Learn ➜ Course ➜ Part 2
A Framework for Assessing Any Risk in Your Home
Dr. Meg Christensen is the founder of Interior Medicine, a physician-created resource on non-toxic home products and household exposures. Her layer-by-layer analysis of materials and products draws on her background in medicine, biochemistry, epidemiology, and clinical research.
Published May 2026 | Updated May 2026
What We Really Mean When We Say “Non-Toxic”
When we want to know if something is toxic or non-toxic, we're really asking about risk. Right? We want to know if something is risky, or not risky, to our health.
The problem is, risk gets confused with 5 other things constantly. Every argument about whether something is toxic or not— every viral post, every outraged comment, every condescending eye-roll — comes down to this: someone is confusing one of these five other things for risk.
Those things are: hazard, exposure, dose, dose-response, and susceptibility. They are each an important part of the pathway that leads to risk, but none of them are risk on their own.
Some real-life examples of how we get risk wrong, to fire you up before we dive in:
The EWG rates cleaning products on hazard, not risk. The IARC classifies carcinogens based on hazard, not risk, too. But a hazard can’t harm you if you’re not exposed to it.
Prop 65 is confusing because it covers potential exposure, not risk. This is why a completely organic couch could have a Prop 65 sticker for the wood dust the craftsman are exposed to. It’s not about your risk.
Assuming that sunscreen chemicals detected in blood are automatically a risk is a dose mistake. Your bloodstream is constantly filtering substances from our world, and dose says nothing about whether your body clears them easily, struggles with them, or are a risk at all.
But dismissing those same sunscreen chemicals as zero risk is a mistake, too, because we’re still learning about their dose-response curves. Low dose doesn't always mean something will have a low risk: endocrine disruptors can cause a response at very low doses. Even more interestingly, vitamin D levels, UV radiation exposure from the sun, and the chemicals in sunscreen each have a different dose-response.
And last, susceptibility: a healthy adult man that waves off flame retardant risk from a mattress isn’t considering that pregnant women and babies are more susceptible than he is.
Whew! Here is how these things are related visually. Then we’ll dive into each one in more detail.
Hazard
If you're not already familiar, the Environmental Working Group (EWG) has databases rating thousands of personal care, sunscreen, and cleaning products for safety. But instead of going through all six steps of the above framework to arrive at a risk score for each ingredient — considering exposure, dose-response curves, and susceptibility — their ratings are based entirely on the first step: hazard. Hazard is the intrinsic property for something to cause harm in some situation, and hazard isn’t the same as risk. For example, arsenic in a sealed bottle high on a shelf in another country is a very high hazard. But, without exposure, it poses zero risk to you.
The EWG lists cause a lot of controversy because most people don't know this detail. Even fewer people know the difference between risk and hazard in the first place. Many doctors and scientists say the lists make people unnecessarily worried, because cleaning products that have high hazard scores look really risky, even when they aren’t.
The same thing happens with carcinogenic hazard classifications from the International Agency for Research on Cancer, the IARC. If you’ve seen phrases like, "Group 2B Carcinogen," you’ve probably felt worried. But again, these are hazard assessments, not risk assessments. Aloe vera is in Group 2B, meaning it’s a "possibly carcinogenic" hazard. This is because ingesting the yellow latex sap can cause cancer in rats. It does not mean that this has been shown to cause cancer in humans, and most people don’t drink the yellow sap part of aloe vera. The clear gel portion of aloe vera is generally considered safe to drink and apply to the skin after sunburn. It just means that the yellow gel portion of aloe vera warrants further study to find out if it drinking it could cause cancer in people.
Context matters a lot, and hazard classifications don’t give us context.
None of this means EWG or IARC are bad. Hazard identification is an important first step. EWG started flagging PFAS as hazardous in products in 2001, a solid 20 years before the rest of the country showed interest. And EWG does acknowledge that its ratings "do not account for the level of exposure or individual susceptibility, factors that determine actual health risks." So they do state they're measuring hazard, not risk. IARC says the same.
When we treat hazard lists as risk lists, that's the mistake.
Here is your Hazard reference card, with an official definition. You can download the full set at the bottom of this page for easy reference later:
When IARC classifies formaldehyde as a Group 1 carcinogen, or when the EWG rates a cleaning product with an F score, that's a hazard identification. They are not telling you what your risk is under real-world conditions.
Exposure
Without exposure, even the most hazardous substance poses no risk.
Prop 65 warnings confuse just about everyone, and even the creating agency has acknowledged they're difficult to interpret. When you see one of these labels on a product, it can feel like: “oh no, this product is exposing me to something dangerous!” But the warning requirement is actually about someone’s potential exposure to a carcinogen or reproductive toxicant. It covers not just you, the end user, but anyone involved in making the product. This is why solid wood furniture can carry a Prop 65 warning for wood dust: it's a hazard to the workers breathing in wood dust particles, not necessarily to you reading in your living room.
A Prop 65 warning tells you a listed chemical simply exists somewhere in a product's chain. It does not tell you whether that chemical can leave the product, reach your body, by what route, at what concentration, or whether any meaningful contact is occurring at all. Companies applying these warnings are accounting for worst-case exposure scenarios. They are legally protecting themselves, asking, what if someone misuses the product in an extreme way and we get sued? They are not considering your actual use. Neither Prop 65, nor the companies abiding by it, are describing your personal risk.
In Part 6, we'll get deeper into how hazards actually become exposures. For now, here is your Exposure card:
Exposure has four important dimensions: route (ingestion, inhalation, dermal absorption), frequency (once vs. daily), duration (acute vs. chronic), and timing (during fetal development vs. adulthood).
Exposure is the lever you actually control. You cannot change a substance's hazard, but you can often change your exposure to it.
Dose
Dose is how much of something is in your body. It says nothing about whether it's safe, harmful, or what health response might follow. It doesn't state your risk. Dose just says: it's there, and this is how much there is.
The FDA published 2 studies between 2019 and 2020 showing that chemical sunscreen ingredients enter the bloodstream after a single day of use, and some of those chemicals stay for at least 21 days. The blood concentrations were above 0.5ng/mL, which is the level at which the FDA requires more studies need to be done to thoroughly assess safety. Importantly, the study did not look into whether or not these doses were harmful — just that the doses existed in the body. They concluded only that they deserve more investigation.
Social media treats this two ways. One way is the confirmation that chemical sunscreen harms people. The other way is the dismissive statement, "the dose is so tiny, you're being ridiculous." But both are wrong, because neither takes dose-response into account, as you’ll see next. First, here is your Dose card:
Accumulation and excretion: Some substances (lead in bone, PFAS in blood and tissue) accumulate over years. The total body burden — the amount stored in tissues over time — can be far higher than any single day's dose suggests. Conversely, other substances are efficiently excreted by your liver, kidneys, or gut, and do not accumulate meaningfully in healthy individuals.
Dose-Response
Detecting a chemical in blood doesn't automatically mean harm. We are porous beings, and a lot of the world passes through our bloodstreams. Sometimes our bodies know exactly how to clear that out, and it doesn’t cause a problem. Sometimes our body doesn’t protect itself well, and health consequences follow. That’s what this step covers: your body’s response to a dose. It’s the most technically difficult one, so take you’re time if it’s all new to you.
There are three ways our body responds to doses. To describe each one, I’ll carry over the theme of the sun from the Dose section above. We’ll use Vitamin D synthesis, UV radiation from sun exposure, and chemical sunscreens to explore each. We’ll start with the Threshold Model. It’s the trickiest for health-minded people, so we spend the longest on it:
1. Threshold Model
The threshold model means that below a certain level, the body has pathways designed and ready to clear a substance without harm. Oral vitamin D supplements are a perfect example. Your body easily manages normal doses of vitamin D, regulating storage to keep levels in a healthy range. But push past a certain limit (the threshold) with long-term high-dose supplementation, and your body gets overwhelmed. Toxicity follows. Simple, right?
It is simple — for people who aren’t worried about toxicants. But, since you are taking a course about non-toxicity, you might see it as more complicated. You might be wondering: isn't the threshold model just a polite way of dismissing low-level exposures we should still be worried about? Shouldn't we treat every exposure as carrying some risk, even at doses below the threshold limit? Even if a threshold technically exists, isn't it still better to minimize the use of detox pathways, because any processing is stress on the system?
I have a three-part answer for you:
First, Vitamin D, like other essential substances (water, oxygen, salt, iron), is required for life. You need some vitamin D the way you need some water, oxygen, salt, and iron. They are truly harmless at low doses, and can kill you at high doses. Low doses are good, high doses are bad. It's not that low doses are a little bad and high doses are really bad. This is great news: our physiology is elegantly designed, and I think it's worth pausing to appreciate.
Second, you might think, fine, the body handles the basics, but for anything non-essential, like nickel, let's reduce our detox burden as much as possible. I agree with that instinct, but it's worth using the rest of this framework to see if that thinking is useful in your specific situation. Nickel is a good example. Most people are exposed to small amounts daily through food (it's naturally present in legumes, nuts, chocolate, and whole grains), water, and contact with stainless steel cookware and jewelry. For most people, the body handles these exposures and clears nickel without issue. The threshold for nickel toxicity is much higher than typical daily exposure. So aggressively avoiding nickel for the average person buys you very little, and costs you something (the mental load of vigilance, the difficulty of finding alternatives, the loss of foods you might enjoy). On the other hand, if you have nickel sensitivity, or are an industry worker breathing in nickel dust daily, reduction may very well be the right call. The point is that the threshold response itself doesn't tell you what to do. It tells you the relationship between dose and outcome. What you do with that information depends on the rest of the framework.
Third, and most sinister, establishing a threshold is enormously useful for industry. If you can claim a chemical you're producing has a threshold, you don't need to protect people below it. As David Michaels, the epidemiologist and former head of OSHA, has documented extensively, the pressure to maintain threshold thinking in industry and regulation is not purely based on science or health. Some of it is, but some of it is industry working hard to keep its products on the market. Within the scientific community, this idea is rejected— they state that there is no safe threshold for carcinogenic exposures (which we'll get to in the LNT model, next). Some skepticism about how thresholds are set for novel synthetic carcinogens and other chemicals is warranted.
So where does this leave you? Three things to take away: the threshold model is real, and for substances your body has evolved pathways for, it's actually good news. For non-essential substances, the model is useful as a starting point, but only if you also factor in your exposure level and susceptibility. And, the model is sometimes deployed in bad faith by regulators and industry to justify exposure levels that aren't actually safe, particularly for carcinogens. Hold all three at once, and take a pause to settle — this is a lot of dense information.
Threshold
2. Linear No-Threshold (LNT) Model
There are times when little doses are a little bad, and high doses are very bad. These are substances that follow a linear no-threshold (LNT) dose-response model. The model usually applies to carcinogens, where no amount is totally safe.
The clearest example is ionizing radiation like X-rays and gamma rays. There is no dose low enough to produce zero cancer risk, and as cumulative exposure rises, so does that risk. UV radiation from sun exposure follows this pattern for skin cancer risk. Some researchers argue that because our cells have DNA repair mechanisms, a safe threshold exists. They say that below a certain dose, repair keeps pace with damage entirely. But, the dominant position, based on big epidemiological studies, is that repair is imperfect. Some damage always escapes and that unrepaired fraction accumulates over a lifetime. No safe UV threshold for skin cancer has been identified. This is why childhood sunburns still show up in lifetime cancer risk decades later.
-
The same sun exposure produces two completely different dose-response relationships, depending on which biological effect you're tracking.
As you know, you can also get your vitamin D through sun exposure. Your skin has a self-limiting photochemical process that kicks in and shuts further production down once you reach your threshold. So again, you need a certain amount of sun, and vitamin D follows a threshold response.
But, the exact same sun exposure can cause DNA damage, and that situation follows a linear no-threshold relationship. Every photon from the sun that reaches a DNA strand is a chance for a mutation, and repair doesn't catch all of them.
So the same square inch of skin, in the same afternoon, is running a threshold response and an LNT response simultaneously, on two different endpoints. You need the sun to survive, and it's also a carcinogen. It's mind-boggling because different effects of the same exposure follow different rules. Some of the fighting on social media about sun exposure and sunscreen could be solved by knowing this, but it’s nuanced and requires a lot of work (work that you’re doing right now).
It happens in other ways, too, not just with the sun: every cell in your body uses oxygen to make ATP, the molecule your metabolism runs on. A small percentage of that oxygen doesn't fully convert to water, and escapes as reactive oxygen species, which damage DNA, proteins, and cell membranes. So in the same breath, on the same chemical pathway, oxygen is keeping you alive and oxidatively damaging you. The dose-response is both a threshold one and an LNT-style one: the cellular damage accumulates over a lifetime. This is one of the major theories of biological aging.
The substances that make us also unmake us. Life is a wild contradiction. Just a moment of appreciation for how mysterious this all is!
Linear no-threshold (LNT)
3. Non-Monotonic Dose-Response (NMDR)
Non-monotonic dose-response (NMDR) curves might already be your favorite dose-response. If you’ve spent time in the non-toxic world, you might know what these are already, and that endocrine disruptors might follow them.
NMDRs have a U-shape because low doses can produce a high response, and higher doses can produce less of response. This has been documented for a growing class of chemicals, particularly endocrine disruptors. The Endocrine Society's 2025 position statement notes that this pattern is now well enough established that it has significant consequences for how we do risk assessment. This means it’s no longer considered a fringe position, as it once was.
Which brings us back to the sunscreen studies we started talking about in the Dose section. After the 2020 FDA paper about sunscreen dose in the blood came out, a handful of studies followed, looking at the dose-response of these sunscreen chemicals. They said, ok, at this level of oxybenzone in the blood, what happens?
What they found in a 2023 review that looked at 254 studies on oxybenzone was interesting. After someone applies sunscreen to their whole body once, the amount of oxybenzone that shows up in their blood can reach the same levels that disrupt hormones (specifically, estrogen and testosterone activity) in lab and animal studies. Whether that actually causes hormone problems in people is still an open question. Frustratingly, three years on, we still don't have much human data to answer it.
So, when people raise panic about sunscreen in blood, that's not necessarily useful. We may find there isn’t a meaningful hormone risk in humans. But assuming we won’t, or dismissing concerns entirely, isn't useful or accurate either.
Non-monotonic
A more honest and accurate version of, "sunscreen chemicals are in your blood and that's toxic!" would be: “we don't yet know which dose-response model applies, and we'd rather not expose ourselves to something with unresolved endocrine-disrupting potential.”
A more honest version of "the dose is too low to matter," would be: “we don't know, but the risk of skin cancer is better established and more serious, so the tradeoff still favors using it.”
I used sunscreen and sun exposure for the dose-response section because it's such a cultural flashpoint right now, and I think it illustrates how amazingly complicated this all is. But of course, we can apply this thinking to any toxicant, in home products or otherwise. Here is your Dose-Response card:
Threshold model: Below a certain dose, the body can detoxify or repair and no measurable harm occurs. This applies to most non-carcinogenic exposures. This is where the phrase "the dose makes the poison" comes from.
Linear no-threshold (LNT) model: Any dose carries some probability of harm, however small. This is often applied to many carcinogens, because there is no dose at which the probability of a mutation drops to absolute zero.
Non-monotonic (U-shaped) curves: For some endocrine disruptors, low doses produce different effects than high doses, sometimes the opposite effect. This challenges the assumption that lower is always safer and is an active area of research.
Susceptibility
What's toxic for one person isn't necessarily toxic for all people. A chemical that's a modest concern for a healthy adult can be a much bigger deal for a pregnant woman or a baby.
Flame retardants in furniture foam are a good example. Studies have found toddler exposure levels nearly five times higher than their mothers', partly because young children spend more time on the floor and put their hands and toys in their mouths, which puts them in close contact with the dust that carries these chemicals. The timing matters as much as the dose: their brains and thyroids are actively forming, and exposure during this window has been linked to neurodevelopmental effects.
The complication of susceptibility is one of the main reasons EWG, IARC, and Prop 65 can't produce true risk assessments. Accounting for life stage, genetics, pre-existing conditions, sex, and body size makes risk assessment hard to do at scale. The framework you're learning lets you do it for yourself.
With that, here is your Susceptibility card. It's the last one before we get to Risk:
Key susceptibility factors include:
Life stage: The developing fetus and infant are profoundly more susceptible to neurotoxicants (lead, mercury, PCBs) because the blood-brain barrier is incomplete and neural circuits are actively forming. Exposures that cause no detectable effect in an adult can cause irreversible developmental harm.
Genetic variation: Polymorphisms in metabolizing enzymes (CYP450s, GSTs) mean some people activate or detoxify substances faster or slower. These inherited differences can shift individual risk significantly.
Pre-existing conditions: Compromised kidney or liver function reduces the body's ability to eliminate substances. Asthma increases sensitivity to airway irritants. Chronic inflammation may affect how substances are processed.
Body size: Smaller bodies receive a higher dose per kilogram of body weight from the same absolute exposure, which is one reason children are not just small adults in toxicology.
Risk
This is where everything above clicks into place!
Risk is the final output of all five previous steps: If you’re exposed to a hazard and experience a high enough dose of it in your body, its dose-response curve means harm at that level, and you’re susceptible, then it’s a risk.
Risk belongs to you specifically, not to the substance in isolation.
This is where most non-toxic conversations get stuck. It's much easier to argue about a single concept (hazard, exposure, dose) than to walk through all six. This framework is the harder, slower path, and it's also the one that gets you closest to a real answer.
For now, here is your risk card:
Risk exists on a spectrum and is almost always a probability, not a certainty. A 1-in-100,000 lifetime excess cancer risk means something very different than a 1-in-10 risk.
Recap
To evaluate whether any product in your home is a risk, all six steps matter. Here's the short version to refer back to anytime:
Hazard: the chemical or substance itself, and its inherent capacity to cause harm. A hazard sitting in a product isn't doing anything yet.
Exposure: what happens when the hazard leaves the product and reaches you. If there is no exposure, there is no risk.
Dose: how much of it actually gets into your body. Your body’s defenses intercept a lot, but aren’t perfect.
Dose-response: the relationship between the dose and what effect it causes. More isn't always worse, and less isn't always safe. It depends on the substance.
Susceptibility: how vulnerable you are. Age, health status, genetics, and timing all affect how your body handles the same dose that someone else might handle differently.
Risk: what you actually get when you put everything together. Change any one variable and you change the risk.
You can download a PDF with all 6 of the reference cards from above:
The framework only works if the science going into it is solid. Often, it’s not, and it can be really hard to tell at first. That’s what the next section is all about!
Enter the giveaway below, or start Part 3 ➜
Enter the Giveaway
You finished Part 2! You’re eligible to add an entry into the giveaway.
Fill out the form below to submit your feedback and add 1 entry to the giveaway. Each of the 10 sections has its own form, so you can earn up to 10 entries total. Drawing is June 8, 2026. See full giveaway details and prize list here.
Part 2 References
EWG, IARC, and the difference between hazard and risk
Toxicology Education Foundation. Hazard vs. Risk.
The dose-response framework
Toxicology Education Foundation. Basics of Dose-Response.
FDA findings on chemical sunscreen ingredients in the bloodstream
Matta, M. K.; Florian, J.; Zusterzeel, R.; Pilli, N. R.; Patel, V.; Volpe, D. A.; Yang, Y.; Oh, L.; Bashaw, E.; Zineh, I.; Sanabria, C.; Kemp, S.; Godfrey, A.; Adah, S.; Coelho, S.; Wang, J.; Furlong, L.-A.; Ganley, C.; Michele, T.; Strauss, D. G. Effect of Sunscreen Application on Plasma Concentration of Sunscreen Active Ingredients: A Randomized Clinical Trial. JAMA 2020, 323 (3), 256-257.
Vitamin D regulation and the threshold model
Vieth, R. Vitamin D Toxicity, Policy, and Science. J. Bone Miner. Res. 2007, 22 (S2), V64–V68. PMID: 18290725.
Threshold model useful for industry
Michaels, D. The Triumph of Doubt: Dark Money and the Science of Deception; Oxford University Press: New York, 2020
Tabuchi, H. Starting With Formaldehyde, Trump Administration Reassesses Chemical Risks. The New York Times, December 10, 2025.
UV radiation, skin cancer, and no safe threshold
National Academies of Sciences, Engineering, and Medicine. Health Risks from Exposure to Low Levels of Ionizing Radiation: BEIR VII Phase 2. National Academies Press: Washington, DC, 2006.
Juzeniene, A.; Grigalavicius, M.; Baturaite, Z.; Moan, J. Minimal and Maximal Incidence Rates of Skin Cancer in Caucasians Estimated by Use of Sigmoidal UV Dose-Incidence Curves. Int. J. Hyg. Environ. Health 2014, 217(8):839-44. PMID: 25023193.
Vechtomova, Y. L.; Telegina, T. A.; Buglak, A. A.; Kritsky, M. S. UV Radiation in DNA Damage and Repair Involving DNA-Photolyases and Cryptochromes. Biomedicines 2021, 9 (11), 1564.
Lergenmuller, S.; Rueegg, C. S.; Perrier, F.; Robsahm, T. E.; Green, A. C.; Lund, E.; Ghiasvand, R.; Veierød, M. B. Lifetime Sunburn Trajectories and Associated Risks of Cutaneous Melanoma and Squamous Cell Carcinoma Among a Cohort of Norwegian Women. JAMA Dermatol. 2022, 158 (12), 1367–1377. PMID: 36197657.
Oxygen, reactive oxygen species, and biological aging
Liguori, I.; Russo, G.; Curcio, F.; Bulli, G.; Aran, L.; Della-Morte, D.; Gargiulo, G.; Testa, G.; Cacciatore, F.; Bonaduce, D.; Abete, P. Oxidative Stress, Aging, and Diseases. Clin. Interv. Aging 2018, 13, 757–772.
Endocrine disruption and non-monotonic dose-response curves
Vandenberg, L. N.; Colborn, T.; Hayes, T. B.; Heindel, J. J.; Jacobs, D. R., Jr.; Lee, D.-H.; Shioda, T.; Soto, A. M.; vom Saal, F. S.; Welshons, W. V.; Zoeller, R. T.; Myers, J. P. Hormones and Endocrine-Disrupting Chemicals: Low-Dose Effects and Nonmonotonic Dose Responses. Endocr. Rev. 2012, 33 (3), 378–455.
Endocrine Society. Position Statement: Endocrine-Disrupting Chemicals. Endocrine Society; 2025.
Oxybenzone (scientific name: benzophenone-3, BP-3) in blood and endocrine-disrupting effects
Mustieles, V.; Balogh, R. K.; Axelstad, M.; Montazeri, P.; Márquez, S.; Vrijheid, M.; Draskau, M. K.; Taxvig, C.; Peinado, F. M.; Berman, T.; Frederiksen, H.; Fernández, M. F.; Vinggaard, A. M.; Andersson, A.-M. Benzophenone-3: Comprehensive Review of the Toxicological and Human Evidence with Meta-Analysis of Human Biomonitoring Studies. Environ. Int. 2023, 173, 107739.
Flame retardants and developmental susceptibility
Castorina, R.; Bradman, A.; Stapleton, H. M.; Butt, C.; Avery, D.; Harley, K. G.; Gunier, R. B.; Holland, N.; Eskenazi, B. Current-Use Flame Retardants: Maternal Exposure and Neurodevelopment in Children of the CHAMACOS Cohort. Chemosphere 2017, 189, 574–580. PMID: 28963974.
Cheng, X.; Lu, Q.; Lin, N.; Mao, D.; Yin, S.; Gao, Y.; Tian, Y. Prenatal Exposure to a Mixture of Organophosphate Flame Retardants and Infant Neurodevelopment: A Prospective Cohort Study in Shandong, China. Int. J. Hyg. Environ. Health 2024, 258, 114336.
Explore More Interior Medicine
Product Reviews
Tools and References
About Interior Medicine
Material Health Guides
Learn ➜ Course ➜ Part 2