustan HP: Clinical–physiological correlations in the development of hypertensive heart disease. Circulation 1971;44:446 – 455. Merillon JP, Motte G, Masquet C, Azancot I, Guiomard A, Gourgon R: Relationship between physical properties of the arterial system and left ventricular performance in the course of aging and arterial hypertension. Eur Heart J 1998;3(Suppl A):95–102. Bouthier JD, De Luca N, Safar ME, Simon AC: Cardiac hypertrophy and arterial distensibility in essential hypertension. Am Heart J 1985;109:1345–1352. Boutouyrie P, Laurent S, Girerd X, Benetos A, Lacolley P, Abergel E, Safar M: Common carotid artery stiffness and patterns of left ventricular hypertrophy in hypertensive patients. Hypertension 1995;25:651– 659. Asmar R, Benetos A, Topouchian J, Laurent P, Pannier B, Brisac A-M, Target R, Levy BI: Assessment of arterial distensibility by automatic pulse wave velocity measurement. Validation and clinical application studies. Hypertension 1995;26:485– 490. Hashimoto J, Chonan K, Aoki Y, Nishimura T, Ohkubo T, Hozawa A, Suzuki M, Matsubara M, Araki T, Imai Y: Pulse wave velocity and the second derivative of the finger photoplethysmogram in treated hypertensive patients: their relationship and associating factors. J Hypertens 2002;20:2415–2422. Lehmann ED: Pulse wave velocity as a marker of vascular disease. Lancet 1996;348:741. Blacher J, Asmar R, Djane S, London GM, Safer ME: Aortic pulse wave velocity as a marker of cardiovascular risk in hypertensive patients. Hypertension 1990;33:1111–1117. Yamashina A, Tomiyama H, Takeda K, Tsuda H, Arai T, Hirose K, Koji Y, Hori S, Yamamoto Y: Validity, reproducibility, and clinical significance of noninvasive brachial-ankle pulse wave velocity measurement. Hypertens Res 2002;25:359 –364. Hashimoto J, Watabe D, Kimura A, Takahashi H, Ohkubo T, Totsune K, Imai Y: Determinants of the second derivative of the finger photoplethysmogram and brachial–ankle pulse-wave velocity: the Ohasama study. Am J Hypertens 2005;18:477– 485. Imai Y, Nagai K, Sakuma M, Sakuma H, Nakatsuka H, Satoh H, Minami N, Munakata M, Hashimoto J, Yamagishi T, Watanabe N, Yabe T, Nishiyama A, Abe K: Ambulatory blood pressure of adults in Ohasama, Japan. Hypertension 1993;22:900 –912. Dahlof B, Devereux RB, de Faire U, Fyhrquist F, Hedner T, Ibsen H, Julius S, Kjeldsen S, Kristianson K, Lederballe-Pedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H: The Losartan Intervention For Endpoint Reduction (LIFE) in Hypertension Study. Rationale, design, and methods. Am J Hypertens 1997; 10:705 Dahl FB 713. Molloy TJ, Okin PM, Devereux RB, Kligfield P: Electrocardiographic detection of left ventricular hypertrophy by the simple QRS voltage– duration product. J Am Coll Cardiol 1992;20:1180 –1186. Okin PM, Roman MJ, Devereux RB, Kligfield P: Gender differences and the electrocardiogram in left ventricular hypertrophy. Hypertension 1995;25:242–249. Okin PM, Roman MJ, Devereux RB, Kligfield P: Electrocardiographic identification of increased left ventricular mass by simple voltage duration products. J Am Coll Cardiol 1995;25:417– 423. Sokolow M, Lyon TO: The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and link leads. Am Heart J 1949;37:161–186. Dahl FB, Devereux RB, Julius S, Kjeldsen SE, Beevers G, de Faire U, Fyhrquist F, Hedner T, Ibsen H, Kristianson K, LederballePedersen O, Lindholm LH, Nieminen MS, Omvik P, Oparil S, Wedel H: Characteristics of 9194 patients with left ventricular hypertrophy: The LIFE study. Hypertension 1998;32:989 –997.
is associated with LVH(⫹). This is likely because the timing and amplitude of wave reflection, and hence late systolic pressure augmentation as a LV load, depend not only on central PWV but also on peripheral PWV, considering that peripheral artery–arteriole junctions are major reflecting sites.32 Based on these explanations, using the baPWV might be advantageous. The present study has some limitations. First, LVH was diagnosed using ECG instead of echocardiography. The ECG is simpler than echocardiography for screening a general population. However, ECG is undoubtedly less sensitive than echocardiography in diagnosing LVH.33 The findings of this study would have been more generalizable by clarifying the relationship of PWV to echocardiography-derived LVH as well as to ECG-derived LVH, because the two methods could provide somewhat different information about LVH. Second, we constructed a t-baPWV based on the LVH(⫺) status on ECG rather than on normalcy. Some of the LVH(⫺) subjects actually had high BP, had received treatment for hypertension, or might have had anatomical LVH. Therefore, the currently used method for calculating t-baPWV may not be applicable to a different population. Establishment of a more precise t-baPWV (namely, a normal reference value of baPWV for each individual) requires a further study aimed at the subjects who have no history of hypertension or evidence of LVH by echocardiography. Third, we did not measure central BP. Although baPWV was related to LVH(⫹), independent of peripheral brachial BP, it remains to be determined whether baPWV would still be a LVH predictor that is independent of central aortic BP. Further studies are necessary to clarify such an important relationship among arterial stiffness, central BP, and LVH. Finally, the possibility cannot be completely ruled out that the reported relationship between arterial stiffness and LVH(⫹) is not necessarily independent of BP, but is only as good as the adjustment for BP. Sampling variability and the limitations of modeling could influence the effectiveness of the adjustment. Prospective studies are still required to resolve this key issue. In conclusion, the present study showed that arterial stiffness might be related to electrocardiographically determined LVH, independent of peripheral brachial BP, in the general population. Thus, stiffening in large arteries appears to significantly contribute to the pathogenesis of LVH, together with elevated BP. Further longitudinal studies are necessary to clarify causal relationship between arterial stiffening and LVH development.
3.
4.
5.
6.
7.
8.
9. 10.
11.
12.
13.
14.
15.
16.
17.
18.
References 19. 1.
Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP: Prognostic implications of echocardiographically determined left ventricular mass in the Framingham Heart Study. N Engl J Med 1990;322:1561–1566.
AJH–December 2006 –VOL. 19, NO. 12
ELECTROCARDIOGRAPHIC LVH AND ARTERIAL STIFFNESS
20. Casale PN, Devereux RB, Alonso DR, Campo E, Kligfield P: Improved sex-specific criteria of left ventricular hypertrophy for clinical and computer electrocardiogram interpretation. Circulation 1987;75:565–572. 21. Norman JE Jr, Levy D: Improved electrocardiographic left ventricular hypertrophy: results of a correlated data base approach. J Am Coll Cardiol 1995;26:1022–1029. 22. Devereux RB, Bella J, Boman K, Gerdts E, Nieminen MS, Rokkedal J, Papademetriou V, Wachtell K, Wright J, Paranicas M, Okin PM, Roman MJ, Smith G, Dahl FB: Echocardiographic left ventricular geometry in hypertensive patients with electrocardiographic left ventricular hypertrophy: the LIFE study. Blood Press 2001;10:74 – 82. 23. 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. 24. Nichols WW, O’Rourke MF: Vascular impedance, in Nichols WW, O’Rourke MF, (eds): McDonald Blood Flow in Arteries: Theoretical, Experimental and Clinical Principles, 4th ed. London, Edward Arnold, 1998, pp 54 –97, 243–283, 347–395. 25. Kelly RP, O’Rourke MF: Evaluation of arterial wave forms in hypertension and normotension, in Laragh JH, Brenner BM (eds): Hypertension: Pathophysiology, Diagnosis, and Management, New York, Laven Press, 1995, pp 343–364. 26. Hashimoto J, Watabe D, Hatanaka R, Hanasawa T, Metoki H, Asayama K, Ohkubo T, Totsune K, Imai Y: Enhanced radial late-
27.
28.
29.
30.
31.
32.
33.
1205
systolic pressure augmentation in hypertensive patients with left ventricular hypertrophy. Am J Hypertens 2006;19:27–32. Saba PS, Roman MJ, Pini R, Spitzer M, Ganau A, Devereux RB: Relation of arterial pressure waveform to left ventricular and carotid anatomy in normotensive subjects. J Am Coll Cardiol 1993;22: 1873–1880. Marchais SJ, Guerin AP, Pannier BM, Levy BI, Safar ME, London GM: Wave reflections and cardiac hypertrophy in chronic uremia. Influence of body size. Hypertension 1993;22:876 – 883. Stergiopulos N, Meister JJ, Westerhof N: Simple and accurate way for estimating total and segmental arterial compliance: the pulse pressure method. Ann Biomed Eng 1994;22:392–397. Bella JN, Devereux RB, Roman MJ, Palmieri V, Liu JE, Paranicas M, Welty TK, Lee ET, Fabsitz RR, Howard BV: Separate and joint effects of systemic hypertension and diabetes mellitus on left ventricular structure and function in American Indians (the strong heart study). Am J Cardiol 2001;87:1260 –1265. Raison J, Rudnichi A, Safar ME: Effects of atorvastatin on aortic pulse wave velocity in patients with hypertension and hypercholesterolaemia: a preliminary study. J Hum Hypertens 2002;16:705–710. Nichols WW, O’Rourke MF: McDonald Blood Flow in Arteries: Theoretical, Experimental and Clinical Principles, 5th ed. London, Arnold, 2005. Reichek N, Devereux RB: Left ventricular hypertrophy: relationship of anatomic, echocardiographic and electrocardiographic findings. Circulation 1981;63:1391–1398.