TORONTO - What is your risk of having a stroke after having your neck manipulated by a chiropractor? According to the latest research on the controversial question, it's beyond remote. An article published Tuesday in the Canadian Medical Association Journal says that strokes occur in one out of every 5.85 million manipulations.

"The take-away message, I think, is that it is highly safe," said co-author Dr. Paul Carey, a chiropractor in Stratford, Ont.

"I mean, getting something with an incident rate of one in five million is astronomically into the ballpark of safety."

An accompanying commentary questions the validity of that number, but admits for most people, the risk is very small.

"It's much less risky than having angiography. It's less risky than having surgery. It's less risky than some drugs - many common drugs. But we don't have a really good estimate of what that risk is because it's so bloody difficult to study," said Susan Bondy, an epidemiologist at the Institute for Clinical Evaluative Sciences in Toronto.

"There is some risk. It's small. It's not zero."

The problem of defining the risk is further aggravated by the fact that while experts believe some people are more at risk than others - people who have congenital anomalies they may not know about, people who have had some previous soft tissue damage - there is not yet any good way of identifying who is at increased risk.

"It's a pretty much well established fact that this can happen. And chiropractors who are ethical and well-trained of course do everything in their power to minimize the risk. But they cannot eliminate the risk because they cannot know who . . . has those pre-existing problems," Bondy said.

The thinking is that the twisting or pulling of the neck frequently done by chiropractors, physiotherapists and osteopaths can damage arteries, leading occasionally to stroke.

Just how frequently that happens is a hotly contested issue. Chiropractic treatment has a goodly share of detractors.

"There are extremes of view," admitted Bondy. "There are clearly people who are blind advocates of chiropractic care, and there are people who are blind detractors of chiropractic care."

Several lawsuits have been filed against chiropractors by families of individuals who died of a stroke following cervical manipulation. In 1998, a coroner's inquest into the death of a Saskatchewan woman called for more research into chiropractic manipulations and stroke.

But the research to date has been inconclusive and will remain so, predicted Dr. Moira Kapral, Bondy's co-author on the commentary in the journal.

"It's such a rare event that it's going to be very difficult to do a good quality study to give an accurate estimate of risk," said Kapral, a specialist in internal medicine at Toronto's University Health Network and an associate at the Institute for Clinical Evaluative Sciences.

It's impossible to get a firm number, Bondy and others explain, because you can't structure a proper medical study with a control group when the event you want to study happens so rarely. As consequence, studies like this one are at best estimates.

Carey and his colleagues came up with their number by comparing the number of malpractice claims for strokes made to the Canadian Chiropractic Protective Association for the period from 1988 to 1997 to the number of cervical manipulations - 134 million - estimated to have been administered in that decade.

But their figure of one to every 5.85 million is dramatically smaller than the figure of 1.3 cases per 100,000 manipulations that Bondy and a colleague from ICES reported in a study published earlier this year in the journal Stroke.

Kapral and Bondy believe Carey's number is an underestimate, because they don't think all strokes would have led to a malpractice claim against a chiropractor. And Kapral said the 1.3 per 100,000 figure is probably closer to the mark that one per 5.85 million.

But Bondy argued such numbers mean nothing to average people who really just want to know if they are taking their lives into their hands by seeing a chiropractor for neck pain.

"Frankly, whether it's one in 5,000 or one in a million, it's rare - unless a person has other problems that we don't know how to find out if they have them."