Patients with a very low or a very high hematocrit are at higher risk for death after a stroke, and anemic patients are at greatest risk, according to a study presented at the International Stroke Conference sponsored by the American Heart Association on Feb. 2.

Previous studies have shown that extremes of hematocrit increase mortality after myocardial infarction, congestive heart failure, and kidney disease. Dr. Jason J. Sico of the VA Connecticut Healthcare System and researchers from the Department of Veterans Affairs medical system explored whether there was a similar association in stroke. Previous stroke studies had not adjusted for stroke severity or a large number of comorbidities.

They found that “among stroke patients, severe anemia is a potent predictor of dying throughout the first year after a stroke,” said Dr. Sico, who is also an assistant professor of neurology at Yale University, New Haven, Conn.

The researchers abstracted medical records for a sample from 131 Veterans Health Administration (VHA) hospitals of 3,965 patients admitted for a confirmed diagnosis of ischemic stroke in fiscal 2007. Patients with unavailable hematocrits, those who received thrombolytics, or those whose charts had inconsistent death dates were also excluded.

“Among stroke patients, severe anemia is a potent predictor of dying throughout the first year after a stroke.”

The hematocrit, taken from 24 hours of admission, was divided into six tiers: less than or equal to 27% (defined as severe anemia); 28%-32% (moderate anemia); 33%-37% (mild anemia); 38%-42% (normal); 43%-47% (normal); and greater than or equal to 48% (polycythemia).

Researchers adjusted for age, National Institutes of Health Stroke Scale (NIHSS) score, comorbidities (including pneumonia, hypertension, hypercholesterolemia, diabetes, and history of cancer and heart disease), and Acute Physiology and Chronic Health Evaluation (APACHE)-III scores.

A total of 2% of the 3,750 patients analyzed had severe anemia, 6.2% had moderate anemia, and 17.9% had mild anemia. About 64% were in the normal categories. A total of 9% had a high hematocrit of greater than or equal to 48%.

People with lower hematocrits tended to be older and have higher APACHE scores, a higher Charlson index, a history of heart disease, and were more likely to have diabetes.

The risk of death was 2.5 to 3.5 times higher for patients with severe anemia (P = .013 for in-hospital and 30-day mortality; P = .002 at 6 months and P = .001 at 1 year). A high hematocrit was independently associated only with in-hospital mortality (OR 2.9, P = .004).

The study showed that having a history of severe anemia put stroke patients at a higher risk of death than having a history of other conditions, including cancer and heart disease, said Dr. Sico.

There are several potential mechanisms to explain why anemia might increase the risk of death, he said in an interview. With lower hematocrits, less blood and less oxygen circulate to various parts of the body. Long-term anemia also impairs the ability of the brain’s blood vessels to respond appropriately to a stroke.

A higher hematocrit also decreases blood flow to brain, said Dr. Sico. It causes a more turbulent blood flow to the brain, which could predispose the stroke patient to having a bad outcome.

Dr. Sico said that the study’s findings were limited to men because no women had been analyzed.

But he concluded that stroke patients with low or high hematocrits should be evaluated for potentially reversible causes, and that they should also be closely monitored during the post-stroke period.

Dr. Sico reported having no financial disclosures. The study was funded by the VA Health Services Research and Development Quality Enhancement Research Initiative.