Are “safe” exposure levels really “safe”?

Health risk assessment is the method that regulators use to determine whether chemicals in the environment pose a risk to people’s health and how much exposure to the chemical produces these effects. Unfortunately, current human health risk assessments don’t capture everyone’s risk level.

Our recent paper, Application of Probabilistic Methods to Address Variability and Uncertainty in Estimating Risks for Non-Cancer Health Effects, examines how updated methods can better account for health risks from environmental exposures and thus protect the public’s health.

Regulators Currently Assess Cancer and non-Cancer Health Risks Differently – why?

Health risks have traditionally been divided into two categories: risks of developing cancer and risks of developing all other adverse health effects, including delays in neurodevelopment in children, nerve damage, impaired reproductive function, compromised immunological function, kidney and liver damage, and many others. Risk assessments are done differently for chemicals that increase risk of cancer and chemicals that increase risk of harmful health effects other than cancer. But scientific experts have been pushing for a unified method because biological changes linking chemical exposures and non-cancer diseases can be measured and acted upon using the same methods that are applied to cancer risk assessments. Additionally, the same biological processes that increase risk of cancer can also increase risk of non-cancer health effects.

For cancer-causing chemicals, risk assessments assume that any exposure to the chemical increases the risk of developing cancer. For chemicals that don’t cause cancer but cause other devastating diseases, risk assessments do not use this approach. Instead non-cancer risk assessments rely on a reference value, which is an assumption that there is a level of the chemical in the environment that people can be exposed to every day without being expected to develop health problems from exposure to that chemical. Reference values are typically thought of as “safe” exposure levels, meaning that if the amount of a chemical in your drinking water, air, food, or your work or home environment is below that reference value, then you are not at risk of developing health effects.

So-called “safe” levels are outdated

Authoritative scientific bodies, including the National Academy of Sciences, have said that methods used to identify “safe” levels of chemical exposure are outdated and fail to acknowledge risk at the population level. Identifying “safe” levels does not reflect what is known about how chemicals affect people, nor how people are actually exposed to chemicals. Furthermore, the risk of developing an adverse health effect from exposure to chemicals in the environment is a function of who you are and where you live and work, among other factors.

For example, your age can impact your risk. Children are generally more susceptible to health effects of chemical exposures because their body systems, such as the brain and immune system, are rapidly developing and thus susceptible to abnormal development. They are also often more exposed to chemicals than adults because their normal activities, such as crawling and putting things in their mouths, bring them in closer contact with exposure sources like dust that often contain harmful chemicals.

Contextual factors, like where you live and work, can change both exposure to chemicals and the risk of developing diseases. The result is that, at a population level, no “safe” level of exposure exists without increased risk of developing disease for vulnerable members of the population. Concerningly, the reference value approach does not even consider that there is a probability of a health risk below this supposed “safe” reference value.

Updated Methods Let Us Measure Health Risks to People who are Most Susceptible

Measuring and describing health risks for different levels of exposure, including above and below existing reference values, has become increasingly important to accurately reflect and act on health risks stemming from chemical exposures. In collaboration with the Science Action Network and public health scientists across the country, our recently published paper reviews methods to quantify health risks that consider the real-world conditions people experience when exposed to environmental chemicals.

These methods are referred to as ‘probabilistic approaches’ because they use data and methods to measure the probability or risk of developing a disease at any exposure level, including below the reference value. We used perchloroethylene (also called PCE), a chemical applied as a stain remover in dry cleaning and a high priority substance by the U.S. Environmental Protection Agency as a case study, and showed how updated risk assessment methods can be used to better capture and measure health risks at varying exposure levels, including those at levels considered to be “safe.”

Assessing Risk to a Chemical Used in Dry Cleaning

Perchloroethylene (PCE) is a solvent used in dry cleaning and a variety of other industries that has been well studied for at least 25 years, with evidence that it affects brain function. Studies of dry cleaning workers and communities exposed to perchloroethylene show a connection between long-term exposure and decreased performance on neurobehavioral tests that measure different brain functions (among other adverse health effects).

The risk assessment we conducted quantifies the effects of increasing PCE exposure on visual memory function. Visual memory refers to the ability to learn and remember information that people are presented visually and is an important way that we learn. For example, learning to read and write relies heavily on visual memory because it requires readers to remember images of words and writers to recall what each letter looks like. Our probabilistic assessment shows that visual memory function is adversely affected at exposure levels below EPA’s “safe” level of 0.0059 ppm and that it could negatively impact 1 in 1000 people. This means that if your exposure is at the reference value or “safe” level, there is a reasonable likelihood that your visual memory performance will be negatively affected.  

By comparison, EPA regulates chemicals that increase risk of cancer to exposure levels that are not expected to cause more than a 1 in 100,000 chance of increased cancer risk.

Providing More Information for EPA to Protect Health

Probabilistic approaches are easy to use and provide essential information to better understand health risks across diverse populations, as our paper demonstrates. They also provide really useful data for risk managers who make decisions about risk for waste site cleanup and pollution prevention. Government regulatory processes can leverage information from probabilistic risk assessments to perform cost benefit analyses that attach dollar values to the health benefits of reducing environmental exposures and improving population health.

This post is fourth in a series about our Roadmap papers, which provide scientific recommendations to update chemical risk assessment to better reflect exposures and hazards, guide decision-making, and protect the public’s health. Lead authors of the papers will discuss their recommendations on a webinar on Thursday, February 2nd. Register here:

Other blogs in this series:

About the Authors

Greylin Nielsen is a PhD Candidate in the Department of Environmental Health at Boston University School of Public Health, where she uses human-relevant systems to understand the effects of PFAS on liver function.


Wendy Heiger-Bernays, PhD is a clinical professor in the Department of Environmental Health at the Boston University School of Public Health, where she applies her training in molecular toxicology to practical questions about the impact of industrial chemicals, consumer products and pharmaceuticals in water and waste streams on people’s health. 

Roberta White, PhD is professor emeritus and former Chair in the Department of Environmental Health at the Boston University School of Public Health, where she uses her expertise as a neuropsychologist and neuroscientist to conduct research on the effects of chemicals on the brain and nervous system across the lifespan.