A man walks past a coal-powered steel plant in Tianjin, China, on May 29, 2015. (Andy Wong/AP)

Air pollution isn’t just bad for the environment — it’s a danger to human health too. And new research is giving us a clearer idea of just how much. A study published Wednesday in the journal Nature concludes that outdoor air pollution leads to more than 3 million premature deaths around the world every year.

While ozone concentrations play a role in these deaths, the biggest culprit is a type of pollution known as “fine particulate matter,” or tiny particles too small to be seen. These toxic particles can be composed of many different chemical compounds and come from many different sources, including the burning of fossil fuels or organic matter, and even emissions from fertilizer used in agriculture. They can be responsible for a variety of potentially fatal illnesses, including respiratory and heart disease.

Their size is part of what makes them so dangerous, said George Thurston, a professor of environmental medicine and population health at New York University’s School of Medicine, who was not involved in the new study. “They’re so small that they can actually get deep into the lungs, into the alveoli, and get into your bloodstream.”

Worries about particulate matter aren’t new, and many studies have linked outdoor pollution with increased risk of premature death. In fact, Thurston recently led and published a study linking particulate matter pollution with an increased risk of death in the U.S., notably demonstrating a 10 percent increase in risk of death by heart disease and a 27 percent increase in risk of death by respiratory disease in non-smokers.

But the new study takes the research in another direction. Led by Jos Lelieveld, a professor at the Max Planck Institute for Chemistry in Germany, the researchers used an atmospheric chemistry model, armed with global health and population data, to investigate the link between premature mortality around the world and outdoor air pollution from a set of seven pollution sources: residential and commercial energy use, agriculture, power generation, industry, biomass burning, land traffic and natural sources such as dust storms.

By identifying which sources cause the most mortalities in various parts of the world, the scientists “hope to provide information that helps develop better [pollution] control measures in different countries and regions,” said Lelieveld during a Tuesday teleconference.

They found that the greatest number of mortalities were concentrated in the Western Pacific region, followed by the East Mediterranean and Southeast Asia. According to their analysis, China suffers the greatest number of deaths linked to outdoor air pollution each year, followed by India.

When examining the different sources of outdoor air pollution, the researchers concluded that energy emissions have the greatest impact on premature mortality worldwide, causing about a third of air pollution-related deaths. The second biggest single source was agriculture, which releases ammonia particles that can form other compounds, like ammonium sulfate and nitrate.

Regionally, these sources varied much more. While residential energy use was the biggest contributor in many parts of Asia, natural sources of air pollution were found to be dominant in much of northern Africa and the Middle East. Power generation was the biggest source throughout much of the U.S., and agriculture was the biggest factor in Europe.

“With regard to agriculture, it is quite surprising that [these particles] are so important,” Lelieveld said at the teleconference. He said he had assumed, as most people probably would, that industry and traffic would probably play a much larger role in outdoor air pollution. Agriculture’s apparently large role in particulate pollution worldwide suggests that policymakers should pay more attention to regulating this source.

However, these conclusions rely on a very important assumption: that particles from all sources are equally toxic, meaning their contributions to mortality depend solely on how much of them are produced. In reality, however, there’s some debate about whether that’s actually true.

The authors note that the World Health Organization’s guidelines treat all fine particles as equally toxic. But, they add, some recent evidence has suggested that carbon-based particles are more toxic than others. If true, differing toxicity levels among particles would mean some emissions sources are more (or less) important than the new study makes them out to be.

Michael Jerrett, chair of the Fielding School of Public Health at the University of California Los Angeles, pointed this out in an opinion article also published in Wednesday’s Nature. The finding about agriculture “assumes that ammonium nitrate and sulfate have the same toxicity as other constituents of the atmospheric particle mixture,” he wrote, pointing out that research has suggested they’re actually not that potent at the levels they tend to be found in the air. “It could therefore be that Lelieveld et al. overestimate the effects of particles from agricultural sources,” he wrote.

Thurston, the New York University professor, said he agrees. “We’re learning … that particles are not all the same, and that they have very different compositions. Logically, we know that they’re going to have different health implications when we breathe,” Thurston said. “That’s where the problem lies.”

The authors of the study did perform some additional calculations to see how their results would change if they assumed that carbonaceous particles were five times more toxic than inorganic particles. These results suggest that agriculture would have a reduced impact and sources such as traffic would play larger roles in premature mortality in places like the central U.S. and Europe.

In addition, the researchers made some projections for how outdoor air pollution-related mortalities would change in the future. They based their projections off a business-as-usual scenario, meaning they only factored in the global pollution regulations that are already in place and assumed no new ones would be implemented.

They found that premature mortality related to outdoor air pollution could double by 2050, a finding that Jerrett wrote should “sound alarm bells for public-health agencies around the world.” The projection highlights the need for more intensive control measures, particularly in Asia, as well as further research into the differing toxicity of fine particles, Lelieveld said.

More intensive regulations on outdoor air pollution could also go hand-in-hand with another environmental cause: reducing greenhouse gas emissions to combat global climate change. Cutting down on carbon dioxide output tends to have the “co-benefit” of also reducing fine particulate pollution, Thurston said.

Another study, also published today in Nature, provides an example of this effect. The study shows that cutting down on deforestation in Brazil, where land is often cleared by burning, has caused particulate matter concentrations to decline by about 30 percent during the dry season, preventing anywhere from 400 t0 1,700 premature deaths. Burning trees is also known to release large amounts of carbon into the atmosphere, and deforestation is a considered a contributor to global climate change. So reducing deforestation rates in Brazil seems to have achieved two positive effects.

Lelieveld pointed out that the benefits often work both ways. Cutting down on carbon emissions generally requires an investment in clean technology, like solar- or wind-generated energy, which will also cut down on sources of particulate air pollution, he said. In the same way, introducing cleaner technologies designed to reduce particulate air pollution tend to have the bonus effect of reducing the output of greenhouse gases. “It will be a win-win situation in both directions,” Lelieveld said.

So cutting down on pollution could save the world in more than one way. It’s an important consideration, and — as the study points out — one that could mean the difference between life and death for millions of people every year.