Freedom from adverse coronary heart disease events according to renovascular disease (RVD) status. RDS indicates renal duplex sonography.
Edwards MS, Craven TE, Burke GL, Dean RH, Hansen KJ. Renovascular Disease and the Risk of Adverse Coronary Events in the ElderlyA Prospective, Population-Based Study. Arch Intern Med. 2005;165(2):207-213. doi:10.1001/archinte.165.2.207
Renovascular disease is a cause of secondary hypertension and renal insufficiency and is suspected to contribute to morbidity and mortality of coronary heart disease. This investigation prospectively examined associations between renovascular disease and adverse coronary events among a population-based sample of elderly Americans.
The Cardiovascular Health Study is a prospective, multicenter cohort study of cardiovascular disease risk factors, morbidity, and mortality among Americans older than 65 years. Renal duplex sonography was performed on 870 individuals between January 1995 and February 1997. Renovascular disease was defined as any focal peak systolic velocity of 1.8 m/s or greater (renal artery stenosis) or the absence of a Doppler-shifted signal from an imaged artery (renal artery occlusion). Adverse coronary events were defined as hospitalized angina, fatal or nonfatal myocardial infarction, and coronary revascularization.
During a mean follow-up of 14 months, 68 participants experienced incident or recurrent adverse coronary events. The presence of renovascular disease demonstrated a significant relationship with adverse coronary events (hazard ratio, 1.96; 95% confidence interval, 1.00-3.83; P = .05) that remained after controlling for the effects of coexisting atherosclerotic risk factors and prevalent cardiovascular disease. The relationship between renovascular disease and adverse coronary events was not dependent on the effects of increased blood pressure.
The presence of renovascular disease was associated with an increase in the risk of adverse coronary events in this sample. The increment in risk was not dependent on the effects of associated atherosclerotic risk factors, other prevalent cardiovascular disease, or increased blood pressure.
Renovascular disease (RVD) is a term used to describe lesions of the renal artery, including stenoses and occlusions, that can result in significant reductions in renal parenchymal perfusion.1 In the elderly, atherosclerotic RVD may be asymptomatic without clinical manifestations, or it may contribute to severe secondary hypertension (ie, renovascular hypertension) and decreased excretory renal function (ie, ischemic nephropathy).2- 4 It is widely assumed that these clinical manifestations may result in increased morbidity and mortality secondary to coronary heart disease. Although unproven, these assumptions provide the rationale for current treatment efforts including open surgical and percutaneous methods of renal revascularization.5- 9
Previous investigations in selected populations of individuals with RVD have demonstrated an increased risk of incident and/or recurrent adverse coronary events and mortality.10- 13 However, it remains unclear whether the incremental risk observed was secondary to an independent contribution by RVD and/or its clinical manifestations, or due to a shared relationship with generalized atherosclerosis and coexisting cardiovascular disease (CVD).
This investigation used renal duplex sonography (RDS) to detect the presence of anatomic RVD among participants in the Forsyth County, North Carolina, cohort of the Cardiovascular Health Study (CHS). Participants were then followed up prospectively and assessed for the occurrence of adverse coronary events. The application of RDS to this cohort provided a unique opportunity to investigate the associations between RVD and adverse coronary events among an unselected, population-based sample of community-dwelling elderly Americans. We hypothesized that the presence of RVD would be associated with an incremental risk of adverse coronary events that was dependent on its primary clinical manifestations (associated hypertension and/or excretory renal insufficiency) but independent of the effects of atherosclerotic risk factors and other prevalent CVD.
The design of the CHS has been previously described.14 The CHS is a longitudinal, multicenter, cohort study of CVD risk factors, morbidity, and mortality among Americans older than 65 years. The initial CHS cohort was recruited from a randomly selected sample of Medicare-eligible individuals in 4 US communities (Forsyth County, North Carolina; Sacramento County, California; Washington County, Maryland; and Allegheny County, Pennsylvania). Initial recruitment was performed between April 1989 and May 1990. A supplemental cohort of predominantly African American participants was recruited by the same methods from June 1992 to June 1993 to allow for racial subgroup analyses.
As part of an ancillary study funded through the National Institute of Diabetes and Digestive and Kidney Diseases, CHS participants in the Forsyth County cohort were examined with RDS between January 1995 and February 1997. This study was approved by the Wake Forest University Human Subjects Review Committee, Winston-Salem, NC. Participants scheduled for routine annual examination were contacted by telephone and informed of this ancillary project. During the ancillary study interval, 1245 Forsyth County participants returned to the CHS Field Center for their annual examination. Among returning participants, 870 (69.9%) consented to RDS examination. Of the 870 RDS examinations performed, 834 (95.9%) were technically adequate to define the presence or absence of RVD. Participants with technically adequate RDS examinations made up the study cohort for this report.
All CHS participants underwent a baseline examination at the time of enrollment consisting of a detailed medical history and clinical examination. Clinical examination included physical examination, phlebotomy, electrocardiography, and echocardiography. Diabetes mellitus was defined as a fasting serum glucose level of 125 mg/dL (6.9 mmol/L) or more, a serum glucose level 2 hours after oral glucose load of 200 mg/dL (11.1 mmol/L) or more, or a history of diabetes mellitus in conjunction with oral hypoglycemic or insulin use. Clinical hypertension was defined as a systolic blood pressure of 140 mm Hg or more, a diastolic blood pressure of 90 mm Hg or more, or a history of hypertension in conjunction with antihypertensive medication use. Excretory renal insufficiency was defined as an estimated glomerular filtration rate of less than 60 mL/min per 1.73 m2, calculated by means of the Modification of Diet in Renal Disease Study abbreviated formula.15 All self-reported cardiovascular conditions were confirmed by components of the baseline examination and/or by a defined validation protocol using medical records and/or information from the treating physician.16
On the basis of the results of physical examination, medical history, and noninvasive studies, participants were defined as having no apparent CVD, clinical CVD, and/or subclinical CVD. The methods of classification and measurement for these variables in the CHS have been previously described.16- 22 All ultrasound and echocardiographic interpretation was performed at central CHS reading centers. Carotid artery intimal-medial thickness was measured by means of an average of near- to far-wall B-mode ultrasound distance measurements.20,21 Increased internal and common carotid artery intimal-medial thickness was defined as a maximum wall thickness exceeding the 80th percentile for the entire CHS cohort (1.18 mm for the common carotid artery and 1.81 mm for the internal carotid artery).21 Internal carotid artery stenosis was estimated by means of B-mode ultrasound images and Doppler-derived flow velocities, with significant internal carotid artery stenosis defined as one that reduces diameter by 25% or more.19 Echocardiographic abnormalities were defined as an ejection fraction of 45% or less and/or the presence of significant wall motion abnormalities on M-mode images.18 Major electrocardiographic abnormalities were classified according to the Minnesota code and included ventricular conduction defects, major Q/QS-wave abnormalities, left ventricular hypertrophy, isolated major ST/T-wave abnormalities, atrial fibrillation, and first-degree atrioventricular block.17,23,24 Ankle-arm indexes were determined by means of a ratio of the highest obtained posterior tibial blood pressure divided by the right brachial artery blood pressure.22
Each participant was prospectively followed up through biannual telephone contact and annual clinic visits. At each annual visit, the medical history was updated and anthropometric and blood pressure measurements were repeated. Phlebotomy, electrocardiography, carotid duplex sonography, and echocardiographic imaging were repeated at previously defined intervals.14
Duplex sonography was performed (Ultramark-9 HDI Ultrasound System; Advanced Technologies Laboratories, Bothell, Wash) by 2 registered vascular technologists with extensive experience in renal artery evaluation. Written informed consent was obtained from CHS participants on return for their annual examination. During that same visit, consenting participants were studied with RDS at the CHS Forsyth County Field Center. As part of the annual examination, CHS participants had fasted overnight. The technique of RDS has been described previously.25- 27 All B-mode and Doppler spectral data were collected on super-VHS tape and transferred to an electronic database. This process was repeated and the data were compared for agreement. A 3% discordance in electronic data was adjudicated by review of the original duplex study.
An RDS study was considered negative, positive, or inadequate for interpretation of RVD status according to the following criteria: (1) RDS was negative for RVD when renal artery peak systolic velocity from aortic origin to renal hilum was less than 1.8 m/s; (2) RDS was positive for RVD when there was a focal increase in renal artery peak systolic velocity to 1.8 m/s or greater (ie, hemodynamically significant renal artery stenosis) or no Doppler signal was obtained from an imaged artery (ie, renal artery occlusion); and (3) RDS was technically inadequate to interpret for the presence or absence of RVD when renal artery Doppler signals were not obtained from the entire renal artery from aortic origin to renal hilum. These RDS criteria for hemodynamically significant renal artery stenosis and occlusion have been validated prospectively in comparison with conventional renal angiography.25,27
Biannual telephone contact and annual clinical examinations were performed to update participants’ medical histories and assess for incident and recurrent adverse coronary events. All adverse coronary events occurring during surveillance were confirmed by means of a defined protocol.28 For all reported events, a 6-member panel reviewed hospital records and, where appropriate, death certificates and autopsy results to adjudicate reported events according to previously defined criteria into definite, probable, and no event categories.
For the purposes of this study, adverse coronary events were defined as a combined measure of common coronary heart disease complications. Adverse coronary events were defined as any confirmed (definite) incident or recurrent episode of fatal or nonfatal myocardial infarction, hospitalization for angina, or need for coronary revascularization (either coronary artery bypass or angioplasty/stenting).
After ancillary study data were keyed and verified, RDS results were matched with participant data provided by the CHS Coordinating Center. Demographic, risk factor, and prevalent CVD variables for all participants were defined by means of the most current data available from all CHS examinations before or coincident with the RDS examination.
Differences among participants with and without RVD were assessed by means of χ2 or Fisher exact test for dichotomous factors and unpaired t test for continuous factors. Time from RDS examination to first incident or recurrent adverse event was examined by means of proportional hazards regression modeling.29 Univariate associations with adverse coronary events are reported, as well as associations adjusted for age, race, and sex for all variables considered. Multivariate associations between RVD and incident or recurrent adverse coronary events were also examined. All multivariate models included demographic and atherosclerotic risk factors as candidate covariates, as well as coexisting measures of prevalent clinical and subclinical CVD. Variables considered are listed in Table 1 and Table 2. After the evaluation of potential 2-way interaction terms, a model was constructed with a forward stepwise variable selection where, beginning with the most significant, individual variables with P<.10 were selected one by one for model inclusion. All variables remaining significant at α = .05 were included in the final model.
Demographics and medical risk factors for the RDS ancillary study cohort are presented in Table 1. The cohort included 525 women and 309 men (194 African American and 640 white) participants with a mean age of 77.2 ± 4.9 years. Overall, 57 participants (6.8%) had anatomic RVD demonstrated by RDS. Previous analysis demonstrated that participants with RVD were older, were more likely to exhibit clinical hypertension and/or excretory renal insufficiency, and demonstrated higher mean systolic blood pressures and lower mean serum high-density lipoprotein cholesterol levels than participants without RVD.25
Prevalent clinical and subclinical manifestations of CVD at the time of RDS examination are presented in Table 2. In general, participants with RVD demonstrated a higher prevalence of previous myocardial infarction and angina, as well as a higher prevalence of subclinical CVD than participants without RVD.
During an average post-RDS follow-up of 13.9 months (range, 0-38 months), 68 individuals experienced 122 incident or recurrent adverse coronary events. These events included 63 cases of angina with hospitalization, 32 cases of myocardial infarction, and 27 cases of coronary revascularization.
Univariate and multivariate hazard ratios for the associations between examined risk factors and adverse coronary events are presented in Table 3. The presence of RVD demonstrated a significant univariate association with the occurrence of adverse coronary events (hazard ratio, 2.92; 95% confidence interval, 1.53-5.57) that was largely unchanged by adjustments for age, race, and sex (hazard ratio, 2.70; 95% confidence interval, 1.40-5.19). The strength of association between RVD and adverse coronary events was similar to those observed for other prevalent clinical and subclinical manifestations of CVD.
The results of the multivariate modeling procedure that included demographics, identified atherosclerotic risk factors, and prevalent manifestations of CVD as candidate covariates are also summarized in Table 3. In the multivariate analysis, the presence of RVD demonstrated a significant association with adverse coronary events (hazard ratio, 1.96; 95% confidence interval, 1.00-3.83; P = .05). The observed association between the presence of RVD and subsequent adverse coronary events remained after adjustment for identified atherosclerotic risk factors and other prevalent CVD. No significant 2-way interactions between the presence of RVD and increased blood pressure or decreased estimated glomerular filtration rate in the prediction of adverse coronary events were observed. The Figure depicts the freedom from adverse coronary events according to the presence or absence of RVD. Other conditions demonstrating significant and independent associations with adverse coronary events included the presence of excretory renal insufficiency and abdominal aortic aneurysm as well as a history of claudication, lower-extremity revascularization, stroke, transient ischemic attack, abdominal aortic aneurysm repair, or angina.
This investigation provides, to our knowledge, the first prospective population-based data concerning the risk of adverse coronary events in the elderly associated with the presence or absence of RVD. During a follow-up of 14 months, a 96% increase in the odds of subsequent adverse coronary events associated with the presence of RVD was observed among this unselected sample of free-living elderly African American and white participants. The observed relationship remained even after adjustment for the effects of associated atherosclerotic risk factors and other prevalent CVD. Interestingly, no significant interactions were demonstrated between the presence of RVD and its principal clinical manifestations (ie, elevated blood pressure or decreased estimated glomerular filtration rate) in the prediction of subsequent adverse coronary events. However, a significant and independent relationship was observed between the presence of defined excretory renal insufficiency at baseline and the occurrence of subsequent adverse coronary events, consistent with reports by other authors.30- 33 The significance of this finding as it relates to RVD is difficult to assess given the relative scarcity of study participants demonstrating both conditions. Collectively, these findings suggest that (1) the presence of RVD was associated with an increase in subsequent adverse coronary events and thus represents an important identifiable indicator of individuals at increased cardiac risk, (2) the increased risk of adverse coronary events observed among individuals with anatomic RVD was not mediated solely by the effects of associated clinical and/or subclinical CVD, and (3) the observed increase in risk observed in this free-living elderly sample was not mediated by the effects of elevated blood pressure.
Despite the widespread application of renal artery intervention for RVD, it is not clear that intervention decreases the risk of adverse coronary events. Consequently, it is unclear (1) how affected individuals should be selected for intervention, (2) at what point they should undergo intervention, and (3) what method of intervention should be used. The only existing data demonstrating an improvement in dialysis-free survival among individuals with RVD after renal artery intervention come from patients selected for surgical revascularization because of the presence of severe hypertension with or without ischemic nephropathy.5,6 These reports demonstrated that severe preoperative hypertension, cure of hypertension, and at least 20% improvement in excretory renal function after intervention were significant and independent predictors of improved dialysis-free survival among such selected individuals. On the basis of the data presented herein, however, such severely affected individuals appear to be relatively uncommon among the general elderly population. Furthermore, the results after intervention in less severely affected individuals are unknown. Given the findings presented herein that increased systolic blood pressure did not confer additional risk of adverse coronary events among individuals with RVD at a population level, there appears to be no indication for the treatment of atherosclerotic RVD solely on the basis of lesion presence (to prevent subsequent hypertension) or in the setting of easily controlled hypertension.
Although this investigation represents, to our knowledge, the first prospective, population-based, follow-up study of its kind, a number of significant limitations exist that deserve comment. The recruitment strategy used by the CHS did not produce a true randomly selected population-based study sample. Index participants were selected for initial contact on the basis of a random sampling of Medicare eligibility lists. Age-eligible members of the same household were recruited as well. As a result, a significant number of participants were enrolled through shared household recruitment. In addition, exclusion criteria eliminated all mentally impaired and/or institutionalized individuals from consideration for enrollment. These strategies of recruitment were used to enhance initial enrollment and longitudinal follow-up. However, they unavoidably introduced some degree of selection bias, as significant differences existed among the randomly selected individuals who chose to enroll in the CHS as opposed to those who declined enrollment. In general, participants enrolling in the CHS were younger, more highly educated, and more likely to have previously quit smoking.34 Similar trends were noted among the 70% of Forsyth County CHS participants who consented to RDS examination.26 The resulting “healthy cohort” effect may have limited the number of adverse coronary events and individuals with clinically significant RVD for analysis. An additional factor that may have contributed to the relatively low number of adverse coronary events for analysis was the brief follow-up interval (mean, 14 months) available for analysis. Collectively, these limitations may have reduced the statistical power to define additional associations between RVD and subsequent adverse coronary events.
With these limitations aside, the Forsyth County CHS cohort represented a unique opportunity to examine issues related to RVD and subsequent coronary events in a large sample of well-characterized, elderly, African American and white men and women. The findings presented herein have important implications for the general health and management of RVD among such individuals. Previous investigation of this cohort demonstrated a 6.8% prevalence of RVD.26 Applying this prevalence estimate to year 2000 US census data, an estimated 2.6 million elderly Americans may have RVD, and this number is expected to grow as the general population ages during the next 2 decades. These results suggest that such individuals are at a significantly increased risk of adverse coronary events, much like individuals with other previously identified markers of generalized atherosclerosis such as depressed ankle-arm index, increased carotid intimal-medial thickness, electrocardiographic abnormalities, and echocardiographic abnormalities.21,22,35- 37 Thus, individuals with RVD represent an important target population for aggressive secondary prevention efforts.
In summary, the presence of RVD was significantly and independently associated with incident and recurrent adverse coronary events in a population-based sample of free-living elderly Americans. The presence of RVD imparted a 96% increase in the risk of adverse events, and the incremental risk was not due to shared associations with atherosclerotic risk factors, other prevalent manifestations of CVD, or associated hypertension.
Correspondence: Kimberley J. Hansen, MD, Division of Surgical Sciences, Department of General Surgery, Wake Forest University School of Medicine, Medical Center Boulevard, Winston-Salem, NC 27157-1095 (firstname.lastname@example.org).
Accepted for Publication: August 26, 2004.
Financial Disclosure: None.
Funding/Support: This study was supported by grants from the National Institute of Diabetes and Digestive and Kidney Diseases (1-R01-DK-47414) and the National Heart, Lung, and Blood Institute (N-01-HL-85079), Bethesda, Md.
Role of the Sponsor: The study sponsors provided initial peer review of the study design but performed no other active roles in the collection, analysis, interpretation, or reporting of these data.
Previous Presentation: This study was presented at the American Heart Association Scientific Sessions; November 11, 2003; Orlando, Fla.
Acknowledgment: We acknowledge the technical assistance and expertise of Scott Reavis, RVT, and Jill Leighton, RVT, whose efforts were indispensable to this project.