May 1, 1999
Exposure to second-hand smoke increases the risk of coronary heart disease by 25 percent, according to a study published by researchers from Tulane's School of Public Health and Tropical Medicine. Jiang He, principal investigator of the study and assistant professor of epidemiology, based the findings on a "meta-analysis" of 18 studies on the effects of passive smoking in humans.
The results of the meta-analysis, which combines the research results from multiple studies, were published in the March 25 issue of The New England Journal of Medicine.
"We focused on coronary heart disease because most of the studies on passive smoking have focused on the association with lung cancer," He says. "The relationship between second-hand smoking and lung cancer is pretty clear, but the relationship between coronary heart disease and second-hand smoking is still controversial."
The number of subjects in the 18 studies evaluated ranged from 513 (a study of non-smoking wives in Evans County, Ga.) to 479,680 (an American Cancer Society study). The 18 studies included data from 643,750 individuals.
To identify the studies, the researchers performed a computer search to find all published and unpublished research on the topic, including master's theses and doctoral dissertations. Of the 18 used in the meta-analysis, three were unpublished.
"For a meta-analysis, it's important to find all of the studies including small studies and studies reporting a negative finding," He says. "We also used statistical methods to figure out any publication bias, which our analysis did not suggest."
Combining the results of the 18 studies, the researchers showed "pretty consistent findings of risk," he says. "All show a 25-percent increase in risk of coronary hearts disease associated with second-hand smoke, among different populations and by different study designs."
The study also showed exposure to a greater amount of second-hand smoke--20 or more cigarettes a day--increased the risk of heart disease to 31 percent. The risks were not significantly different for men or women or for exposure in the home and exposure in the workplace, the study's authors reported. These results have implications for public health policy, he says.
According to a national survey, about 43 percent of nonsmoking children and 37 percent of nonsmoking adults are exposed to tobacco smoke in this country.
"To achieve a meaningful reduction in the burden to society of coronary heart disease, both passive and active cigarette smoking must be targeted," the authors say in their report.
The publication of the research in The New England Journal of Medicine has led to some confusion in the general media, He says, largely due to an accompanying editorial that criticized meta-analysis and the findings of Tulane's study.
Among the comments of the editorial's author, John C. Bailar III, professor of health studies at the University of Chicago, was a criticism of meta-analysis as a research tool. Bailar cites a study where meta-analyses didn't agree with subsequent large, randomized trials.
"Good epidemiologists disagree with him" on the drawbacks of meta-analysis, He says. "Meta-analysis is a very subjective way to summarize data."
Bailar also questions the size of the effect of second-hand smoke on nonsmokers (25 percent, according to He's analysis) compared to the increased risk of heart disease on active smokers (commonly reported as 75 percent). Bailar contrasts these conclusions with data from other studies on the increased risk of lung disease in active smokers (1,200 percent) and those exposed to second-hand smoke (25 percent).
The nonsmoker's risks associated with lung disease and coronary heart disease should be more proportional with the smoker's added risk, Bailar argues. "Smoking causes lung cancer and coronary heart disease by different biological mechanisms," He says. "Very light exposure to smoking can damage the cardiovascular system."
He and his fellow authors wrote a letter to the journal addressing Bailar's criticisms. Other authors of the study are Paul Whelton, dean of the public health school; Janet Hughes, professor of biostatistics; and epidemiology graduate students Suma Vupputuri, Krista Allen and Monica Prerost.
Tulane University, New Orleans, LA 70118 504-865-5000 email@example.com