The intensive study, the first to systematically monitor ill health and indoor pollution levels over a large range of exposure values, involved nearly 80 households encompassing 400-500 people in central Kenya. The study showed that particulate matter pollution levels inside homes that use traditional open fires can be tens of times greater than those in western industrialized countries.

"One-third of the world's population - almost two billion people - use wood, charcoal, dung or crop residue as cooking fuel, which is an important cause of respiratory illness, one of the most common diseases worldwide," said study coauthor Daniel Kammen, professor of energy and resources at UC Berkeley and director of the Renewable and Appropriate Energy Laboratory.

Midway through the study, Kammen and his colleague Majid Ezzati introduced the households to simple, cleaner-burning stoves and found a sharp reduction in the level of indoor pollution. When used with charcoal, these inexpensive stoves made by local craftsmen reduced particulates by more than a factor of 10 and cut respiratory illness nearly in half.

The report by Kammen and Ezzati, currently with the Center for Risk Management at Resources for The Future in Washington, D.C., will be published in the August 25 issue of The Lancet. A related article by Ezzati and Kammen appeared in the May 2001 issue of Environmental Health Perspectives. Ezzati until recently was with the Epidemiology and Burden of Disease Unit of The World Health Organization in Geneva.

Average particulate matter concentrations of 5,000 to 10,000 micrograms per cubic meter were not uncommon in the homes studied. The latest Environmental Protection Agency (EPA) standards state that individuals should not be exposed to more than 150 micrograms per cubic meter for a 24-hour period. The EPA limits average annual exposure to less than 50 micrograms per cubic meter per day.

As anticipated, the researchers found that higher particulate levels were correlated with higher rates of respiratory illness for both men and women, a relationship which persisted after controlling for other potential causes of the disease. At the highest average levels of exposure, children were sick 20 percent of the time, on average, with upper respiratory infections such as tonsillitis or cough.

"This study makes an important contribution to our understanding of acute respiratory illness due to indoor air pollution," said Kirk R. Smith, professor of environmental health sciences in the School of Public Health at UC Berkeley, who was not involved with the study. One of the world's authorities on indoor air pollution, Smith and colleagues in four countries are currently engaged in studying the connection between pollution from indoor cooking and childhood pneumonia in Guatemala in the first randomized trial in air pollution history.

Smith noted, for example, that the Kenya researchers found that brief exposure to high levels of particulates, as when the stove is lit, significantly increases the incidence of acute lower respiratory illness over and above what would be expected based on average exposure. This suggests that finding ways to cut peak exposures could have a big impact on people's health, Smith said.

Many studies have documented the acute and chronic respiratory infections resulting from open fires in developing countries. Acute lower respiratory infections, which commonly lead to pneumonia, are the leading cause of mortality from infectious diseases, with an estimated 3.5-4 million deaths worldwide each year, according to the World Health Organization in 2000.

The current study was conducted on a large, 22,000 hectare (80 square mile) ranch in central Kenya, northwest of Mount Kenya, donated as a research station to Princeton University and other U.S. and Kenyan institutions by Princeton alumnus George Small. The Mpala Ranch/Research Centre has numerous small villages whose families are employed to tend the ranch's 3,000 head of cattle and other livestock.

Kammen initiated the study when he chaired Princeton's Science, Technology and Environmental Policy Program (STEP), before moving to UC Berkeley in 1999. The similarity among villagers living in bomas - clusters of cylindrical mud houses with conical straw roofs - made the population ideal for studying the dose-response relationship between particulate levels from smoke and the rate of respiratory illness. Most women cooked meals indoors balancing a pot on three stones with a wood fire in the middle. The insides of most homes were dense with smoke, Kammen said.

With the help of four assistants and two local nurses, Ezzati set about measuring the levels of particulates inside the homes and the amount of respiratory illness among household members. Only particles under 10 microns in diameter were assessed, since these are the most likely to travel deep into the lungs and cause severe illness, including pneumonia.

Each family in the study was visited once every seven to 10 days and given medical exams, with nurses recording all illnesses or problems, not just respiratory illness. Indoor pollution levels and individual exposure levels were monitored simultaneously.

They found that young and adult women who regularly participate in cooking have exposure levels two to five times higher than men, with daily averages about 5,000 micrograms of particulates per cubic meter. Peak exposures were as high as 50,000 micrograms per cubic meter.

Not surprisingly, women also were, on average, twice as likely as men to be diagnosed with a case of acute respiratory infection.

Young children under five years of age were, on average, exposed to approximately 1,500 micrograms per cubic meter - 30 times the EPA's exposure limit. Such children were found to be sick with respiratory illness on one of every five weekly exams and to have more severe acute lower respiratory illness on 6-8 percent of visits.

After a year of collecting data among the households, Ezzati and Kammen offered each household a free manufactured stove - a simple, ventilated clay bowl about 18 inches in diameter, sufficient to support a pot and burning either wood or charcoal. Though made locally and inexpensive (about $3-$5 U.S. or 200-400 Kenyan shillings each), the cost represents up to several days' pay for such a family. Because of the cost and lack of distribution infrastructure in rural areas, most households had only the simplest stove - a triangle of three rocks.

After the households started using the more efficient wood stoves, exposure levels dropped 40 percent. When using charcoal, exposure dropped 90 percent compared to traditional wood burning stoves. The results indicate that, if maintained, this would result in a 20-45 percent reduction in lower respiratory infections in children alone.

"This stove, which was developed 15 years ago, has been a big success story in improving health and people's livelihood," Kammen said.

Kammen acknowledges that this small study doesn't have the statistical power of larger studies like that underway in Guatemala, but he said he believes the results are ironclad.

"This shows the large potential benefits that could result from efforts to develop, promote and sustainably commercialize improved stoves and improved fuels that can dramatically improve the health, economic and environmental situation among the poorest households in the world."