RETHINKING STROKE RISKS:
LIFELONG HYPERTENSION AND VALVULAR HEART DISEASE EMERGE AS STRONG STROKE PREDICTORS

NEW ORLEANS—New research may prompt a re-examination of two well-known stroke risk factors—high blood pressure and heart disease. At the 25th International Stroke Conference, Sudha Seshadri, MD, reported that past periods of high blood pressure are a strong predictor of stroke risk in elderly patients even if current blood pressure is normal. Also reporting at the same meeting, George W. Petty, MD, presented evidence suggesting that valvular heart disease is a stronger predictor of stroke than had previously been recognized.

PREVIOUS HYPERTENSION RAISES STROKE RISK

Current blood pressure is not the only thing that counts, said Dr. Seshadri, a researcher with the Framingham Heart Study. "We all know that hypertension is the most important risk factor for stroke," Dr. Seshadri said. She noted that risk is usually calculated using current blood pressure levels, and this is based on the assumption that previous levels are inconsequential. However, stroke is primarily a disease of the elderly, and up to one third of these strokes occur in normotensive subjects. The Framingham investigators suspected that this might reflect hypertensive damage from earlier periods. "The aim of this study was to evaluate the effect of past blood pressure levels on the future risk of ischemic stroke," Dr. Seshadri said.

The study sample included the 5,197 subjects in the original Framingham cohort, on whom there is 50 years of information. Patients were free of stroke at baseline. Blood pressure was recorded once every two years, and there was continuous surveillance for stroke throughout this period.

The investigators classified subjects by their age at baseline. The main exposure variable was blood pressure. Dr. Seshadri examined both current blood pressure and past blood pressure. "For the past blood pressure we looked at the previous decade as well as earlier decades and averaged all the available blood pressure records in the decade of interest," she said. The outcome variable was the time to first completed ischemic stroke. Transient ischemic attacks (TIAs) were not counted as end points.

Dr. Seshadri's analysis of the 10-year risk of stroke for 70-year-old subjects included separate analyses of systolic blood pressure, diastolic blood pressure, and pulse pressure. "Since these are 70-year-olds, we examined past blood pressure between the ages of 60 to 69, and 50 to 59, and adjusted for smoking and diabetes," she said.

Dr. Seshadri reported that 2,668 Framingham participants reached the baseline age of 70 alive and free of stroke. Over the subsequent 10 years, 114 of 1,634 women and 92 of 1,034 men had an ischemic stroke.

Current blood pressure determines the risk of stroke over the next 10 years in women at age 70, and the relative risk is 1.46 to 1.61, Dr. Seshadri said. "After adjusting for the current blood pressure, when we looked at blood pressure in the previous 10 years we found that there was an additional 40% to 64% increase in the risk of stroke per standard deviation increment in the blood pressure. Similarly, for the period 10 to 20 years before the base period, when the subject was between 50 and 59, after adjusting for current blood pressure there was still a 37% to 45% increase in the risk of stroke per standard deviation increment in blood pressure."

Similarly, for men at age 70, after adjusting for current blood pressure, blood pressure in the previous decade showed a further 10% to 28% increase in the risk per standard deviation increment. Blood pressure 10 to 20 years earlier, while the subject was in his fifties, showed a 38% to 45% increase in the risk of stroke.

"Past blood pressure influenced the future risk of ischemic stroke, even after adjusting for current blood pressure. This is a robust effect that is seen in both sexes," Dr. Seshadri said. "It is also seen at baseline ages of 60 and 80. It is seen for all the blood pressure components we looked at. We did a subgroup analysis of people who at the base age had normotensive blood pressure, less than 140/90 mm Hg, and found the same effect."

In conclusion, Dr. Seshadri said that antecedent blood pressure is important over and above current blood pressure in determining the risk of ischemic stroke in older adults. Although this is an observational study, she added, it suggests that effective prevention of stroke in later life is likely to require better control of blood pressure throughout life.

VALVULAR HEART DISEASE—AN UNDERESTIMATED STROKE RISK?

Valvular heart disease increases the risk of stroke and transient ischemic attack (TIA) threefold, reported Dr. Petty. His population-based study of rates and predictors of cerebrovascular events in patients with valvular heart disease also indicated that aortic stenosis may be either an underrecognized source of cardiac embolism or a marker for underlying vascular disease.

"Despite the decline in rheumatic valve disease and recent advances in medical and surgical therapy for cardiac disease, our study demonstrates that the relative risk of cerebrovascular events in valve disease patients remains high: 3.2 times the age- and sex-adjusted rates in the general population," said Dr. Petty, an Associate Professor of Neurology at the Mayo Medical School, Rochester, Minnesota.

All residents of Olmsted County, Minnesota, who had a first 2-D color Doppler echo diagnosis of moderate or severe mitral stenosis, mitral regurgitation, aortic stenosis, aortic regurgitation, or any combination thereof, between 1985 and 1992, and who did not have a history of cerebrovascular disease, were included in the cohort.

Inclusion criteria were met by 729 subjects (age range, 13 to 107). The cohort was followed for 2,694 person-years; the mean follow-up was 3.5 years. During that time, 356 patients died and 98 had cerebrovascular events: amaurosis fugax (four), TIA (20), ischemic stroke (68), and intracerebral hemorrhage (six). Syncope alone was not considered an end point.

"Kaplan-Meier estimates of rates of development of stroke and TIA for the entire cohort were significantly greater than the corresponding age- and sex-adjusted rates for the community, with a standardized morbidity ratio of 3.2," Dr. Petty said. "Rates of cerebrovascular events for the four major categories of valve disease adjusted for age and sex were not significantly different, but this information doesn't take into consideration differences in the hemodynamic severity of the valve disease."

Age and atrial fibrillation were predictors for cerebrovascular events. Atrial fibrillation conferred a greater risk of cerebrovascular events at a younger age. After adjustment for age and gender in atrial fibrillation, severe aortic stenosis emerged as an independent risk factor (risk ratio, 3.5) for cerebrovascular events. A clinical diagnosis of valve disease prior to the echo diagnosis was not a significant determinant risk, he said.

During follow-up, 122 members of the cohort (17%) underwent aortic or mitral valve repair or replacement. Having had such a procedure was not a predictor for survival free of cerebrovascular events. Approximately 50% of the cohort received anticoagulation therapy after the diagnosis of valvular disease. A nested case control analysis adjusted for age and atrial fibrillation showed that neither current nor prior warfarin sodium had any effect on the occurrence of cerebrovascular events.

"Previous studies of the relationship between valve disease and stroke derived from patients followed during earlier decades and focused on the importance of mitral stenosis as a source of thromboembolism," noted Dr. Petty. "The presumption has been that aortic stenosis is not commonly associated with stroke, although stroke occasionally may be due to calcific embolization. In our study, aortic stenosis was associated with rates of development of cerebrovascular events similar to those for mitral disease and was an independent determinant for the development of cerebrovascular events, even after adjusting for age and atrial fibrillation."

—Janis Kelly
Contributing Writer

Return to table of contents