Susilo sat attached to an oxygen tank instead of tending his vegetables. The 62-year-old farmer, who like many Indonesians goes by one name, had a long history of respiratory trouble, but during the 2015 fire and haze crisis things got worse. As the landscape smoldered around his village in western Borneo, his attacks came so frequently that he could hardly work. This was his 37th visit to a local clinic for the life-saving treatment of a simple nebulizer.

Indonesia’s vast peat swamp zones have been widely drained and dried for agriculture, rendering them highly flammable, and they often burn on a massive scale, blanketing the country and its neighbors in smoke. A recent survey on perceptions of the fires showed that while different groups – from small farmers to industrial agriculturists to people at all scales of government – have varying levels of concern about forest loss or carbon emissions, everyone agrees that protecting public health is a top priority. However, the first step to solving a problem is to agree on how critical the issue is.

Two independent studies have estimated that the 2015 Southeast Asian haze crisis caused somewhere between 11,880 and 100,300 premature deaths. However, because these estimates were based on remote sensing and models – and not “hard data” – they were roundly rejected by the governments of Indonesia and Malaysia. There, officials maintain, only 24 Indonesian and zero Malaysian deaths can directly be attributed to the 2015 haze. The rest of the health issues, they have claimed, were just temporary respiratory irritations that cleared up when the haze lifted.

Nailing down how severely the fires affect health has been notoriously difficult. Accurate numbers for hospital admissions and mortality are not freely available, or in many cases are nonexistent. In addition, academic research has tended to focus on environmental conservation and carbon emissions rather than public health, leaving a data gap in our understanding. As a result, different models based on solid, but varying, assumptions can provide different conclusions, which can lead people to cherrypick results based on their agenda – or reject them all outright.

However, while some study results may contradict one another, the body of science as a whole is less hazy on the conclusions: smoke from uncontrolled fires is a deadly threat to Southeast Asia whose victims number in the thousands, not dozens.

Toxic clouds

While emissions from burning peatlands are less well-studied than other types, recent research has shown that they contain potent carcinogens and over 90 different gases, some of which are highly toxic. Among those, formaldehyde, acrolein, benzene, carbon monoxide and nitrogen dioxide are the most concerning, given their negative health effects in even relatively small doses. As such, the World Health Organization (WHO) and most other regulatory agencies have established recommendations for safe exposure levels to these chemicals – levels that are regularly exceeded by peat smoke and resultant haze, as shown by a comprehensive 2016 meta-analysis of 375 fire and health studies published between 1970-2014 from different fuel types around the world.

Formaldehyde and acrolein are both known carcinogens, while benzene, a potent organic compound, can negatively affect the blood, brain and immune systems. Firefighter exposure to formaldehyde at prescribed burns in the U.S. has been recorded as high as 3,700 percent of the recommended maximum, and measurements of peat fire production of the gas have lead researchers to caution that exposure to the local population will likely exceed recommended levels.

Carbon monoxide (CO) is a highly toxic blood poison that binds with hemoglobin, preventing effective oxygen uptake. The amount of CO a fire produces varies widely, and although it readily dissipates in the atmosphere, acute exposure at the source can have deadly consequences. While most studies find exposure levels from forest fires to be below regulatory guidelines, the slow smoldering nature of peat can produce high levels of CO not typically seen from other fires. In 2015, outdoor levels of CO in Kalimantan, the Indonesian part of Borneo, were measured at 3-6 times higher than what the U.S. Environmental Protection Agency (EPA) considers dangerous. Further, CO can linger in the bloodstream for several hours, adding to the level of CO already in the environment from vehicle exhaust and other pollution.

Finally, nitrogen dioxide (NO2) has been shown to affect lung function and is particularly threatening to individuals who already suffer from asthma or chronic obstructive pulmonary disease (COPD). Although long-term, low-level exposure to NO2 may be relatively safe, short-term exposure to high concentrations of NO2 correlates with increased mortality in several studies. Firefighter exposure to NO2 has been recorded at 2.5 times the acceptable occupational limits.

As noted, these chemicals can all have severe health effects in high doses, but less is known about long-term, repeated exposure. Further, the established guidelines typically do not consider the cumulative effects of multiple carcinogens, irritants and toxins bombarding the body at the same time for an extended period.

Killer Particles

Beyond the highly toxic gases listed above, however, a more threatening byproduct of peat fires appears to be fine particulate matter (PM). These airborne particles are classified as either PM10 (between 2.5 and 10 micrometers, roughly the size of dust, pollen or mold) or PM2.5 (less than 2.5 micrometers, only visible with an electron microscope). According to the WHO, both are responsible for acute respiratory issues such as asthma, while PM2.5 is increasingly linked to mortality from heart and lung disease.

The ultrafine PM2.5 particles are small enough to penetrate deep into the lungs, coating the tiny air sacs known as alveoli and even crossing into the bloodstream. PM2.5 is produced by vehicle exhaust, wood burning and most other types of combustion, and while all particulates can have negative health effects, different emission sources produce different toxins. A 2013 study of Indonesian peat fire smoke found that the carcinogenic metals cadmium, chromium, nickel and cobalt where being produced at 16, 9, 8, and 13 times the rate of background pollution.

The health effects of PM2.5 have been thoroughly researched in laboratory settings as well as in population-based cohort and ecological studies, and even short-term exposure is linked to cardiovascular and pulmonary diseases.

A 2014 meta-analysis of 110 published studies of PM2.5 concluded that although there was some variability among the data, the general consensus was that death rates rise with even short-term exposure to ultrafine particulate matter. Taken as a whole, the literature shows that increasing short-term exposure to PM2.5 by 10 micrograms per cubic-meter (μg/m3) will result in a 1.04 percent increase in mortality risk. This is observed across all age groups in all parts of the world, with the rate varying from 0.25-2.08 percent. The highest mortality increases were reported from studies in Chile, Mexico and Brazil. (The analysis included no studies from Southeast Asia.)

To put the numbers in perspective, the WHO sets the standard for 24-hour average exposure to PM2.5 at 25 μg/m3. During the 2015 haze event 24-hour averages in Singapore regularly topped 100 μg/m3, with one station recording a high of 471 μg/m3.

It is important to note that these acute spikes in particulate matter during haze events occur on top of daily background exposures that already well exceed safe standards. Over 90 percent of the populations of Indonesia and Malaysia and 100 percent of Singapore’s population are consistently exposed to PM2.5 levels above WHO recommendations. By contrast, less than 10 percent of the U.S. population ever experiences levels above the WHO guideline.

However, even in the U.S. – where advanced healthcare is readily available and background particulate matter is relatively low – the effects from forest fires can still be deadly. A recent analysis by the U.S. EPA and Australian researchers found that short-term exposure to PM2.5 produced by wildfires in the U.S. between 2008-2012 was likely responsible for 1,880 premature deaths per year. Further, they estimate that long-term exposure to fine particulate matter likely caused 10,940-24,600 premature deaths per year for the same period.

Although these numbers are lower than the 11,800-100,300 premature deaths predicted for Southeast Asia as a result of the 2015 blaze, one must consider all confounding factors including exposure times and intensity as well as access to medical care and population health before drawing any comparisons between the two.

Vulnerable Populations

One issue with using broad population-wide studies – like the majority of those above – is that they are typically unable to tease out which groups might be at a higher risk, or to separate immediate causes from chronic conditions. For example, the effect of PM2.5 exposure on pregnancies, newborns and infants is still poorly understood, but at least one study reached a tragic conclusion.

Northwestern University economist Seema Jayachandran analyzed Indonesia’s 2000 census data and found that there were over 16,000 children “missing” from the census that models otherwise expected. Further, regional decreases in the number of children correlated significantly with exposure levels during the 1997 haze crisis in different provinces.

Other studies have found that exposure to haze during pregnancy can negatively impact a developing fetus. An analysis of 886,034 births during four years in Southern California found that fetuses who’s parents were exposed to wildfire smoke weighed 6.1 grams less than normal at birth. This weight loss may be a result of reduced oxygen in the bloodstream during pregnancy, or even the mother’s stress during haze events.

Although the effect found in this study is slight, the researchers note that increased frequency of haze exposure resulted in increased negative effects. Further, they point out, if enough pregnant women are affected, no matter how slight, this could significantly alter birth outcomes for the population – particularly among those populations already at risk for low birth weight.

At the other end of the lifeline, a new paper from the Universiti Kebangsaan Malaysia Medical Centre published this month reports that cases of acute lung disease increased during the 2015 haze episode, as did the number of new lung cancer patients. Although the latter is likely due to more undiagnosed cases seeking treatment for breathing difficulty, the center also reports cancer survival time decreased by 36 percent for those exposed to the haze.

Clinic ASRI in Indonesia’s West Kalimantan province said via email that during the same period, their respiratory cases increased by 47 percent, and their oxygen use tripled. They did not report any haze-related deaths – which means Susilo’s nebulizer treatment must have been successful.

And, while it is true that the lack of “hard data” could mean the exact toll that fires take on the people of Southeast Asia may never be known, the growing body of science well-documents the health effects of forest fire smoke on both individuals and populations.

Meanwhile, for patients like Mr. Susilo, the only “hard data” that truly matters is the fact that he made it back home to his family and his garden – just in time for the rainy season.