Boosting “good” HDL cholesterol did not protect against heart attacks and strokes in an eagerly awaited study that was abruptly stopped when the surprising outcome became clear, officials said Thursday.

The federally funded study of more than 3,400 U.S. adults at high risk for heart attacks and strokes was halted 18 months early after researchers realized that the HDL-boosting drug niacin failed to cut the risk — as was hoped and suggested by many earlier, smaller studies, officials said. In fact, there was a hint that niacin might increase the risk for one type of stroke.

The findings are the latest blow to what had been one of the most promising hopes for reducing the toll from cardiovascular disease, which kills about 800,000 Americans each year, making it the leading cause of death in the United States.

“This sends us a bit back to the drawing board,” said Susan B. Shurin, acting director of the National Heart, Lung and Blood Institute, which funded the $52.7 million study.

Other experts agreed.

“This is surprising, given some of the preliminary data,” said Ralph Sacco, president of the American Heart Association. “We are always looking for new, more effective ways to reduce cardiovascular risks. It’s disappointing.”

The researchers emphasized, however, that they were not ready to abandon the approach. Raising good cholesterol with another drug might be beneficial or higher levels might help other types of patients, such as those who are less stable than those in the study, they said.

“We don’t know if the same outcome would have occurred in a different population or with a different drug,” said Jeffrey Probstfield, a professor of medicine and epidemiology at the University of Washington in Seattle who helped lead the study.

The risk for heart attacks and strokes can be reduced by cutting “bad” LDL cholesterol, which can clog arteries. Researchers hoped that a new generation of drugs that boost HDL cholesterol, which is thought to clear LDL, would provide an important new tool. A large body of research had found that people with low HDL levels were more likely to have heart attacks and strokes.

Doubts about the HDL-boosting hypothesis began to emerge, however, when the HDL-raising drug fenofibrate failed in 2005 to cut heart attacks and strokes among diabetics. Another study the next year found that a different HDL-raising drug, torcetrapib, increased the risk. But researchers held out hope that niacin might be beneficial.

To put that strategy to the first rigorous test, researchers launched the AIM-HIGH study involving 3,414 people with a history of heart disease. All were taking the statin Zocor to keep their LDL levels low but remained at high risk because they had low HDL and high triglycerides.

The subjects took either a high-dose, extended-release prescription version of niacin known as Niaspan or a placebo. Some also took another cholesterol-lowering drug called Zetia. HDL levels rose and triglycerides dropped as expected among those taking niacin, which is also known as vitamin B3. But an independent panel of experts monitoring the experiment terminated it April 25, after just an average of 32 months of follow-up, when it concluded that those taking niacin were no less likely to have a heart attack or stroke.

“The lack of effect on cardiovascular events is unexpected and a striking contrast to the results of previous trials and observational studies,” Probstfield said.

Moreover, the analysis detected a small increase in “ischemic” strokes among participants taking niacin. Such strokes occur when blood flow to the brain is obstructed. There were 28 strokes among those taking niacin, compared with 12 among those taking a placebo.

It remains unclear whether the high stroke rate was a fluke or related to the drug, the researchers said. Nine of the strokes that occurred among those taking niacin occurred at least two months and up to four years after they stopped taking the drug.

Study participants were immediately informed about the findings so they could schedule a doctor’s examination within the next three months. They will continue to be followed for at least a year.

In a statement, Abbott Laboratories, which makes Niaspan, said it would be premature to conclude that the findings apply to all patients.

“The relevance of these results to patients outside the study population is currently unknown and it would be premature to extrapolate these results to a broader patient population at this time,” the company said.

In a statement posted on its Web site, the Food and Drug Administration said it was not recommending any changes in niacin use yet based on the study, but would analyze the data when it becomes available.