Everyday we rely on governmental regulatory systems and animal testing to tell us if the chemicals we’re exposed to will or won’t hurt us. But what if AI could help us get more ethical and more accurate answers and avoid animal testing altogether?
In today’s episode, Ken is joined by Dr. Thomas Hartung, Ph.D, a professor at John Hopkins Bloomberg School of Public Health, whose work spans the fields of environmental health, engineering, molecular microbiology, and immunology. Hartung’s team are developing cutting edge tools in the world of toxicology: AI and human stem cell models that help us understand our genetic reactions to chemicals and, ultimately, make our lives safer.
Hartung argues that AI has the power to analyze vast data sets of untested or undertested chemicals, “cocktail” combinations, and even different exposure rates and frequencies. All these variations can give us a much more nuanced understanding of the chemicals that permeate our lives, potentially reshaping how chemicals are assessed and regulated.
===
Ken: Hi there. I'm Ken Cook, and I'm having another episode. This episode I'm particularly interested in sharing with you. My guest today is Dr. Thomas Hartung, an MD and PhD who splits his time between environmental health and engineering and molecular microbiology and immunology at Johns Hopkins. I mean, this is exactly how I split my time, so I completely relate to him.
Thomas has spent his career focused on toxicology. He used to run the European Centre for the Validation of Alternative Methods, and for the last 17 years at Hopkins, he's been building some of the most interesting tools I've come across for understanding how chemicals actually affect the human body, tools that don't rely on animal testing, which the field of toxicology has leaned on for decades.
Instead, he and his team are using AI and human stem cell models to try and understand how our genes and our environment interact, especially when it comes to neurodevelopment. Last year, we had a guest, journalist, author, and advocate, Eric Garcia. Eric is autistic, and through his amazing book, "We Are Not Broken: Changing the Autism Conversation", he's done exactly that, and we'll continue to do that on this show.
Dr. Hartung's team is exploring autism and ADHD and rising diagnosis rates for both. It's worth noting that a significant share of that rise reflects broader diagnostic criteria and greater public and clinical awareness. What Thomas' research actually looks at is narrower. It's about whether specific combinations of genetic risk and chemical exposure, including chemical mixtures, might increase susceptibility to certain changes that researchers can see in lab models.
That's a real careful, hypothesis-driven scientific question, and it's a very different thing from the kind of sweeping, unfounded claims about vaccines and Tylenol and autism that have been floating around, especially this past year Claims that science has never supported, no matter how many times the claims get repeated.
So on this show, we celebrate our neurological differences. We're also always going to follow the science and fight for the next generation to be more safe and more accessible for everyone. Thomas, thank you so much for joining. I'm, I'm thrilled you're here. Can you say a little bit about the trajectory of your career?
Because the background is really fascinating, how you found yourself to be now at the forefront of so many important questions at the interface between environment and health.
Dr. Hartung: Yes, sure, Ken. Thanks for having me. My career was not very linear, I would say. I-
Ken: That's right ...
Dr. Hartung: I was interested in science, every type of science, and when doing my civil service, because I didn't wanted to join the military in Germany, the only thing I could study was mathematics informatics, and having worked as a coder.
So that was actually the first partial career before I finally could do what I was longing to, which was biochemistry, and I want to do, use these tools for medical research So I convinced the university to let me study medicine and biochemistry in parallel, which was almost unheard of. They said I will fail anyway.
And so I did not finish the mathematics informatics, but all three have in the end, um, shaped me, became MD PhD in pharmacology and toxicology, only to find out this, uh, for an animal lover, not the ideal profession. Pharmacology and toxicology are areas which are based on a lot of animal testing.
Ken: Yeah.
Dr. Hartung: So I was always trying to find a sweet spot where there is, let's say, either working with humans or was working with cell culture until, yeah, life found for me my path into alternatives to animal testing, leading the European Center for the Validation of such alternative methods, and then, uh, moved, uh, 17 years ago to Hopkins.
As I thought for two or three years inter- experience in international health, and it did grow on me, and I f- I love it, and I, I think I found my place.
Ken: It's not linear, but it does all make sense. So thanks for giving us a little bit of that background. Let's just leap right into the whole question, all of the questions around animal testing.
I have to say, our perspective at Environmental Working Group, we've, we have long wanted to emphasize alternatives to animal testing for all kinds of reasons. The inefficiencies of it. We worry about the gaps between what we can reliably say from animal testing and what applies to human beings. The fact that industry is so excited about it gives us our own anxieties, right?
Mm. The notion that there will be a testing regime that might, um, be more permissive in allowing chemicals onto the market or not take into account, uh, mixtures, which we're not doing now anyway. Mm. I know you wrestle with all of these issues, both in the European context and here. The whole question of, well, what can we do to really understand environmental health issues and ground it in science, even as we reduce the use of animals as the foundation of so much of our toxicology.
So say a little bit about that, that divide, and is my skepticism completely unfounded, or i- is there reason to be a, a little bit concerned that some people may be overstating or, uh, applying alternative methods where it might not be developed enough yet?
Dr. Hartung: I mean, first of all, uh, I'm a physician in the leading school of public health in the world.
I'm not compromising on patient and consumer safety over animal welfare, yeah?
Ken: Yeah.
Dr. Hartung: Science is incredibly progressing. Our knowledge is doubling at the moment about every three years.
Ken: Yeah.
Dr. Hartung: I'm now 62 years old, and the methods in toxicology in particular, where we use these animal tests, they have been introduced when I was not yet born or in kindergarten.
Ken: Yeah, yeah.
Dr. Hartung: It is difficult to understand how one area of science- Yeah ... can stick to methodologies which are clearly outdated.
Ken: Yeah.
Dr. Hartung: And this is something I, I would love to overcome, but only with better methods, yeah? And it is a scandal that so many things we are using every day have never been properly tested.
Ken: Yeah.
Dr. Hartung: In the US, we have 4,500 chemicals which we add to food. 80% of them don't have adequate data to do a risk assessment, which are publicly available. So imagine this, yeah? We're eating this- Yeah ... every day-
Ken: Yeah ...
Dr. Hartung: and nobody has the data.
Ken: And in fact, the regulatory system has rearranged itself or formed itself around that lack of knowledge and, and just allowed the things into the food supply by law and regulation untested.
Dr. Hartung: Yes. We get about 1,000 new chemicals to the market in the US per year. 1,000, yeah? And the capacities to do a proper risk assessment we have probably for around 150. So by finding methods which are faster, cheaper, and human relevant, we're doing everyone a service. I understand your skepticism over industries and, uh, we have seen too often that industries were defending their chemicals too long.
But at the same time, I believe nobody wants to poison their customers. They simply try to not give up on something profitable if the evidence is poor.
Ken: Yeah.
Dr. Hartung: So for me, it is really about creating strong evidence when we have to do something, and avoid this alarmism, everything is bad. And we have chemophobia, yeah?
People are afraid of chemicals, and that's not a good thing. We have tripled life expectancy while all of these chemicals came into our life, yeah? So there must be something good about it as well, yeah?
Ken: Yeah, no, it is not black and white. Yeah, on the industry side we see the sort of the reverse position, which is if, if we don't have evidence or if, uh, it's at a de minimis level, we don't have to look any further.
It's perfectly safe. So, you know, it, it really does cry out for the kind of vanguard work you're trying to do. I can see why it, it grabs you as a, as someone who's fundamentally concerned about human health, but wants to approach it in a way that's m- has all these benefits and advantages to the traditional way we've done it.
Can you give an example? Maybe I want you to describe some of the new methodologies that you're exploring, and then a real world example where examining a chemical or a class of chemicals with these new methodologies- has come to different conclusions than what we've learned from toxicology to date.
Dr. Hartung: What we should first of all, uh, understand is science is delivering. There is more methodologies which actually can do the job, and this is very, very important because the industries are changing. So for example, about half of the new drugs we are getting are no longer synthesized by a chemist as small molecules where you might have the hope at least that animals show a similar reaction to humans.
About half of them are now antibodies or proteins of humans which are just working in humans. They're not working, not even in chimps, not even in non-human primates of any type. Yeah?
Ken: Yeah.
Dr. Hartung: So we need for these type of products something which is based on human cell systems. There's also this need for assessing so many chemicals.
Ken: Yeah.
Dr. Hartung: Chemists have synthesized more than 200 million chemicals so far. 350,000 chemicals are in commercial use. That's the number of registered chemicals in the 19 most developed countries. But we have data not even for 10% of them.
Ken: Yeah.
Dr. Hartung: We have- More or less sufficient data for about 1% of them. Uh, but if you ask, for example, do we know whether they produce cancer or not, it's one in 1,000, 0.1%, which has this- Yeah
has this type of data, yeah? And we will- Yeah ... not solve this problem with millions of animals made available. No. We need different types of technology, and that's what science is delivering at this moment, yeah.
Ken: Yeah. And so what, for example, would the methodology be to examine a family of chemicals? I know there are many of them that are evolving in your field, but what would be an example of how you would examine anew a chemical or f- a family of chemicals for their carcinogenicity?
Dr. Hartung: Yeah.
Ken: The method you're talking about versus the way we used to do it. Mm-hmm. I know I'm asking you to maybe give a toxicology 101 lecture, but I think it would be helpful for my audience to hear about that.
Dr. Hartung: Yeah. There's, at the moment, actually two technologies which are at the forefront of making a difference.
One of them is artificial intelligence, AI.
Ken: Yeah.
Dr. Hartung: It is really most remarkable, yeah? AI is progressing so fast. Since deep learning was introduced in 2010, it is doubling capacity every three months. So this year's AI- Yeah ... is eight times more powerful than last year's. Whenever we have a situation where we know enough about similar chemistry, AI is fantastic to make connections, yeah?
And to help us find out what is known about the slightly different molecules, and then close the gaps and, and make predictions. This is working out extremely well and helps us to say, "This smells like problem. This does not."
Ken: Uh-huh.
Dr. Hartung: Then you can focus your energy on the things which smell like problem, yeah?
Ken: And what does that focus entail? Let's say the, the AI step, looking at, uh, comparable chemicals or chemicals in the same family, structurally similar. Then what happens? What's the next step?
Dr. Hartung: The step would be that a regulator or also somebody working in this area in, in, in the industry can start saying, "Okay, this seems to be a group of chemicals which is linked to liver problems or to cancer.
Now we have to rule it out. Now we have to look into this." It is, at the moment, mostly useful for priority setting. It's not taking a decision. Yeah? It is not saying- Yeah,
Ken: yeah ...
Dr. Hartung: it's thumb up or thumb down, but it's saying here, "If you cannot study them all, then better look into these."
Ken: Yeah.
Dr. Hartung: Okay. The other side is that we do have better and better cell culture systems.
And this comes mainly out of the development of stem cell technology. It was only in 2006, so 20 years ago, that Yamanaka, uh, found a methodology how we can produce stem cells from skin or blood cells from a human, and then science delivered. We can now produce tiny little brains, tiny little livers and kidneys-
Ken: Right
Dr. Hartung: and we can study how well they work. I like to call them living biopsies. They're really reproducing some of the architecture and the function of these organs, and you can study them. If we know somebody has Let's say a problem in the brain like Alzheimer's disease, the little brain organoids we can do with their stem cells also have a problem.
They also show some features of, of Alzheimer. In this way, we can start producing models of the disease of these people which are linked to, we call this phenotype of these people. And then this is the starting point for understanding the disease, for understanding are there chemicals which are aggravating Alzheimer?
Can we find chemicals which are disturbing neurodevelopment and might be linked to autism? Yeah. And we will never find it with an animal test, which- Yeah ... in case of developmental neurotoxicity takes you two years, 1.4 million for one chemical, and uses more than 1,000 animals. This is not acceptable ethically, economically, time-wise, and nobody can test the tens of thousands of substance which could possibly be blamed.
Yeah.
Ken: Yeah. You mentioned two technologies. One is AI that's come along. What's the other one?
Dr. Hartung: The other one is really the stem cell-based bioengineering- Okay ... of, um, human relevant test systems. So we have now organoids, which are really not just one cell, yeah? Take the lung, we have more than 40 cell types in the lung, yeah?
Ken: Yeah.
Dr. Hartung: All together they give lung function, yeah? And we, with the stem cell technologies, we can make pretty complete models, but we don't have to stop there. We can say, "Okay, nice to have a lung, but if the liver is, uh, metabolizing the drug and the toxic agent is the metabolite, then we need to co-culture it with the liver."
And in the meantime we can put a lung, and a liver, and a kidney, and perhaps a brain on a chip, yeah? And, uh, perfuse them. It's really plug and play, and, uh- ... yeah, not yet have the entire human body on a chip, but the important organs which we want to understand how they interact.
Ken: Yeah. The organs that are often the target for toxic substances that can cause problems.
Dr. Hartung: Yeah. For example, liver is the, the typical problem organ for many drug developments. We call this drug-induced liver injury. This costs pharmaceutical industry billions per year. Yeah, because they develop these drugs very far and then only the clinical trials or sometimes even only after they deploy to the market, they find out that an unacceptable number of people develops liver problems, yeah?
And having filters which say, "Ah, uh, this one smells like problem," better early on look into liver problems. These are saving this industry billions, yeah.
Ken: And making drug development much, much more efficient and-
Dr. Hartung: Yeah ...
Ken: more targeted.
Dr. Hartung: Yeah. The AI is not only finding the problems, it's also helping to build effective substances.
We have AI-designed drugs, uh, roughly since 2019. In 2024, so only five years later, 18 AI-first drugs, how we call them now, so drugs designed- Mm-hmm ... by AI-
Ken: Yeah ...
Dr. Hartung: made it into human trials. In 2025, the first of them already made it after the first patient trial was successful to make them a large patient trial.
This is a value proposition which is enormous for the industry.
Ken: No, no question about it. So say a little bit about the work your team is doing around the exposome. Yeah. Give me your definition of exposome too.
Dr. Hartung: Yeah. So the, the exposome is, uh, at the moment essentially defined as everything which is not the genome.
Ken: Yeah. Yeah.
Dr. Hartung: Over the last 40 years, we had a major, major development in biomedicine that we were mapping the human genome. Yes. And, uh, over these 40 years, we have seen how the hu- genome became part of our approach to diagnostics, to treatments. Our entire understanding how drugs work, how disease is, uh, advancing, is shaped by genetics.
You also saw the enormous progress of, of the technology. The first genome, that cost us $3 billion. Yeah? Yeah. Now you can get a full human genome in days for $200.
Ken: It's mind-blowing, really.
Dr. Hartung: This is because, uh, as soon as something becomes an engineering problem and there's a market, this will advance.
Ken: Yeah.
Dr. Hartung: The same holds for finding chemicals, let's say, in your blood. The important factor here is how sensitive is your methodology. How small amounts can you still find? Over the last 30 years, uh, we have become 1 million times more sensitive. Yeah?
Ken: Yeah.
Dr. Hartung: It's exponential- Yeah ... growth. It is not at all worrying that we find more and more chemicals in the blood of people because we can find smaller and smaller amounts.
Ken: Yeah.
Dr. Hartung: Yeah.
Ken: Yeah.
Dr. Hartung: People have calculated that in, uh, about five to 10 years we should be able to find a single molecule in a sample. They're not producing diseases, yeah, because, uh, Paracelsus said already for 500 years, it's only the dose which makes the poison, yeah?
Ken: Right. Right.
Dr. Hartung: But they're extremely helpful because we can now say, "Okay, the chemicals are there, so do-" They might have been more at some point.
Uh, we also can measure in our blood what has changed. Yeah. Why is this molecule, which normally is at this level, so much elevated in this patient? Mm-hmm. The combination of these things, uh, lead us to understanding, okay, possible exposure here to these chemicals and a change in biology, uh, which is a perturbation, which might be indicative of developing a condition, a disease which you want to prevent.
So the idea of the exposome is can we make, uh, create such hypothesis of connections, chemicals and disease, and then we start verifying them, then we start building a catalog of these. By doing so, we do something very different to, uh, replacing one animal test by the other. We are changing the entire narrative to, so no longer the art of poisoning rats, but the, uh, exposure side of disease.
Building a full-blown understanding of how we are influencing the things which are happening in our body by the exposure to chemicals. And this sounded 20 years ago when Chris Wild, at the time leading the International Agency for Research on Cancer in Lyon-
Ken: Mm-hmm ...
Dr. Hartung: he said, "We need a human exposome to complement the human genome."
And everybody said, "Yeah, yeah, dream." All this stuff we are-- you are exposed to. But we are coming closer and closer that not really everything you have ever done for you, but relevant exposure for a group of people, for patients, whether healthy people, uh, can be analyzed and we can make sense of them, and that's what we call exposomics.
Ken: And exposomics has, for some years now, begun to shape how we think about the enormous task of trying to understand the exposures. The next step, of course, is to understand if the exposures are meaningful. But we also know that One person's exposure might cause no problem at all, whereas someone else's exposure might in fact, uh, trigger a, a response that's of, of concern.
That's where I see the marriage of exposomics with the new assessment technologies built on AI and stem cells. That's where they converge. Do I have that about right?
Dr. Hartung: You have it completely right. So the, the first thing is we are talking about enormous amounts of data, but AI is coming to our help, yeah?
Yeah. AI is, is the tool to make sense out of enormous data sets, yeah? But by saying the exposome is everything which is, uh, not the genome, you're giving the wrong impression that it is something parallel. Uh, it is really about understanding them in context.
Ken: Yes.
Dr. Hartung: We know so much now about what risk genes are driving something.
We know there is a gene-environment interaction, yeah? Yeah. So nobody is, is seriously proposing just look into the exposome and not take advantage of 40 years of genomic work.
Ken: Right.
Dr. Hartung: And we know there is, uh, this interplay. Uh, we call this the epigenome, so- Mm-hmm ... epigenetics. Uh, because- Yeah ... each and every cell of your body has the same genes Yeah.
Each and every one. Yeah. And they're very different. A neuron has not much in common with a skin cell or a blood cell, an immune cell, yeah? They can do so because they are using different parts of the genome. Different parts is, uh, are active, different parts are silenced. This is the epigenetic part of it.
And a lot of our memory of the past, yeah, what exposure did to us, is actually inscribed in epigenetics. Yeah? So the wrong genes are turned up, the wrong genes are turned down, and this makes you more sensitive for the next hit. Or protects you, yeah? But understanding these- Yeah ... yeah, is now critical. So it is really not a call for, yeah, now forget about the genome, now we study the exposome.
Right. Right. It is about bringing the two together and, and understanding also the interplay, the epigenome.
Ken: Can you give me an e- an example of where we've changed our mind about a family of chemicals or a specific chemical that we might have once thought was harmful and now we are less concerned about it?
Or the reverse example, or some- Mm ... something we, we thought maybe was benign, now we're more worried about it.
Dr. Hartung: Yeah. I mean, from my point of view, there's certain fields of particular interest. This is where the frequency of disease is changing. Autism is a, a prime example. Mm. Or also ADHD, so hyperactivity syndromes, yeah?
They are- Yeah ... clearly increasing, yeah? And this was in the beginning a matter of awareness and setting the diagnosis or not, but-
Ken: Yeah. Yeah ...
Dr. Hartung: over the last 20 years they have tripled. We also know that there is some genetic components to this, but none of them strong enough to explain it. But the field, and, and we also in our own research are, are pu- trying to pursue is, how can we understand that perhaps a sensitive embryo, yeah, is exposed- Mm-hmm
to a chemical or not? Because this brings the two together, so it's about susceptibility.
Ken: Mm-hmm.
Dr. Hartung: We managed, four years ago, to find the first pair of a risk gene and a chemical. These two synergize, and suddenly produce in our cell cultures an aggravation of, uh, of this change towards what you typically see in patients.
Mm-hmm. SHAT8 is the risk gene, and we combined it with a pesticide, chlorpyrifos- Ah. Mm ... which was, was interestingly around the time of our study banned by EPA because of the possible link to autism. And if you put them together, it's like nitro and glycerol and- Yeah ... things explode, yeah? So our cells react to the risk gene and to the pesticide in producing a change in neurotransmitters and things which we see in patients, but if you give them together, poof, yeah- Yeah
we have a much stronger reaction. And this was so convincing that the NIH funded now a study, uh, with $12 million, and we used these brain organoids to show is this a true association or not. And this is a new approach. No animal experiment in the world would allow us to do this.
Ken: So fascinating. It's gonna shake up regulatory science.
Some chemicals that we were worried about will s- maybe seem less worrisome, and the reverse will probably also be true, things that we thought were not to be concerned about. Now let's add the next complication, 'cause, uh, you're a complications guy. So let's go now to mixtures.
Dr. Hartung: Okay. Yeah.
Ken: The working assumption is that if it's, uh, set at a legal limit, if it's considered a de minimis exposure in the regulatory world, or if we don't have any information about it, there's no limit to the number of chemicals we can add to people.
Yeah. That's the, that's kind of the working assumption. That's what we're living with now. How are those new technologies useful or relevant, or how will they be useful and relevant to looking at chemical mixtures, which is really everyday life? We don't live in a one chemical exposure world at a time.
How do we take that on, Thomas?
Dr. Hartung: Yeah. No, I mean, you're completely right. This is the holy grail of, of toxicology, yeah? Yeah. To understand how things come together, yeah? What we're doing at the moment is the best we can do. Yeah. Which is we are, we are using extremely high doses of these, of these chemicals to say if they do nothing at these high doses, they will never do anything, yeah?
Yeah. That's more or less the assumption. Very often we are testing in mice and rats things which are many thousand times more, uh, than what any human would be exposed to. But then we get really strange results, yeah? For example- Yeah ... uh, Coca-Cola had to change their recipe after 140 years because, uh, one of the caramel dyes they are using It was found to produce cancer in rats.
And then the Delaney Clause in California says you must not add something known to be a carcinogen to food.
Ken: Yeah.
Dr. Hartung: So they had to change. But the amount of Coke would translate to 8,500 Cokes per day you have to drink- Yeah ... to reach the concentrat- uh, reach the dose which produced this cancer in rats, yeah.
Ken: Right.
Dr. Hartung: I mean, this is ridiculous. We know it, uh, but, uh, we cannot do anything against a law which is very strict, yeah? So our goal would be really to come to something where we have a mechanistic understanding and can say, "Yeah, okay, 8,500 Cokes. Nobody is drinking this," yeah?
Ken: Yeah.
Dr. Hartung: But is it a mechanism which is relevant?
Or is it one which only shows i- in rats, yeah? For example, it is only in rats that, uh, kidney cancer develops with aspartame, one of the common sugar substitutes. And the reason is that at these very high concentrations, there's crystals forming in the kidney, and these crystals- Mm-hmm ... are producing inflammation.
Inflammation gets cancer up, yeah?
Ken: Mm-hmm.
Dr. Hartung: Has nothing to do, because at the levels we use, there is no crystals in the, in the kidney, yeah? Yeah.
Ken: And
Dr. Hartung: like this, we need to understand when does it make sense to translate and when does it not make sense. We can understand quite well if something is adding up, and this is quite often the case.
If you have, let's say, different endocrine disruptors and they're all hitting the estrogen receptor, you can simply say they are adding. More estrogen disruptors will produce a stronger effect. Uh, but very often we have situations where we don't know what these chemicals are doing, yeah? They might, on the contrary, uh, hit on the androgen receptor.
This could be counteracting the effect of- Yeah ... of something hitting the estrogen receptor, yeah? Or they could block each other because they're competing for the same receptor, and they're blocking the receptor- Yeah ... for the, for the real thing, yeah? So it is really important that we start producing data and that we allow AI to start understand when do these more than additive effects happen.
And the more of these examples we have, we can apply it to our knowledge, and we start knowing, okay, perhaps chemicals which are helping to transport other chemicals to go through a barrier, yeah, to go through the blood-brain barrier and hit the brain.
Ken: Yeah, yeah.
Dr. Hartung: Yeah. They could be important mixtures. You know, we call them cocktail effects, yeah?
Ken: Yeah.
Dr. Hartung: Yeah. Cocktail effects are describing that sometimes, uh, it is not just the amount of alcohol, it is also the combination of things which gets you tipsy or, or, or drunk, yeah? Yeah, yeah, yeah. So damage can come from the combination of things, yeah? We need to produce data and analyze them properly to understand when these cocktail effects come into play.
And this is what science starts to deliver. We have more and more technologies which have high throughput, robotized testing, so we can really test many, many combination of things. We can then ask why is suddenly here a so much stronger signal than we would expect, yeah?
Ken: Mm-hmm.
Dr. Hartung: But this would require a systematic analysis.
The data in the world we have are not enough on mixtures where we have data on the individual substances and the mixture at the same time.
Ken: Yeah, and these are not problems that could ever even wildly be solved with animal testing.
Dr. Hartung: That's the point, yeah. Because it's- Yeah ... you know, it is not just if you take only two chemicals, yeah, imagine how many combination of doses you have if you just think- Yes, yes
about the amounts, yeah? But it's not only about the amounts, it's also about sequence, yeah?
Ken: Yeah.
Dr. Hartung: What is the distance? Is it an hour after the other? It's a day? It's a week, yeah? There's so many impossibilities. Chemicals change us.
Ken: Yeah.
Dr. Hartung: There's this nice saying of what, what doesn't kill you makes you strong.
We actually see this sometimes in chemicals.
Ken: Yeah.
Dr. Hartung: They can produce something which we call resilience, that the second hit is, uh, much weaker than the first because, yeah, the cells, the animals or whatever, they are prepared for, for what's coming. They have started a defense system.
Ken: Mm-hmm.
Dr. Hartung: So we really need to understand when do they do this?
How do they do this? Is this only against the same chemical hitting or is it also something which is preparing you for a completely different hit? So there's a lot to study here, but I very much believe that we can, with all of these tools of characterizing how the system changes, mechanistic analysis, robotized testing, sensor technologies, all of these omics of transcriptomics and metabolomics and...
So we can really produce enormous sets of data and then have the AI work for us to identify.
Ken: Is the so-called windows of vulnerability or the timing when these mixtures come together at their different levels, so it, it, it might have a very different effect on a, an embryo or a developing child than it would have on Someone who's 150 years old like myself.
Dr. Hartung: Yeah, good for you. I just want to say that there must be some cosmetics you're using for 80 years at least. But you're right. This is the next layer of complication, yeah? There are certain things which only happen in an embryo. They are switched off in adults, yeah? Yeah. And so if a chemical is hitting exactly this, yeah, bad luck, yeah?
Ken: Yeah.
Dr. Hartung: You need to, uh, develop an understanding of vulnerable targets and when they appear, and which model system is reflecting them.
Ken: Yeah.
Dr. Hartung: And in a similar way, uh, yes, older people have less resilience, have less capacity. They're doing well, but a little infection blows them- Yeah ... away.
Ken: Yeah.
Dr. Hartung: And probably it's the same, they're doing well, but they are more vulnerable to certain chemical exposures than others are.
But current systems don't allow us to study this.
Ken: Yeah.
Dr. Hartung: All animal tests are done in young adults, yeah? We use the reasonably mature animals which are cheap to produce, so they're typically an equivalent of an adolescent, yeah? That's, that's what we are using. Yeah.
Ken: I know I'm not performing like an adolescent anymore.
I just went mountain biking, and, uh, I'm still feeling it, Thomas. I really appreciate your time, and I, I look forward to learning more about what you're doing, and I know the EWG scientists feel the same way, so thank you so much.
Dr. Hartung: Yeah, no, I really enjoyed always over the last 15 years working together.
Ken: Thank you so much for listening. If you'd like to learn more, be sure to check out our show notes for additional links to take a deeper dive into today's discussion. Make sure to follow our show on Instagram at KenCooksPodcast, and if you're interested in learning more about EWG, head on over to ewg.org or check out the EWG Instagram account at Environmental Working Group.
Now, if this episode resonated with you or you think someone you know would benefit from it, please send it along. The best way to make positive change is to start as a community with your community. Today's episode was produced by the extraordinary Beth Rowe and Mary Kelly. Our show's theme music is by Moby.
Thank you, Moby, and thanks again to all of you for listening.