Renal Duplex Parameters, Blood Pressure, and Renal Function in Elderly People Jeffrey D. Pearce, MD, Matthew S. Edwards, MD, Timothy E. Craven, MSPH, William P. English, MD, Matthew M. Mondi, MD, Scott W. Reavis, RVT, and Kimberley J. Hansen, MD ● Background: Changes in renal artery and renal parenchyma perfusion are believed to correlate with severity of hypertension and worsened renal function, but population-based studies of these associations are not available. This study examines relationships between parameters derived from renal duplex sonography (RDS), blood pressure (BP), and excretory renal function in a population-based cohort of elderly Americans. Methods: Through an ancillary study to the Cardiovascular Health Study, 758 participants (37% men; mean age, 77 years) underwent RDS in which flow velocities and frequency shifts were determined from spectral analysis of Doppler-shifted signals obtained from the renal artery and parenchyma. Associations of these duplex parameters with BP and inverse serum creatinine were examined by using multivariate regression techniques. Results: Main renal artery peak systolic flow velocity (PSV) showed independent associations with BP, with an SD increase in PSV (0.53 m/s) associated with a 3.3–mm Hg increase in systolic BP (SBP) and a 2.4 –mm Hg decrease in diastolic BP (DBP). An SD decrease in end-diastolic frequency shift (EDF; 131 kHz) was associated with a 6.0 –mm Hg increase in SBP, a 4.2–mm Hg decrease in DBP, and a significant 3.7% decrease in inverse serum creatinine. Conclusion: Increases in renal artery PSV and decreases in parenchymal EDF are associated with increased SBP and decreased DBP. Moreover, decreased parenchymal EDF showed significant associations with impaired excretory renal function. These results suggest that renal duplex parameters are associated with renal parenchymal changes caused by hypertension and progressive renal dysfunction in elderly people. Am J Kidney Dis 45:842-850. © 2005 by the National Kidney Foundation, Inc. INDEX WORDS: Cardiovascular Heath Study (CHS); population-based cohort; renal duplex sonography; blood pressure (BP); renal artery velocity; end-diastolic velocity; resistive index.
A
THEROSCLEROTIC RENAL vascular disease (RVD) of the main renal artery is a recognized cause of secondary hypertension and excretory renal insufficiency (ie, ischemic nephropathy).1-3 Recently, noninvasive screening for RVD has been aided by the development of renal duplex sonography (RDS).2,4 RDS is an
From the Division of Surgical Sciences, Section on Vascular Surgery; and Department of Public Health Sciences, Wake Forest University School of Medicine, Winston-Salem, NC. Received August 12, 2004; accepted in revised form January 14, 2005. Originally published online as doi:10.1053/j.ajkd.2005.01.028 on March 3, 2005. Supported in part by grant no. 1R01DK47414 from the National Institute of Diabetes and Digestive and Kidney Diseases and contracts no. N01-HC-85079 through N01-HC85086, N01-HC-35129, and N01 HC-15103 from the National Heart, Lung, and Blood Institute. Address reprint requests to Kimberley J. Hansen, MD, Professor of Surgery, Department of General Surgery, Division of Surgical Sciences, Wake Forest University School of Medicine, Winston-Salem, NC 27157-1095. E-mail:
[email protected] © 2005 by the National Kidney Foundation, Inc. 0272-6386/05/4505-0007$30.00/0 doi:10.1053/j.ajkd.2005.01.028 842
accurate noninvasive means to diagnose RVD in patients with severe hypertension and/or renal insufficiency that requires no special preparation and poses no risk to renal function. Furthermore, numerous reports have highlighted the relationship between renal parenchymal disease and changes in renal parenchymal resistance reflected by changes in Doppler-shifted frequencies obtained from RDS examination. Recent reports suggested that Doppler-shifted frequencies obtained from the renal artery and renal parenchyma may predict outcomes after renal artery intervention.5-8 Previous investigators showed associations between RDS estimates of parenchymal blood flow and clinical response (improved blood pressure [BP] control and/or excretory renal function) after repair of hemodynamically significant RVD. These investigators speculated that decreased diastolic flow in the renal artery and parenchyma observed in patients with poor clinical response after repair of RVD resulted from increased intrarenal resistance secondary to renal parenchymal disease. This relationship between Doppler-shifted signals and renal parenchymal disease has been supported by studies of resistive index (RI). RI represents a composite value derived from the
American Journal of Kidney Diseases, Vol 45, No 5 (May), 2005: pp 842-850
DUPLEX PARAMETERS, BP, AND RENAL FUNCTION IN OLDER PEOPLE
Doppler-shifted signals obtained from the region of the interlobar-arcuate arteries. Investigators reported that measures of RI correlated with changes in age, sympathetic tone, and hypertensive nephrosclerosis.9-11 In addition, increased RI was associated with increased plasma renin level and decreased creatinine clearance in select groups of patients.10,11 Moreover, patients with diabetes with excretory renal insufficiency were more likely to have an increased RI compared with individuals with renal insufficiency related to other causes.12-14 Consequently, RI may be considered a noninvasive measure of disease affecting the renal microcirculation. This concept is supported further by a recent study of patients with excretory renal insufficiency showing that an RI exceeding 0.8 was more predictive of a decline in excretory renal function or death than 24-hour creatinine clearance or proteinuria.15 Despite clinical evidence for the association between RDS Doppler-derived parameters and renal parenchymal disease, relationships between RDS parameters, hypertension, and excretory renal function have not been well described in free-living adults. Recently, we showed strong associations between the presence of RVD (defined dichotomously as a main renal artery peak systolic velocity [PSV] ⬎ 1.8 m/s) and clinical hypertension and excretory renal insufficiency in the Forsyth County cohort of the Cardiovascular Health Study (CHS).16 The CHS is a longitudinal, prospective, population-based study of coronary heart disease and stroke in elderly men and women. RDS was performed on members of the CHS Forsyth County cohort to estimate the prevalence of renal vascular disease in elderly Americans. This study examines relationships between Doppler parameters derived from the renal artery and renal parenchyma, BP, and estimates of excretory renal function. METHODS
The Cardiovascular Health Study The CHS is a longitudinal, multicenter, observational cohort study of cardiovascular disease risk factors, morbidity, and mortality in Americans aged 65 years and older. The initial CHS cohort was recruited from a randomly selected sample of Medicare-eligible individuals in 4 US communities (Forsyth County, NC; Sacramento County, CA; Washington County, MD; and Allegheny County, PA). Initial recruitment was performed between April 1989 and May 1990,
843
with a supplemental cohort of predominantly AfricanAmerican participants recruited using the same method from June 1992 to June 1993. All CHS enrollees underwent a baseline examination consisting of a detailed medical history and clinical examination. Clinical examination included physical examination, phlebotomy, electrocardiography, and pulmonary function testing. Annual follow-up examinations were performed to update medical data, assess for the occurrence of cardiovascular disease events, and repeat portions of the clinical examination at previously defined intervals.17 BP measurements were obtained at each annual follow-up, and repeated phlebotomy with measurement of serum creatinine was performed at postenrollment years 2 and 5 and at the time of the renal duplex examination.
Renal Duplex Sonography RDS examination was performed as part of an ancillary study to the CHS funded through the National Institute of Diabetes and Digestive and Kidney Diseases and approved by the Wake Forest University Human Subjects Review Committee (Winston-Salem, NC). CHS participants in the Forsyth County cohort were studied by means of RDS between January 1995 and February 1997. RDS was performed as previously described using an Ultramark-9 HDI Ultrasound System (Advanced Technologies Laboratories, Bothell, WA).2,4,16,18 Briefly, RDS was performed during the participant’s return visit for their annual examination by 2 registered vascular technologists with extensive experience in renal artery evaluation. As part of the annual examination, CHS participants had fasted overnight. After written informed consent was obtained, the participant was placed in the supine position and a 2.25- or 3.0-MHz ultrasound probe was coupled to the abdominal skin with acoustic gel 3 or 4 cm inferior to the xiphoid process. Sagittal B-mode scan images were obtained of the upper abdominal aorta, celiac axis, and superior mesenteric arteries. Identification of these 3 vessels was confirmed by the characteristic fasting waveforms from each vessel. After a sagittal aortic and superior mesenteric artery signal was obtained, the probe was rotated 90° to obtain a B-mode scan image of the aorta and proximal superior mesenteric artery in cross-section. The left renal vein was identified in longitudinal section. Using the left renal vein as a reference, aortic origins of the main renal arteries were identified. While maintaining an angle of insonation less than 60°, Doppler samples were obtained from each renal artery from aortic origin to the renal hilum, for a total of approximately 10 Doppler sample sites per renal artery. The estimated angle of insonation was combined with spectral analysis from the Doppler-shifted signals to obtain the renal artery PSV (meters per second) and end-diastolic velocity (EDV; meters per second). After Doppler interrogation in the supine position, renal artery PSV was confirmed by using a flank approach with the participant in the right or left lateral decubitus position, along with B-mode scan imaging of each kidney to determine the greatest longitudinal kidney length. While in the decubitus position, color Doppler was used to identify the corticomedullary junction. Using a 3-mm Doppler sample, the intrarenal peak systolic frequency shift (PSF; kilohertz) and end-diastolic frequency shift (EDF; kilohertz) were
844
PEARCE ET AL
obtained. RI then was calculated from these Doppler-shifted frequencies as follows5-7,12,13: RI ⫽ 1 ⫺ EDF ⁄ PSF All B-mode and Doppler spectral data were collected on super VHS tape and transferred to an electronic database. This process was repeated and data were compared for agreement. Three percent discordance in electronic data was adjudicated from review of the original duplex study.
Statistical Analysis After ancillary study data were keyed and verified, RDS results were matched with participant data provided by the CHS Coordinating Center. Data from the closest annual CHS visit before or within 3 months of RDS were used when measurements were missing from the visit coinciding with the RDS examination. Because additional quantitative measures of excretory renal function were not available for CHS participants, inverse serum creatinine ⫻ 100 (100/SCr) was chosen as the renal function outcome and is a surrogate for creatinine clearance19 when other variables included in the predictive equations for creatinine clearance are included in multivariate modeling. Univariate associations between RDS measures of main renal artery flow velocities (PSV and EDV), intrarenal parenchymal flow velocities (PSF, EDF, and RI), and systolic BP (SBP), diastolic BP (DBP), and 100/SCr were examined by using Spearman’s rank correlations because of non-normality of some variables. Linear regression analyses were used to build multivariate predictive models for SBP, DBP, and 100/SCr by using these RDS parameters and other covariates. Empirical distributions of model outcomes were examined to evaluate normality assumptions. A backwards variable elimination procedure was used to choose a suitable “best” model. Model selection began with the inclusion of potential 2-way interaction terms involving RDS parameters. Terms were eliminated sequentially by removing the least significant predictor with P greater than 0.10, then re-estimating model parameters. Stepping continued until all remaining factors were significant at an ␣ level of 10%.20 Only hierarchical models were considered (ie, steps containing 2-way interactions always included main effects involved in those interaction terms, as well). All model selection procedures used the following list of candidate variables: (1) a measure of intrarenal flow velocity (either 2-kidney maximum intrarenal EDF and maximum intrarenal PSF or 2-kidney minimum intrarenal RI); (2) measures of renal artery flow velocity (2-kidney maximum renal artery PSV and maximum renal artery EDV); and (3) other covariates hypothesized to be related to renal artery blood flow and/or renal function: SBP, DBP, age, black race, male sex, height, weight, and diabetes. Because of their known influence on both BP and renal function, individual indicators of antihypertensive medication use (angiotensin-converting enzyme inhibitor, vasodilator, calcium channel blocker, -blocker, and diuretic use) were forced into all models.
RESULTS
Between January 1995 and February 1997, a total of 1,245 Forsyth County participants returned to the CHS Field Center for their annual examination. Of returning participants, 870 participants (70%) consented to RDS examination. Of 870 RDS examinations performed, 758 examinations (87%) were complete studies. Participants with complete RDS examinations made up the study cohort for this investigation. Table 1 lists demographics of the study cohort. The cohort consisted of 474 women (63%) and 284 men (37%) with a mean age of 77.0 ⫾ 4.8 years. Seventy-seven percent of participants were white (n ⫽ 584), 21% were African American (n ⫽ 158), and 2% were classified as other (n ⫽ 16). Table 2 lists results of the cross-sectional univariate analysis of RDS-derived parameters with SBP, DBP, and inverse serum creatinine. Univariate analysis showed a weak positive association between RI with increased SBP (P ⫽ 0.002). EDV and EDF showed weak positive associations (P ⬍ 0.05) with increased DBP, whereas PSV, PSF, and RI showed weak negative associations with increased DBP (P ⬍ 0.05). There were no significant associations between main renal artery (PSV and EDV) or intrarenal parameters (PSF and EDF) and SBP on univariate analysis. Cross-sectional univariate analysis of RDS parameters and continuous inverse serum creatinine showed weak positive associations with that measurement and PSV, EDV, PSF, and EDF (P ⬍ 0.05). There was no significant univariate association between inverse serum creatinine and RI from spectral analysis. Table 3 lists results of the multivariate analysis of associations between RDS parameters and continuous measures of SBP and DBP. In crosssectional analysis, increases in PSV, PSF, and RI, as well as decreases in EDV and EDF, showed strong independent associations with increases in SBP (P ⬍ 0.05). Multivariate analysis of associations between RDS parameters and DBP showed that increases in PSV, PSF, and RI, as well as decreases in EDV and EDF, showed strong independent associations with decreases in DBP (P ⬍ 0.01). The only other significant associations with SBP and DBP shown in multi-
DUPLEX PARAMETERS, BP, AND RENAL FUNCTION IN OLDER PEOPLE
845
Table 1. Demographics and Medical Comorbidities Variable
Age* (y) Race† White or other African American Sex† Female Male SBP* (mm Hg) DBP* (mm Hg) Height* (cm) Weight* (lb) Estimated GFR*‡ (mL/min/1.73 m2) Inverse serum creatinine Hypertensive medication* ACE inhibitor use* Vasodilator use* Calcium channel blocker use* -blocker use* Diuretic use* Excretory renal insufficiency* Diabetes mellitus*
Definition
77.0 ⫾ 4.8 600 (79) 158 (21)
Serum creatinine⫺1 ⫻ 100 (dL/mg) Use of any antihypertensive medication at time of RDS Use of ACE inhibitor class of medication at time of RDS Use of vasodilator class of medication at time of RDS Use of calcium channel blocker class of medication at time of RDS Use of -blocker class of medication at time of RDS Use of diuretic class of medication at time of RDS GFR ⬍ 60 mL/min/1.73 m2‡ Fasting serum glucose ⬎ 140 mg/dL, 2-h post–glucose load serum glucose ⬎ 200 mg/dL, or insulin/oral hypoglycemic agent use
Diet controlled Regular insulin injections
474 (63) 284 (37) 136 ⫾ 20 72 ⫾ 10 165 ⫾ 9 160 ⫾ 31 70 ⫾ 16 104 ⫾ 23 416 (55) 86 (11.3) 54 (7) 161 (21) 109 (14) 210 (28) 202 (26) 177 (23) 117 (66) 26 (15)
NOTE. N ⫽ 758. Values expressed as mean ⫾ SD or number (percent). To convert glucose in mg/dL to mmol/L, multiply by 0.05551. Abbreviation: ACE, angiotensin-converting enzyme. *Parameter recorded/updated at nearest visit before or concurrent with RDS. †Parameter recorded at initial baseline examination. ‡Calculated using the Modification of Diet in Renal Disease study equation.
variate modeling included age, height, concurrent SBP or DBP, and use of a calcium channel blocker. Use of a vasodilating medication showed a significant association with SBP only. No additional associations with BP were shown, including race or the use of other antihypertensive medications. Table 4 lists results of multivariate analysis of associations between RDS parameters with in-
verse serum creatinine, a surrogate measure for creatinine clearance. For all RDS participants, increases in parenchymal EDF and decreases in RI were independently associated with increases in inverse serum creatinine (P ⬍ 0.05). Relationships between EDF and inverse serum creatinine are shown graphically in Fig 1. However, there were no associations with PSV, EDV, or PSF and inverse serum creatinine.
Table 2. Univariate Analysis of RDS Parameters, BP, and Inverse Serum Creatinine Variable (mean ⫾ SD)
Renal artery PSV* (1.26 ⫾ 0.53 m/s) Renal artery EDV* (0.31 ⫾ 0.16 m/s) Intrarenal PSF* (1,544 ⫾ 539 kHz) Intrarenal EDF* (378 ⫾ 131 kHz) RI‡ (0.73 ⫾ 0.06)
SBP
0.051 ⫺0.033 0.037 ⫺0.017 0.111†
NOTE. N ⫽ 758. Spearman’s correlation coefficients. *Maximum value obtained per patient. †P ⬍ 0.05. ‡Minimum value obtained per patient.
DBP
Inverse Serum Creatinine
⫺0.070† 0.122† ⫺0.134† 0.116† ⫺0.312†
0.117† 0.115† 0.270† 0.266† 0.002
846
PEARCE ET AL Table 3. Multivariate Analysis of RDS Parameters: BP SBP (R2 ⫽ 0.438)
DBP (R2 ⫽ 0.471)
Variable (modeled using RDS frequencies)
Regression Parameter Estimate
SE
P
Regression Parameter Estimate
SE
P
Intercept Renal artery PSV* Renal artery EDV* Intrarenal PSF* Intrarenal EDF* SBP* DBP* Vasodilator medication Calcium channel blocking medication Age* Height*
47.81 3.30 ⫺2.07 6.49 ⫺5.97 — 12.48 5.65 6.73 2.37 ⫺2.51
15.91 1.03 1.01 1.00 1.01 — 0.59 2.22 1.38 0.61 0.57
⬍0.01 0.04 ⬍0.01 ⬍0.01 — ⬍0.01 0.01 ⬍0.01 ⬍0.01 ⬍0.01
26.46 ⫺2.41 2.44 ⫺4.45 4.23 5.99 — ⫺0.05 ⫺1.39 ⫺0.78 0.99
7.78 0.50 0.49 0.48 0.48 0.29 — 1.09 0.69 0.30 0.28
⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 — NS 0.04 0.01 ⬍0.01
SBP (R2 ⫽ 0.445)
DBP (R2 ⫽ 0.474)
Variable (modeled using RI)
Regression Parameter Estimate
SE
P
Regression Parameter Estimate
SE
P
Intercept RI* Renal artery PSV* Renal artery EDV* SBP* DBP* Vasodilator medication Calcium channel blocking medication Age* Height*
⫺9.29 4.67 3.35 ⫺1.62 — 12.55 5.50 6.82 2.16 ⫺2.43
17.57 0.64 1.01 1.01 — 0.59 2.20 1.37 0.59 0.56
⬍0.01 ⬍0.01 NS — ⬍0.01 0.01 ⬍0.01 ⬍0.01 ⬍0.01
63.61 ⫺3.02 ⫺2.52 2.24 6.06 — ⫺0.02 ⫺1.52 ⫺0.69 0.95
8.31 0.31 0.49 0.49 0.28 — 1.09 0.68 0.39 0.28
⬍0.01 ⬍0.01 ⬍0.01 ⬍0.01 — NS 0.03 0.02 ⬍0.01
NOTE. N ⫽ 758. Abbreviation: NS, not significant. *Parameter estimates based on 1 SD change.
Table 4 also lists results of multivariate analysis of association between RDS parameters and a subgroup of RDS participants with moderate excretory renal insufficiency (glomerular filtration rate [GFR] ⬍ 60 mL/min/1.73 m2, estimated by means of Modification of Diet in Renal Disease19 calculation). Increased EDF and decreased RI maintained strong independent associations with inverse serum creatinine in participants with excretory renal insufficiency. Again, there were no associations between PSV, EDV, or PSF and inverse serum creatinine in the population with excretory renal insufficiency. DISCUSSION
This population-based study examines the relationships between RDS parameters derived from spectral analysis of the Doppler-shifted signal in the renal artery and renal parenchyma
with BP and excretory renal function in a freeliving elderly cohort. RDS estimates of main renal artery PSV and EDV showed strong associations with SBP and DBP. Moreover, particularly strong associations with BP and excretory renal function were shown with renal parenchymal EDF and RI. These observed relationships were independent of age, sex, race, weight, height, diabetic status, and antihypertensive medications. Data from this and other studies suggest that Doppler-derived parameters may reflect changes in the microcirculation of the kidney. A number of studies support the association between RDS parameters derived from renal parenchymal Doppler-shifted signal and hypertension.9-11 RI, a commonly reported measure of Dopplerderived diastolic frequency shift relative to systolic shift, has been associated with increasing
DUPLEX PARAMETERS, BP, AND RENAL FUNCTION IN OLDER PEOPLE
847
Table 4. Multivariate Analysis of RDS Parameters: Inverse Serum Creatinine
Inverse Serum Creatinine: All Participants (n ⫽ 755) (R2 ⫽ 0.390)
Inverse Serum Creatinine: Renal Insufficiency (n ⫽ 205) (R2 ⫽ 0.671)
Variable (modeled using RDS frequencies)
Regression Parameter Estimate
SE
P
Regression Parameter Estimate
SE
P
Intercept Renal artery PSV* Renal artery EDV* Intrarenal PSF* Intrarenal EDF* Age* Height* Weight* Black race Male sex Positive history of diabetes mellitus Vasodilator medication use Calcium channel blocker medication use
197.99 — — — 3.85 ⫺3.17 ⫺1.83 ⫺2.86 ⫺3.54 ⫺19.90 8.65 5.37 ⫺3.10
23.55 — — — 0.72 0.73 1.09 0.82 1.67 2.16 2.53 2.60 1.61
NS NS NS ⬍0.01 ⬍0.01 NS ⬍0.01 0.03 ⬍0.01 ⬍0.01 0.04 0.05
91.88 — — — 2.56 ⫺0.47 0.33 ⫺0.67 ⫺14.39 ⫺22.00 0.78 0.63 ⫺1.20
23.00 — — — 0.75 0.65 1.12 0.79 2.16 2.18 3.01 3.15 1.46
NS NS NS ⬍0.01 NS NS NS ⬍0.01 ⬍0.01 NS NS NS
Inverse Serum Creatinine: All Participants (n ⫽ 755) (R2 ⫽ 0.370)
Variable (modeled using RI)
Intercept RI* Renal artery PSV* Renal artery EDV* Age* Height* Weight* Black race Male sex Positive history of diabetes mellitus Vasodilator medication use Calcium channel blocker medication use
Inverse Serum Creatinine: Renal Insufficiency (n ⫽ 205) (R2 ⫽ .658)
Regression Parameter Estimate
SE
P
Regression Parameter Estimate
SE
P
244.56 ⫺1.55 — — ⫺3.83 ⫺2.15 ⫺2.74 ⫺3.81 ⫺20.97 8.63 5.21 ⫺3.26
25.02 0.70 — — 0.73 1.12 0.84 1.70 2.19 2.60 2.64 1.63
0.03 NS NS ⬍0.01 0.06 ⬍0.01 0.03 ⬍0.01 ⬍0.01 0.05 0.05
118.26 ⫺1.68 — — ⫺3.82 0.27 ⫺0.27 ⫺15.26 ⫺23.32 0.89 0.46 ⫺0.88
24.31 0.67 — — 0.67 1.14 0.81 2.18 2.19 3.06 3.19 1.49
0.01 NS NS NS NS NS ⬍0.01 ⬍0.01 NS NS NS
NOTE. Inverse serum creatinine: (serum creatinine⫺1 ⫻ 100. N ⫽ 755). Renal insufficiency indicates GFR less than 60 mL/min/1.73 m2 calculated using the Modification of Diet in Renal Disease Study Equation. Abbreviation: NS, not significant. *Parameter estimates based on 1 SD change.
severity and duration of hypertension.10 In this study, a single SD increase in RI was associated with an increase in SBP of 4.7 mm Hg. Moreover, parenchymal EDF and PSF were independently associated with increases in SBP and more predictive of SBP than RI. An SD change in EDF and PSF conferred a 6.0–mm Hg decrease and 6.5–mm Hg increase in SBP, respectively. These results in this population-based cohort show that Doppler-derived measures of renal perfusion are associated with changes in SBP and DBP. Moreover, these observed associa-
tions between Doppler-derived frequencies from the renal artery and parenchyma may be related to worsening hypertensive nephrosclerosis, with increased renal parenchymal resistance to perfusion. In addition to the strong associations noted between RDS and systemic BP, we observed a significant association between Doppler-derived frequency shift from the renal parenchyma and inverse serum creatinine. Increased intraparenchymal EDF and reduced RI showed strong independent associations with increased inverse
848
PEARCE ET AL
Fig 1. Graphic depiction of associations between EDF and inverse serum creatinine in (top line) all RDS participants and (bottom line) patients with excretory renal insufficiency (estimated GFR < 60 mL/min/1.73 m2 based on the Modification of Diet in Renal Disease Study).
serum creatinine. Associations with EDF and RI remained strong for patients with moderate excretory renal insufficiency (estimated GFR ⬍ 60 mL/min/1.73 m2). Our observations support those of other investigators that increased renal parenchyma resistance is associated with worsening of excretory renal function.11-14 However, these observed associations with inverse serum creatinine were stronger for EDF than RI in both the entire cohort and those with excretory renal insufficiency, and we found no association with PSF, a component of the RI calculation. Thus, these data suggest that intraparenchymal EDF is a better predictor of excretory renal function than RI and may provide the best Doppler-derived estimate of parenchymal disease. Moreover, these observed associations further support the relationship between Doppler-derived frequencies from the renal parenchyma and such parenchymal diseases as nephrosclerosis and its related worsening excretory renal function. The observed relationship between diabetes and serum creatinine level in this cohort was
unexpected. Although there were no statistically significant differences in serum creatinine levels between participants with and without diabetes (1.00 versus 1.02 mg/dL, respectively; P ⫽ 0.39), a history of diabetes was predictive of increased inverse serum creatinine, a surrogate for estimated creatinine clearance using multivariate modeling (Table 4). This finding may be explained in part by the cohort age. On average, RDS participants with diabetes were 14 years older than the national average of patients with diabetes requiring renal replacement therapy because of diabetic nephropathy.21 In addition, 66% of patients with diabetes had disease controlled by means of diet, and only 15% were insulin dependent. These characteristics suggest that participants with diabetes in this study, as a whole, had mild clinical disease at an older age of onset. Hence, the participants with diabetes may have been relatively free of diabetic nephropathy and associated renal parenchymal disease. Recent reports have focused on the relationship between pulse pressure (PP) and cardiovas-
DUPLEX PARAMETERS, BP, AND RENAL FUNCTION IN OLDER PEOPLE
cular events. Although dependent on changes in SBP and DBP, investigators have shown that increased PP was an independent predictor of myocardial infarction and cardiovascular mortality.22-24 Chae et al23 showed that increased PP was independently associated with an increased risk for heart failure in an elderly communitybased cohort. In this study, RDS measures of extrarenal and intrarenal systolic flow were associated with increased SBP and decreased DBP. Furthermore, on multivariate analysis (data not shown), increases in renal artery Doppler-derived parameters (PSV and PSF) and RI were both strongly predictive of increases in PP. Conversely, increased EDV and EDF were independently associated with decreased SBP, increased DBP, and reduced PP. These data suggest that changes in renal artery and renal parenchymal Doppler-shifted frequencies also may be associated with changes in PP. Considered collectively, these data support the relationship between worsening hypertension and other factors, such as vascular compliance, that may be related to increased renal parenchymal resistance. Unfortunately, this study does not allow for hypothesis testing to define the nature of these relationships; however, the observed changes in PP and BP associated with RDS measures of parenchymal resistance may be predictive of subsequent cardiovascular events. Recently, the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial showed that small changes in SBP were associated with significant increases in risk for heart failure and stroke.25 Thus, the observed increases in SBP and PP noted with increases in PSV and measures of parenchymal resistance (RI and EDF) may be clinically relevant in predicting subsequent cardiovascular mortality in individual patients undergoing a screening examination for secondary hypertension. Although this study uses a carefully constructed free-living community-dwelling elderly cohort, several study weaknesses deserve comment.18 The recruitment strategy used by the CHS did not produce a true randomly selected population-based sample. Index participants were selected for initial contact based on a random sampling of Medicare eligibility lists. Ageeligible members of the same household also were recruited. As a result, a significant number
849
(30%) of participants were enrolled as a result of shared household recruitment. This recruitment strategy was designed to enhance recruitment and longitudinal follow-up. However, it unavoidably introduced selection bias because significant differences existed among the randomly selected individuals who chose to enroll in the CHS compared with those who declined. In general, participants enrolling were younger, more highly educated, more likely to be married, and more likely to have previously quit smoking.26 This healthy-cohort effect, combined with a similar trend among RDS ancillary study participants and nonparticipants,3 may have resulted in a lower prevalence of RVD, hypertension, and renal insufficiency among studied individuals, reducing the statistical power to identify other associations with RDS. In summary, this study shows strong associations between RDS parameters derived from Doppler spectral analysis of the renal artery and renal parenchyma with BP and inverse serum creatinine, a surrogate for creatinine clearance when used in multivariate modeling. Both main renal artery– and parenchymal-derived velocities were associated with BP in this free-living elderly cohort. Moreover, parenchymal-derived EDF showed the greatest overall associations with BP and changes in inverse serum creatinine. These associations support the relationship between Doppler-derived duplex parameters associated with renal parenchymal disease and progressive renal dysfunction in elderly people. Additional study may delineate whether these strong RDS associations with BP, PP, and excretory renal function also may predict subsequent cardiovascular-related mortality. REFERENCES 1. Dean RH, Kieffer RW, Smith BM, et al: Renovascular hypertension: Anatomic and renal function changes during drug therapy. Arch Surg 116:1408-1415, 1981 2. Hansen KJ, Tribble RW, Reavis SV, et al: Renal duplex sonography: Evaluation of clinical utility. J Vasc Surg 12:227236, 1990 3. Hansen KJ, Cherr GS, Dean RH: Dialysis-free survival after surgical repair of ischemic nephropathy. Cardiovasc Surg 10:400-404, 2002 4. Motew SJ, Cherr GS, Craven TE, et al: Renal duplex sonography: Main renal artery versus hilar analysis. J Vasc Surg 32:462-471, 2000 5. Cohn JE, Benjamin ME, Sandager GP, Lilly MP, Killewich LA, Flinn WR: Can intrarenal duplex waveform
850
analysis predict successful renal artery revascularization? J Vasc Surg 28:471-481, 1998 6. Frauchiger B, Zierler R, Bergelin RO, Isaacson JA, Strandness DE: Prognostic significance of intrarenal resistance indices in patients with renal artery interventions: A preliminary duplex sonographic study. Cardiovasc Surg 4:324-330, 1996 7. Radermacher J, Chavan A, Bleck J, et al: Use of Doppler ultrasonography to predict the outcome of therapy for renal-artery stenosis. N Engl J Med 344:410-417, 2001 8. Mukherjee D, Bhatt DL, Robbins M, et al: Renal artery end-diastolic velocity and renal artery resistance index as predictors of outcome after renal stenting. Am J Cardiol 15:1064-1066, 2001 9. Knapp R, Plotzeneder A, Frauscher F, et al: Variability of Doppler parameters in the healthy kidney: An anatomicphysiologic correlation. J Ultrasound Med 14:427-429, 1995 10. Galesic K, Brkljacic B, Sabljar-Matovinovic M, Morovic-Vergles J, Cvitkovic-Kuzmic A, Bozikov V: Renal vascular resistance in essential hypertension: DuplexDoppler ultrasonographic evaluation. Angiology 51:667675, 2000 11. Aikimbaev KS, Canataroglu A, Ozbek S, Usal A: Renal vascular resistance in progressive systemic sclerosis: Evaluation with duplex Doppler ultrasound. Angiology 52: 697-701, 2001 12. Matsumoto N, Ishimura E, Taniwaki H, et al: Diabetes mellitus worsens intrarenal hemodynamic abnormalities in nondialyzed patients with chronic renal failure. Nephron 86:44-51, 2000 13. Sari A, Dinc H, Zibandeh A, Telatar M, Gumele HR: Value of resistive index in patients with clinical diabetic nephropathy. Invest Radiol 34:718-721, 1999 14. Platt JF, Rubin JM, Ellis JH: Diabetic nephropathy: Evaluation with renal duplex Doppler US. Radiology 190: 343-346, 1994 15. Radermacher J, Ellis S, Haller H: Renal resistive index and progression of renal disease. Hypertension 39:699703, 2002
PEARCE ET AL
16. Edwards MS, Hansen KJ, Craven TE, et al: Relationships between renovascular disease, blood pressure, and renal function in the elderly: A population-based study. Am J Kidney Dis 41:990-996, 2003 17. Fried LP, Borhani NO, Enright P, et al: The Cardiovascular Health Study: Design and rationale. Ann Epidemiol 1:263-276, 1991 18. Hansen KJ, Edwards MS, Craven TE, et al: Prevalence of renovascular disease in the elderly: A populationbased study. J Vasc Surg 36:443-451, 2002 19. National Kidney Foundation: K/DOQI Clinical Practice Guidelines for Chronic Kidney Disease. Am J Kidney Dis 39:S76-S110, 2002 (suppl 1) 20. SAS Institute Inc: SAS/STAT User’s Guide, version 8. Cary, NC, SAS Institute Inc, 1999 21. US Renal Data System: USRDS 2002 Annual Data Report. The National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, Bethesda, MD, 2002 22. Domanski M, Mitchell G, Pfeffer M, et al: Pulse pressure and cardiovascular disease-related mortality. JAMA 287:2677-2683, 2002 23. Chae CU, Pfeffer MA, Glynn RJ, Mitchell GF, Taylor JO, Hennekens CH: Increased pulse pressure and risk of heart failure in the elderly. JAMA 281:634-639, 1999 24. Henry RMA, Kostense PJ, Spijkerman AMW, et al: Arterial stiffness increases with deteriorating glucose tolerance status. Circulation 107:2089-2095, 2003 25. Furberg CD, Wright JT, Davis RB, et al: Major outcomes in high-risk hypertensive patients randomized to angiotensin converting enzyme inhibitor or calcium channel blocker vs diuretic: The Antihypertensive and Lipid Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA 288:2981-2997, 2002 26. Tell GS, Fried LP, Hermanson B, Manolio TA, Newman AB, Borhani NO: Recruitment of adults 65 years and older as participants in the Cardiovascular Health Study. Ann Epidemiol 3:358-366, 1993