INTERSTROKE: Ten Modifiable Risk Factors Explain 90% of Stroke Risk

June 18, 2010

June 18, 2010 (Beijing, China) — A large case-control study evaluating risk factors for stroke has shown that 10 risk factors are associated with 90% of the risk of stroke and that of these modifiable risk factors, hypertension is the most important for all stroke subtypes and is a particularly dangerous risk factor for intracerebral hemorrhage [1].

Dr Martin O'Donnell

"A lot of the evidence for stroke comes from inference from heart disease, and certainly the risk factors are the same, but the relative importance of the risk factors is different," lead investigator Dr Martin O'Donnell (McMaster University, Hamilton, ON) told heartwire . "We've gone through phases where we've thought the risk factors for heart disease and stroke were different, or where we thought they were the same. The important message is that these identified risk factors are important for reducing the risk of stroke, and some appear to be more important than anticipated."

The results of the study, known as INTERSTROKE, were presented this week here at the World Congress of Cardiology (WCC) and are published online June 18, 2010 in the Lancet. Not unlike the previously published INTERHEART study of coronary heart disease, a trial led by Dr Salim Yusuf (McMaster University, Hamilton, ON) that identified nine modifiable risk factors accounting for 90% of disease, this study showed that many strokes can be predicted and that relatively simple measures, such as blood-pressure control, could reduce the burden of disease.

"If you take myocardial infarction, where, say, 50% of the risk is just the lipids, in low-income settings you have to take and analyze the blood, which might not always be available," said O'Donnell. "Blood pressure is different. You can tell the patient they have a risk factor, which can be measured in any setting, and it doesn't require a lot of medical expertise. It can be modified with generic medications, and it can also be modified at the population level by implementing such policies as those aimed at reducing salt intake."


At the WCC meeting, organizers and presenters have highlighted the global burden of stroke, noting that countries of low and middle income are disproportionately affected by the disease. These low- and middle-income countries, for example, account for more than 85% of stroke mortality worldwide. The contribution of various risk factors to stroke burden is not entirely known, however, particularly in lower-income countries, because most of the data from clinical trials are derived from developed or Westernized countries.

INTERSTROKE is a standardized, case-control study looking at the importance of established and emerging risk factors for the common stroke subtypes in different regions. In total, 3000 first acute-stroke cases and 3000 controls from 22 countries were included in the analysis. Of the stroke patients, just 14% were from a high-income country, while 81% were from Southeast Asia, India, or Africa.

Overall, self-reported hypertension was the strongest risk factor for stroke and was stronger for intracerebral hemorrhage than for ischemic stroke. A history of hypertension was associated with a more than 2.5-fold increased risk of stroke. When a stricter definition of hypertension was used--blood pressure >160/90 mm Hg--the strength of the association increased.

Along with hypertension, current smoking, abdominal obesity, diet, and physical activity accounted for 80% of the global risk of stroke, explaining 80% of the risk of ischemic stroke and 90% of the risk of hemorrhagic strokes. When additional risk factors were included in the model, including diabetes mellitus, alcohol intake, psychosocial factors, the ratio of apolipoprotein B to A1, and cardiac causes (atrial fibrillation or flutter, previous MI, and valve disease), these 10 risk factors accounted for 90% of the risk of stroke. Hypertension, smoking, abdominal obesity, diet, and alcohol intake were the most important risk factors for intracerebral hemorrhagic stroke.

INTERSTROKE: Population-Attributable Risk for Common Risk Factors

Risk factor Population-attributable risk, % (99% CI)
Hypertension 34.6 (30.4–39.1)
Smoking 18.9 (15.3–23.1)
Waist-to-hip ratio (tertile 2 vs tertile 1) 26.5 (18.8–36.0)
Dietary risk score (tertile 2 vs tertile 1) 18.8 (11.2–29.7)
Regular physical activity 28.5 (14.5–48.5)
Diabetes 5.0 (2.6–9.5)
Alcohol intake 3.8 (0.9–14.4)
Cardiac causes 6.7 (4.8–9.1)
Ratio of apolipoprotein B to A1 (tertile 2 vs tertile 1) 24.9 (15.7–37.1)
Psychological factors  
Stress 4.6 (2.1–9.6)
Depression 5.2 (2.7–9.8)

"The issue for us was whether or not the same nine risk factors were as important for stroke," said O'Donnell. "Before INTERHEART was done, there was the belief that 30% to 40% of the risk of myocardial infarction was unexplained. Similarly, some people believe that about 30% of stroke isn't explained and have been pursuing all sorts of other markers, genetics, and so on. Here, we show that these risk factors, known or proposed to be important, have now been quantified and can be extended to other regions of the world."

To heartwire , O'Donnell noted that epidemiological studies have failed to show a consistent relationship between total cholesterol and stroke risk, a finding confirmed in the INTERSTROKE study. In this analysis, the researchers found no association with total and non-HDL cholesterol for ischemic stroke risk but did observe a strong association between apolipoprotein and HDL-cholesterol levels and the risk of ischemic stroke. Interestingly, the group observed that the reduction in risk of ischemic stroke associated with elevated apolipoprotein A1 and HDL cholesterol was larger than the increase in risk associated with increased levels of apolipoprotein B or non-HDL cholesterol.

Just Like Your Mom Told You

O'Donnell said the most important message from the study is the importance of blood pressure. In the INTERHEART study, the risk of coronary artery disease attributable to hypertension was 18% but was nearly double that in stroke patients. He advises clinicians to treat blood pressure to target and urges the implementation of policies to lower blood pressure at the population level. In addition to these risk factors, quitting smoking, losing weight, getting active, and eating well--"everything that your mom told you to do"--is highly recommended, he said.

In an editorial accompanying the published study [2], Dr Jack Tu (University of Toronto, ON) noted that identifying the cause of stroke across diverse regions is a difficult task, a thought O'Donnell also echoed. The INTERSTROKE researchers weren't sure initially whether the study could be conducted, because the confirmation of stroke and the type of stroke requires access to imaging technology.

Tu noted that INTERSTROKE confirms that hypertension, a well-known risk factor for stroke in developed countries, is also a risk factor in developing nations. "This finding is particularly relevant, because it highlights the need for health authorities in these regions to screen the general population for high blood pressure and, if necessary, offer affordable treatment to reduce the burden of stroke," writes Tu.