Exposome Perspectives Blog by Robert O. Wright, MD, MPH
I believe most people think that Ernest Hemingway would take down Oscar Wilde in a boxing match, but I want to offer a more nuanced prediction of what would happen. In a spur-of-the-moment fight, yes—but if it were a challenge and Wilde had time to prepare, I think he’d win. Oscar Wilde deconstructed social norms. Hemingway, in contrast, romanticized masculinity and emphasized a code of honor. Wilde looked for the hidden weakness. Hemingway would likely insist that they shake hands before the match. Wilde would know this—that’s when he’d get him: a sucker punch to the nose. Hemingway would never know what hit him. That’s how it would’ve gone down.
So how does this relate to science communication? Before a deeper analysis, I might argue that Hemingway’s style of writing is better suited to science communication. However, Wilde’s gift for revealing what is hidden is also key to public health messaging. After giving it some thought, I believe it’s just as important to be Oscar as it is to be Ernest.
We live in an age of distrust—phishing, spam phone calls, recurring subscriptions we can’t remember signing up for, everything we don’t want seems impossible to cancel. Alongside these societal changes, science has become “Big.” It’s big data and big business. We are also clearly getting sicker—so it’s a big deal. Public health messages have not kept pace with societal or scientific complexity. We know that our messages can’t be long-winded or no one will listen, but if the science is complex, simple messages run the risk of being untrue. Trying to explain complicated health trends and conflicting data or conflicting interpretations of the same data takes some guts to be honest. I follow a number of people on Substack who are really good at it. I’m still learning. The implications of this dichotomy may be vast. More than ever, we may be prone to manipulation by simple messaging that leaves out crucial caveats because our brains prefer the simple. The messenger may be well-intentioned, or they may have a vested interest in getting their version of the science to the public to sell a product or to achieve notoriety in the press. Furthermore, well-intentioned scientists are often poor or unprepared messengers—myself included—because we don’t put an equal amount of thought into our messages as we do our study designs.
Messaging in the Age of Big Data
We live in the age of “omics” and artificial intelligence—genomics, epigenomics, proteomics, and even exposomics are conducting increasingly complex studies with highly nuanced results—we report polygenic risk scores nowadays and sum up all the small effects of genetic variation in the hope they better predict something about our health. Science is also big business. There’s money to be made from drug development or widely adopted blood tests. I may even have a motive—maybe I want more research funding for my line of work, for example (yes, I do, all scientists do—if they say otherwise, they are lying). Scientists rarely speak directly about their work except to other scientists. There may be 10 scientific disciplines studying autism or Alzheimer’s disease, but they rarely talk to one another, and too often, even if they tried, the other discipline wouldn’t listen. I doubt the public understands how political and siloed scientists really are. By political, I am referring to the belief that our work is the most important and that other disciplines don’t understand it and are therefore “lesser”. I find myself editing this blog repeatedly to try to remove judgments about the motives of those who disagree with me—especially on whether we should focus on genetics instead of gene-environment interaction. I don’t actually know what motivates them, and I shouldn’t presume to know—I should only point out facts and explain what facts drive my perspective. I’m not particularly religious, but perhaps there is a Biblical reference that says it well.
“A wicked messenger falleth into mischief: but a faithful ambassador is health.” Proverbs 13:17
Back to the dilemma: how does one distill extremely complicated “big data” research into a public health message that is both accessible and accurate? I believe we resort to the oversimplified “nature vs nurture” arguments in part because that message is so easy to understand, not because it reflects reality or even what the science shows. But I’ve written a lot about Nature vs Nature in this blog recently. I’ll try to deconstruct a different science message instead.
Toxicology in the 15th Century: the dose makes the poison
One of the most famous toxicology messages is “the dose makes the poison.” First coined by the Swiss physician Paracelsus to show that every substance can be both toxic and benign—it’s how much you ingest or breathe that matters. Arsenic can be a medicine at “low” doses (e.g., for cancer treatment) and a poison at “higher” doses. Even water, which is essential to life, can be toxic if the “dose” is “high” enough. Every vitamin and nutrient can be toxic if enough is taken. Given that this was about 500 years ago, and science communication was in an embryonic stage, his statement was actually a bit more long-winded: “What is there that is not poison? All things are poison, and nothing is without poison. Solely the dose determines that a thing is not a poison.”
Does dose matter the most? Or is that just a boast?
Slightly shorter than Paracelsus but I am clearly still in the learning stages of science communication. If I were a wicked messenger, I could use “the dose makes the poison” to imply that small doses of anything are always safe, i.e., “So what if there is a little microplastic in your blood—it’s such a small amount, what could it do?” Dose is not the full story on toxicity though. The phrase can also be used as a precautionary tool, i.e., each chemical is different, and sometimes small doses matter. Alternatively, it can be used as a way to chastise those who act with caution: “Seriously, how can that hurt you? I was exposed to lead in gasoline as a kid. I went to college. I’m fine. You’re being ridiculous. After all, the dose makes the poison.” The natural reaction may be to counter with an example of a chemical that is very toxic at low doses by showing low doses can indeed hurt you. “Well, the dose of botulinum toxin needed to kill you is very small—a small vial could wipe out a city. They may counter with: “And the dose of water needed to kill you is very high. What’s your point?” We all were exposed to lead if we were alive in the 1980s, but not many of us were exposed to botulinum toxin. By equating the two poisons in a discussion, we are not addressing the issue of lead toxicity directly. I am advocating for a different approach—explain the nuance and the role of context in lead poisoning.
Can the toxic and beneficial dose of the same chemical vary greatly in different people?
Dose-response curves were not designed to be universal. They represent the average dose response curve across a population and only accurately reflect those nearest the average. Many people predictably have different dose-response curves for the same chemical. Over the last 50 years, we’ve started examining the dose response curves by subgroups—pregnant women, children, and the elderly have received the most attention, but there are other contextual factors that can dramatically impact the dose response curve. Intuitively, once you point them out, it’s not difficult to understand that different people will react differently to the same exposure in predictable ways.
Our messages still need to be succinct, but instead of oversimplifying, we should figure out how to explain the complexity without being obtuse or boring. Let’s start with a well-known poison—carbon monoxide (CO) as an example—no one disputes its toxicity. It’s a deadly, odorless gas that binds to hemoglobin in our blood preventing oxygen transport. The dose that will kill is surprisingly small. If it reaches 1000 parts per million (ppm) in the air, it is very fatal quickly. That means death will come when only 0.1% of the air is CO and the other 99.99% is mostly nitrogen and oxygen, just like in normal air. At a lower dose, such as 200 ppm, CO takes a few hours to kill you, but it can still kill you. So, chronicity of exposure is an important context. OK, chronicity and dose make the poison.
What about people with a disease? If I have heart disease, might the dose of carbon monoxide that will kill me be lower than someone without heart disease? Chronicity, dose, and health status make the poison. Still succinct. Or what if I am an infant still developing? Does that change the dose of concern? Or what if I have a genetic anemia and can’t carry enough oxygen in my blood under normal circumstances, can I die from a lower dose? And that’s just differences in different people—what about geography, like altitude? Is the fatal amount of CO the same if I live 8,000 feet above sea level? Or occupation—what if I am a painter? Paint strippers contain methylene chloride, which is metabolized to CO in the liver. I may have a chronic low level CO exposure from work, does that change the dose in the air needed to kill me? What about mixtures? If I am exposed to a low dose of CO and I take multiple doses of aspirin because of the headache it gives me, does that affect the lethal dose because aspirin toxicity works similarly to CO? Ok, I give up—maybe everything makes the poison?!
While the dose may make the poison in all these examples, it is not the same dose making CO poisonous. If we just say the dose makes the poison and stop there, are we really explaining how a poison works? Isn’t greater understanding the whole point of science communication? I should point out that there are protective contexts, too. I might have a genetic variant that reduces the binding affinity of CO to hemoglobin. I may be a high-performance athlete (I wish…) and can deliver oxygen to my tissues more efficiently than someone who is mostly sedentary. Perhaps the reason someone thinks lead toxicity is overblown is because the toxic impacts are contextual. It may be small in some people and large in others, making it seem hidden. How you live your life—diet, exercise, and social engagement can also mitigate the impact of an environmental toxicant. Long-term tutoring might overcome childhood lead poisoning—unfortunately, we’ve never systematically tried that idea. If you replace CO with lead in the preceding paragraph, the conversation would be very similar. The key point is human agency—knowing yourself (i.e., your context), what has happened to you, and what you can do about it, including choosing not to be exposed. If you believe you have a right to know what you are eating, drinking, and breathing, you should care about accurate science communication.
The context makes the poison
This is my attempt at a 21st century message on toxicology and science communication. Dose itself is a form of context and therefore it fits with the larger idea that context matters. It’s not only big vs. little doses or short vs. long-term doses—while those are important contexts, context doesn’t end there. It’s the overall context, which we can now measure with the exposome. During the Renaissance, Paracelsus was not thinking about gene-environment interaction, vulnerable populations, the obesity epidemic, the developmental origins of health and disease, or even the exposome. The dose makes the poison was for a different age. We need to update the science communication around toxicology to fit both what we know and what the public can understand.
We should care about how we message science because research findings can impact our lives, and exposure should be a matter of choice. We all want to make the most informed choices in what we eat, drink, purchase, or breathe. Do parents have enough information to make an informed choice about chemical exposures during pregnancy in food, clothing, or water? If not, we should fix that problem and give them the information they deserve and want. They can then decide—because that is their right. Corporations can make their products, and we can choose to purchase the ones we feel are safe, but that system requires transparency, which is lacking these days. One way to help would be if more scientists learned how to explain their research to an educated, interested audience that wants to understand.
Alcohol is a good example, because I know the sources in my diet and I can make choices about whether to ingest it. I recently cut back on alcohol consumption. I had heart surgery about a year ago, and I decided I didn’t want the low-grade cardiotoxicity of alcohol impacting my heart, which already has diminished circulation. When I was 25, I mostly worried about drinking and driving or whether it would make me do something stupid I would later regret (that’s about as confessional as I intend to get). Back then, a big dose made the poison. Today, low doses are enough to make the poison for me. My context changed; alcohol’s effects did not. The choice is mine, though. If a company were to sneak alcohol into my morning coffee and I didn’t know about it, I’d be upset because they took away my ability to choose. Suppose I want to avoid microplastics, forever chemicals, or pesticides. That is also my right—and as one of the millions of Americans with a chronic disease, I want to make informed choices—but I need accurate information—not judgment or obfuscation. The more we inform the public about what is in our air, water, food, perfume, cookware, etc, the better off we all are. If that information isn’t provided and the health effects are glossed over as a one-size-fits-all, such as “the dose makes the poison,” we are being deceived.
Effective science communication avoids oversimplification, undue certainty in either direction, and judgmental tones. The audience should be treated with respect. Messages that seem really simple, are expressed with certainty, and interspersed with judgmental tones should make us suspicious. We need to embrace the scientific picture as one of complexity and context, and not condescend. We are more than capable of providing accurate information by including all the uncertainty and caveats, and the public is capable of understanding uncertainty. In today’s world, understanding context is critical, and oversimplification is too often used to mislead or avoid discussing all the caveats in a way that informs choices and behaviors. Risk communication should be as simple as possible, but not simpler than possible. Context matters deeply in science—in every discipline. I believe we avoid it too often, and this is causing a lot of confusion. The context makes the poison. We should talk about that.
“There was a wicked messenger
From Eli he did come
With a mind that multiplied the smallest matter
When questioned who had sent for him
He answered with his thumb
For his tongue it could not speak, but only flatter”
– Bob Dylan