‘Chemical soup’: the everyday exposures affecting our health – with Tracey Woodruff and Thomas Hartung

Jennifer Dixon:

There are rising concerns about the 350,000 synthetic chemicals in use today. Only a small fraction have been tested for their impact on human health, and there are few restrictions and no post-market surveillance for long-term adverse impacts. So what's going on and how do we reduce the risk of harm to health in future? 

Well, with me to discuss all this, I'm pleased to welcome two experts on the subject. Tracey Woodruff, who is Professor at University of Berkeley at San Francisco, and Thomas Hartung, who is Professor at Johns Hopkins University and expert in toxicology. Welcome both. 

Can you describe, both of you, the kind of chemical environment we now live in and also what do you think we should be most concerned about?

Tracey Woodruff:

We are living in a chemical soup. We've been documenting both in the United States as well as globally growth in over fifteen-fold since really the 1950s in chemical manufacturing and production. There was a paper that was published identifying about 350,000 chemicals that are registered across different countries. And many of them we don't actually have very good information because of confidential business information reasons, which is basically businesses don't have to tell us about their chemicals and other factors. And for most of the chemicals that are used in commerce, we actually have very little data about their exposures, where they're used and health effects. A lot of these different industrial chemicals are used in plastics or are part of plastics, are essential to plastics. And we know that plastic production doubled since 2008.

And the really big thing that's concerning for us as researchers is that the plastic production is projected to triple by 2060, and that means that we're going to be exposed to a lot more toxic chemicals, which we've already documented through many, many studies measuring chemicals, both in air, water, food, soil, and we measure them in people. And then we also see this concurrently increase in chronic health conditions, both in the general public but also in children. And children are supposed to be healthy, right? They're like at the beginning of their lives, but there's increasing rates of many different types of childhood chronic diseases, which is raising concern, even more concern about the role of these toxic chemicals and exposures in people's lives.

Thomas Hartung:

The number of chemicals is mind-blowing. And for all of these chemicals, not even 10% have any type of safety data, and not 1% is well tested. I mean, the good news is in these 150 years since we started [the] chemical industry, life expectancy has tripled and there's quite a few diseases which are going down. But there's some, and I'm particularly worried about the development of the human brain. We call this developmental neurotoxicity, something which is linked to things like autism and ADHD, and these are skyrocketing. In the '70s, one in 10,000 children was diagnosed with autism. So this is really something which cannot be explained by any genetic change. There must be lifestyle factors, there might be chemicals.

Jennifer Dixon:

Do we have any high concern chemicals?

Tracey Woodruff:

So for sure microplastics are a really important place to be focusing both, the research, but also policy efforts or intervention efforts because many of the chemicals that people have studied in individual groups are part of plastic gemish. So things like PFAS are used in plastics, so there's a perfluorinated chemicals, another group of chemicals that we know can increase the risk of adverse health effects are phthalates. There's a lot of different phthalates that are used in plastics. They're often used to make hard plastic, soft and flexible. And it's used in many different types of consumer product applications. So everything from medical IV bags to vinyl flooring to some paint. Another group of chemicals that's also well known to people in part of plastic are bisphenols. I mean, people probably know mostly the chemical Bisphenol A, which is one of the original poster childs of chemicals that's used in plastic and a chemical that has a variety of adverse health effects.

But I think what we find with these chemicals is that you have one chemical like BPA, which for example, state of California has identified it as a reproductive toxin. It can affect the ovarian quality. The chemical industry, what it does is it like, Okay, BPA is maybe not as popular. We see it's going down in exposures, but it's replaced with another chemical that is essentially its sibling. Like BPLs or Bisphenol F. So you have these clusters of chemicals, these families of chemicals that some of the people may know one sibling, but not all of them. So those are some of the big ones I would say, that we know about, but to emphasise what Thomas was saying is, a lot of them, we don't know if they're going to be big or not, but there is some requirement in Europe, let me just say that, but in the United States, which I'm most familiar with, is there's no requirement for these chemicals that are already on the market, for the chemical companies, to provide a complete data set on how they may be adversely influencing human health.

Jennifer Dixon:

And Thomas said that was a list of candidates there. Are there others that you would add?

Thomas Hartung:

I mean, the crazy thing is we don't know. It is not clear, and I think the culprit is the animal testing. If you want to know whether a substance is impairing the development of the brain, so possibly linking it to autism, the study you have to do uses more than a thousand animals, takes 2 years, and you pay about $1.4 million. And we know it does not find a lot. It finds the very strong things, which we know and have been anyway, but it has not really helped us to identify anything new since we have this test for more than 20 years now. And so I think it is really a strong demand for something new. This work on rats and mice, which we have been doing for quite exactly a hundred years now, has given us as much headache as it has given us some solutions.

Jennifer Dixon:

Can we go on then to the evidence to date of the impact of chemicals on health? Tracey, I know you focus a lot on children, but also adults. One of the key trends here, we had a podcast back in November about the rising rates of cancer in the under-fifties. We had someone from Dana-Farber Cancer Institute saying she had a 12-year-old in clinic who had bowel cancer, which was unheard of 20 years ago. And the rising incidents in successive cohorts. Tracey, I think this is your backyard, isn't it? Can you just give some contours for us on this?

Tracey Woodruff:

I think what is most striking globally is that we have seen a shift in the global burden of disease, which used to be in the 1990s, mostly from communicable diseases. Diseases that can be transmitted from person to person, so malaria or AIDS or tuberculosis. What we've seen is that global burden of disease has declined, but it replaced with basically what's called non-communicable diseases. Diseases that are caused by these external factors that we're talking about, like clinical exposure, so things you said, cancer is definitely going up, metabolic disorder. So in the United States, we're seeing explosive growth in diabetes even in younger children, and excessive weight, that's also been increasing. You know, every time we turn around, a clinician at UCSF is saying, ‘Wow, I really see a lot more,’ for example, ‘non-alcoholic fatty liver disease.’

And not just in the adult population, but in children and in kids. And that's basically a disease of the liver that can lead to cirrhosis and liver cancer, a liver condition that you get when you drink, but we're talking about people who are not drinking or kids, so we're seeing that. Gestational diabetes, so that's a diabetes during pregnancy is also going up. And the shift now is that the largest global burden of disease is from these different types of health conditions, and they've increased at a rate greater than the communicable diseases have gone down. So we're really looking at these exogenous, external factors that are driving this, and we as a medical health, public health community have not really shifted our attention to better addressing the burden of both toxic chemicals and then their concurrent use in many different products.

Jennifer Dixon:

And lots of those diseases you mentioned actually are vaguely metabolic, aren't they, including the rising cancers? The discussion we had there was not a lot, but not all of the rising cancer was in something to do with the gastrointestinal tract related liver, bowel, small bowel stomach, some lung cancer of course. But Thomas, you also mentioned some mental health issues, the rise in autism. Is there anything you would add there about the kinds of conditions that you're worried about?

Thomas Hartung:

I think we have to be careful to make a differentiated picture here because first of all, I have to say cancer for example, is going down, in total. There's a few types of cancer which are going up, and these are things which are worrying me. And certainly the transition to more childhood diseases, childhood asthma is another good example, is certainly something which is worrying. I mean, if we talk about obesity, I think the culprit is portion sizes, very often. It is at least not clear evidence that chemicals are making us gain weight. I'm with you that there's lifestyle things we have to address, but lifestyle encompasses not only chemical exposure, it is really what we eat, how much we are physically active. It also has a very strong psychological component. Wellbeing is much more than chemical exposure.

Jennifer Dixon:

Although some of the chemicals in the food are thought to be addictive, aren't they?

Thomas Hartung:

We have 4,500 chemicals which are added to food. Please, more than 80% don't have public data adequate to describe their risk. Europe, on the other hand, is dealing with just 450, one-tenth of it, and they have been risk assessed twice in the last 20 years quite extensively. There's really completely different approach to chemicals and food on both sides of the Atlantic.

Tracey Woodruff:

Yeah, I mean, I completely agree that Europe has a more basic standard of data call-in for chemicals that are on the market, and that the US could definitely use that. But I would say that there have been high-quality, gold-standard reviews on some of these chemicals we've been talking about, like phthalates and BPA. One of them done by the Environmental Agency for Food Safety, finding that these chemicals can increase metabolic disorders. Which can be linked to obesity. I agree that there's many factors going on in the increasing weight of at least of people in the United States, but for sure to have a lot of these chemicals in food. The state of California has set up a new commission on ultra-processed food, which is both addictive, but also the list of chemicals used in ultra-processed food is enormous, and many as you know are a health problem. And yet we actually have done nothing to really address that component of how that can be contributing to adverse health outcomes.

Jennifer Dixon:

So let's then move on to how we get better at the evidence. Now, Thomas, you just mentioned the studies are costly and also both you and Tracey in your New England Journal piece mentioned that historically we test one chemical at a time and we're not so used to testing multiple small exposures accumulating over time. Thomas, could you just explain how toxicology can be modernised to help?

Thomas Hartung:

I mean, the crazy thing is I'm now past 60 and the animal tests we use in toxicology have been introduced when I was not yet born. So this explains quite nicely the backlog of adapting to new methods. But you have to see a good cell culture we have starting in the late '60s, '70s, it was still something you would do in your small lab. Real infrastructure for good cell culture, which allows us to work in a standardised way, emerged in the '80s, '90s. Computers have not been good enough to do really high-quality predictions of toxicity until the turn of the century. And these things have now turned out to be disrupted technologies. They are really changing how we can do things.

And we can do them faster, we can do them more human relevant. And this is something which is meeting much more the pace of development for a drug, a pesticide of any product that is going to the market. If you want, for example, to assess whether something is producing cancer, as you discussed cancer a little bit earlier, you get the animal results in 4 to 5 years, and for a million dollar per chemical. So if you press a button on a computer, you can get them in minutes, even if the quality was not as good or there was a difference, which I would challenge it is giving you the opportunity to really check large parts of the chemical universe.

Jennifer Dixon:

So this sort of smoke signal idea, using data linked AI and so on, we held a seminar here actually stimulated by one of your papers, Thomas, on this to say, how could we get better over here? And I got a lot of scientists together and AI people as well as data people, and the main message was the data aren't good enough for us to spot a needle in a haystack or suspicious candidates. And I left feeling quite deflated after this. Do you think that the data are good enough, certainly in the US now to be able to make some progress here?

Thomas Hartung:

The surprising thing is that I'm talking about the animal data and there's a lot of issues on reproducibility and how few we have, but the incredible thing is if you start digging, you find tremendous treasure troves of data altogether. We have just built a database which has data on 117 million chemicals, and we have 254 million tests run on these 117 million chemicals. And we are trying to use AI, which has only become good enough over the last handful of years to really make predictions and we are now approaching 90% predictivity. This is really mind-blowing. 

Jennifer Dixon:

What kind of health data are you putting into these models? That was the scientists here saying, the data, even the electronic patient record or the survey data we have, the new big cohort studies we have over here, even they just aren't good enough at person level. So what makes you so hopeful, Thomas, and what's the health data you're adding in here?

Thomas Hartung:

Really the game changer is artificial intelligence. And I was myself until some 10 years ago, always preaching trash in, trash out. From bad data you don't get good answers. And the astonishing thing is that if the data are only big enough, the AI finds its way: it gives weight to the things which are valuable, often finding patterns we humans can't see. The problem is at the moment that we don't trust this data, and it is my point of view at the moment, more about building trust to these technologies and giving them a rightful place in our strategies, and not to take decisions on our behalf, but to find the data so that we can understand what is known about a given substance and its similars, and then to try to integrate so diverse things like cell culture work, computer predictions, animal work, human studies, case studies, epidemiological findings. We are on the brink of doing exactly this, what we call a [inaudible 00:16:11] project.

Tracey Woodruff:

I just want to say, first of all, that one animal studies have been done since, for many, many years, and they've been very useful for identifying carcinogens. And that has led us to a lot of really important findings. I think some of the challenges with animal studies, which I think as you're alluding to in the BPA study, in the BPA analysis is, and what we've found through work that we've done adapting systematic reviews is that there's variability. And if you have poor methods that can lead to inadequate not good findings from these animal studies. So what we find when we look at higher quality animal studies, they can be predictive.

That being said, I think these invitro methods, and some of them we've been using here with different types of cell cultures like yeast assays or C. elegans worms, fruit flies, placental cells, are very good at identifying many different types of reproductive and developmental adverse health effects. And I think one of the challenges that we see in the field is that the tools to interpret these data's are not well established, and they've been misused by regulatory agencies to dismiss sometimes human evidence, for example, around neurodevelopmental effects of things like pesticides, like chlorpyrifos. So I feel like we can quickly identify chemicals that are of concern, rule them in so that actions can be taken, but those that we don't know, we will still have to use animal testing to verify that are correctly identified.

So I agree that we have these tools to move forward to do a better job of... I mean, we have been doing this ourselves and published on this, identifying these many, many, many chemicals which regulatory agencies in the United States are not addressing at all, part of this, working with these regulatory agencies on how can they appropriately use this to identify health effects in humans, and then take the actions that they need to take. And the reality is there's just going to be a lot of chemicals on the market that are causing health effects. And the chemical industry is a really big actor in this space and they fight the listing and regulatory activities to get rid of these chemicals from the marketplace. And that's where there's a big sticking point.

Jennifer Dixon:

Can I just ask a specific question on data linkage? Have you been attaching electronic patient records, healthcare records or other public health data? But I suppose what I'm trying to get at is the epidemiological approach, where you're looking at trends in diseases across the population that... Or even you can't really see them, but they could be there hidden by noise. Do you think, both of you, that AI and data linkage could help to find the needle in the haystack?

Thomas Hartung:

I think it's not traditional epidemiology, which is helping us, let's take cancer again. Typically, the exposure which started your cancer is 10 years to 15 years before you're diagnosed, and nobody knows you've been exposed to it. The new kid on the block is what we call exposomics, which is trying to use modern technologies to see what is changing in the blood, in the urine of patients and forming what we call an exposure hypothesis because you cannot only find some of the chemicals these people are exposed to, but also the imprint which was made by them. All the small molecules which have changed the gene expression, which has changed... And this new type of science comes to a hypothesis, and then you can follow it up in a targeted way in experimental systems as well in larger populations. 

I would like to give you an example what we have done in our own group. We have identified in the human brain organoids. These are small clumps of cells from stem cells, which are representing a bit of human brain, but half a millimetre large. And we use these systems to show that a certain gene and a certain chemical chlorpyrifos, a pesticide, which was already mentioned, synergise in producing something which looks very much like autism. And the NIH has now funded our autism excellence centre in Hopkins, and we are monitoring with 18 partners, 175,000 children for this type of interaction of the chemical and the gene. And whenever they come up with a candidate, we are using these brain organoids to verify this. And this shows your mechanistic approach, how you come to a hypothesis and then you follow it up in larger populations.

Tracey Woodruff:

Where we are moving and I think the field needs to move looking at these key mechanistic events. So like for example, the international agency on research on carcinogens uses key characteristics, mechanisms that are based on already known carcinogens. What are the mechanisms that lead to those cancer risks like inflammation effects on the immune system? But we can use those to do the evaluation of the exposures and do they change these different mechanisms and use that to identify cancer risk, rather than waiting until someone gets cancer. That is very old school. And we actually used this approach when we did a systematic review on microplastics because it hasn't... How this is evaluated and used in decision making is very at the nascent field, but we were able to use this to identify that there are these increases in inflammatory responses from exposure to microplastic in the colon gut as well as in the lung.

And then based on the already identified key characteristics, we could say, well, this is likely to increase risk of cancer. Let's say the state of California has been also using this, they declared one of the PFAS chemicals as a carcinogen based on using key characteristics and they listed bisphenol S, which is BPA's sibling as a reproductive toxicant in part based on key characteristics of reproductive toxicity. So I think you're seeing these more forward-thinking government entities starting to look to this as a better way. We should really be using this and not waiting for animals to get sick as long as we can agree on how to interpret it.

Jennifer Dixon:

And in your New England Journal piece, you do stress very importantly the regulator and legal environment, but also the need to really change our approach. So it's more precautionary than it has been, at least in the United States if the regulators are not so active or can't be. What about the role of the legal environment with class actions and so on, to then force change legally? Well, I think the example you used was Bayer was fined in 2023, I think it was, for weed killer chemicals, 1.6 billion.

Thomas Hartung:

And this was glyphosate, which was developed by Monsanto, but Bayer did unfortunately for them buy this company and inherited a tremendous problem. Glyphosate is not the worst pesticide, but it has to come into focus very much in possible association with cancer. I'm not fond of the American system, of these class action acts and billions sometimes paid for in these cases, bringing down entire companies. I have an approach where I think, no company wants to poison their customers. And the problem is they want clear evidence before they make changes, which also cost hundreds of millions of dollars to exchange one substance against another, often more than this. And so we need better tools, and if we have better tools we can talk with each other. The regulated communities can talk on facts, but they cannot talk on things which are not scientifically sound enough. And this is where the academic science has to deliver. And I really think that saying goodbye to many of these old-fashioned and outdated animal tests is a very important step to open up for better tools and better evidence to talk to each other.

Tracey Woodruff:

I think it is actually shown that the industry wants to protect their product and it has been documented that they will lie about the toxicity of their product in order to protect their market share. So that's where I think it's really important. I agree that trial lawyer situations are complicated and it's probably not the optimal mechanism for enforcement in the United States, but it has been a really important tool when the regulatory system has failed. And that is where we are, and I think that we can see that happening in real time. The rules that EPA put in place to try and protect people from very toxic chemicals like Trichloroethylene and perchloroethylene and methylene chloride are now being pulled back by this new administration and is being run in an office by somebody who came from the American Chemistry Council, which is basically a consortium of chemical companies. So I think you have to have some checks and balances.

Jennifer Dixon:

Thomas, can you say a bit more about the situation in Europe then just to make us feel a bit better here?

Thomas Hartung:

I have a chair at Johns Hopkins University, obviously in the US, which is my main employer. But I still have a professorship in Konstanz, Germany and many, many activities. So I really compare these two systems. I was very much involved in creating the chemical legislation Reach when I was working for the European Commission. And I have to say, the regulation, the pacemaker is now Europe. There's much more regulation being done, and also the scientific experience I've made in the US is second to none. If there's one thing in the US which you don't have to make it great again, it is science. And astonishingly, science is under attack at the moment. It is really bizarre, as let's say, I still consider myself a bit of a bystander, a visitor. I cannot grasp it. Why could you possibly try to bring down the scientific machinery exactly with the progeny of this? The people who have benefited from AI and technology and all of these things are at the forefront of now attacking science.

Tracey Woodruff:

The other thing about AI is, AI needs data and science produces data. I mean, there's other countries and we haven't talked about it, we're talking about Europe and the United States, but China has been doing a lot of really important science investments. They've done a lot of work on microplastics, for example, a lot. When we did our microplastic systematic review on the health effects, most of the studies are coming out China and they're good. You can see where when countries invest in their science, they get better information about being able to address health harms.

Jennifer Dixon:

Yes. And just going back to this EU Reach legislation, can you just for listeners who don't know anything about it, and actually the UK still has its own version of Reach. Thomas, can you just say a little bit more about the main contours of it? I mean, how much weight does it put on companies and how much weight does it put on the regulators to police this?

Thomas Hartung:

Both the US and Europe did essentially the same. In the late '70s they said we have to somehow control chemicals properly the Toxic Substance Control Act in the US and the dangerous substance directive in Europe, they took the same approach. They said, let's start new. We forget about we grandfather, all the things which are on the market, but from now on, do it well. We can argue how well we did, but in around the turn of the century, we understood more and more that the old chemicals were 99% of the volume of chemicals produced and 97% of the trade of chemicals. So it was not the new stuff which was dominating. And so Europe was faster.

In 2007, the [inaudible 00:28:19] legislation said, okay, everything at least, which we have more than a tonne on the market per year needs to get some safety assessment. The US came about 10 years later with the Lautenberg Chemical Safety Act, which was remarkably a bipartisan legislation which was signed into law by Obama. It is a Reach light, it's a very, very light version of it where about a hundred substances are prioritised any given time and then being assessed, compared to 23,000 chemicals which have received safety assessments in Europe under the reach legislation.

Jennifer Dixon:

I've got two last quick questions for you both and they're going to put you on the spot, I suppose. If you were president or, let's say, the main politician in the country, you had charge over making policy in this area to reduce human harm from chemicals, what are the top things that you would do? Maybe one or two. And I think the second question is, if you're a punter listening to this podcast, is there anything that you would advise them to do to protect themselves in their daily lifestyle from the harmful effects of chemicals? Is there anything easy they can do?

Thomas Hartung:

The most important thing is we have to build trust again, that our regulations are protecting people. And this is not the case, at the moment people fear the health risks of chemicals more than they fear mortal diseases. So there's chemophobia and this chemophobia can only be solved when we have an investment into fair assessment of substances, which is comprehensive. So I really would like to encourage the larger structures which are not based on this piecemeal of an experiment published here, an experiment published there, but large data gatherings which feed into the new technologies of AI and others to make sense of these and prioritise what we have to work on first, and do this in an evidence-based factual way. Not as a war of faith, I think all chemicals are bad, or all animal tests are terrible, and this really, this conflict between different types of NGOs and different types of scientific approaches, is not helping us.

Jennifer Dixon:

And Tracey, on the policy side, you've got a lot of recommendations in your excellent New England Journal article, but what are the priorities there?

Tracey Woodruff:

We've actually published a set of priorities that we've collaborated with scientists and experts from the public health community on our website. So the first one is to remove financial conflicts of interests and uplift scientific integrity in the regulatory process. And that means government funding of research to identify and account for the fact that studies, and this has been shown empirically produced by those companies and corporations that have financial conflict of interest are going to be biassed, compared to those that are not. So that's a really important part to increase trust in the government, improve the scientific quality of the database and the quality of decision making. The second thing is to reorient the responsibility for providing the funds for the people who manufacture these chemicals should be providing funds and the data that on these chemicals that are in market already. And there would be a minimum data set about the health harms and they would have to also tell us where the chemicals are being used, and the basic information. And then we also have a set of recommendations about the way that the science is evaluated, including these new approach methods, that those methods need to be upgraded. They haven't changed very much since the 1970s. So the evidence evaluation of systematic review quality and then methods to identify the actual health harms for these non-communicable diseases that are not cancer. And I think that we will see, as we've already written, that chemicals are an important component of the growing burden of chronic disease.

Jennifer Dixon:

Is there anything you do in your personal life to try and avoid some of this exposures and what would you advise others?

Tracey Woodruff:

We'd work with our clinical partners. We put out a set of recommendations under the title, Toxic Matters, and there are evidence-based tips about how people can redo their exposures to toxic chemicals as well as uplifting their resiliency to them. And I mean, many of these things are very basic and they're free, but some things we recommend are like, don't microwave in plastic. Wash your hands before you eat. Eat more food prepared inside, that you prepare, lower on the food chain. These have all been shown to reduce your exposures to microplastics. And we have other tips on our website.

Jennifer Dixon:

And Thomas, what about you? Is there any steps that you take and would advise?

Thomas Hartung.

My wife, Lena Smirnova, is professor working with me. And the funny thing is, she is taking care of all of the chemical stuff, and I'm mostly worried about the infection hygiene parts in our life because I have also an infectious disease background. What we do, we are avoiding processed food because I think too many things have not been properly done. We simply cook ourselves from fresh ingredients. We also follow some of the guidance for, let's say, shopping sustainable stuff. But it is such a small part of what we understand and what we know about, and very often we are exchanging really something we know is bad against something which is less tested. So I think we are actually largely blind, in the dark here.

Tracey Woodruff:

I think it's important for people to realise that it's not their fault when we're talking about these exposures. I think it's frustrating. I know I'm frustrated by it. And that really it's important, your podcast and other people talk about the role of do we elect people to address this. People think the government's dealing with toxic chemicals, but they're not. And we need to continue to reinforce how important it is to address this systemically because there's many exposures that we don't know about, as Thomas said, or we can't control, and that's their job.

Thomas Hartung:

It is a herculean task to really deal with these thousands of chemicals coming to the market all the time. But there is hope. We see that there is technologies emerging which enable us to do things we have not been able to do even a few years ago. And we now embrace these because they give us the potential to get some of this under control.

Jennifer Dixon:

Yes, and we'll be watching what you do there, Thomas, with great interest. So we must leave it there. Many thanks to Tracey Woodruff and Thomas Hartung for an absolutely fascinating discussion. Please join us next month, we'll be looking at where the health mission work has got to. And meantime, many thanks for listening and thanks to Sean and Leo at the Health Foundation and Paddy and team at Malt Productions for their help. And it's goodbye from me, Jennifer Dixon.