Renal dysfunction and risk of ischemic stroke or TIA in patients with cardiovascular disease

Study sample.

Patients with documented CHD (n = 15,524) were screened between February 1990 and October 1992 for inclusion in the Bezafibrate Infarction Prevention (BIP) Study, a placebo controlled secondary prevention randomized clinical trial, that assessed the effect of bezafibrate on the risk of mortality or recurrent events.¹⁵ The age range of the screened patients was within a predefined range of 45 to 74 years, with the exception of 0.3% of younger patients. During the first physician visit, records were obtained on medical history, conventional reported cardiovascular risk factors, and medications used, and a complete physical examination was carried out. Patients with total cholesterol ≤7.0 mmol/L (270 mg/dL), high-density liproprotein cholesterol (HDL-C) ≤1.17 mmol/L (45 mg/dL), and serum triglycerides ≤3.39 mmol/L (300 mg/dL) at the time of the first visit were invited to a second visit after a 2-month diet (n = 6,993). For those patients, several blood levels, including serum creatinine, were measured. Patients meeting selection criteria who provided an informed consent were included in the BIP randomized clinical trial (total 3,090). Inclusion criteria for the trial comprised the following: age 45 to 74 years, history of either myocardial infarction ≥6 months but <5 years before enrollment into the study or stable angina pectoris confirmed by coronary angiography, radionuclear studies, or standard exercise tests. At this stage, a more restrictive lipid profile of serum total cholesterol between 180 to 250 mg/dL, low-density lipoprotein cholesterol (LDL-C) ≤180 mg/dL (≤160 mg/dL for patients <50 years), HDL-C ≤45 mg/dL, and triglycerides ≤300 mg/dL was required. The main exclusion criteria were insulin-dependent diabetes mellitus, severe heart failure, unstable angina pectoris, disabling stroke, and hepatic or renal failure. Within the current analysis, patients included in the RCT had similar characteristics to those not included with the exception of their lipid profiles (because of the lipid selection criteria). For example, mean ages for the non-included and included patients were 59.4 vs 59.8 years, systolic blood pressure 134 and 133 mm Hg, diastolic blood pressure 81 mm Hg for both groups, and body mass index 26.8 and 26.7 kg/m². Bezafibrate users and non-users in this trial exhibited comparable rates of stroke (4.6% vs 5%, p = 0.66).¹⁵ In order to account for any differences in management and treatment between patients included or not in the clinical trial, we adjusted all analyses for inclusion in the clinical trial as well as for use of bezafibrate or any other lipid modifying drug at baseline. We excluded patients with a history of prior stroke or TIA in order to assess the risk of a first clinically manifest cerebrovascular event associated with renal dysfunction. We also excluded patients without known mortality status (who were neither citizens nor permanent resident), or without event dates (ICD-9 code of post stroke but no record of stroke in the hospital records) or with missing data in some important variable (creatinine, lipids). Of patients with measures of creatinine clearance, 270 non-citizens did not have survival status, 12 did not have event date, and 10 lacked one or more variables required for the analysis. The total number of patients included in the present analysis was 6,685.

Laboratory measurements were all performed at a central study laboratory (the Physiologic and Hygiene Laboratory at the Wolfson Medical Center). Blood samples were taken after at least 12 hours of fasting. All analyses were performed with a Boehringer-Hitachi 704 random access analyzer using Boehringer diagnostic kits. Accuracy and precision for total and LDL-C as well as triglyceride were under periodic surveillance by the Centers for Disease Control and Prevention service in Atlanta, GA. Serum creatinine levels were measured employing the Jaffe method without deproteinization.

Patients included in the BIP randomized clinical trial were routinely followed up every 4 months, during which data on the occurrence of new cerebrovascular events were obtained as a part of the study procedures. For all other patients we obtained computerized data files (based on national ID number and name) from hospitals participating in the study screening process up to the end of 1998. Hospitalizations with a diagnosis of cerebrovascular disease (ICD-9 codes 430-438 or code 38.1) were identified. We also matched the patients against a registry of all hospitals but one and the Clalit Health Services (insuring over 60% of population) participating in the screening process. Mortality data were obtained through January 1999 from the Israel Population Registry, with cause of death coded according to ICD-9 codes. The registry is virtually complete for mortality.

Patients were identified and attainable medical records and hospital discharge summaries were reviewed systematically. Data were collected on history, findings on neurologic examination, brain CT, and ancillary examinations as available. A study stroke neurologist reviewed all cases. Stroke was defined according to World Health Organization (WHO) criteria.¹⁶ Ischemic stroke and intracerebral hemorrhage were differentiated by the results of brain CT performed at the acute stage. Ischemic stroke was diagnosed if the patient had an appropriate clinical event and had a brain CT that showed a compatible low-density lesion or was normal, or had findings compatible with hemorrhagic conversion of a cerebral infarct. Events resolving completely within <24 hours were diagnosed as TIA. Of 451 cases considered having any incident ischemic cerebrovascular disease by ICD-9 codes, in 287 patients the diagnosis of ischemic stroke or TIA was confirmed after review of medical records, and these cases were regarded as the endpoint for this analysis. Remaining patients were admitted for carotid endarterectomy surgery for asymptomatic disease, or medical records were not available for review, or brain CT was not performed, so that the type of cerebrovascular disease could not be determined.

Statistical methods.

Data were analyzed with SPSS software version 11.0. Glomerular filtration rate (GFR) was estimated using the Cockroft-Gault equation: GFR (mL/minute) = [(140-age) × weight (kg)]/[serum creatinine (mg/dL) × 72], corrected in women by a factor of 0.85.¹⁷ A secondary confirmatory estimation used the four-component Modification of Diet in Renal Disease (MDRD) equation,^18,19 incorporating age, race, sex, and serum creatinine level: GFR = 186 × (serum creatinine level [mg/dL])^-1.154 × (age [in years])^-0.203. For women, the product of this equation was multiplied by a correction factor of 0.742. GFR ≤60 mL/minute/1.73 m² defined CKD as defined by the National Kidney Foundation and American Heart Association.^2,3 In addition, we categorized serum creatinine levels as <0.9, 0.9 to 1.10, 1.10 to 1.29, 1.30 to 1.49, 1.50 to 1.69, and ≥1.70 mg/dL to evaluate the presence of a trend with serum creatinine and cerebrovascular event risk.^4,6,20 Serum creatinine and GFR were introduced into the models once as categorical and once as continuous variables. Hazard ratios (HRs) and 95% CIs for incident ischemic stroke or TIA by serum creatinine, GFR, and renal dysfunction were estimated using Cox proportional hazard models adjusting for potential confounders. Conventional risk factors were selected for the model based on an association with stroke in univariate analysis or findings of significant associations in previous studies. The log rank test was calculated to compare the cumulative event-free curves by categories of serum creatinine and by presence of CKD, adjusting for potential confounders.

Strength and limitations.

The current study highlights the significance of mild degrees of renal dysfunction in the prediction of ischemic cerebrovascular events in a large prospective cohort of patients with stable atherothrombotic disease. It has the limitation of using estimations of GFR rather than direct measurement and evaluation at a single time-point, without addressing change over time. Data on microalbuminuria^24,25 or other renal-specific factors were not available for this cohort. Data on the incidence of ischemic stroke or TIA were obtained through medical records from participating hospitals. Patients with minor strokes not admitted to hospital and those who had been re-admitted to a nonparticipating hospital would have been missed. Incidence rates, however, were comparable in patients followed up routinely as part of a randomized clinical trial¹⁵ and patients for whom data were obtained through medical records. Complete medical records were not available in all cases, but comparable findings were observed also for the endpoint of all ischemic cerebrovascular disease (n = 451) defined by ICD codes. Finally, our findings are based on a group of patients with pre-existing stable atherothrombotic disease and defined blood lipid boundaries, and caution should be used in generalization of these findings.

CKD is common in patients with atherosclerotic vascular disease and associated with a high prevalence of risk factors, suggesting that attention to risk factor management may affect the risk of stroke in these patients, already at high risk. Risk factors for atherosclerosis are often inadequately addressed in patients with CKD. In one study, 35% of those with CKD and established cardiovascular disease had a blood pressure >140/90 mm Hg; 45% were receiving aspirin, and only 50% of those with hyperlipidemia were on statins.²⁶ The National Kidney Foundation Task Force on Cardiovascular Disease and other authoritative guidelines recommended that patients with CKD be considered in the highest risk group for cardiovascular events.^2,3 Our study supports this recommendation and expands it also for ischemic stroke.