Toxic at Any Level

Why Our Safety Standards Are Failing Us

When Does Disease Begin?

Let me take you back to medical school, where I learned to see disease in binary terms: you either had it or you didn’t. Over time, I came to see how wrong that was. Most diseases exist on a continuum. When does cancer begin? When a person starts smoking? When a rogue cell mutates? When the immune system fails to keep it in check? Or only when a doctor makes the diagnosis?

I was also taught that toxic chemicals had “safe” exposure levels—and if exposures stayed below those thresholds, we had nothing to worry about.

Cause and effect were simple: one agent, one disease, one safe level. An agent could be a microbe, a toxic chemical, or a missing nutrient. Case closed.

But life, like medicine, is messier than that.

Most chronic diseases and conditions—premature birth, heart disease, lung cancer—don’t arise from a single event. They accumulate, quietly and gradually, like junk in a garage. And the most studied toxic chemicals show a disturbing pattern: even tiny exposures can do measurable harm, especially when they’re widespread.

The Strange Curve of Toxicity

Regulatory science rests on two key assumptions:

· Non-carcinogens are safe below a threshold. The linear, threshold model (Fig 2A)
· Carcinogens follow a linear model—double the dose, double the risk. This is the linear, no-threshold model (Fig 2B)

Turns out, both assumptions are wrong for some of the most well-studied and widely disseminated toxic chemicals and pollutants.

Figure 2: Different exposure-response curves

For decades, researchers have studied toxic chemicals and chronic disease. Again and again, the same pattern has emerged: the greatest harm occurs at the lowest levels of exposure. Imagine a graph that rises steeply, then flattens. That’s a decelerating exposure-response curve. (Fig 2C)

At first, I thought it was a blip in my own data. But the same pattern kept showing up—in study after study, chemical after chemical.

Lead and the Myth of “Safe”

Lead is the poster child of industrial-era pollution. Early studies found that blood lead levels below 10 micrograms per deciliter—once considered safe or “acceptable”—were linked to significant IQ loss (Fig 3A). The biggest drop happened at the lowest exposures.

Figure 3: Exposure-response curves: Lead and IQ, PM2.5 and Mortality, Benzene and Leukemia.

A rise in blood lead from under 1 µg/dL to 30 µg/dL was linked to a 9-point drop in IQ. But two-thirds of that loss occurred before the level even reached 10 µg/dL.

Translation: most of the damage occurred at levels considered low by today’s standards. This finding has been replicated more than a dozen times. No safe level of lead has ever been identified.

The Air We Breathe

Another culprit is fine particulate air pollution—known as PM2.5, for particles less than 2.5 microns in diameter. These particles are small enough to penetrate deep into the lungs and enter the bloodstream. Arden Pope at Brigham Young University found the same steep curve when linking PM_2.5 to mortality (Fig 3B). The steepest increase in risk occurred at the lower levels of exposure.

Going from “clean” air to moderately polluted air is more dangerous than going from moderately polluted to filthy air. Yet we tend to set air quality standards at “moderate.”

A Canadian study found increased mortality at PM_2.5 levels as low as 1 µg/m³. That’s almost nothing—and still harmful.

When cities ban indoor smoking or reduce air pollution, health improves almost immediately: fewer heart attacks, fewer preterm births, even among nonsmokers. Low-level exposures aren’t subtle after all.

Benzene, Asbestos, Radon—Same Story

We’ve seen the same pattern with other toxic chemicals:

· Benzene: causes leukemia; risk is steepest at low exposures (Fig 3C).
· Asbestos: causes mesothelioma; low exposures still kill.
· Radon: raises lung cancer risk; steepest rise at low levels.

These aren’t outliers. They’re part of a pattern.

But Why?

The biology isn’t fully understood. With benzene, toxic metabolite production slows at higher doses—like the body saying “enough.” For other chemicals, we don’t know why the dose-response curve bends.

Is it a fluke? Unlikely. This curve appears across chemicals, studies, populations, and continents. Some critics blame confounding or poor measurement. But those critiques don’t hold up—not when the evidence is this consistent.

The Prevention Paradox

If low-level exposures to the most widely-disseminated toxic chemicals cause the most harm, our current system of risk assessment needs a major overhaul.

Take lead again. David Bellinger at Harvard estimated that U.S. children born over six years lost 23 million IQ points due to lead. But if we only protect kids whose blood lead exceeds 5 µg/dL—the WHO’s action level—we prevent just 18% of the damage.

In other words, most harm happens below the threshold.

This is Geoffrey Rose’s “prevention paradox”: the greatest number of cases comes from people at low or moderate risk—not the high-risk few.

We think of lead as a threat to children’s brains, but the damage extends further. Even in adults, low-level lead exposure increases risk of cardiovascular disease. In a national study of 14,000 U.S. adults, we estimated that nearly 400,000 deaths annually—more than one in six—could be attributed to lead, including 250,000 from heart disease.

And again, most of those deaths occurred at blood lead levels below 5 µg/dL—levels still widely considered acceptable.

Why the Silence?

Why haven’t regulators responded?

Because accepting that low-level exposures cause widespread harm would open the floodgates. It would mean admitting that millions more people are at risk. That we need near-zero exposure standards for lead, PM_2.5, benzene, asbestos, and radon.

That’s politically and logistically daunting.

Regulators walk a fine line: protect public health, but also reassure the public. Those aren’t always the same thing. Saying “no level of lead is safe” makes headlines—and headaches.

Then there’s cost. Reducing exposures is expensive. The benefits—fewer heart attacks, smarter kids, less cancer—are real, but spread out over time and across populations. In a market-driven health system, that’s a hard sell.

What We Need to Do

We need to abandon the myth that low exposures are harmless.

That means:

· Rethinking chemical regulations for air, water, and consumer products.
· Protecting people with even low-to-moderate exposures.
· Recognizing that for some toxins, zero may be the only safe number.

It won’t be easy. But the alternative is to keep pretending a little poison is no big deal—until, of course, it is.