Factors Associated with Incident Ischemic Stroke in Hospitalized Heart Failure Patients: A Pilot Study

289 Hypertens Res Vol.31 (2008) No.2 p.289-294 Original Article Factors Associated with Incident Ischemic Stroke in Hospitalized Heart Failure Patients: A Pilot Study Takahiro KOMORI1),2), Kazuo EGUCHI2), Hidenori TOMIZAWA2), Joji ISHIKAWA2), Satoshi HOSHIDE2), Kazuyuki SHIMADA2), and Kazuomi KARIO2) Stroke is sometimes seen in patients with congestive heart failure (CHF). The factors that best predict incident stroke in hospitalized CHF patients are not well known. We performed this pilot study to explore the clinical markers of incident stroke in CHF patients. We studied 111 hospitalized patients with CHF (mean age, 67 ± 11 years). Ambulatory blood pressure (BP) monitoring, blood tests, and echocardiography were performed in these patients just before they left the hospital, and all cardiovascular events during the study period were followed for an average of 18 ± 9 months. Cox regression analysis was performed to explore the predictors of incident stroke using age, sex, body mass index (BMI), casual and ambulatory systolic BP (SBP), and brain-type natriuretic peptide (BNP). There were 10 stroke events (9%) during the follow-up period. The stroke group had higher nocturnal SBP and plasma BNP levels than the non-stroke group. With Cox regression analysis, both nocturnal SBP and BNP were significant predictors of incident stroke independent of other covariates. When nocturnal BP of 120 mmHg and BNP of 600 pg/mL (75th percentile) were used as cutoffs, nocturnal SBP ≥ 120 mmHg was associated with a 7-fold increase in the risk of incident stroke, while BNP ≥ 600 pg/mL was associated with a 46.6-fold increase. However, abnormal circadian BP patterns were not associated with incident stroke. In this pilot study, elevated nocturnal BP and high plasma BNP just before patients left the hospital were significant predictors of stroke events in CHF patients. Further study is needed to confirm this hypothesis. (Hypertens Res 2008; 31: 289–294) Key Words: congestive heart failure, stroke, nocturnal blood pressure, brain-type natriuretic peptide Introduction Congestive heart failure (CHF) is commonly seen in elderly populations (1), and recent epidemiological studies have shown that its prevalence has been increasing in a growing number of elderly populations (2). Despite progress in antihypertensive treatment, the incidence of CHF has not decreased (3). Incident stroke, which sometimes accompanies CHF (4, 5), is one of the most serious complications in the clinical course of CHF. CHF itself has been reported to be associated with a two- to three-fold increased risk of incident stroke (6). Various heart diseases that include CHF are described as stroke risk factors in the American Heart Association Stroke Guidelines (7). The mechanisms of stroke seen in CHF patients may include intracardiac thrombus caused by left ventricular (LV) dysfunction (8), atrial fibrillation, or an atherosclerotic risk factor such as diabetes, hypertension, or hyperlipidemia (9). However, the clinical factors most closely associated with incident stroke in hospitalized CHF patients remain unclear. Thus, we performed this pilot study to explore the clinical From the 1)Department of Cardiovascular Medicine, Utsunomiya Social Insurance Hospital, Utsunomiya, Japan; and 2)Division of Cardiovascular Medicine, Jichi Medical University School of Medicine, Shimotsuke, Japan. Address for Reprints: Kazuomi Kario, M.D., COE (Center of Excellence) Program, Division of Cardiovascular Medicine, Jichi Medical University School of Medicine, 3311–1, Yakushiji, Shimotsuke 329–0498, Japan. E-mail: Received July 12, 2007; Accepted in revised form September 13, 2007. 290 Hypertens Res Vol. 31, No. 2 (2008) predictors of stroke. Methods Study Population We completed baseline examinations in 117 patients (mean age: 67 ±11 years; 73 men, 44 women) who were currently hospitalized with a diagnosis of CHF at Jichi Medical University Hospital, Tochigi, Japan. The baseline examinations were performed when the patient condition was stabilized, and in most cases occurred just before the patient left the hospital. The study was performed from July 2002 to March 2004. All of the patients were seen by cardiologists. We excluded patients with renal failure (serum creatinine > 3 mg/dL), dementia, cancer, or other severe noncardiovascular disease. Informed consent was obtained from all participants. Six patients lost to follow-up were excluded from the study, and finally 111 patients were analyzed. Diabetes mellitus was defined by either fasting glucose level ≥ 126 mg/dL, a random nonfasting glucose level ≥ 200 mg/dL, or the use of antidiabetic drugs or insulin (10). Hyperlipidemia was defined as total cholesterol level > 220 mg/dL or the use of an oral lipid-lowering drug. Body mass index (BMI) was calculated as weight (kg)/height2 (m2). Electrocardiographically verified LV hypertrophy was defined as an abnormally high voltage of QRS complex (R in V5 plus S in V1 > 3.5 mV) associated with either flat T-waves (< 10% of the R-wave) or ST-segment depression and biphasic T-waves (11). Twenty-Four–Hour Ambulatory Blood Pressure Monitoring When CHF was stabilized, noninvasive ambulatory blood pressure (BP) was monitored by an automatic system using electric cuff inflation (TM-2425, A&D Co., Tokyo, Japan), which recorded both BP (by the oscillometric method) and pulse rate every 30 min for 24 h. The accuracy of this device was validated previously (12). Nocturnal BP was defined as the average of BP measurements during the time the patient was in bed, and awake BP was defined as the average of BP measurements recorded during the rest of the day. We subclassified the patients as either dippers, nondippers, or risers according to the percentage of nocturnal systolic BP (SBP) reduction (100 × (1 − nocturnal SBP/awake SBP)), as follows: in dippers, the fall was between 10% and 20%; in nondippers it was between 0% and 10%; and in risers it was less than 0%. Other Examinations Blood was drawn after 10 min rest in the supine position. Brain-type natriuretic peptide (BNP) was measured from unextracted plasma using highly sensitive, noncompetitive immunoradiometric assays (Shiono-RIA; Shionogi, Osaka, Japan). Transthoracic two-dimensional echocardiography (Sonos 5500, Philips, Andover, USA) was performed in all subjects. The left ventricular ejection fraction (LVEF) was calculated by the Teichholz method (13). Follow-Up and Outcome A follow-up study was performed from September 1 to December 31, 2005. The mean follow-up period was 18±9 months. The patients’ medical records were reviewed when they were followed in the same hospital. When the patients were followed in other hospitals, we interviewed them by mail or telephone. Stroke events were the main outcome of this study, including cerebral infarction, cerebral hemorrhage, and subarachnoid hemorrhage. There were no significant differences between the stroke and non-stroke groups in the durati (...truncated)