Stiffness index derived from digital volume pulse as a marker of target organ damage in untreated hypertension

Abstract

Objective. An index of large artery stiffness (SI_DVP) simply derived from the digital volume pulse (DVP) was developed recently. However, the role of the SI_DVP in untreated hypertensive patients was not well elucidated. Methods. We enrolled 124 untreated hypertensive patients (mean age 55.4±13.1 years, 57 men). The DVP was measured in right index finger by a photoplethysmography. The SI_DVP was formulated as body height divided by transition time from early systolic peak to the inflection point of reflection wave. Two functional indices of aortic compliance, stiffness index (SI) and distensibility (DI), were also used for measurement of aortic stiffness. Results. The SI_DVP was significantly correlated with blood urea nitrogen (BUN), and left ventricular mass index (LVMI). Patients with vascular diseases had higher level of SI_DVP (10.12±2.97 vs 8.45±1.78, p<0.001), SI (13.76±7.63 vs 10.87±8.88, p = 0.116), BUN (28.4±24.7 vs 14.5±4.6, p<0.001) and lower level of DI (1.34±0.88 vs 1.93±1.12, p = 0.010) than those without vascular diseases. By multivariate analysis, only the SI_DVP was significantly associated with vascular diseases (OR 1.39, 95% CI 1.06–1.82, p = 0.016). Conclusions. SI_DVP, SI and DI were significantly correlated with target organ damage in untreated hypertension. However, only the SI_DVP was independently associated with presence of vascular diseases. SI_DVP simply derived from the DVP can be used as a marker for risk stratification in untreated hypertensive patients.

• Aortic stiffness
• digital volume pulse
• hypertension
• stiffness index

Introduction

Increased stiffness of large arteries is associated with an increased cardiovascular risk in hypertension [1]. Pulse wave velocity (PWV) is now recognized as a standard method for measurement of aortic stiffness [2], [3]. An elevated aortic PWV was not only associated with increased cardiovascular risks in hypertension [3], but also could be a prognostic factor in end‐stage renal disease patients [4]. Aortic stiffness increases with age [5–8] and cardiovascular risk factors [9]. An index of large artery stiffness (SI_DVP) derived from the digital volume pulse (DVP) measured by photoplethysmography was well correlated with PWV and age [10–13]. Photoplethysmography can measure DVP easily by using transmission of infrared light through the finger pad [11], [14]. However, the role of the SI_DVP in hypertensive patients was not well elucidated. This study used SI_DVP derived from DVP and two aortic function indices derived from M‐mode echocardiography to investigate the clinical significances of these indexes in untreated hypertension and to compare their differences in the associated role for vascular complications.

Methods

Subjects

One hundred and twenty‐four consecutive patients (mean age 55.4±13.1 years, 57 men) with untreated essential hypertension from a hypertension clinic were included in this study. Hypertension was diagnosed if blood pressure more than 140/90 mmHg on two separate occasions. All patients received the appropriate evaluation to exclude secondary hypertension and did not take any anti‐hypertension agents before the study. All patients received detail examination for the presence of end organ damage and vascular diseases. Vascular diseases were diagnosed if patients had coronary artery disease (defined as patients with any history of angina pectoris, myocardial infarction, coronary revascularization or congestive heart failure), cerebral vascular diseases (defined as patients with any history of transient ischemic attack, ischemic stroke or cerebral hemorrhage) or renal disease (defined as plasma creatinine concentration >2.0 mg/dl).

Acquisition of the SI_DVP

The DVP was measured in right index finger by a photoplethysmography (Micro Medical, Gillingham, UK) transmitting infrared light at 940 nm. Frequency response of the photoplethysmography was flat to 10 Hz. Digital output from the photoplethysmography was recorded through an analogue‐to‐digital converter (12 bits, sampling frequency 100 Hz). The DVP waveforms were recorded for 10 s and SI_DVP was formulated automatically by computer as body height (m) divided by transition time (s) from the first systolic peak to the inflection point of reflection waveform (Figure 1). All measurements were made with the subject supine in a temperature‐controlled laboratory at 25±1°C. All subjects were allowed to acclimatize to this temperature for at least 10 min before recordings commenced [13].

Figure 1. Digital volume pulse (DVP) was recorded from a photoplethysmography. ΔT represented transitional time from first peak to second peak. Stiffness index derived from DVP (SI_DVP) was calculated from body height divided by Δt.

Determination of stiffness index (SI) and distensibility (DI)

All patients received trans‐thoracic echocardiography. The studies were performed by means of a commercially available ultrasound system (Hewlett‐Packard Sonos 1000, Andover, MA, USA) with a 2.5‐MHz phased array transducer. Two functional indices of aortic compliance, SI and DI, were derived from measurements of aortic diameters at 3 cm above aortic valve by M‐mode echocardiography and sphygmomanometric brachial artery blood pressure. The SI and DI were calculated from following formulas: SI = ln(systolic blood pressure/diastolic blood pressure)/(changes in aortic diameter/diastolic diameter); DI = 2(changes in aortic diameter)/(diastolic diameter)(pulse pressure) [15–17].

Determination of left ventricular mass index (LVMI)

We assessed left ventricular mass by using a standard method, which was previously reported [18]. Two skilled cardiologists performed two‐dimensional and M‐mode echocardiography, using a Hewlett Packard Sonos 1000 with a 2.5‐MHz transducer. A single observer who was unaware of the results of blood pressure measurements read the echocardiograms and videotape recordings. Measurements of end‐systolic and end‐diastolic left ventricular internal dimensions, end‐diastolic interventricular septal thickness, end‐diastolic posterior wall thickness, aortic root dimension, and left atrial dimension were made according to the American Society of Echocardiography guidelines [19]. Left ventricular mass was determined by the method of Devereux et al. [20], with LVMI calculated by dividing left ventricular mass by body surface area.

Assessment of transmitral Doppler flow velocity

The heart was scanned in the standard apical four‐chamber view so that the diameter of the mitral annulus was maximized. This imaging study was followed by Doppler interrogation of the mitral inflow velocities with the sample volume at the tip of mitral leaflets [21]. The following parameters were measured from the mitral flow velocity spectrum: peak early filling velocity, peak atrial velocity and early filling to atrial velocity ratio (E/A). E/A was used as an index of left ventricular diastolic function.

Statistical analysis

The relation between measurements of aortic stiffness and target organ damage was assessed using Pearson's correlation. Differences between patients with or without vascular diseases were compared with Student's t‐test, for continuous variables, or the chi‐square test, for categorical variables. Multiple logistic regression analysis was used for assessment of the most independent marker for vascular diseases. All data are presented as the mean±standard deviation (SD). A p‐value of less than 0.05 was considered statistically significant. All analysis was performed with the SPSS 10.0 for Windows (SPSS Institute, Chicago, IL, USA).

Results

The SI_DVP was significantly correlated with SI (r = 0.314, p<0.001) and DI (r = −0.340, p<0.001). Among risk factors, SI_DVP was significantly correlated with systolic blood pressure (r = 0.162, p = 0.063), diastolic blood pressure (r = 0.202, p = 0.024) and serum triglyceride level (r = 0.275, p = 0.008). SI was significantly correlated with systolic blood pressure (r = 0.171, p = 0.050), diastolic blood pressure (r = −0.219, p = 0.015) and duration of hypertension (r = 0.357, p<0.001). DI was significantly correlated with duration of hypertension (r = −0.291, p = 0.002), systolic blood pressure (r = −0.334, p = <0.001) and serum cholesterol level (r = −0.249, p = 0.015). Both SI and DI were significantly correlated with age (r = 0.450, p<0.001; r = −0.514, p<0.001, respectively) (Table I). Sex, smoking, diabetes mellitus or history of hyperlipidemia did not affect SI_DVP, SI and DI (Table II). Among measurements of target organ damage, SI_DVP was significantly correlated with BUN (r = 0.230, p = 0.023) and LVMI (r = 0.177, p = 0.050). SI was significantly correlated with serum creatinine level (r = 0.279, p = 0.005) and E/A (r = 0.279, p = 0.002). DI was significantly correlated with BUN (r = –0.231, p = 0.023) and LVMI (r = –0.239, p = 0.008) (Table III). There were 29 (23%) patients with vascular diseases; these had a higher level of SI_DVP (10.12±2.97 vs 8.45±1.78 m/s, p<0.001), LVMI (114.4±40.6 vs 96.8±35.1, p = 0.026), BUN (28.4±24.7 vs 14.5±4.6 mg/dl, p<0.001) and lower level of DI (1.34±0.88 vs 1.93±1.12×10⁻⁶ cm²/dyne, p = 0.010) than those without vascular diseases (Table IV). By multivariate logistic regression analysis, only SI_DVP was significantly associated with vascular diseases (OR 1.39, 95% CI 1.06–1.82, p = 0.016) (Table V).

Table I. Correlations between clinical characteristics and aortic stiffness.

Characteristics	SIDVP	SI	DI
Age	r = 0.162, p = 0.071	r = 0.450, p<0.001	r = −0.514, p<0.001
Duration of hypertension	r = 0.163, p = 0.081	r = 0.357, p<0.001	r = −0.291, p = 0.002
Systolic blood pressure	r = 0.089, p = 0.324	r = 0.171, p = 0.058	r = −0.334, p<0.001
Diastolic blood pressure	r = 0.202, p = 0.024	r = −0.219, p = 0.015	r = 0.066, p = 0.468
Cholesterol	r = 0.044, p = 0.672	r = 0.195, p = 0.059	r = −0.249, p = 0.015
Triglyceride	r = 0.275, p = 0.008	r = 0.136, p = 0.196	r = −0.074, p = 0.484

SI_DVP, stiffness index derived from digital volume pulse; SI, stiffness index derived from M‐mode echocardiography; DI, distensibility.

Table II. Risk factors and aortic stiffness.

	SIDVP (m/s)	SI	DI (10^−6 cm^2/dyne)
Sex
Man (57)	9.03±2.01	9.58±6.30	1.99±1.10
Woman (67)	8.68±2.38	10.22±9.99	1.63±1.08
Diabetes mellitus
Present (12)	8.78±1.35	15.88±12.54	1.39±0.94
Absent (112)	8.84±2.30	11.08±8.08	1.84±1.11
Smoking
Present (20)	9.26±1.75	12.02±8.41	1.68±1.10
Absent (104)	8.76±2.30	11.45±8.75	1.82±1.10
Hyperlipidemia
Present (68)	9.07±2.40	12.45±8.51	1.64±1.09
Absent (56)	8.57±2.02	10.73±9.04	1.97±1.12

Data was expressed as mean±SD.

SI_DVP, stiffness index derived from digital volume pulse; SI, stiffness index derived from M‐mode echocardiography; DI, distensibility.

Table III. Correlation between aortic stiffness and measurements of end organ damage.

	SIDVP	SI	DI
BUN	r = 0.230, p = 0.023	r = 0.168, p = 0.100	r = −0.231, p = 0.023
Creatinine	r = 0.049, p = 0.628	r = 0.279, p = 0.005	r = −0.174, p = 0.083
LVMI	r = 0.177, p = 0.050	r = 0.117, p = 0.198	r = −0.239, p = 0.008
E/A	r = −0.031, p = 0.734	r = 0.279, p = 0.002	r = −0.102, p = 0.261

BUN, blood urea nitrogen; LVMI, left ventricular mass index; E/A, peak atrial velocity and early filling to atrial velocity ratio.

Table IV. Comparison between patients with or without vascular diseases.

	Vascular diseases (+) (n = 29)	Vascular diseases (−) (n = 95)	p‐value
Age (years)	57.2±13.1	54.9±13.1	0.403
Duration of hypertension (years)	4.2±3.5	3.5±4.5	0.446
Systolic blood pressure (mmHg)	189±26	180±27	0.133
Diastolic blood pressure (mmHg)	113±16	107±24	0.222
Cholesterol (mg/dl)	213.7±38.6	203.9±37.8	0.286
Triglyceride (mg/dl)	140.0±90.9	141.1±73.3	0.955
Fasting glucose (mg/dl)	99.5±19.7	104.2±24.6	0.478
LVMI (g/m^2)	114.4±40.6	96.8±35.1	0.026
E/A	0.91±0.34	1.79±1.49	0.530
SIDVP (m/s)	10.12±2.97	8.45±1.78	<0.001
SI	13.76±7.63	10.87±8.88	0.116
DI (10^−6 cm^2/dyne)	1.34±0.88	1.93±1.12	0.010

Data was expressed as mean±SD.

LVMI, left ventricular mass index; E/A, peak atrial velocity and early filling to atrial velocity ratio; SI_DVP, stiffness index derived from digital volume pulse; SI, stiffness index derived from M‐mode echocardiography; DI, distensibility.

Table V. Multiple logistic analysis for association of vascular diseases.

	Odd ratio (95% CI)	p value
SIDVP	1.389 (1.062–1.817)	0.016
SI	0.977 (0.887–1.076)	0.639
DI	0.494 (0.201–1.216)	0.125
Age	1.017 (0.967–1.070)	0.506
Duration of hypertension	1.027 (0.891–1.184)	0.712
Systolic blood pressure	0.991 (0.967–1.017)	0.498
Diastolic blood pressure	1.019 (0.984–1.054)	0.291

SI_DVP, stiffness index derived from digital volume pulse; SI, stiffness index derived from M‐mode echocardiography; DI, distensibility.

Discussion

The present study showed that all three indices of aortic stiffness were significantly correlated with target organ damage and age. An increased aortic stiffness has been noted to be associated with target organ damage in hypertension [1], [3]. PWV is frequently used for measuring aortic stiffness; however, measuring PWV is time‐consuming and skill‐dependent. Measuring PWV also needs specialized devices [3], [4], making this method difficult to be used extensively in clinical practice. The methods that we used in this study, including SI and DI derived from M‐mode echocardiography and SI_DVP from plethysmography, were more simple, convenient, skill‐independent and time‐saving in comparison with PWV. SI_DVP was noted to be correlated with aging [11]. SI and DI were noted to be associated with microalbuminemia [22] and LVH [23] in essential hypertension. However, this is the first time that the correlation between these indices of aortic stiffness and target organ damage in hypertension has been extensively demonstrated.

This analysis also showed that SI_DVP was significantly associated with vascular diseases but not SI and DI after multivariate analysis controlling blood pressure. One of the possible reasons was that SI and DI were pressure‐dependent measurements [24]. SI, DI and other factors lost their association with vascular diseases when we analyzed them controlling blood pressure in multivariate analysis. The other possible reason was that good image acquisition was needed to measure diameter changes in the ascending aorta from M‐mode echocardiography. SI and DI were more subjective to variation when the image was not perfect.

This study was the first to demonstrate the clinical significance of SI_DVP in hypertension. SI_DVP could be obtained easily and objectively from commercially available photoplethysmography with automatically formulated software [11], [13]. However, good and steady DVP was necessary for this measurement. DVP patients with deformed fingers, arrhythmias or peripheral occlusive artery diseases were difficult to detect by photoplethysmography. SI_DVP could not be applied in these patients.

Our study was limited by a cross‐section design rather than a longitudinal study. However, SI_DVP was still an independent factor associated with vascular diseases after carefully evaluation by multivariate analysis considering all possible factors. Another limitation was that we only included untreated hypertension patients. The influences of anti‐hypertensive drugs on these aortic stiffness indices were not studied. Further large‐scale studies are warranted to study the drug effects on these stiffness indices and their impact on prognosis.

In conclusion, we found that the SI_DVP was independently associated with the presence of vascular diseases among untreated hypertensive patients. The SI_DVP simply derived from the DVP can be used as a marker for risk stratification in untreated hypertension and it is suitable for use in large‐scale trials because the DVP contour analysis is simple, time‐saving, operator‐independent and relatively inexpensive.

Acknowledgements

This study was supported by Grant NSC 90‐2324‐B‐006‐085 from the National Science Council, and Grand A‐19‐B‐FA09‐2‐4 from the Ministry of Education, Executive Yuan, Taipei, Taiwan.