TUESDAY, Nov. 11 (HealthDay News) -- A Danish study links high levels of the blood fats called triglycerides with an increased risk of stroke -- and the way those levels were measured could change a basic medical practice, one of the researchers says.

The 31-year study of almost 14,000 Danish men and women found a direct association between higher triglyceride levels and risk of ischemic stroke, the most common kind that occurs when a clot blocks a brain blood vessel, said the report in the Nov. 12 issue of the Journal of the American Medical Association.

"Our results are really quite clear," said Dr. Borge G. Nordestgaard, a professor of genetic epidemiology at Copenhagen University Hospital. "For the highest level, those with triglycerides above 443 milligrams per deciliter [of blood], they have a three- to four-fold higher risk of ischemic stroke, compared to those with the lowest levels, less than 89 milligrams per deciliter."

That association "is much higher than previously found for elevated cholesterol levels," Nordestgaard noted.

While several studies have linked elevated triglyceride levels with an increased risk of heart attack, this is one of only a few to look at the association with stroke, he said.

And the new study did not measure triglyceride levels in the usual way -- after an eight- to 12-hour fast, Nordestgaard said. Instead, blood readings were taken at any time, regardless of when the last meal had been eaten. These no-fasting readings are better for two reasons, he said: "First, they are simpler for the patient, and second, they are better at defining people at higher risk of stroke and myocardial infarction [heart attack]."

"It potentially has profound importance for clinical practice, because it suggests that fasting before lipid profile measurement is not necessary," Nordestgaard said. "This is important for millions of patients worldwide."

All hospitals in Copenhagen now use the non-fasting method of testing blood triglyceride levels, Nordestgaard said. The change is spreading to other cities in Denmark and possibly to other Scandinavian countries, he said.

The case for non-fasting measurements was made in another paper outlining results of tests of 34,000 people, Nordestgaard said. "We measured standard lipids [fats], looking at how much does the reading change after eating normal food," he said. "The changes were really minimal and of no clinical importance. Based on that finding and the JAMA paper, we believe it might be better to do all testing in a non-fasting state."