Three days after warning that the prescription painkiller Celebrex can increase the risk of heart attacks and strokes, federal health officials yesterday raised questions about the potential for similar problems with a principal over-the-counter alternative, Aleve.

"This is a very confusing situation," said Dr. Sandra Kweder, deputy director of the Food and Drug Administration's Office of New Drugs. But the FDA stopped short of urging patients to seek alternatives to Aleve, its generic versions or its prescription forms Naprosyn and Anaprox, as the agency had done Friday with Celebrex.

Instead, officials urged consumers buying the over-the-counter medication to follow label directions carefully and advised that patients taking the prescription versions consult their physicians.

The announcement about Aleve, which is manufactured by German drug maker Bayer, came after the stock markets had closed.

"Bayer Healthcare LLC consumer care division was notified of the trial suspension today," spokesman William O'Donnell said in a statement. "The company has not yet seen the data and therefore is unable to comment on the study at this time. We are investigating the matter and in the meantime we are in agreement with the FDA's recommendation that consumers carefully follow instructions on the label."

The concerns about Aleve were prompted by preliminary findings from a study by the National Institute on Aging, which sought to determine whether naproxen — the drug's generic name — could help reduce the risk of developing Alzheimer's disease. The study, which began in 2001, followed about 2,400 volunteers ages 70 and older who were considered at risk of developing Alzheimer's because of their family history.

Patients taking Aleve were found to have about a 50% greater chance of suffering strokes or heart attacks than those taking a placebo, said Dr. John Breitner of the Veterans Administration Puget Sound Health Care System, the study's principal investigator.

Breitner described the results as a "weak signal" of problems with naproxen and said that he could not yet comment on whether the findings were statistically significant.

The trial also had sought to test whether Celebrex could help prevent the onset of Alzheimer's. Patients were divided into three groups: some people taking a placebo, some taking Aleve, and some taking Celebrex.

No increased risk of heart problems was found among those taking Celebrex, Breitner said. Yet a separate federal study on cancer prevention found a strong indicator of heart risk among patients taking high doses of Celebrex. After those findings were announced Friday, Breitner said researchers decided to stop giving Celebrex to the patients in the Alzheimer's study.

Then, given the signs of a possible problem with Aleve, investigators decided they could not continue to administer that drug without resolving questions.

"We had to face the dilemma of whether we would continue to treat with naproxen, when naproxen appeared to be giving a more negative signal," Breitner said. "In light of that, we elected to suspend both treatments."

Breitner said the cancer study's findings on Celebrex were stronger than the Alzheimer's study data on Aleve.

But he could not explain why no evidence of problem with Celebrex was detected in the Alzheimer's study. Some experts say that rare side effects sometimes do not emerge even in studies of several thousand patients.

A total of 70 patients in the Alzheimer's study suffered heart attacks or strokes, said Breitner, including those taking Aleve, Celebrex and a placebo. Twenty-three died.

Patients in the study were given 220 milligrams of Aleve twice a day, the same dose recommended for over-the-counter use. Consumers buying Aleve for their own use are advised not to take it for more than 10 days without consulting a doctor.

Steven Chen, an assistant professor at the USC School of Pharmacy, said he thought it would be surprising if naproxen truly increased the risk of heart attacks and strokes. If anything, he said, the drug is supposed to make blood clots less likely because it decreases the stickiness of platelets in the blood.

Chen said that the withdrawal of the painkiller Vioxx in late September due to increased cardiovascular risk did not come as a surprise because studies had already shown that it raised the risk of heart attacks and strokes. The same was not the case for naproxen.

"It's the first time we've seen anything like this with Naprosyn," he said, referring to the drug by its prescription brand. "I think it will likely surprise the medical community, but I don't think anybody will take it to heart unless it can be substantiated" by another study that is designed specifically to look at effects of the drug upon the heart.