NEW YORK - An analysis of 20 years of data on the health of over 900 adults has found that long-term use of traditional nonsteroidal anti-inflammatory drugs (NSAIDs), such as ibuprofen and naproxen, cuts the risk for oral cancer in smokers by half.

However, use of these pain relievers (with the exception of aspirin) for 6 months or more also doubled users' risks for cardiovascular death, according to collaborative research published online Oct. 7 by The Lancet.

The study was conducted by researchers at the Norwegian Radium Hospital and The National Hospital in Oslo; University of Science and Technology, Trondheim, Norway; NewYork-Presbyterian Hospital/Weill Cornell Medical Center in New York City; The University of Texas M. D. Anderson Cancer Center in Houston; and the University of Helsinki.

"Our findings highlight how a commonly used drug can have a benefit from the standpoint of cancer prevention but can also have side effects -- in this case, an increased risk for cardiovascular death," said co-researcher Dr. Andrew Dannenberg, the Henry R. Erle, M.D., Professor of Medicine at Weill Medical College of Cornell University and Director of Cancer Prevention at NewYork-Presbyterian/Weill Cornell.

The findings "also support moves by the U.S. Food and Drug Administration, which recently mandated special 'black box' warning labels on all NSAID pain relievers except aspirin, warning consumers of potential cardiovascular side effects linked to long-term use," added lead researcher Dr. Jon SudbÀ, a senior consultant in the Department of Medical Oncology and Radiotherapy at Norwegian Radium Hospital.

The new FDA-mandated labeling applies to both the over-the-counter and prescription versions of these drugs. The agency's decision followed earlier moves in 2004 and 2005, when two COX-2 enzyme-specific NSAIDs were withdrawn from the market -- first Vioxx was voluntarily recalled by Merck in September 2004, and then the FDA ordered the recall of Bextra earlier this spring.

The findings were presented earlier this year at the annual meeting of the American Association for Cancer Research.

The research team took a retrospective look at data collected prospectively from 1975 to 1995 on more than 123,000 adults participating in the Norwegian Health Survey. The survey tracks participant lifestyles, habits, and long-term health outcomes, and is one of the richest such databases in the world.

Dr. SudbÀ and colleagues narrowed their focus to 454 people with oral cancer and 454 others without such malignancies matched for age and sex. All of the individuals had a history of heavy smoking.

"Specifically, we were looking for associations between the long-term use of traditional, non-COX-2-specific NSAIDs and the risk of oral cancer, since previous work has suggested that these drugs can lower risks for other malignancies, such as colon cancer," explained co-researcher Dr. Scott M. Lippman, Ellen F. Knisely Distinguished Chair and Chairman of the Department of Clinical Cancer Prevention at M. D. Anderson Cancer Center.

In fact, the researchers did find a correlation: Adults who were prescribed NSAIDs such as aspirin, ibuprofen, naproxen, indomethacin, piroxicam, and ketoprofen for 6 months or more (most of them for much longer periods, with 88 percent taking NSAIDs for 5 years or more) were at a 53 percent lowered risk for oral malignancies compared with those who did not take the drugs over the long-term.

"But there was a puzzling finding," Dr. SudbÀ said. "Even though the drugs appeared to protect users from oral cancer, we saw no added benefit overall in terms of prolongation of life or reduced mortality. So something was potentially tipping the balance the other way."

Digging deeper, they identified that "something" as cardiovascular disease: According to the study, long-term use of traditional NSAIDs, except for cardiovascular-dose aspirin, doubled users' risk for cardiovascular death.

This type of elevated heart risk had already been noted with a subclass of NSAIDs called COX-2 inhibitors, drugs like the now-recalled Vioxx and Bextra and a third (still available) painkiller, Celebrex, which target a specific enzyme linked to inflammation.

"But our data support the recent FDA move to put a warning label on the entire class of NSAID medications, except for aspirin -- labeling that alerts consumers that long-term use of these drugs might raise their cardiovascular risk," Dr. Lippman said.

The investigators stressed that the study does have its limitations. First, the NSAIDs used in the study were available to Norwegians via prescription only, and it's not clear whether dosages used by the survey participants were similar in strength to popular American over-the-counter products like Aleve (naproxen), Motrin (ibuprofen), or Advil (ibuprofen).

In addition, Dr. Dannenberg and co-author Dr. J. Jack Lee, Professor of Biostatistics at the M. D. Anderson Cancer Center, stressed that although the data themselves were collected prospectively, the study remains a relatively small, retrospective effort. "It's tough to make sweeping generalizations until the results are confirmed by much larger, prospective trials," they said.

So, might long-term use of traditional NSAIDs still be right for some people? According to the study investigators, a careful risk-benefit assessment is required, and that's a question best left to a patient and his or her doctor.

The study was funded by grants from the U.S. National Cancer Institute, the Center for Cancer Prevention Research, the Norwegian Cancer Society, the Research Foundation of the Norwegian Radium Hospital, Astrid and Birger Torsteds Legat, and the Research Council of Norway.

Co-researchers include Dr. Jan Folkvard Evensen, Wanja Kildal, MSc, Prof. Albrecht Reith, and Dr. Nina Flatner, of Norwegian Radium Hospital, Oslo; Simone Sagen, MPH, of The Research Foundation of the Norwegian Radium Hospital, Oslo; Dr. Jon Mork, of The National Hospital and The Norwegian Cancer Registry, Oslo; Dr. Li Mao and X. Zhou, MSc, of the University of Texas M. D. Anderson Cancer Center, Houston; Prof. A. SudbÀ, of the University of Science and Technology, Trondheim, Norway; and Dr. A. RistmÉki, of the University of Helsinki.