Blood Pressure Goals and Arterial Stiffness in Chronic Kidney Disease

Abstract

The association of optimal blood pressure (BP) control and arterial stiffness was evaluated in 172 patients with chronic kidney disease. The authors compared the augmentation index (AIx) of patients who achieved a recommended BP goal (<130/80 mm Hg) with those who did not (≥130/80 mm Hg). The median age was 57 years, 60% were male, and 70% were Caucasian. One‐third of patients had achieved a BP goal of <130/80 mm Hg. AIx was significantly lower in patients who achieved BP goal than in those who did not (median AIx, 19% vs 23%; P=.04). AIx remained significantly lower in patients who achieved the BP goal, after adjusting for age, sex, and height (mean effect on AIx, −3.3%; 95% confidence interval, −6.1% to −0.4%; P=.03). Achievement of BP goal of <130/80 mm Hg in chronic kidney disease patients is associated with significantly lower AIx and may reflect a reduction in overall arterial stiffness.

Chronic kidney disease (CKD) reflects an overall deterioration in vascular health. CKD is associated with an increased risk for cardiovascular (CV) disease (CVD). CVD is the major cause of death in CKD. ¹ , ² , ³ , ⁴ , ⁵ Vascular changes and progressive kidney dysfunction are probably the end result of several interacting risk factors that act in concert to cause vascular and kidney injury over time. ⁶ Several studies have demonstrated that intensive lowering of blood pressure (BP) in CKD patients improves CV outcomes. ⁷ , ⁸ Based on these data, the Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7) recommended a BP goal of <130/80 mm Hg for patients with CKD compared with a BP goal of <140/90 mm Hg for the general population. ⁹

Lowering peripheral BP has the effect of decreasing target organ BP burden and thus reducing target organ damage. One of the possible mechanisms that results in improved CV outcomes is improvement in arterial stiffness, as a result of BP control. Arterial stiffness is a functional marker of vascular changes or damage as the result of several risk factors including hypertension. Several recent longitudinal studies have confirmed the predictive value of arterial stiffness estimated by noninvasive methods both in individuals with increased CV risk and those with end‐stage renal disease (ESRD). ¹⁰ , ¹¹ , ¹² , ¹³ CKD is known to be associated with higher arterial stiffness ¹⁴ and has been associated with worse CV outcomes. ¹⁵ , ¹⁶ , ¹⁷ Guerin and colleagues ¹⁸ reported that improved arterial stiffness is associated with improved CV outcomes in patients with ESRD.

Peripheral BP is a product of cardiac output and peripheral vascular resistance and provides health‐risk estimates that have been validated in large population‐based studies. In comparison, central vascular measures of arterial stiffness including augmentation index (AIx) are dependent not only on cardiac output and peripheral vascular resistance but also the stiffness of conduit arteries and the timing and magnitude of arterial pressure wave reflections, therefore, providing a more comprehensive assessment of vascular function. ¹⁹

Central aortic BP (which closely approximates the BP experienced by the heart, brain, and kidney) as compared with peripheral BP measurement better quantifies the arterial pressure that these organs undergo. ²⁰ At the level of the heart, aortic systolic BP (SBP) reflects more accurately the load on the left ventricle, while the aortic diastolic BP (DBP) determines perfusion of the coronary arteries.

The physiology and generation of these central vascular measures of arterial stiffness are now better understood with data from clinical studies. With each contraction of the left ventricle during systole, a pulse wave is generated and propagated forward to the peripheral arterial system. The BP wave is then reflected back to the heart from branching points of peripheral arteries. The final pressure waveform at the aortic root is the summation of the forward traveling wave and the reflected wave. In healthy individuals with normal arteries, the reflected wave merges with the forward traveling wave in diastole and augments coronary blood flow. In patients with stiff vascular arteries due to aging or vascular comorbidities, the reflected wave returns faster and merges with the forward wave in systole. This results in an increase in left ventricular afterload and decreased coronary perfusion. ²⁰ AIx is the ratio of augmentation pressure (difference between the second and first systolic peak caused by early wave reflection) to central pulse pressure (Figure). The AIx is thus a composite parameter that represents the reflective properties of the peripheral distal arterial bed and the elastic properties of large arteries. ²¹ A higher AIx implies increased arterial stiffness and has been reported to be strongly associated with an increased risk of CVD. ²² , ²³

Figure.

 Central aortic waveform and augmentation index (AIx). (A) Forward waveform; (B) reflected waveform; (C) summation waveform as the result of early wave reflection in a patient with stiff arteries.

Given the above data that suggest significant potential benefits in achieving lower BP goals in CKD and improving CV outcomes, we sought to examine possible mechanisms that contribute to improved clinical outcomes in CKD patients with good BP control. We hypothesized that individuals with CKD who achieved the recommended lower BP goals (<130/80 mm Hg) have improved arterial stiffness, a factor recognized for its association with improved CV outcomes.

Methods

Study Population

This retrospective cross‐sectional study was approved by the institutional review board of the Cleveland Clinic Foundation. We reviewed the medical records of consecutive patients with CKD by history who had undergone standard BP measurement and applanation tonometry for the estimation of AIx at our institution’s nephrology outpatient clinic during the period of August 1, 2006 to September 31, 2007.

Clinical and demographic characteristics were reviewed including age, sex, race, weight, height, body mass index (BMI), serum creatinine, presence or absence of diabetes mellitus, dyslipidemia, or CVD (history of myocardial infarction, stroke, or peripheral arterial disease), smoking history, number and types of BP medications (angiotensin‐converting enzyme inhibitors [ACEIs] or angiotensin receptor blockers [ARBs], calcium channel blockers [CCBs], β‐blockers, diuretics, nitrates), and statin use.

Patients were classified into the following subgroups: (1) “goal BP group” with BP ≤130/80 mm Hg and (2) “non‐goal BP group” with BP ≥130/80 mm Hg.

Measurement of Arterial BP and AIx

With the patient in the sitting position after 5 minutes of resting, brachial BP (SBP and DBP) was obtained using an automated sphygmomanometer machine. Two consecutive BP measurements were obtained and the second measurement was used for applanation tonometry.

A commercially available computer system linked to the arterial tonometer (SphygmoCor; AtCor Medical Pty Ltd, West Ryde, Australia) was used to measure central BP indices. Brachial artery BP was measured using the automated device. Radial arterial pressure waveforms were obtained by applanation tonometry using a solid‐state high‐fidelity external Millar transducer. Central arterial waveforms and pressures were calculated using a generalized transfer function. These data were then processed using a computer to yield a synthesized central arterial waveform with central aortic SBP and DBP. The pulse wave morphology was recorded within the attached notebook computer (the computer, software, and Doppler tonometer probe form the SphygmoCor apparatus). Since AIx estimates are affected by the heart rate, standardization of the measure was performed to a heart rate of 75 beats per minute (AIx‐75) by the SphygmoCor software. ²⁴

Statistical Analysis

All values were expressed as median (quartiles), mean ± standard deviation, or frequency (percentage). Continuous variables were compared between the goal BP and non‐goal BP groups using Wilcoxon rank sum test. Categoric variables were compared using chi‐square or Fisher exact tests. Multiple linear regression analysis of AIx‐75 against age, sex, height, and BP goal <130/80 mm Hg, and against age, sex, height, and SBP were performed. All statistical tests utilized a significance level of α=.05. Data analysis was performed using SAS 9.1 software (SAS Institute, Cary, NC).

Results

The sample included 172 patients with CKD and 101 (59%) were male and 119 (70%) were Caucasian. Median age was 57 years. All patients were taking antihypertensive medications and 124 (72%) were taking ≥2 medications. Thirty‐three percent had achieved a BP goal of <130/80 mm Hg. Median estimated glomerular filtration rate (eGFR) was 41 mL/min/1.73 m² and median SBP was 128 mm Hg (Table I).

Table I.

 Baseline Characteristics

	Overall	BP Goal <130/80 mm Hg (n=56)	BP Goal ≥130/80 mm Hg (n=116)	P Value a
	Median (Quartiles)	Median (Quartiles)	Median (Quartiles)
Age, y	57 (45, 67)	58 (47, 68)	57 (42, 68)	.52
Male, No. (%)	101 (59%)	34 (61%)	67 (58%)	.71b
White race, No. (%)	119 (70%)	40 (73%)	79 (68%)	.57c
Weight, kg	83.7 (69.2, 97.6)	84.1 (69.7, 97.6)	83.3 (68.9, 97.8)	.85
Height, cm	170 (160, 178)	168 (160, 178)	170 (160, 178)	.70
Body mass index, kg/m2	29.4 (24.8, 33.2)	29.1 (25.8, 33.6)	29.5 (24.5, 33.1)	.68
Serum creatinine, mg/dL	1.7 (1.2, 2.9)	1.8 (1.3, 2.9)	1.7 (1.2, 2.9)	.76
eGFR, mL/min/1.73 m2	41 (24, 65)	38.6 (22.3, 60.4)	41.9 (24.1, 70.0)	.54
SBP, mm Hg	128 (117, 146)	113 (105, 120)	142 (127, 162)	<.001
DBP, mm Hg	78 (70, 86)	68 (61, 74)	82 (76, 90)	<.001
PP, mm Hg	52 (40, 70)	45 (36, 53)	60 (42, 76)	<.001
DM, No. (%)	45 (26%)	17 (30%)	28 (24%)	.38b
Dyslipidemia, No. (%)	111 (65%)	37 (66%)	74 (64%)	.77b
Smoking, No. (%)	20 (12%)	4 (7%)	16 (14%)	.20b
≥2 BP medications, No. (%)	124 (72%)	37 (66%)	87 (75%)	.22b
≥3 BP medications, No. (%)	67 (39%)	16 (29%)	51 (44%)	.052b

Abbreviations: BP, blood pressure; DBP, diastolic BP; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; PP, pulse pressure; SBP, systolic BP. ^aWilcoxon rank sum test. ^bChi‐square test. ^cFisher exact test.

Baseline characteristics (age and sex distribution, height, BMI, eGFR) were similar in both the goal BP group as well as the non‐goal BP group (Table I). There was a trend toward more aggressive antihypertensive therapy (more than 3 antihypertensive medications) in those with BP ≥130/80 mm Hg (44% vs 29%; P=.052).

The AIx‐75 in patients with a BP goal <130/80 mm Hg was significantly lower than that of patients who did not achieve this goal (median AIx‐75 was 19% [quartiles 12, 27] vs 23% [quartiles 15, 29]; P=.043) (Table II). AIx‐75 in the goal BP group remained significantly lower than in the non‐goal BP group after adjusting for age, sex, and height (mean effect on AIx‐75 was −3.3% [95% confidence interval, −6.1 to −0.4]; P=.027) (Table III). A higher AIx‐75 remained associated with a higher SBP after adjusting for age, sex, and height (Table IV).

Table II.

 Blood Pressure and Augmentation Index by BP Groups

	Goal BP Group Median (Quartiles)	Non‐Goal BP Group Median (Quartiles)	P Value a
SBP, mm Hg	113 (105, 120)	142 (128, 162)	<.001
DBP, mm Hg	68 (61, 74)	82 (76, 90)	<.001
PP, mm Hg	45 (36, 53)	60 (42, 76)	<.001
Central SBP, mm Hg	101 (94, 109)	129 (116, 143)	<.001
Central DBP, mm Hg	69 (62, 75)	83 (78, 91)	<.001
Central PP, mm Hg	34 (27, 39)	44 (31, 61)	<.001
Augmentation index, %	19 (12, 27)	23 (15, 29)	.043

Abbreviations: BP, blood pressure; DBP, diastolic blood pressure; PP, pulse pressure; SBP, systolic BP. ^aWilcoxon rank sum test.

Table III.

 Multivariable Regression of AIx on Age, Sex, Height, and BP Goals

	Average Effect	95% Confidence Interval		P Value
Age per 10 y	1.3	0.5	2.2	.002
Male vs female	−6.7	−10.2	−3.1	<.001
Height per 10 cm	−1.8	−3.4	−0.2	.028
BP <130/80 mm Hg	−3.3	−6.1	−0.4	.027

Abbreviations: AIx, augmentation index; BP, blood pressure.

Table IV.

 Multivariable Regression of AIx on Age, Sex, Height, and SBP

	Average Effect	95% Confidence Interval		P Value
Age per 10 y	1.1	0.3	1.9	.010
Male vs female	−6.9	−10.4	−3.4	<.001
Height per 10 cm	−1.6	−3.2	−0.003	.050
SBP per 10 mm Hg	0.8	0.3	1.4	.005

Abbreviations: AIx, augmentation index; SBP, systolic blood pressure.

Discussion

In the Multiple Risk Factor Intervention Trial (MRFIT), ²⁵ elevation in BP >120/80 mm Hg was associated with an increased incidence of ESRD. In patients with CKD, elevated BP is also known to exacerbate complications of diabetes mellitus, thus exposing these patients to higher risks of CV events. In addition, several trials have demonstrated that aggressive BP control with a goal of <130/80 mm Hg in CKD patients is associated with improved CVD outcomes. ²⁶ There are also preliminary data which demonstrate that improvement of arterial stiffness is associated with improved CVD and survival in ESRD. However, a paucity of data on the relationship of aggressive BP control and arterial stiffness in the CKD population exists. We proposed that aggressive BP control and management are associated with lowering of arterial stiffness, which could be a mechanism possibly contributing to improved CVD outcomes in patients with CKD. In this study cohort comprised of predominantly non‐ESRD patients with CKD (median eGFR of 41 mL/min/1.73 m²), aggressive BP management to a goal BP of <130/80 mm Hg was associated with a significantly lower AIx indicative of better vascular health. This association remained after adjusting for variables including age, sex, and height.

Several studies have investigated the effects of various antihypertensive medications on AIx. The majority of these trials have demonstrated that ACEIs, ARBs, and CCBs improve the AIx as compared with atenolol, a β‐blocker. ²⁷ In the Preterax in Regression of Arterial Stiffness in a Controlled Double‐Blind Study (REASON), ²⁸ 471 patients with hypertension were treated with perindopril/indapamide or atenolol and followed for 12 months. There was a 3% reduction in AIx (P<.01) in the perindopril/indapamide arm from baseline. In the Conduit Artery Function Evaluation (CAFÉ) substudy, the CCB arm demonstrated a net difference in AIx of 6.5% (P<.0001) when compared with patients taking an atenolol‐based regimen. Improved measures of arterial stiffness including AIx have been proposed as the explanation for better clinical outcomes in the CCB arm of the main Anglo‐Scandinavian Cardiac Outcomes Trial (ASCOT). ¹²

The optimal level of BP has been debated since it became known that effective treatment and management of hypertension offers protection against CV events like coronary artery disease and stroke and slowing progression of renal disease. A meta‐analysis of 65 randomized controlled trials that included more than a million patients demonstrated that an increased risk for CV events begins at a BP level of 115/75 mm Hg and that the relationship between BP and CV events was continuous above this level. ²⁹ Although peripheral BP has been shown to correlate with CVD, it appears to be less sensitive than central vascular measures in predicting CVD or vascular health. This is because central BP is a surrogate for arterial stiffness and more accurately reflects cardiac afterload, coronary perfusion, and cerebral perfusion. ¹¹ Further, de Luca and colleagues ³⁰ showed that regression of left ventricular hypertrophy correlated better with central SBP than peripheral SBP.

Peripheral BP measurement that is routinely obtained in the office is a “snapshot” of vascular health and varies with physical activity and sympathetic drive that affects heart rate, cardiac output, and peripheral vascular resistance. In contrast, central vascular measures additionally take into account the stiffness of conduit arteries and the timing and magnitude of pulse wave reflections, which, in turn, are dependent on chronic vascular disease processes and vascular health. Thus, central vascular measures obtained under standardized conditions are less likely to be altered by acute changes in physical activity and sympathetic output.

The association between BP and arterial stiffness has been suggested to be akin to a 2‐way street where high BP can result in increased arterial stiffness and vice versa. ³¹ Benetos and colleagues ³² clearly demonstrated that elevated BP is associated with increased arterial stiffness over time, and Liao and colleagues ³³ reported that increased arterial stiffness preceded hypertension. Despite the strong association between elevated BP and arterial stiffness, limited data exist on the relationship between BP control, arterial stiffness, and CVD outcomes. Results from the CAFÉ trial, a substudy of the ASCOT trial, suggested that the better CV outcomes noted in the amlodipine‐based regimen group as compared with the atenolol‐based treatment group were likely related to improvement in central BP and vascular indices. ¹² Data supporting a desired office BP goal of <130/80 mm Hg in patients with CKD are derived from studies that showed improved clinical outcomes with this BP goal of <130/80 mm Hg as compared with an SBP goal of 140 mm Hg. ²⁶ A recent study by Howard and colleagues ³⁴ of a cohort of Native Americans with diabetes mellitus also using an office SBP goal of ≤115 mm Hg showed that improved left ventricular hypertrophy was attributable to the lower BP goal.

Hypertension management studies in the past 4 decades have consistently emphasized and reiterated that good BP management strategies require customized and individualized therapy. Such an approach would require risk profiling of patients, definition of optimal BP goals, and institution of multilevel antihypertensive therapy including nonpharmacologic as well as pharmacologic measures. Despite the current recommendation for targeting BP goal of <130/80 mm Hg in the CKD patient, the percentage of CKD patients achieving this BP goal is <15%. ³⁵ Our study results suggest that lowering arterial stiffness is the possible mechanism for improved CVD outcome in CKD patients who achieved a lower BP goal of <130/80 mm Hg. The finding of lower AIx may, in turn, reflect improved vascular health. We postulate that this is likely due to effective BP management of these CKD patients over time that results in improved arterial stiffness and vascular health.

Limitations

A limitation of this study is its retrospective nature. Data collection may not have included all of the factors that could have potentially contributed to and accounted for better vascular indices in this study cohort. It is also possible that CKD patients who achieved the lower BP goal had an inherently better vascular profile and therefore BP was well controlled, although baseline characteristics were similar as given in Table I. Given the retrospective nature of the study, results are limited to the association relationship between BP goal and arterial stiffness measures.

Conclusions

Our study findings suggest that lowering BP to the recommended goal of <130/80 mm Hg in patients with CKD is associated with improved AIx, which reflects reduced arterial stiffness. Improved arterial stiffness is a potential mechanism for a reduction in CV morbidity and mortality in patients with CKD. Future studies of a prospective design are needed to evaluate the response of arterial stiffness measures to aggressive BP treatment and achievement of lower BP goals in patients with CKD.

Acknowledgment:  The authors thank Sandra Bronoff for her assistance and editorial support in this paper.