TORONTO - A ban on smoking in Toronto restaurants may be paying big health dividends, a new study suggests.

Rates of hospitalization for heart attacks as well as respiratory and heart disease conditions plummeted after the city's ban on smoking in restaurants came into effect, according to the study, published Monday in the Canadian Medical Association Journal.

Over a 10-year period that straddled the coming into force of the ban, there was a 17.4-per-cent drop in the rates of people being hospitalized with heart attacks, a 39-per-cent decline in admissions for cardiovascular conditions overall and a 33-per-cent decrease in hospital admissions for respiratory conditions such as asthma or emphysema flare-ups or pneumonia.

"This study's findings are consistent with the understanding that secondhand smoke has detrimental health impacts and legitimizes efforts to further reduce exposure," said Dr. Alisa Naiman, lead author of the study.

But the figures were met with a mixed reaction. A leading Toronto-based tobacco researcher said the findings mirrored those seen elsewhere. But a senior epidemiologist said the numbers were "too good to be true."

"The smoking ban is great. That's for sure. Smoking is harmful. Secondhand smoke is harmful as well and it should not be imposed on people," said Dr. Samy Suissa, director of clinical epidemiology at the Jewish General Hospital in Montreal.

"But to say that by having a smoking ban you've wiped out 40 per cent of the cardiovascular events, hospitalized cardiovascular events, is I think quite optimistic."

Naiman is a family physician and a researcher with Toronto's Institute for Clinical Evaluative Sciences. Her co-authors are also researchers with ICES, where usage data for Ontario's massive health-care system have proved to be a rich source of scientific knowledge.

As is generally the case with these studies, the researchers did not test a theory by exposing or not exposing subjects to something -- in this case tobacco smoke. Instead they looked at the admissions data for Toronto hospitals from January 1996 to March 2006, and compared them to data from two regions of Ontario where a restaurant smoking ban was not put in place.

Toronto's restaurant smoking ban went into force in June 2001.

This type of study has to be done carefully, because it can ascribe to something gains or risks that were really caused by something else. And Naiman herself said the researchers do not believe the large effect they saw was solely due to the restaurant smoking ban.

"There was a huge drop that was seen. And a lot of factors are related to it. But this is only one," she said.

Suissa said the study type, called an ecological study, is often avoided because of the risk of producing false results. An oft-cited example is a study that seems to show birth rates in humans drop when the number of stork nests declines.

He said that it is hard to believe banning smoking in restaurants led to a 17 per cent decline in heart attack rates across the entire population, regardless of whether they smoked or frequented restaurants.

"The best medications we have on the market are associated with a 10, 12, eight per cent decreases. But not 17 per cent. And these are directly in patients who are taking these medications carefully," he said.

"The numbers don't add up. Scientifically they don't add up."

But other research points to big gains from this kind of move to limit the public's exposure to cigarette smoke, countered Roberta Ferrence, executive director of the Ontario Tobacco Research Unit at the University of Toronto and the Centre for Addiction and Mental Health.

"There may be other reasons that we haven't thought of that could explain such a large effect. But it is not physiologically impossible that this would happen," she said.

"Levels (of smoke) in hospitality venues can be much higher than in any other sort of venue -- at home or at friends' houses and so forth.... And people who have asthma or people who have heart conditions are much more likely to be affected at these higher levels."

Ferrence noted a study in Bowling Green, Ohio showed a 39 per cent drop in hospitalizations for smoking related diseases a year after a ban on smoking in workplaces and public locales was instituted there.