Noninvasive central pulse pressure is an independent determinant of renal function

Abstract

The purpose of this study was to investigate the prognostic properties of different BP measurements for renal function decrement and early chronic kidney disease (CKD) in community‐dwelling populations with normal renal function at baseline. A total of 1426 participants were included and followed for a median of 4.8 years (interquartile range, 4.5‐5.2), and central hemodynamic profile and estimated glomerular filtration rate (eGFR) were evaluated. One main outcome was the rapid eGFR decline defined as a decline in eGFR of greater than 3 mL/min per 1.73 m² per year; the other was the new incidence of CKD. At the end of follow‐up, mean eGFR decreased from 93.39 ± 13.46 mL/min per 1.73 m² to 85.72 ± 14.81 mL/min per 1.73 m², and the incidence of rapid eGFR decline and CKD were 20.7% and 5.6%, respectively. In multivariate linear regression analysis, central pulse pressure (PP), age, fasting blood glucose, and concentration of homocysteine were independent determinants of the change in renal function. Not only in the prediction of rapid eGFR decline but also in the incident of CKD, baseline central PP was the only BP component that consistently independently associated with both outcomes after adjustment for various confounders. When compared with subjects in the lowest quartile of central PP, those in the highest quartile demonstrated a significantly increased risk of CKD (hazard ratio [HR], 1.57; 95% confidence interval [CI], 1.08‐2.96; P = .006). The study showed that central PP emerged as an independent predictor of the decline in renal function.

1. INTRODUCTION

Chronic kidney disease is a major public health concern due to the increasing prevalence1 and associated cardiovascular (CV) morbidity and mortality. Kidney dysfunction is especially prevalent among elderly, affecting 10%‐15% of population older than 70 year of age.2, 3 Efforts to reduce the burden of kidney disease require an identifying of kidney disease modifiable risk factors. Hypertension is one of the most important aggravating factors and enhances the age‐associated decline in renal function.4 In a longitudinal study conducted in normotensive subjects and hypertensive participants who remained untreated, after a follow‐up period of 5.2 years, blood pressure emerged as the main determinant of the decline in the glomerular filtration rate and effective renal plasma flow measured by isotopic urinary clearance techniques in the whole population.5 In contrast, the association between BP and early decline in kidney function is not clear. In addition, the component of BP most responsible for kidney disease is unknown.

Subclinical vascular dysfunction and arterial stiffening are the most characteristic features that are already observed in those with the early stages of CKD. With aging, the arterial stiffening is more pronounced in the aorta and central arteries than in peripheral conduit arteries.6 Central BP can now be assessed noninvasively based on radial tonometry. Given that central BP (ie, the BP in the ascending aorta) creates the effective perfusion pressure of the most vital organs—the brain, heart, and the kidneys—greater precision in hypertension management may be achieved and central BP may potentially be a better predictor of clinical outcome than brachial BP.7, 8, 9 However, the kidneys are distant from central aorta pressure, and whether this superiority of central over peripheral BP at predicting clinical outcomes is also true for the loss of renal function is not well established.

We hypothesize that central BP is the main determinant of renal function decrement and can independently predict future kidney disease progression more strongly than peripheral BP. Thus, we aimed to prospectively investigate the associations of different BP measures with the risk of renal function aggravation in an apparently healthy Chinese community–dwelling cohort with normal renal function (estimated glomerular filtration rate [eGFR] ≥60 mL/min per 1.73 m²) at baseline.

2. METHODS

2.1. Study population

This current study reports the results of a community‐based cohort study of people living in the Pingguoyuan communities of Shijingshan district in Beijing, China. Briefly, between September 2007 and January 2009, a total of 1631 without clinically apparent cardiovascular disease (CVD) participants were initially recruited for cross‐sectional analysis of CVD risk factors. In this enrolled population, we further excluded 30 participants, who already had CKD at baseline. After the initial assessment, enrolled subjects were contacted every 2 years for follow‐up, and the last follow‐up visits were conducted through September 30, 2013, and the median follow‐up was 4.8 (interquartile range 4.5‐5.2) years. During the period between the initiation of the study and the follow‐up, 52 subjects died (cause of death consists of 23 cardio‐cerebrovascular diseases, 13 pulmonary diseases, 9 malignant tumor, 1 hematopathy and 6 were unknown), and 175 participants were lost to follow‐up and were excluded from the analysis. No differences other that baseline risk factors were noted in those who completed baseline and follow‐up assessments. Complete follow‐up data were obtained from 1426 participants (return rate, 89.1%). The study was approved by the ethics committee of the Chinese People's Liberation Army General Hospital, and all participants gave written informed consent at time of enrollment.

2.2. Data collection

2.2.1. Clinical data

Baseline data for these analyses were collected using self‐reported standardized questionnaires that included demographics, life‐style information, medical history, and medication use. The medication use was validated by examination of pills and prescriptions brought to the clinic for that purpose. Anthropometrics were evaluated by trained medical doctors, and the body mass index (BMI) was derived from height and weight measured with participants wearing light clothing and no shoes. Blood pressures were measured manually using a mercury sphygmomanometer (Yuwell medical equipment & supply Co., Ltd) and a standard‐sized cuff by trained personnel. Two or more measurements separated by at least 5 minutes were taken from the right arm after 5 minutes of rest in a sitting position, and the average of at least 2 readings were used for further analysis. Brachial PP was calculated as (SBP − DBP), and brachial mean blood pressure was calculated as (DBP + [PP/3]).

2.2.2. Biomarker assays

Baseline blood samples were collected from participants between 8 AM and 10 AM after a fast of at least 12 hours at enrollment. Follow‐up measures were performed on blood samples collected 4‐5.5 years later. Serum or plasma samples were separated within 30 minutes of collection and were frozen at −80°C until analysis in a central laboratory were performed. Serum creatinine was measured using an enzymatic method by the Roche enzymatic assays (Roche Diagnostics GmbH) on a Hitachi 7600 automatic analyzer (Hitachi). Concentrations of glucose (fasting and 120 minutes after an oral glucose load), total cholesterol (TC), triglycerides (TGs), high‐density lipoprotein cholesterol (HDL‐C), low‐density lipoprotein cholesterol (LDL‐C), uric acid, and homocysteine levels were measured by routine laboratory analysis on a Roche autoanalyzer. All testing was performed by well‐trained personnel blinded to clinical data in the Department of Biochemistry of Chinese PLA General Hospital.

2.2.3. Assessment of central pulse wave and arterial stiffness

Measurement of arterial properties was conducted in the morning, in a quiet environment, at a stable temperature. All subjects were asked to abstain from caffeine, smoking, alcohol, and taking vasoactive medication for at least 12 hours before this assessment. These parameters were determined on one occasion both at baseline and at the end of the study.

Central pulse wave analysis

After brachial BP and anthropometric measurements, central arterial BP and wave reflection were assessed sitting position after ≥5 minutes of rest using the SphygmoCor pulse wave analysis system (AtCor Medical) via the right radial artery at the same sites. All of the operators were trained and certified to perform the central pressure measurements using radial artery tonometry. The probe was placed at the site of the strongest radial artery pulse to record a stable pulse wave. After 20 sequential waveforms were captured, a validated, generalized transfer function was used to generate the corresponding central aortic pressure waveform, using a Food and Drug Administration (FDA)‐approved high‐fidelity transducer. The central BPs were automatically calculated, and also, we obtained parameters about wave reflection (eg, augmentation index [AIx]).

Arterial stiffness

After participants rested in the supine position for 5‐10 minutes, pulse wave velocity (PWV) was determined using a Complior SP device (Artech Medical), which allows online pulse wave recording and automatic calculation of PWV. The details of the method and mechanism have been described and validated previously.10 Carotid‐femoral PWV is a well‐established index of aortic arterial stiffness.11

2.3. Definition of variables

The eGFR was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation, details of the computing equation have been published elsewhere.12 Changes in eGFR over time were calculated as the difference between the eGFR at follow‐up and baseline, and positive values indicate a decline of eGFR from baseline to the end of follow‐up.

One main outcome of interest was the rapid eGFR decline, defined as a decline in eGFR of greater than 3 mL/min per 1.73 m² per year; the other was the new incidence of CKD, defined as a decrease in eGFR of 20% or more and to a level of below 60 mL/min per 1.73 m² at the end of follow‐up.

2.4. Statistical analysis

Continuous variables are reported as mean ± standard deviation (SD) or median (interquartile range) and categorical variables as count and percentages; homocysteine level and other biomarkers were normalized by natural logarithm transformation, as necessary. Differences in the clinical characteristics and arterial properties between baseline and a median follow‐up period of 4.8 years were analyzed using a t test for continuous variables and a chi‐square analysis for categorical variables. First, multivariate linear regression models were used to evaluate the independent determinants for change in renal function and changes in eGFR as a continuous variable.

In addition, to better understand the effects of BP indices on the occurrence of rapid eGFR decline, logistic regression models were used. Odds ratios (ORs) and 95% confidence intervals (CIs) were computed across quartiles of BP variables, using the lowest quartile as the referent group. We conducted 3 types of models for each analysis: an unadjusted model (Model 1); a model adjusted for age, sex, current smoking, BMI, brachial mean arterial pressure, fasting blood glucose, HDL‐C, LDL‐C, homocysteine, eGFR at baseline, the presence of hypertension or diabetes mellitus, and antihypertensive medications use (Model 2); and Model 3 was adjusted for model 2 plus central SBP and carotid‐femoral PWV.

Furthermore, multivariable Cox proportional hazards regression models were used to determine associations of baseline BP categories with incident CKD. Cox regression models were adjusted for the relevant variables as described in the logistic regression models. The adjusted hazard ratios (HRs) are reported with 95% CIs, and values of P < .05 (2‐tailed) were considered statistically significant.

All data were analyzed using the SPSS statistical package software (version 17.0; SPSS Inc). A two‐sided value of P < .05 was considered statistically significant.

3. RESULTS

3.1. Participant characteristics at baseline and follow‐up

Characteristics of the population at baseline and after a median follow‐up period of 4.8 years (interquartile range, 4.5‐5.2) are summarized in Table 1. Among 1426 participants at enrollment, the mean (±SD) age was 61.3 ± 11.3 years and 57.4% were women. Hypertension was present in 45.8% of all participants, whereas 18.5% had diabetes mellitus. Mean baseline central SBP and central PP were 119.49 ± 17.48 mm Hg and 42.56 ± 13.37 mm Hg, respectively; at the end of follow‐up, some increase in central PP and decrease in central DBP were observed.

Table 1.

Population characteristics at baseline and after a median follow‐up period of 4.8 y

Variable	Baseline	Follow‐up	P value
Age (y)	61.3 ± 11.3	66.2 ± 11.4	<.001
Male, no. (%)	607 (42.57)
BMI (kg/m2)	25.49 ± 3.37	25.66 ± 5.45	.374
Current smoking, no. (%)	260 (18.2)	244 (17.1)	.132
Current alcohol use, no. (%)	278 (19.5)	267 (18.7)	.45
Brachial SBP (mm Hg)	128.61 ± 17.64	130.96 ± 17.74	<.001
Brachial DBP (mm Hg)	77.02 ± 10.21	75.06 ± 10.58	<.001
Brachial PP (mm Hg)	51.56 ± 14.06	55.99 ± 15.17	<.001
Resting heart rate (beats/min)	75.12 ± 9.89	75.12 ± 10.99	1.0
Fasting blood glucose (mmol/L)	5.37 ± 1.57	5.62 ± 1.61	<.001
postprandial blood glucose (mmol/L)	7.40 ± 3.79	7.79 ± 3.34	.005
Total cholesterol (mmol/L)	5.02 ± 0.93	5.19 ± 1.07	<.001
HDL‐cholesterol (mmol/L)	1.39 ± 0.36	1.43 ± 0.57	.024
LDL‐cholesterol (mmol/L)	2.91 ± 0.71	3.19 ± 1.04	<.001
Triglycerides (mmol/L)	1.81 ± 1.25	1.52 ± 1.01	<.001
Homocysteine (μmol/L)	16.60 (13.50, 20.80)	14.90 (12.20, 18.50)	.012
Serum creatinine (μmol/L)	66.78 ± 15.70	73.66 ± 16.10	<.001
eGFR (mL/min per 1.73 m2)	93.39 ± 13.46	85.72 ± 14.81	<.001
Central SBP (mm Hg)	119.49 ± 17.48	120.04 ± 16.25	.397
Central DBP (mm Hg)	76.93 ± 11.28	75.57 ± 10.75	.001
Central PP (mm Hg)	42.56 ± 13.37	44.54 ± 14.12	<.001
Carotid‐femoral PWV(m/s)	11.11 ± 2.82	11.75 ± 2.91	<.001
Aortic augmentation index (%)	25.82 ± 9.87	26.20 ± 9.93	.323
Antihypertensive use, no. (%)	460 (32.25)	575 (40.32)	<.01
Hypoglycemic use, no. (%)	154 (10.80)	210 (14.73)	<.01
Lipid‐lowering use, no. (%)	224 (15.71)	229 (16.06)	.618
Antiplatelet use, no. (%)	352 (24.68)	381 (26.72)	.153

Values are reported as no. (%), mean ± SD, or median (IQR).

Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; PP, pulse pressure; PWV, pulse wave velocity; SBP, systolic blood pressure.

During the observation period, the eGFR decreased by 8.2%, corresponding to a yearly change of 1.60 mL/min per 1.73 m² per year, 80 participants experienced new‐onset CKD. The incidence of rapid eGFR decline, and CKD were 20.7% and 5.6%, respectively.

3.2. Independent determinants of the change in renal function

To determine independent predictors of change in renal function, multivariate linear regression analysis was used and changes in eGFR as a continuous variable. As shown in Table 2, the baseline central PP was the sole component of BPs that correlated to the decrease in eGFR during follow‐up. Other parameters entered in the model were age, fasting blood glucose, and concentration of homocysteine. No correlation between the change in eGFR and the concomitant changes in other BP components was observed.

Table 2.

Multivariate linear analysis between the change in eGFR and various baseline parameters

Parameter	Standardized β	t	P value
Age (y)	0.145	4.590	<.001
Central PP (mm Hg)	0.130	3.971	<.001
Fasting blood glucose (mmol/L)	0.086	3.130	.002
Ln homocysteine	0.081	2.828	.005

β and P values are shown only when P < .05.

Abbreviations: Ln, natural logarithm‐transformed; PP, pulse pressure; β, regression coefficient.

3.3. Rapid renal function decline by initial blood pressure level

Table 3 displays the results of multivariate logistic regression for effects of baseline SBP and PP on rapid eGFR decline. Here, the population was categorized into quartiles by baseline BPs. However, when adjusting for various baseline parameters, only the association between central PP and a rapid decline in eGFR remained strong and graded. Compared to the lowest category, the multivariable‐adjusted ORs (95% CI) for rapid eGFR decline were 2.06 (1.34‐3.16) for the highest category of central PP (>50 mm Hg) and 1.93 (1.20‐3.08) for the third category of central PP (41‐50 mm Hg).

Table 3.

Association of baseline brachial and central BPs with rapid renal function decline

Variable	Model 1		Model 2		Model 3
	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P
Brachial SBP
≤116	Reference		Reference		Reference
117‐127	1.13 (0.78‐1.64)	.517	1.13 (0.75‐1.70)	.554	1.12 (0.71‐1.78)	.617
128‐140	1.13 (0.78‐1.64)	.517	0.94 (0.62‐1.42)	.771	0.93 (0.54‐1.61)	.800
>140	1.47 (1.02‐2.12)	.038	1.27 (0.83‐1.92)	.269	1.25 (0.61‐2.56)	.549
P for trend			0.446		0.565
Brachial PP
≤41	Reference		Reference		Reference
42‐49	0.86 (0.58‐1.26)	.425	0.88 (0.59‐1.31)	.823	0.86 (0.57‐1.28)	.448
50‐60	1.23 (0.86‐1.77)	.266	1.03 (0.70‐1.53)	.872	0.99 (0.65‐1.50)	.955
>60	1.32 (0.92‐1.88)	.133	1.05 (0.69‐1.59)	.822	0.98 (0.61‐1.57)	.921
P for trend			0.822		0.858
Central SBP
≤108	Reference		Reference		Reference
109‐118	1.04 (0.71‐1.54)	.834	0.93 (0.62‐1.39)	.727	0.95 (0.62‐1.45)	.801
119‐130	1.42 (0.99‐2.03)	.058	1.23 (0.84‐1.80)	.289	1.28 (0.83‐1.96)	.267
>130	1.55 (1.07‐2.23)	.019	1.30 (0.87‐1.94)	.198	1.37 (0.82‐2.27)	.227
P for trend			0.295		0.348
Central PP
≤33	Reference		Reference		Reference
34‐40	1.38 (0.92‐2.08)	.123	1.24 (0.79‐1.95)	.342	1.25 (0.81‐1.94)	.320
41‐50	2.24 (1.52‐3.29)	<.001	1.82 (1.14‐2.91)	.012	1.93 (1.20‐3.08)	.006
>50	2.45 (1.66‐3.63)	<.001	2.02 (1.31‐3.10)	.001	2.06 (1.34‐3.16)	.001
P for trend			0.005		0.003

Models are defined as follows: Model 1 is unadjusted; Model 2 is adjusted for age, sex, current smoking, body mass index, brachial mean arterial pressure, fasting blood glucose, HDL‐ cholesterol, LDL‐cholesterol, homocysteine, eGFR at baseline, the presence of hypertension or diabetes mellitus, and antihypertensive medications use; Model 3 is additionally adjusted for central systolic blood pressure and carotid‐femoral pulse wave velocity.

Abbreviations: CI, confidence interval; OR, odds ratio; PP, pulse pressure; SBP, systolic blood pressure.

3.4. Blood pressure categories and risk of chronic kidney disease

The estimated hazards ratio (95% CI) of incident CKD during follow‐up according to BP categories is shown in Table 4. The risk of CKD occurrence was associated with both brachial and central BPs in crude model, but the relationship was not statistically significant after adjusting for various confounders except for central PP. When compared with subjects in the lowest quartile of central PP, those in the highest quartile (without those in the third quartile) demonstrated a significantly increased risk of CKD (HR, 1.57; 95% CI, 1.08‐2.96; P = .006).

Table 4.

Multivariate Cox regression analysis of the determinants of new occurrence of chronic kidney disease

Variable	Model 1		Model 2		Model 3
	HR (95% CI)	P	HR (95% CI)	P	HR (95% CI)	P
Brachial SBP
≤116	Reference		Reference		Reference
117‐127	2.79 (1.29‐6.07)	.009	1.32 (0.37‐4.66)	.665	1.08 (0.49‐2.37)	.844
128‐140	2.48 (1.13‐5.45)	.024	1.11 (0.39‐3.13)	.850	0.95 (0.42‐2.13)	.891
>140	3.49 (1.65‐7.40)	.001	1.49 (0.68‐3.28)	.315	1.37 (0.60‐2.97)	.499
P for trend			0.473		0.667
Brachial PP
≤41	Reference		Reference		Reference
42‐49	1.11 (0.45‐2.74)	.516	1.07 (0.44‐2.24)	.823	1.01 (0.48‐2.08)	.991
50‐60	2.14 (1.06‐4.35)	.031	1.36 (0.92‐2.83)	.174	1.16 (0.71‐2.27)	.357
>60	2.56 (1.27‐5.15)	<.001	1.52 (0.98‐3.37)	.056	1.25 (0.73‐2.46)	.132
P for trend			0.262		0.468
Central SBP
≤108	Reference		Reference		Reference
109‐118	2.02 (1.17‐3.48)	.011	1.46 (0.82‐2.61)	.199	1.19 (0.68‐2.07)	.548
119‐130	2.21 (1.31‐3.73)	.003	1.49 (0.83‐2.69)	.180	1.28 (0.74‐2.23)	.379
>130	2.84 (1.69‐4.74)	<.001	1.56 (1.04‐2.95)	.025	1.33 (0.75‐2.35)	.129
P for trend			0.073		0.12
Central PP
≤33	Reference		Reference		Reference
34‐40	1.87 (0.87‐4.02)	.107	1.69 (0.80‐3.59)	.169	1.26 (0.67‐2.47)	.480
41‐50	2.52 (1.16‐5.46)	.009	2.12 (1.02‐4.42)	.045	1.35 (0.91‐2.71)	.144
>50	3.34 (1.40‐7.94)	<.001	2.36 (1.13‐4.94)	<.001	1.57 (1.08‐2.96)	.006
P for trend			<0.001		0.017

Multivariable Cox regression models were adjusted for the covariates in models 1, 2, and 3 (the same as described in Table 3).

Abbreviations: CI, confidence interval; HR, hazard ratio; PP, pulse pressure; SBP, systolic blood pressure.

Similar associations were observed in stratified analyses of hypertensive patients (Table S1) and normotensive participants (Table S2). However, the relationship between central PP and the risk of new incident CKD was statistically significant only in crude model but not again after adjusting for various confounders because of the low incidence of CKD in each group (Table S3 and Table S4).

4. DISCUSSION

The main objective of the present study was to evaluate which BP indices is valuable in the prediction of renal function aggravation and incident CKD in a community‐based study of subjects with initial eGFR >60 mL/min per1.73 m² and without clinically evident CVD at entry. Central PP (pulsatile component of central BP) indicated a clear association with both renal function decline and the occurrence of CKD in this prospective cohort study. Also, these risk prediction persisted even after adjustment for traditional cardiovascular risk factors as well as adjustment for other BP parameters. Therefore, central PP may be an useful tool in clinical for prediction of progressive renal disease.

Chronic kidney disease is a worldwide health burden due to the high prevalence and associated cardiovascular morbidity and mortality.13 Hypertension, diabetes, and proteinuria are well‐recognized risk factors for progressive kidney function loss.14 In a longitudinal study conducted in essential hypertensive patients who initially untreated, within a mean follow‐up period of 14 years, blood pressure emerged as the main determinant of the decline in renal function.15 Furthermore, hypertension is among the leading causes of end‐stage renal disease (ESRD) in Europe and the United States, second only to diabetes mellitus. However, uncertainty remains as to which BP measures may best predict the risk of early stage of milder renal dysfunction.

To the best of our knowledge, few prospective studies have examined the relationship of central BPs with the decline of renal function or the risk of CKD. Whether central BP offers significant improvement in CKD risk assessment and stratification compared with brachial (peripheral) BP is debated. In the Framingham Heart Study, central PP (OR, 1.08; 95% CI, 0.71‐1.64; P = .71) was not associated with incident CKD during a 7‐ to 10‐year period in multivariable‐adjusted models.16 Briet et al17 also found that central PP was not associated with the progression of CKD, but the HR for the risk of ESRD is 1.24 (95% CI, 1.03‐1.49; P = .02) for a 10 mm Hg increase of carotid PP in 180 CKD patients (mean measured eGFR, 32 mL/min per 1.73 m² at baseline) during a mean follow‐up of 3.1 years. In contrast, the greater impact of central PP compared with other measures of central pulse wave (including central SBP, augmentation pressure, and aortic PWV) in ESRD progression was reported in CRIC (Chronic Renal Insufficiency Cohort) study.18 In fact, in a Chinese community‐based population study with normal kidney function at baseline, Fan et al demonstrated that, compared with peripheral SBP, central SBP was a stronger predictor for early kidney function decline but not incident CKD.19 In this Chinese study, central BP was not measured by classically valid method and only central SBP can be obtained, so they failed to estimate whether there is a superiority of central PP compared with central SBP. Our observations using widely accepted renal function assessment method and the standard array of tonometry measures suggest that central PP is independently related to renal function decline and the development of CKD but not to brachial BPs.

Pulse pressure is a measure of pulsatile character, it is possible that PP captures more dynamic information through the cardiac cycle than does a PWV alone, which measures distance over time. It has been recognized that PP is a marker of vascular aging, generates direct cyclic stress on conduit vessels, and targets organs such as carotid arteries and kidneys.20 The Framingham Study focused attention toward PP as the best measure of CV risk, at least in older subjects.21 PP plays an important role in the prediction of CVD, CKD, and heart failure.8, 22, 23 From the physiological point of view, target organs, such as the heart and large arteries, are directly exposed to hemodynamic stress of central BP rather than peripheral. Furthermore, accumulating evidence has demonstrated that the measurement of large artery stiffness by central PP is more important than peripheral BP in the prediction of clinical events across different populations.24, 25, 26

The exact mechanisms underlying the relationship between elevated PP and kidney disease progression are not fully understood, and there are several potential explanations for this finding. One mechanism is that PP is a surrogate measure of arterial stiffness and reflects the pulsatile nature of the cardiac cycle, and wider PP could lead to increased cardiac afterload and decreased coronary perfusion. Along with central PP rises, the low resistance and impedance of afferent arterioles exposes the renal microcirculation to increasing pulsatile stress and a high hydrostatic pressure, leading to preglomerular vasoconstriction and increased resistance.27, 28 In addition, high PP damages intra‐renal autoregulation by decreasing baroreflex sensitivity and arterial wall elasticity, which could reduce renal blood flow and enhance vascular damage.29 Second, pulsatile pressure may result in deleterious chronic renal damage through altered myogenic tone, and Bidani et al30 have already showed that myogenic tone is altered by chronically increased in pulsatile pressure, thereby resulting in higher dissipation of pulsatile energy in the microcirculation and exacerbating subsequent glomerular damage. Finally, high aortic PP was found to correlate with an increase in beat‐to‐beat BP variability, 31 which may induce ischemia‐reperfusion type renal damage.32 Collectively, these mechanisms progressively induce to substantial damage and structural remodeling, with subsequent decrement of renal function.

Our study also found that older age, blood glucose and homocysteine were associated with renal function decline besides central PP. Older age is a well‐known risk factor for renal function deterioration. As with CVD risk, associations between various BP measures and risk of kidney dysfunction may also vary by age. In old age, renal function will be compromised as a result of progressive loss of glomeruli and decline in renal blood flow, especially in those with high BP.33 Diabetes is the leading cause of kidney disease in the developed world, the chronically elevated levels of blood glucose lead to the formation of advanced glycation end products with posterior hyperfiltration causing glomerular hypertrophy34; hyperglycemia with its gluco‐toxicity cause arterioles and microvascular hyaline degeneration and fibrinoid degeneration which seems to be triggering a vicious cycle and induces damage to kidney.35 In addition to the traditional risk factors, other factors are likely also important. Homocysteine may directly act on glomerular cells to induce glomerular dysfunction and consequent glomerular sclerosis, leading to renal failure. 36 Evidence is accumulating that serum homocysteine levels are a significant risk factor for accelerated renal function decline and for the development of CKD.37, 38 A post hoc analysis of the Framingham Offspring Study 39 suggested that adding homocysteine measurement to a model including several other covariates significantly improved its ability to predict CKD. Furthermore, our previous study has showed that homocysteine may interact with hypertension to produce synergistic effects on arterial stiffness.40 In this cohort, central PP remained independent prediction of research end points in the final model after additional adjustments for these confounding factors.

In the present study, central PP exhibits the best predictive potential for the prediction of renal function progression in the general population. Recent study additionally has demonstrated that even in patients with CKD,17 large arteries PP also independently associate with faster decline of renal function and progression to ESRD. Therefore, maintenance of the central PP within a rational range is a substaintial step in maintaining renal blood flow and glomerular filtration without inducing vascular damage of stress. The relationships between central BPs and the risk of new CKD were not presented in the subgroup analyses of hypertensive and normotensive participants separately. The potential reason may be the low incidence of CKD (overall incidence 5.61%) in each group. Further studies are needed to compare the usefulness of these parameters among different populations with differing clinical characteristics.

Our study also has several limitations. First, the study population was restricted to Chinese community‐based residents of Han origin with mean age of 61.3 years at baseline, and its generalizability to other demographic groups should be done with caution. Second, urinary albumin excretion rate, an important marker of kidney damage, was not measured during follow‐up; therefore, it could not be performed associated analysis. Third, we did not performed detection using cystatin C, a potentially more accurate measure of kidney function in the elderly than serum creatinine. So we cannot rule out the influence of lower muscle mass (the source of serum creatinine) on renal function evaluation, particularly in older people. Fourth, without detection of ultrasonography, renal biopsy, and so on, we have no information about type of underlying the CKD, but these diagnostic procedures are not considered feasible for a cohort study recruited from a general population. Furthermore, many participants received several antihypertensive medications. Though there was no significant difference in the percentages of each anti‐hypertensive agents used across the central PP quartiles (Table S5), the influence of these on aortic hemodynamics and renal function progression is complex and differ between drug classes, and it was very difficult to discriminate the role of individual drugs on these data.

In summary, in a prospective cohort of community‐dwelling population, we found that noninvasive measures of central PP were independently associated with various measures of change in renal function, including rapid renal function decline and incident CKD. The results suggest that central PP will help stratify subjects at risk for early kidney disease progression in apparently healthy general population. Further investigations, such as medications known to improve vascular pressure pulsatility, should be considered to evaluate their effect on kidney disease progression.

CONFLICT OF INTEREST

The authors declare that they have no conflict of interest.

Supporting information

 

Xiao W, Wen Y, Ye P, et al. Noninvasive central pulse pressure is an independent determinant of renal function. J Clin Hypertens. 2020;22:234–242. 10.1111/jch.13792