Woe be to the Environmental Protection Agency. If President Trump gets his way, the federal agency will lose 31 percent of its annual budget—about $3 billion. Supporters of Trump’s 2018 budget proposal call it a “back to basics” approach, carving away what they see as the agency’s regulatory overreach. Opponents are similarly pithy: The EPA’s former director labeled Trump’s proposal a “scorched Earth budget.”
Disagreements over the utility of the EPA's programs typically fall along the partisan lines you might expect. But not all of them.
At least one of Trump's proposed budget cuts targets a program that some scientists have been questioning for decades: residential radon risk. That program, which sets the standards for risk and protection from radon gas, would be completely zeroed out. Now, according to the EPA, the WHO, and many other big public health organizations, radon is second only to cigarette smoking as a leading cause for lung cancer. The EPA says radon gas causes 21,000 deaths every year. Yet some critics in the health care profession say that’s all baloney.
Trump's budget hasn't passed. It still has to survive the congressional gauntlets, full of many senators and representatives who have already voiced opposition to many of the EPA cuts, including to the radon program. But some scientists might actually want the EPA to back off on this one.
The issue here isn't whether radon causes cancer. It does, certainly. But critics say scientists can only prove that radon is carcinogenic at high doses. Like, rates typically found in mine shafts, not suburban homes. They add that the established scientific method for establishing low-dose radon risk assessments is simply unscientific. Some even go so far as to say low doses of radiation might help your body fight cancer. Follow that logic through, and they say the EPA's low dose radon risk assessment could actually be harming people.
Radon is radioactive gas emitted when uranium decays. And because most rocks contain traces of uranium, it leaks from the soil virtually everywhere on Earth. Like uranium, radon decays, too. If it does this in your lungs, the alpha particles it releases can harm the DNA in your cells. "When we talk about radon risk, we really mean the risk from radon and its decay products," says Phillip Price, a physicist at Lawrence Berkeley National Laboratory. If enough DNA gets damaged, the mutations pile up and … lung cancer.
Radon is everywhere, but it doesn't usually pose much of a risk because it disperses into the open air. "What’s unnatural are the radon levels that accumulate indoors,” says James McLaughlin, president of the European Radon Association.
Scientists began to recognize the threat in the 1970s, when epidemiologists found that miners were getting lung cancer at higher-than-average rates. The tunnels they worked in were full of radon gas. But the agency's real work started after 1984, when an engineer at the Limerick Nuclear Power Plant in Pennsylvania set off radiation alarms as he was coming in to work. After some sleuthing, public health experts figured out that his house was filled with radon. Like, uranium mine levels of radon. This set off a national panic—homeowners feared their home might have similarly high levels.
By 1986, the EPA established an indoor radon risk exposure level—4 picocuries per liter of air. It's similar to that recommended by the World Health Organization and the European Union. That number isn't tied to any regulations: Instead, the EPA recommends that people test their homes (which can cost between $20 and $200) and take action if necessary—say, by installing fans and vents to improve airflow. Hiring a contractor to ventilate your basement can run into the thousands.
That number, 4 picocuries, is a funny one. Because here's the thing: The scientific model used to come up with that limit says there's no acceptable level of radon exposure. None.
It's called the linear no-threshold model, and it's used to estimate risk for low doses of carcinogenic materials. See, it's not hugely controversial to link lung cancer to high radon levels, like those found in a mineshaft. But it's impossible to directly establish a link between low levels of radon and cancer. There are just too many other possible carcinogens, like air pollution, second hand smoke, and even random DNA mutations.
So, the linear no-threshold model takes the cancer rates in highly-exposed populations—those uranium miners—and extends them to lower doses of radiation. It then multiplies that very small risk across a huge number of people in order to estimate the total number of future cancers. That's how the EPA estimated that radon causes 21,000 annual lung cancer deaths.
Radiation risk professionals use the linear no-threshold model to determine the exposure risk from all sorts of other radioactive and carcinogenic material. "The idea is that twice as much radiation means twice as many cells with damaged DNA, and if each cell with damaged DNA has a certain chance of initiating a cancer, then ceteris paribus you have the linear no-threshold model," says Price. "That's not crazy."
But if there's no threshold below which radon doesn't increase cancer risk, how did the EPA settle on 4 picocuries per liter of indoor air as its recommended exposure limit? For one thing, radon becomes both more difficult to detect and remove at lower doses. That translates into dollars. When the EPA established the 4 picocurie action level in 1986, it estimated each saved life cost around $700,000. Lower that to 3 picocuries, and the cost per life more than doubles, to $1.7 million. At 2 picocuries, $2.4 million.
Critics contend that money probably isn't saving anyone. "I personally think the linear no-threshold model is over conservative, and does waste a lot of money that could be better spent," says Cynthia McCollough, a radiologist. "Using the linear no-threshold model to predict future cancers is very bad science, and all radiation protection agencies agree on that." In fact, the Health Physics Society (a large scientific organization for radiation safety specialists) has officially stated that low doses of ionizing radiation produce no statistically measurable health effects. The HPS's official position on radon, however, still falls in line with the radon risk prescribed by the linear no threshold model2.
The EPA sort of lives with the same dichotomy. While it holds fast to the linear no threshold model for establishing radon risk, its its 2003 assessment of household radon risk, acknowledged the model had its faults1: "The [committee tasked with establishing radon risk] adopted the linear no-threshold assumption based on our current understanding of the mechanisms of radon-induced lung cancer, but recognized that this understanding is incomplete and that therefore the evidence for this assumption is not conclusive." (Emphasis added.)
Despite that admission, the EPA went ahead with the linear no-threshold model for radon out of an abundance of caution. And that caution isn't entirely unwarranted. A World Health Organization meta-analysis of three large, previous epidemiological radon risk studies suggested lung cancer cases at 5.4 picocuries per liter. The EPA notes that these risk projections do not require extrapolation from high dose rates to low1.
But some critics say the agency's overprotective action might actually be harming people. "The linear no-threshold model is based on the concept that increased radiation and increased mutations cause cancer. That assumption is fundamentally wrong," says Mohan Doss, a physician at the Fox Chase Cancer Center in Philadelphia. That's because, Doss says, it ignores biology.
Humans develop DNA mutations all the time. Most of them do not turn into cancer. One theory why, says Doss, is that the immune system moves in and gets rid of those afflicted cells early on. He goes so far as to say that low-dose radiation might prevent lung cancer. Doss points to animal studies showing that low doses of radiation might help the body fight cancer by activating the immune system to clear away mutated DNA. "These boosted defenses actually reduce the DNA damage that would have occurred in absence of radon exposure," he says. If he's right, that means the EPA's guidance levels aren't preventing cancer—they're causing it.
So, why doesn't the EPA listen to these low limit soldiers? Well, first of all because linear no-threshold-based radon risk is endorsed by the National Academy of Sciences' Committee on the Biological Effects of Ionizing Radiation, the WHO, the CDC, the US Surgeon General, and the American Lung Association1. Also, consider the precautionary principle: It’s better to overprotect than under-protect. And the total indoor radon program in the EPA's 2017 budget called for just $3.5 million. Also, remember how hard it is to establish radiation risk at low levels? That sword cuts both ways. "There’s some shaky empirical evidence for a weak protective effect at 2 picocuries per liter compared to 0," says Price. But whether you're looking at health benefits or risks, all the evidence is pretty shaky below an exposure of about 8 picocuries.
Finally, consider inertia. The linear no-threshold model underpins not just the EPA's radon recommendation, but its risk exposure estimates for other radiation sources.
Most of the experts I spoke with didn't have a good idea about what effect, if any, Trump's promise to nix the EPA's radon program might have. McCollough said the real hit happened under Obama's watch, when the Department of Energy's Office of Science killed a program studying the biological effects of low dose radiation.
At least one, though, is rooting for the program's demise. "If the Trump administration nixes the EPA radon program, there would be little adverse impact on public health," says Doss. "But in fact there would be improvement of public health, by reducing lung cancers." Talk about scorched Earth.
1 UPDATE: 8/22/2017 1:30pm ET — This article has been updated with data justifying the EPA's position on radon risk.
2 UPDATE: 8/22/2017 1:30pm ET — This article has been updated to differentiate the Health Physics Society's statements on ionizing radiation risk generally, and its position on radon risk.
