Air Quality Matters
Air Quality Matters inside our buildings and out.
This Podcast is about Indoor Air Quality, Outdoor Air Quality, Ventilation, and Health in our homes, workplaces, and education settings.
And we already have many of the tools we need to make a difference.
The conversations we have and how we share this knowledge is the key to our success.
We speak with the leaders at the heart of this sector about them and their work, innovation and where this is all going.
Air quality is the single most significant environmental risk we face to our health and wellbeing, and its impacts on us, our friends, our families, and society are profound.
From housing to the workplace, education to healthcare, the quality of the air we breathe matters.
Air Quality Matters
Air Quality Matters
#34 - Corinne Mandin: The Science Behind Indoor Air Quality - Epidemiology, Public Health, and the Exposome
A conversation with Corrine Mandine
I couldn't think of a better person to talk to about epidemiology and its place in the air quality discussion.
We discussed this fascinating area of science, and if you have been following the podcast, you know we have had several conversations about the potential of an air quality observatory in the UK, and Corrine has direct experience with just this.
She has been working on human exposure to chemical substances and physical agents and the related health effects, first at INERIS (French national institute for industrial environment and risks) for 8 years, and then at CSTB (French scientific and technical center for building) for 13 years.
At CSTB, she coordinated the French Indoor Air Quality Observatory, a public research program created in 2001 to carry out nationwide surveys on air quality in buildings.
In 2022, she joined the French Institute for Radiation Protection and Nuclear Safety (IRSN), where she leads the radiation epidemiology group. This group conducts research on human health effects of occupational, medical or environmental exposure to low doses of ionizing radiation through large epidemiological studies.
She has been involved in various European and international projects and expert committees, including at the World Health Organization and the European Joint Research Center. She is currently chairing the expert committee dedicated to outdoor and indoor air quality at the French Agency for
Food, Environmental and Occupational Health and Safety (Anses).
She was president of the International Society for Indoor Air Quality and Climate (ISIAQ) from 2020 to 2022. In 2022, she coedited the Handbook of Indoor Air Quality (Springer).
Corrine Mandin - LinkedIn
IRSN
Indoor Air Quality Observatory
Check out the Air Quality Matters website for more information, updates and more.
This Podcast is brought to you in partnership with.
21 Degrees
Lindab
Aico
Ultra Protect
InBiot
All great companies that share the podcast's passion for better air quality in the built environment. Supporting them helps support the show.
Air Quality Matters inside our buildings and out, and we already have many of the tools we need to make a difference. The conversations we have and how we share this knowledge is key. I'm Simon Jones, and this is episode 34 of the Air Quality Matters podcast, coming up a conversation with Corrine Mandin. She's been working on the human exposure to chemical substances and physical agents and its related health effects for years For eight years at the French National Institute for Industrial Environment and Risk and then for 13 years at the CSTB, the French Scientific and Technical Centre for Buildings. At the CSTB, she coordinated the French Indoor Air Quality Observatory, a public research programme created in 2001 to carry out national surveys on air quality in buildings. And in 2022, she joined the French Institute for Radiation Protection and Nuclear Safety, where she leads the Radiation Epidemiology Group. This group conducts research on the human health effects of occupational, medical and environmental exposure to low doses of ionising radiation through large epidemiological studies. She's been involved in various European and international projects and expert committees, including at the WHO and the European Joint Research Centre. She is currently chairing an expert committee dedicated to outdoor and indoor air quality at the French Agency for Food, environment and Occupational Health and Safety.
Speaker 1:I couldn't think of a better person to talk to about epidemiology and its place in the air quality discussion. We discussed this fascinating area of science and if you've been following the podcast, you know we've had several conversations about the potential of an air quality observatory in the UK, and Corinne has direct experience of just this In her current role. She is also involved in a major study on radon, another subject of interest for me personally. So I couldn't wait for this chat and I think you'll enjoy it too. Thanks, as always, for listening. Do check out the sponsors in the show notes and at airqualitymattersnet.
Speaker 1:This is a conversation with Corinne Mandine. So we'll start with the epidemiology question, because I think we use that term a lot, particularly on this podcast, because of the kind of guests we have on. But also it's framing increasingly the discussion around air quality as we make increasing links between air quality and health outcomes. But I'm not sure I've ever really discussed what epidemiology really is and what it is to the built environment and particularly air quality, and I can't think of a better person to ask than you, corinne. So perhaps, maybe just give us the 101. What is epidemiology and how does it fit within this part of our environment the air quality piece.
Speaker 2:It's a study of the disease in a geographical area and over time. So descriptive evaluation of the diseases, human diseases, and, moreover, epidemiology is also the study of the human disease determinants, trying to understand what are the causes of the disease that appear in humans. Is it genetic causes or is it environmental causes? So epidemiology is a science of understanding the diseases. That's why it's very important for environmental health, because it's understanding the determinants of the diseases. Is it occupational exposure? Is it domestic exposure? This is what epidemiology tries to understand. So it's one of the fundamental sciences to understand the effects on humans of chemicals or physical agents or biological agents. There are two main disciplines. There are toxicology, experimental toxicology. In a lab you control everything. You expose your chemicals to cells, sell to sell. You expose cells to chemicals or you expose animals mice, for example, rats to chemicals and then you see the type of effects that appear in on these cells or on these animals. So with toxicology toxicology you control everything. That's why it's, in a way, it's better. You know exactly the quantity of the chemical you expose your animal to, but it's not representative of humans. This is not life. Experimental animals are not humans. So that's why epidemiology is coming. This is the second main science to understand the effects of chemicals or other risk factors.
Speaker 2:In epidemiology you study the humans themselves, so you are close to the real life. You really, yeah, you are not simulating anything. You observe the reality. But the counterpart is that it's very difficult to control everything. It's very difficult to understand all the chemicals to which people are exposed.
Speaker 2:To Imagine that to explain the disease appearing in the population now you need to know what the people were exposed to across the course of their life, since they are born and even before, because we know that in utero exposure may have an effect on else when you're a child or an adult. So epidemiology is real life, but it's very complex because you need to reconstruct, to rebuild the exposure and that's why it's not so easy to be used. It's challenging. Epidemiology was largely used regarding ambient air quality, outdoor air quality, because when you have outdoor air quality is monitored, there is a regulation is regulated, so you have a fixed monitoring station. So once you measure a concentration of PM2.5, for example, or nitrogen dioxide outdoor, you have the exposure of millions of people around, thousands of people around. But regarding indoor air quality, it was more challenging because every room, every dwelling is different, so doing epidemiological studies for indoor air quality it was more challenging because every room, every dwelling is different, so doing epidemiological studies for indoor air quality has been hard and is limited to date.
Speaker 1:And will it always be limited by that? Is that just a fundamental factor of epidemiology in general, of epidemiology in general? Is that, depending on what you're trying to understand, there will be more or less confounding factors, and do you reach a threshold, at some point, where it just becomes impossible to draw conclusions?
Speaker 2:it's just too complex a picture to understand yeah, yeah, sometimes when you are studying a rare disease, it's very hard because you need a large group of people to observe a rare disease. That's why there are different types of epidemiological studies. There are what we call the ecological studies. In ecological studies you are not looking at each person in the group of population. You're just having a group on a geographical area and you compare the diseases that are reported in this area and the exposure outdoor air quality, for example but you're not studying each person. So ecological studies are more easy to perform, to carry out, because you're not going to each person. You're just looking at geographic, geographic, regional area, for example. But it's limited because you don't control what each person is doing, if the person is smoking, for example, if it has genetic factors, for example. So ecological studies are limited. So we tend not to use ecological studies anymore. The two other types of studies are cohort studies. So you have a large group of people and you follow them over time. You describe their exposure over time and some people will become sick and some other will not declare the disease. So at the end, after 10 years of follow-up or sometimes more, 15 years of follow-up, you compare the exposure of those who have become sick and those who have not become sick. And you try to understand if those who became sick have had specific exposure, specific chemicals or specific biological agents compared to those people who did not become sick. So it's the cost that the limit is the cost because you need to follow a large group of person over a long period. So it's very expensive and sometimes people are disappearing. You lost them if they move, for example, or if they decided to stop the study.
Speaker 2:So the third type of epidemiological studies are the case control. In a case control study you recruit the people who are sick so that these studies are used for rare disease. So, for example, you go to hospitals and you recruit people having lung cancer and you pair each of these persons with what we call a control. A control is somebody who has the same age, approximately the same sex and the same socioeconomic status. And then you have this case, the case people with the disease and the control.
Speaker 2:Sometimes you have several controls, up to five controls per case, and then you compare their exposure. Do people with lung cancer at 40 years old, a man with lung cancer at 40 years old, what was his exposure to radon in the past? And you try to understand, based on either measurement or modeling, and you compare with the exposure of the person the same man, approximately same age, with no lung cancer and you try to understand if he was exposed to radon. This is how we discovered that radon was causing lung cancer because, observing a group of people with lung cancer, it was observed that they were more exposed to radon over their life compared to people without lung cancer, taking into account their tobacco status, of course, because we know that tobacco is also causing lung cancer.
Speaker 2:So the case-control studies are cheaper or easier to perform. Once you know if you have a disease, a specific disease to study, the challenge in epidemiological studies is to characterize the exposure. Ideally, the best is to measure personal exposure, having people wearing sensors or doing urine measurement or blood measurement. But we know that first, this is expensive and second, this is very intrusive. It's not easy to have a blood sample of thousands of people. So if you cannot measure personal exposure, you measure exposure in the ambient air. This is what is done for outdoor air quality, for indoor air quality. Also radon the case control studies on radon people were not wearing dosimeters. There was a dosimeter in their house and this is how we measure indoor concentration. And this indoor concentration and this indoor concentration is the exposure concentration. But even measuring in the ambient air indoor air is is expensive. Sometimes it's difficult when you need to have a pump to sample air. It's not, it's not easy. You need to send the sample to the lab, so it's it's expensive.
Speaker 2:So the third option is having questionnaire. You ask people, for example, um if in in epidemiological studies of children having leukemia. You ask to the parents if they were exposed to cleaning products when they were, when they were when they were younger, or if they were exposed to pesticides. Is there a? Is there a field with a pesticide outdoor? Based on questionnaire? But the limitation I would say is that there is the bias of memory. You don't always remember what you hit last year, what you did two years ago, where you traveled a few years ago. It's hard to remember and people who are sick tend to remember differently that if you're not sick, you forget. If you are sick, you try to imagine that maybe these things you eat or these products you used in the past could be the cause of your disease. So there is limitation with questionnaire also. So you see, it's very challenging to perform epidemiological studies, but this is why it's interesting to perform epidemiological studies.
Speaker 1:But uh, but this is why it's interesting. Yeah, because out of those three kind of approaches the ecological, the, the, the cohort and the case study there's a thousand questions around each of them that immediately spring to mind because it's so fraught with difficulty. Is it fair to say, then, that epidemiology is a relatively new science, or does it go back in some form, quite a way? And if it's new, do we understand enough about it to understand what is a good study and what is a poor study? Now do you know what I mean? Because I imagine it's the quality of epidemiological studies must be really important, and because there's because it's so investigative, there's so many confounding factors. Potentially, a poor study could be really poor and a good study could be very good, or you could just be lucky, you know.
Speaker 2:Yeah, sometimes when you have a large group, you make many statistical tests and some tests are due to chance. It's by chance that you observe some relationship. That's why there are some indicators in epidemiological studies. There is what we call the p-value. The p-value is a sort of indicator that says you if it's just by chance that you observe something or if it cannot be due only to chance. So there are some indicators and this p-value. There is a limit, generally 5%, and if you have multiple tests, this value is evolving to take into account that if you have multiple tests, this value is evolving to take into account that if you do multiple tests, you may just observe an association by chance. So this is taken into account.
Speaker 2:And coming back to your question about if it's new or not, it is not new, basically because it was discovered by John Snow in the UK in the 50s. He observed that people having cholera were people close to specific wells. That's why he discovered that there was an association between the distance to a well contaminated and the diseases. And this is the first studies that were performed in the last century. But it is not new. But it's improving with new techniques to characterize human exposure, from questionnaire to measurement in the environment to now personal measurement, and what is developing more and more is human biomonitoring. The European Commission is promoting human biomonitoring measuring in the urine or the blood. You, you directly have people.
Speaker 1:exposure, provided the the contaminants can be, can be, can be measured in urine or blood and I suppose the larger the data set, the more chance you've got to finding something you couldn't imagine, because I imagine the there must be. A part of this is stuff comes out of these studies that you weren't expecting and there you have to really check that the difference between causation and correlation and things like that. But I'm guessing the bigger the study, the more chances are you find a pattern or something that tells a story actually, the bigger it is, the more variables you can take, take into account.
Speaker 2:You can. Because if, if you have a small group of people, you take into account those who smoke, those who live close to a road or not close to a road, then you make some categories and the more categories you have, the smaller is the number of people in each category. That's, and you cannot do anything regarding statistics if you have small groups in different categories. So it needs to be large to have enough group of people, enough smokers and not smokers, enough different ages. So that's why you need a large group of people.
Speaker 2:But you raise an important point between correlation and causality. Epidemiological studies generally allow us to identify some association or correlation between a risk factor and a disease, but it does not mean there is a causality. There can be a third variable that was not studied and that is connected both to the exposure and both to the disease. That's why we still need to date some experimental studies in the lab to understand, on cells or on tissue or on animals, the mechanism and once the mechanism is plausible between the risk factor and the disease, the causality is confirmed. But generally we observe association, not causality.
Speaker 2:So, yeah, this is something that people generally don't know and sometimes some epidemiological studies may conclude things that are not causal. That's why we need also to understand the biological mechanism to be sure that the risk factor is explaining a disease is not causal. That's why we need we need also to understand the mechanic mechanic, the biological mechanism, to to be sure that the risk factor is explaining a disease and is that why you hear the term epidemiological and toxicological studies talked about together?
Speaker 1:so much is that they go hand in hand often that either you understand something toxicologically and you try to find that pattern through epidemiological studies, or the other way around you see a pattern or something in an epidemiological study and you you ask the question is there a cause there that we found? And you would use the toxicological studies then to figure out is that plausible? Is that why you hear the two mentioned together so often?
Speaker 2:ideally. Ideally both needs to discuss together to yeah, to yeah to understand the mechanism and or to check the mechanism that that can be plausible. It's not easy, you know, everybody is in its own discipline and sometimes not easy to break the silos. But but when the public agencies then make a study on a given chemicals or biological agent, they compile all the data from biological data, from experimental studies and epidemiological data. Anyway, at the end, epa or a group in the european commission both data are useful to do risk assessment.
Speaker 2:For example, when coming back to indoor air quality, today there are indoor air quality guidelines that are values that are used to interpret a result. When you do a measurement, if you want to know that this concentration is there a risk for the people breathing this indoor air, you compare the measured concentration with a guideline indoor air quality guidelines. So WHO, the World Health Organization, has proposed some guidelines for chemicals. These guidelines are based on health data either from toxicological studies or from epidemiological studies studies or from epidemiological studies. So this is the example where both disciplines, the results of both studies, are used to set these guidelines.
Speaker 1:Yeah.
Speaker 2:What I forgot also to mention a lot of we mentioned epidemiological studies in general population, but there are specific epidemiological studies on occupational group of workers and some information that we know on benzene, for example, or on solvents, are from epidemiological studies done in plants, in industrial plants, or done on workers, not only on general population, because workers are sometimes exposed to specific compounds. That's why it's easier to identify a correlation between a disease and a compound If they're all their career. They were exposed to asbestos or to benzene or to given pesticides. The information on health effects of pesticides are largely from epidemiological studies in workers, showing, for example, a relation between glyphosate and health effects in workers.
Speaker 1:So sometimes this data observed in workers can be extrapolated to general population how do you, as an epidemiologist, map out the idea of an epidemiological study?
Speaker 1:Because I imagine, if you've got a good imagination and an understanding of a particular discipline, you're starting with a blank piece of paper and you go I think this might a risk, or I've seen a toxicological study that says there might be this impact. And then you have to sit down and think right, okay, how do I frame this? Not on just the three different types of epidemiological studies, but, like you say, is it on a particular cohort within a particular industry? Or on a particular industry within a particular area or in a? You know, at some point you've got to try and understand how best to create what could be quite an expensive undertaking to get a result that's valuable. So I imagine, as an epidemiologist, you must have to spend a lot of time figuring out what it is you're trying to ask, what the question is and how to do that in a way to get something meaningful out the other side. I imagine the structure is incredibly important in that.
Speaker 2:Exactly. You mentioned the good word, the question. What is the initial question? Do we have a question on a disease? Do we have a question on the risk factor?
Speaker 2:For example, in a cohort study that is ongoing in French daycare center, the question at the beginning was that we observed respiratory disease in people who are cleaning cleaning workers, cleaning workers in hospitals, cleaning workers in hospitals, cleaning workers in offices they have more asthma. So the hypothesis was that these people have more asthma because of, among other reasons, because they are using cleaning products. Cleaning products have a lot of chemicals inside and disinfecting products may have biocides. So we observed this in workers. So we had the idea is it the same in children? Because we observe more and more asthma and respiratory disease and allergies among young people, among children. So the initial question was can we observe? Is there a link? In daycare center, where there are a lot of use of cleaning products, it must be clean. The parents want that the daycare center, where there are a lot of use of cleaning products, it must be clean. The parents want that the daycare center are very clean. So in this specific environment we have the same. Could also the cleaning products be associated to asthma in children. That's why we set a cohort study. We recruited by random selection under the daycare centre in Paris and around Paris and then we recruited the children. We asked the parents do you agree to? So? There are a lot of authorisations to obtain before leading a pedagogical study. So then we asked the parents do they want to participate? And once we had all the authorization, we did measurement of indoor air quality in this daycare center and in parallel we collected the age data with the parents. Not the children are too small to fill in the questionnaire. We had an initial question is asthma observed in children due to cleaning products, especially cleaning products used in the daycare center?
Speaker 2:And for another epidemiological studies we are carrying at RSN, my institute, there is in France, the Constance cohort study. Constance cohort studies is a group of 220,000 French people randomly selected. Ten years ago we agreed to participate in this cohort and so all these people are now. There is a follow-up of these people and these people are regularly asked to provide information through questionnaire. So we are working on radon and trying to understand the effect of radon.
Speaker 2:So we had the idea of okay, let's use this large cohort to ask all these people where they lived over the course of their life and we will assess the radon exposure all throughout their life, which is unique, which has never been done today, and we will see if disease that we observed in this group of people now could be associated with radon exposure in their childhood. So here the initial question was more an opportunity having this large group of people. Ok, let's add radon to this study and let's study, let's see, if radon could be associated with the health effect that these people are reporting and taking into account, of course, all the other exposure. That's why it's difficult. We need to take into account chemical exposure, biological exposure, physical exposure. So, yeah, at the beginning you need a question related to a disease or related to a disease or related to a risk factor.
Speaker 1:So a disease, asthma or dementia or neurological development or a risk factor a pesticide, radon or asbestos and like that and I imagine it must often be the case that epidemiological studies lead to further, epidemiological studies lead to further, epidemiological studies lead to further. You know that you find a pattern and you go oh I really we must look at this in this way, and that creates a new study and, like a lot of academia, you know further research and further investigation.
Speaker 2:This is often the conclusion. Further studies are needed. Sometimes we have new ideas and new hypotheses came when we are doing the study. So that's true. Sometimes larger studies can be performed. The epidemiological study I mentioned on the ACR Center was performed only in Paris metropolitan area. We could imagine having a larger cohort at the scale of France, for example, or a country.
Speaker 1:Yeah, I imagine data quality and data labeling must be really important in epidemiological studies, especially if you because I imagine the meta meta studies, the studies that look at the studies and try and find patterns and things like that that, if there's good rigor behind how information is collected, how it's sorted and managed and you know, air quality being a classic example of that that I imagine future generations could be using the data collected from current epidemiological studies and thinking of things that we haven't thought of, or combining data sets, a new data set with an old study. Is there quite a bit of that that goes on, that kind of joining up?
Speaker 2:Completely About the data quality. This is fundamental. If you don't have good data, you don't have good results. So this is fundamental and sometimes the founders often forget that we need data managers. We need people cleaning the data and checking that the data are reliable. If questionnaire, sometimes people answer with contradictory responses, so we need to clean everything, maybe to call back the people to check the answers. So this is always long. This is why epidemiological studies are sometimes long long because of sometimes the follow-up, but also long to clean the data, to check that they are reliable, to exclude sometimes some people in the sample. So data cleaning is fundamental.
Speaker 2:And, yes, now we observe that there is a large wave of data sharing, making data as openable as possible and as confidential as requested. So there is a trend to open sciences. So the idea is really to share the data and to be able, if people want, to reuse them as far as possible. Sometimes the limitation of epidemiological studies is that we are working on health of people. Health data on health data, on lifestyle factors are considered as sensible data regarding the GDPR general protection directive, the regulation on data protection. So there is also this issue to be addressed that this is confidential data. We need to be sure not to identify any people when we are sharing data and we need anonymized database, for example.
Speaker 2:But there are many, many epidemiological studies in the world that are available that can be used again in the future. For example, there was a European study on indoor air quality in schools and else stages of children called the Symphony Studies schools and else stages of children called the symphony studies. This database is now open to, not available online, but available upon request to the coordinator. If any research team wants to use the data, it's, it's possible yeah, I had this conversation recently.
Speaker 1:I actually it was at um, an indoor air quality observatory meeting, which was the first one. I didn't see you at um, which was a shame. But, um, I was talking to the guys that manage the born in bradford study, which is a big cohort study that takes in very sensitive data of people's health and their careers and criminal records and social welfare, and you know everything. But the point was made to me that the most sensitive databases that you can imagine are already being managed quite securely anyway. So actually it's mostly about structure and how you access that, and GDPR isn't as big a barrier as you'd think because we already collect a lot of incredibly hypersensitive data and there are already institutions that have very high levels of trust associated with the management of that data. So it can be done. It can be done, of course, yeah of. So it can be done, it can be done.
Speaker 2:Yeah, of course it can be done. It requests a bit of effort and energy to have all the authorization to do it and to share, but now for sure there is a solution now where you have the data on a secured cloud and you can access the data distantly and the data can remain totally confidential. It's possible of, but it's it's something that people who want to start an epi studies or to use data from epidemiological study have to keep in mind yeah, how closely linked is social science to epidemiology?
Speaker 1:epidemiology do you think I I my instinct is there's quite a bit of crossover, certainly on the behavior and habits and impacts that that might have an impact on epidemiological outcomes.
Speaker 2:I'd have thought yeah, yeah, you're right, we need to include more and more human sciences in our epidemiological studies. We need to, as we said earlier, and we need to to break the silos, to have more communication between art sciences and human sciences. Because when when we were doing when I was doing, um, some studies at the french indoor air quality observatories, we were working with people from human science to build a questionnaire, because sometimes the way we ask the question must be very neutral, not to induce the reply, not to make that the people understand in a wrong or correct way. So the questionnaire, if the epistudies is based on questionnaire, the questionnaire must be very well done to be understood, to be neutral, to be not intrusive. So this is important among other topics. Human sciences are important on this aspect.
Speaker 1:So you've started to describe. I mean I think you've described very well epidemiology. I mean I've learned loads out of that. You hear it's a name that's thrown around a lot, but I don't think we, unless you're in that field. Certainly it's mind-meltingly complicated.
Speaker 2:Uh, I guess it must be quite multi-disciplined in its nature epidemiological studies very difficult to undertake with a very narrow focus yeah, we need statisticians to manipulate all this big amount of data, but but also people, health professionals, to work with them when we are recruiting people in a hospital. Yes, this is very multidisciplinary analysis or other type of analysis in lab or analysis in urine or blood, having people developing methods to be able to measure new pollutants, because sometimes I was mentioning the epi studies in the IKEA center we know that some chemicals, the quaternary ammonium, are related to asthma. They are suspected to be related to asthma. Ammonium are related to asthma. They are suspected to be related to asthma.
Speaker 2:But today there are no analytical techniques to measure quaternary ammonium in settled dust. They are not volatile, so they are mainly not in air but in settled dust. That's why we had to collaborate with the lab to develop the techniques to be able to measure this compound. We need to involve many disciplines, absolutely From an epidemiological perspective, to measure this, this compound. So, yeah, we need to involve many, many disciplines absolutely from an epidemiological perspective.
Speaker 1:How have we started to see that science applied to the subject of this podcast, which is indoor air quality predominantly? You said it's difficult because it's we don't necessarily have the big data sets that we have with outdoor air quality monitoring, but it started, you know, I'm guessing at a cohort level and and so on. We're starting to see the results of indoor air quality epidemiological studies yeah, I was.
Speaker 2:I mentioned earlier radon. Right, there were many case control studies on radon in in europe, in european countries, and then the meta joint study of all the many case control studies on radon in European countries and then the meta-joint study of all the case control studies in Europe were merged together. This is a study from Sarah Darby who showed clearly the relationship between lung cancer and exposure to radon. Some similar studies on radon were performed in the US, in China, so radon has been largely studied in epidemiological studies, maybe because radon is easy and cheap to measure. It's a dosimeter that costs about 20 euros. It's passive you just put the dosimeter in the room, you don't touch it during two months, and so it's easy in a way.
Speaker 2:There were some epidemiological studies on mold, mold and asthma. So mold is difficult to measure in air or surfaces, but mold is visible. So not all mold are visible, but if as soon as you have visible mold or visible dampness, many studies epi studies performed in Scandinavian countries have shown clear relationship between mold and dampness in buildings and asthma, especially asthma in children, and WHO has written a report on mold and dampness in buildings. There were epidemiological studies on nitrogen dioxide and respiratory disease in children. Dioxide and respiratory disease in children. Nitrogen dioxide can also be measured rather easily with passing samplers, or nitrogen dioxide is also directly emitted by a gas cooker. So some epi studies based on questionnaire do you have a gas cooker in the kitchen? Yes, no, and I simplify, of course, but some studies, either based on measurement of NO2 or based on questionnaire if people had a gas cooker at home showed the relation with asthma or other respiratory disease in children. There were a little bit of studies for maldehyde showing a relation with asthma. Some studies show especially a study from Carl Gustav Bornhag, published in 2004, was a bit new. It showed the relation between phthalates. So phthalates are chemicals used in plastics. They are plasticizers to make the plastic be smooth and soft, and phthalates can be released little by little by these products containing phthalates. So we find there are a large number of chemicals that can be more or less volatile. Some are present in indoor air, some are present in settled dust and this study carried out in Sweden showed a relation between the EHP of phthalate in settled dust and asthma in children, and the study was repeated in Bulgaria also.
Speaker 2:What is important, what I can add, is also that one epidemiological study does not give the answer on the relation between the risk factor and the disease. We need more. We need a large number of epidemiological studies. We need what we call of epidemiological studies. We need a weight of what we call the weight of evidence. Having several studies showing the same result gives stronger power to the results. If you have only one study, it's not enough. Of course you need to be able to repeat and to observe the same relation. So, replying to your question, yes, there are some epidemiological studies on indoor air quality, but less than the epi studies performed on outdoor air quality.
Speaker 2:The good news is that with the new techniques of low-cost sensors arriving on the market, there are more and more. We will be able in a few years, maybe already now, to measure in many places because it's cheap, in many rooms. Also because it's cheap and on a very large time scale, because when you are measuring radon you need to measure during two months, so you have one concentration over these two months, so you don't have evolution over time. So for radon it's not a problem, but for other VOCs volatile organic compounds that may not be for which the concentration is not very stable depending on consumer products you use, cleaning products. It's good to have the temporal evolution over the time, over the season, and, and with the classical standardized methods, either passive sampler or active sampling during one day or two days, we don't have this temporal evolution, so we are not able to yeah, to study very in fine way the relation with diseases, respiratory disease, for example, and indoor air pollution.
Speaker 2:So with these low-cost sensors, I think, in my opinion, this is a great opportunity to have more epidemiological studies in more buildings for affordable prices. So, provided these sensors are, of course, reliable, they have no drift over time, they need to be able to characterize specific VOCs. To date, these sensors are measuring total volatile organic compounds, which is a controversial indicator, but it's a good way to. So this is a way of improvement for the epidemiological studies, this low-cost sensor. Another promising tool is the machine learning and all these sophisticated models that are more and more used. If we have more data based on the sensor, there are techniques to analyze this data, so this is also very powerful to you to analyze this data.
Speaker 1:So this is also very powerful. And so, yeah, I imagine I imagine low-cost sensors as much consternation as they give to a lot of scientists and academics because of accuracy and drift and what they're actually measuring and algorithms and interpretations that go in on the back of all that stuff. Um, for epidemiological science, because you often have scale in numbers or time, it lends itself much more to that type of technology where you you not saying you don't need the accuracy, but quantity can overcome some challenges you might have with accuracy sometimes, because you can remove outliers and 90th percentiles and all of that good stuff. You know all the statisticians can get their hands on it and do good stuff with it. Um, so I imagine low cost sensors really lend themselves well to large-scale cohort studies or or um, even the, even the ecological studies where you, where you put them into an area or put them into a particular cohort over a long period of time, much more valuable than the very precise um lab type studies that you might have in other parts of science absolutely.
Speaker 2:Sometimes in some studies we just, we, just, we just need to classify people if they are highly exposed or exposed, not exposed or in the middle. So we this is true in some sometimes we don't need the accurate value of the concentration of a given risk factor because we don't, we are not characterizing the place, the environment and we are not comparing to a guidelines or a mandatory value. We are, we just want to classify people and in groups of exposure. So this is true that sometimes relative value is sometimes enough, provided that all the sensors have the same behavior. But if, if they all have the same behavior of overestimating or underestimating, maybe for some epidemiological studies it's not a problem, provided we're able to identify the more exposed and the low exposed or not exposed Exactly.
Speaker 1:There's two questions I want to come back to which I mustn't forget. But first, I imagine, at the pace at which epidemiology moves at which I think is quite slow because of the size and scale often of the studies and the pace at which technology is moving at the moment, it must be incredibly frustrating sometimes that you finish a five-year study or something and look back and wait. Damn, I wish that sensor had been around five years ago when I was imagining this study. Or even even back 10 years ago some of the stuff you're doing in the indoor air quality observatory. How different a picture you might have been able to paint with the tools you have available now. And I imagine, sitting down now, knowing that, trying to imagine how to pull these studies together, god knows what could be available halfway through that study at the moment. The way AI is moving and the way sensor technology is moving at the moment, this is what is exciting.
Speaker 2:We could have regret, but we must see the future and not the past and imagine more exciting studies for example yeah.
Speaker 2:I was mentioning the low-cost sensors to measure the concentration in the environment, but smartphone apps can also be useful, for example, for the cohort studies we had in the ACA Centre in France. We were asking to the parents every month if the child has some diseases over the past month, with a notification on smartphone and two questions. So it's very easy. It's not a big questionnaire of 30 pages and it allows longitudinal follow-up for the studies. So the new technologies can also help to collect easy mass of data very easily, not only the concentration in the environment, but also the behavior you were mentioning and all else events, for example.
Speaker 1:Yeah, I think a lot of people would have been exposed to that for the first time during COVID with a lot of the apps that they were asked to feed information into. I mean the one we had in ireland. You were encouraged every day to log on and tell it whether you were feeling well today or not and if you weren't feeling well, what your symptoms were, and so on and so forth, and those kind of apps make it incredibly easy to give longitudinal feedback in real time of symptoms and feelings and activities or certain behaviors.
Speaker 2:It's it's phenomenal really exactly, and the smartphone in the smartphone itself allows to identify where are the people, are they in which city or in where are they located. So it's's also used for epidemiological studies, for looking at outdoor air quality, to track where the people are and to follow the way if they were walking or if they were in their car. It can also be used in that sense.
Speaker 1:Yeah.
Speaker 2:Another source of data that is new is the health claim, the records, the medical records.
Speaker 2:Medical records in many countries are used to reimburse people when they pay the doctor or they pay the drugs, but these data are recorded and can also be used to identify the diseases of people, and at RSN, my institute, we are starting using this massive data for our participants of cohorts, after having the authorization of the ethics committee and all the committees allowing us to do that. For the person of our cohort, we have access to his medical claims, so we are able to, instead of using a questionnaire to ask people do you have diabetes or do you have other type of diseases, we can look at the medical records and identify in the records over the past years if the people had asthma two years ago or diabetes or had a surgery for cataract or the eyes, for example, and this is very powerful. Also, it's complicated to treat, to manipulate. We need specific statisticians to do that, because it's a massive amount of data and it's a different database to compile and to cross, but it provides very useful information on the diseases of the people.
Speaker 1:And do you think large language models and AI are going to increasingly facilitate this type of large-scale data set, manipulation and characterization?
Speaker 2:Yes, also probably because sometimes it's based on text, so we need to use this text mining information to retrieve the information. So, for sure, this AI is very powerful to help future studies.
Speaker 1:One of the things I made a note which I meant to ask you was it seems to me that epidemiology is, because of its scale and its kind of population health level lens, it's one of the sciences that's most closely aligned to policy, I would have thought, because sometimes some science can be so niche and so specific. It seems such a very long way, niche and so specific. It seems such a very long way from outcomes, from political decisions or regulatory decisions or or something like that. But because of the nature of epidemiological studies you know the dahlies and direct correlations to cost, to health, or you know those kind of things that enable policymakers to make decisions it seems to me to be a science that's very close to the decision-making at a population level. And that must be quite exciting, that if you can present a very strong case or present a very strong idea, you're almost presenting it in a way that enables the people at the levers of power to make an impact to, to change a policy or regulation or a process to get a better outcome.
Speaker 2:That must be quite nice I would say yes and no yeah.
Speaker 2:The example of yes is, for example, the all the radiation protection regulation worldwide is based on the epidemiological studies that were performed after the bombing in Japan, the Hiroshima and Nagasaki bombings.
Speaker 2:The survivors were follow up over many years, they are still follow up, of what we call the lifespan study, what we call the LifePan study, and all the information from these epidemiological studies were used to set the frame of the regulation on ionizing radiation, directly from the epistudies to the regulation. But no, in a way for chemicals, for example, the European Chemical Agency's ECHA, for the regulation of chemicals, most of the data provided by the industrials are chemical, substance by substance, and most of the time they are based on experimental data, experimental studies on animals. So for chemicals the regulation for substances, pesticides or biocides most of the rational data used by regulators are experimental studies. For the outdoor air quality, the regulation on outdoor air quality and is based on there is there will be a new directive signed soon and based on the who guide proposal, who guidelines, and these guidelines are all mainly based also on epi studies, you're right. So it depends on the, it depends on the topic, but depends on the topic.
Speaker 1:Yeah, I mean, a lot of the low emission zones in cities have been driven through both large-scale environmental data and epidemiological data and actually there's some real rigor behind some of those decisions as to where to put low emission zones into cities. I mean, it's a really good example of an application of that kind of science, which is really exciting.
Speaker 2:It's hard at the same time to identify, sometimes in epi studies, to identify one risk factor. That's why this is what we were discussing at the beginning experimental studies in the lab, where you control everything, you have one chemical and you understand the effect of this chemical. Epidemiological studies. It's's good, it's the mixture on the real life, but at the same time sometimes it's hard to identify one specific pollutants associated to the health effect that you are studying just going to grab your attention for a minute.
Speaker 1:I wanted to quickly tell you about lindab, a partner of this podcast. For over 60 years, lindab has been dedicated to improving the climate of buildings. I have known them and some of the great people who work there for as long as I have been in this industry. Lindab offers a broad range of products, from individual components to complete indoor climate solutions. Their systems not only promote better indoor environments, but also deliver economic benefits.
Speaker 1:If you're working on a new building project, lindab's high air tightness products and demand controlled ventilation systems are designed to meet the stringent energy efficiency requirements of today and align with environmental certifications. If you're renovating, lindab's smart units can upgrade existing systems, reducing energy consumption by up to 70%, with minimal impact on the building structure. Lindab's products meet the certification standards for BREEAM, leed, dgnb and many more, ensuring optimal environmental performance. And if you're looking to simplify your design process, their range of ventilation software and tools make product selection, calculation and performance evaluations quite straightforward. Selection calculation and performance evaluations quite straightforward.
Speaker 1:Creating healthy spaces is at the core of Lindab's mission, which is why, at Air Quality Matters, we are so pleased to have them as a partner with this podcast. Do check them out in the podcast notes at airqualitymattersnet and, of course, at lindabie to the podcast. How did you find yourself getting into epidemiology, corinne? Is that something you studied and did you take that with you into the indoor air quality observatory? Maybe give a little background as to your journey into this field, to where you are working today. I think it'd be really interesting for people.
Speaker 2:No, I'm not an epidemiologist by background. I'm a chemist, environmental chemist. I studied organic chemistry and exposure to chemicals. But when you are studying human exposure, you rapidly come to epidemiology and association with health effects. That's why I arrived on the epidemiological studies and leading a group of epidemiologists. But I'm passionate with trying to understand the environmental determinants of human diseases. That's why I arrive on epidemiology. I admit I'm personally more passionate with human studies than studies on experimental animals.
Speaker 1:That's why I?
Speaker 2:arrived on this topic.
Speaker 1:There's an investigator in there, in you. I think there's a lot of that in scientists.
Speaker 2:If they hadn't been scientists.
Speaker 1:They'd have been inspectors or investigators or something.
Speaker 2:You're asking questions the whole time and you're trying to understand the why and there are so many diseases today that that are increasing, like cancer, cancer in children or allergy, alzheimer disease, many diseases that we we don't know the we don't fully know the causes. So it's important to understand if there is an environmental component, environmental determinants explaining these diseases, to be able to, if we understand, if we know, to be able then to prevent from, to raise awareness among people if they need to adjust the behaviours, provide knowledge for the policymaker to regulate. So this is what I like to contribute to, a part of contributing to better knowledge of these environmental determinants.
Speaker 1:And listeners to this podcast would have heard us mention a couple of times. We've had Henry Burridge and Kath Noakes and a few others on talking about this idea and concept of an indoor air quality observatory. But you've lived that. I'd love to get your take on the the kind of the birth of the indoor air quality observatory in france and and practically what that meant during your tenure there.
Speaker 2:Um, what an indoor air quality observatory does or could do yeah, so the french indoor air quality Observatory was set in 2001, so more than 20 years now. At that time, there was a big crisis in France around asbestos because we knew that asbestos was toxic, but it was banned very late. It was banned in 1997. So the policymakers at the end of the 90s in France wanted to have an independent research group collecting data neutral data, disconnected from the industrial stakeholders to describe indoor air pollution. That's why the observatory was set in 2001, to provide knowledge. And the main objective why the observatory was set in 2001, to provide knowledge, and the main objective of the observatory was to collect data regarding indoor air quality, but not only regarding noise in buildings and temperature, also thermal comfort and at the national scale. So the data were collected. The added value of an observatory is to collect data at a large scale and, if possible, in randomly selected buildings, schools or dwellings. If you do random selection, you you will look at all. You have a diversity of situation. You you, if you're based on a voluntary basis, you may have the best and the worst were those who are aware of the problem and those who are in difficult situations. So having a large monitoring across a country with random selection of buildings provides a huge amount of data and this knowledge can be used for many things. It can be used first to provide references.
Speaker 2:We were discussing a bit earlier during our talk about the indoor air quality guidelines proposed by WHO or by different countries. These guidelines based on health data to know if it's exceeded you may have the disease appearing in the people. So it's to interpret the concentration. But unfortunately these indoor air quality guidelines are very scarce today. We have thousands of chemicals in buildings but only about 20 indoor air quality guidelines. So being able, so for the pollutants who don't have any, that don't have any indoor air quality guidelines, having the distribution in dwellings, all the dwellings in France, or all the schools or all the offices help the people doing measurement to interpret the results. If you say you measure the value of 10, but you know that the median value of this pollutant is 20 in dwellings in France or in Germany, you can tell to the people in this office or in this dwelling you are below the median value observed in the housing stock, so it seems it's dwelling. You are below the median value observed in the housing stock, so it seems it's OK for you. On the contrary, if you measure 30 micrograms per cubic meter and that the median value is 20, you can say to these people or the building owner you are over the median and maybe you are over the 95 percentile. So it helps by default If you don't have sanit-health guidelines. It helps to interpret the measured concentration. So the observatory provided for 20 VOCs for the first campaign and more and more over the time, this distribution of concentration in French dwellings, in French schools, in French offices. So it is very helpful then for people doing measurement a municipality, doing measurement in a school or a building owner and doing measurement in an office building to interpret, to know where the measurements can be situated compared to the full stock of buildings. So it's used for that, to providing reference value, reference distribution.
Speaker 2:Since when you're collecting massive data, we were collecting also descriptive data on the buildings, use of concentration materials, the outdoor environment of the buildings, the activities of the occupants of the buildings. So collecting all these descriptive data make possible to then to identify the determinants of some concentration. I had a few, for example, in schools. Here we identified that the classrooms having plastic flooring had more phthalates in settled dust and more phthalates in indoor air. So we identify a source, having this large sample. This allows, as we were mentioning, to avoid not to have confounding factors. If you have a large group of buildings you have a lot of diversity and you can identify these determinants. So the flooring we identify in dwellings that if you have an attached garage you have a higher concentration of benzene and toluene if you attach, if you have an attached garage, you have you have either your car with exhaust gas or you have old paints, all the varnishes, you put everything in your garage and all these solvents may still emit and may contaminate the house. This makes it possible to identify some determinants and then to do prevention. I know that we pass the message to architects If you are building a house, please do not put a door between the garage and the house. Maybe it's put a door between the garage and the house, maybe it's not very convenient for the people living there or put a door very tight to avoid any contamination. So this helps to identify determinants.
Speaker 2:Having an observatory and repeating measurements over time makes it possible to identify the new compounds of interest and, on the contrary, the compounds that there is no need to measure anymore chlorinated compounds, perchloroethylene or trichloroethylene. Over time the indoor concentration decreases. So today we consider it's not necessary anymore to look at these compounds in indoor air. Measurements are expensive, so if you can avoid measuring things that you will never detect, it's better. So chlorinated compounds are disappearing. On the contrary, we observed over the 20 years an increase in terpene concentration, limonene, pinene and we think it's due to the fragrances. We use more and more fragrances in product chemical and cleaning products or air fresheners, so this helps to identify the new emerging compounds.
Speaker 1:Yeah, interesting.
Speaker 2:It helps also for policymakers. In the dwelling survey we identified a group of VOC that were measured in many, many dwellings and coming from building material. That's why the French government decided to have a mandatory labeling scheme on building material and decoration products. So now in France you have a label on carpets, on paint, on varnish. When you want to paint your bedroom you can buy a. There is a label from A+ low emission of these 10 VOCs to C high emission or not evaluated in the lab. So this provided information for policymakers to make the regulation evolve. Based on the basis of the measurement we did in school, the French government decided to implement a regulation to measure indoor air quality in schools. That's why since last year it's now mandatory. As soon as you have an energy retrofit or you have a new extension of a building a big event in a building you need to do some measurements. Not all compounds, only formaldehyde, benzene and CO2. And every year you need to check the windows and the ventilation system, if any, to be sure that the air change rate is okay.
Speaker 2:So an observatory provides knowledge but also provides clear elements for policymakers to be, able to have a science-based decision and I will say that it also provides information to raise awareness on the observatory and also the role to make flyers for the general public to communicate to the general public in newspapers. So having this observatory makes the topic more visible and provides useful information on air pollution for everybody, for the general population. And I know that, not in France but in Germany, the observatory and the campaign they are performing only dwellings are associated to human biomonitoring measurement in urine and blood. So in Germany they are able to quantify, to identify the contribution of indoor pollution to human exposure. They are measuring either semi-volatile organic compounds or lead, for example, are measured.
Speaker 2:And if you add to your observatory health questionnaire, we make it possible to have an EPI study. And during the French first housing survey health questionnaire we have this, make possible to have an epi studies and during the French first housing survey we did measurements in dwellings. But people also had a questionnaire to fill in reporting asthma and COPD and allergies and we identified that the more VOC they had with high concentration, the more risk they had to have asthma or other respiratory diseases. Was there any?
Speaker 1:coordination with academic institutions, with the observatory, to coordinate research, or to stop siloing or duplication. Or to stop siloing or duplication, um, because that that seems to me to be one of the other values is that, uh, I mean air quality in academia has seen an explosion. You know, there's never been more studies being produced than there are right now on indoor air quality. But there seems to be a lack of coordination sometimes, or duplication or siloing, and I can imagine an observatory having a very positive impact on that.
Speaker 2:Exactly, yeah, the French Observatory was collaborating, and still collaborating, with universities to either to develop analytical techniques or to do some analysis chemical analysis or to do some analysis chemical analysis or to do some statistical analysis.
Speaker 2:Yes, and there is a scientific committee with academic people and the studies are discussed and there is a good coordination to avoid overlap with some studies. And I know also that there is an indoor air quality observatory in New Zealand and this is also a group of all the institutes and universities trying to coordinate their efforts. The budget are limited, so it's good to to be efficient and to coordinate and not to duplicate, as you say. So in new zealand the observatory has this objective to coordinate and to to, to organize, to to be more powerful altogether.
Speaker 1:Yeah Well, I mean, the question has to be asked why doesn't every country have one? And that's probably the first question, probably the more realistic question I had for you is what was it? I mean, if you were to describe the observatory, is it a government department? Is it attached to a particular research organization? Is it a big, heavy thing to manage, a costly thing to have within it? I mean, is it possible to describe what it is as an entity, an observatory, what your experience of it was, in France at least?
Speaker 2:It is expensive. That's why it's difficult. You need money. For example, the nationwide survey the first dwelling survey was about 7 million euros to collect all the data and then to analyse all the data and do all the statistical analysis over a course of 10 years maybe. So for one year and for all the people working on the survey it's not so much, but at the end it was several million euros. The school survey costs more than 10 million euros and the second dwelling survey was about 7 million euros. So it's very expensive. That's why it does not exist in all the countries. So the first limitation is maybe the money. The second, maybe the first, actually is the political wish.
Speaker 2:If you want an observatory supported by public money, you need to convince the government. Politicians and people in the ministries need to be convinced that it's important, and sometimes it's not easy to convince people. People really think of outdoor air pollution. This is visible. The car exhaust industries. We speak a lot about water quality also, but indoor air pollution is not really well known. We know in our group of people working on the topic the importance of this issue, but once you make a step back you realize that in environmental sciences indoor air quality is not really well known, even among academics and among policymakers, even more so same in the general public or building owners. So you need to convince the politicians. But when they are convinced they can do nice things. And that's what happened in France the government when it was built in 2001,. There was a politician who really wanted to make things move, and if there is this political wish, then you can find the money.
Speaker 2:So maybe the first is the political wish and then the budget, and this is hard to maintain the budget over a long period. We were discussing the epidemiological studies that are long and expensive. This is the same for building an observatory it's long, the campaigns are long. An observatory is is long, the campaign are long, so sometimes the politicians want the results, uh, the day after. So it's not possible.
Speaker 1:It's, it's a long process or at least within a political cycle. Uh, they want the results.
Speaker 2:Yeah, yeah, for sure it's not a, at least in france, it's not a department in the ministries. There were three ministries environment, health and housing, supporting the scientific and technical center for buildings, cstb. Cstb was mandated to implement this observatory. There was a yearly contract with the ministries, with money and with the decision.
Speaker 2:Every year it was decided okay, next year you're doing that, that you are housing survey or you're analyzing this data yeah, so and cstb was coordinating this observatory with a lot of partners academics, public lab, public or private labs and statisticians, psychologists a group of people around doing the job.
Speaker 1:And size-wise, I mean how many people? At its biggest and smallest was the observatory?
Speaker 2:We were a team between 10 and 15 people technicians, because when you are carrying out a nationwide survey, everywhere you must have the same equipment, same protocols. So there was a technical team looking for equipment, testing the equipment, calibrating the equipment, training the teams on site, on field. So there is this technical team. Then there are the project leaders one person in charge of dwellings, one person in charge of offices, schools, one person in charge of ventilation topic, having cross-cutting advice on all the studies on ventilation. Then we add one data manager. It's important to have a data manager. We mentioned the database and the importance of having clean data, reliable data. So having a data manager is mandatory. And then you need statisticians to do the statistical analysis and we had sometimes a communication person with us to help for the dissemination.
Speaker 2:This is also important to disseminate, to have useful knowledge, not knowledge for knowledge, but knowledge for policymakers, for general public, for other scientists. So it varies over time, depending on the budget, between 10 and 15%.
Speaker 1:Was that big enough? Did it always feel like it was a drop in the ocean of what it should be, or did it feel like you were making good progress? Like is that 15 people doesn't seem like a lot of people for the single biggest environmental risk we face, you know.
Speaker 2:Yes, there were more people. We calculated that the number of people contributed to working group and committees were about 100, not working full time but punctually contributing to this observatory. So, 15, it's not bad. I would say it's good but it's not a lot. You're right. But we had local teams, also that subcontractors on site to do the sampling and private or public labs doing the chemical analysis or biological analysis. So 15 permanent staff, but many people are working around partly for the observatory, and so it's not bad compared to other countries.
Speaker 2:I would say we were lucky in france having this yeah, for sure compared to outdoor air quality, where there are many, many state monitoring station and people in all the region. This is true that compared to outdoor air quality, it's it's not a lot of money, neither a lot of people working on that topic while we spend 90 85 percent in our time in building.
Speaker 1:It's yeah, well, I mean, you know, uh, not to get too tropey, because, uh, we've said that ad nauseum on the podcast, but the vast majority of our exposure to even outdoor air occurs indoors, you know, and it changes indoors, so we do need to know so much more about it. Is it an awkward question to ask? Do you have any sense of how much it costs a year to run an observatory? Was it a public budget for the observatory?
Speaker 2:Yeah, it was only public and still only public. It was about 2 million euros every year. So paying for the surveys and paying for the people. It's not. It's to my opinion, it's reasonable. But even 2 million was sometimes hard to collect.
Speaker 1:Yeah, I was having this argument the other day. You know there are major universities in london that would spend that on research a year on indoor air quality and not even miss a heartbeat, you know. So at a national level that's not even it's presentable doesn't even register, which is a shame because I imagine the return on investment of something like an observatory is almost incalculable from my buyer's perspective.
Speaker 2:Because it's not calculable. It was hard to show that. Sometimes they add in value. You know, yeah, we try to assess the cost of indoor air pollution. In france in 2014, we assessed that for six indoor air pollutants, including radon. By the way, we calculated that it costs uh, it's. It is associated to 20 000 additional premature deaths every year and 28 new diseases, new cases of new disease and the cost of 19 billion euro every year because of so the cost can be calculated because you have health cost when you're sick, you go to hospital and you have drugs that cost, but may also due to the cost of life.
Speaker 2:When you die prematurely not at the life expectancy, at about 80 years old. When you die prematurely, not at the life expectancy at about 80 years old, when you die earlier, you lose some years of life and these years of life cost to the society.
Speaker 1:Of a cost, yeah.
Speaker 2:Give a value of human life and, based on this value of human life, we were able to calculate, based on the diseases, the mean age at which you die prematurely if you are sick, if you have a lung cancer, if there was carbon monoxide poisoning also leukemia. In our study, depending on the mean age and the number of people who die, we were able to calculate the years of life lost at the country level and then to transform this in euros, based on the cost of human life. So that's why we came to this figure of 19 billion euro. That is a lot, of course.
Speaker 2:Yeah, and I think that's part Only for six single pollutants for which we had enough data regarding exposure on one side and health effect on the other side and quantitative relationship between exposure and health effect.
Speaker 1:Sounds very similar to the kind of work that Ben Jones and Max Sherman were doing with the DALIs and these kind of work that Ben Jones and Max Sherman were doing with the DALIs and these kind of five or six major pollutants, none of which are a surprise to anybody in indoor air quality, but nonetheless we can be pretty sure that a fairly defined group of pollutants are causing a vast majority of the trouble, and I think that's half.
Speaker 1:The challenge sometimes in communicating this risk is that the, the costs at a national level or the dali impacts, the disability, adjusted life years, are all so eye-watering that it makes it difficult, or strangely, to pitch a two million quid a year cost of maintaining something where even even having a small impact on a I mean the uk is 20 billion a year, I think the us is 150 billion a year costs, early death, disability, productivity, you know, health care costs, all of these kinds of things. It would appear like it's an absolute slam dunk to invest 20, 30, 100 million in flipping observatories and research and trying to sort this stuff out, because the return on investment to society would be so enormous.
Speaker 2:This is not the same. We're paying, you know, if you ask a building manager to invest in low emission material products. It will cost for this person, but if people are sick in 20 years he will not pay. So this is not the same people who are paying. So, that's why some building managers are reluctant to invest in air quality, because they are not the ones who will pay. They will sell the building and then it will not be their problem if people are sick in a few years or that's why productive.
Speaker 2:That's that's that's why we are. That's what we are told.
Speaker 1:That's it's important to show these figures, but sometimes it does not convince people because they are not paying, you know, and value of life is also something very intangible and this is why I think, within things like the observatory, the communications department is so important, because how we communicate risk, how we change habits and behavior and perception of risk, particularly with something that's often so long-term and chronic in its outcomes, is so critical um we do not need to scare people.
Speaker 2:Also, it is true that when we speak about indoor air pollution, we spend a lot of time in buildings. There, the concentration are higher than outdoors. We are exposed to inhalation, but also thermal contact and injections. There are thousands of chemicals. At the end, people are scared and if you are scared, sometimes you don't act because you're scary. So we need to have yeah, communication is very important and collaboration with human sciences. People is also important to find a good way to communicate and engage people instead of frightening people.
Speaker 1:Yeah, and it like say the complexity is a challenge in its own right. I mean, you mentioned something there which I meant to pick up earlier which we didn't, and that was that you know what is an air quality risk can become a dust risk and a dermal risk and an ingestion risk and like. Risks change over time and in the process as well. Um had charles weschler on the ozone air chemistry guy, and boy does air chemistry get complicated quickly. You were talking about, uh, limonenes and pinenes, which by themselves aren't necessarily very harmful, but put them in contact with ozone and boy do they turn into some nasty things quickly.
Speaker 1:Um, I think that's half the challenge is that it's. It's not the fact, it's not the fact that it's chemistry or that it's air quality, it's just the fact that there are so many, there's so much to comprehend that I think people just shut down ultimately, and if you can't translate that meaningfully into levers that people can control to have an impact on an outcome, they lose the sense of direction very quickly you know, it's really hard yes, you know, when we ask, we'd say people, open the window to remove your indoor air pollutants, but we know that some particles are also coming from outdoor air.
Speaker 2:if outdoor air is polluted, so people can be lost. That's why we need clear messages to really help people and motivate them.
Speaker 1:So what's the IRSN then? What's your role in that today?
Speaker 2:So IRSN is the French Institute for Radiation Protection and Nuclear Safety. We are working on ionizing radiation and effects on humans. Irsn is the French Institute for Radiation Protection and Nuclear Safety. We are working on ionizing radiation and effects on humans through epidemiological studies. So we are working, of course, on radon natural radioactivity but we are also working on its effects due to ionizing radiation at the workplaces. So nuclear workers are exposed. Some medical workers are also exposed to ionizing radiation.
Speaker 2:And space attendance, flight attendance People. When we are flying in the plane we are exposed to cosmic rays and so as passenger it's short exposure because we are not in the plane every day. But for the people working in airplanes there are some studies showing that it could have effects on cancer in flight attendants. So we have a cohort of about 50,000 flight attendants for which this cohort is starting. It's a retrospective cohort, it means we will not follow them in the coming years. We are trying to rebuild, assess their past exposure and looking if today those were dying or were sick because of cancer or other diseases, if it could be due to their past exposure. And their past exposure is assessed based on if they were flying on a short-distance flight or long-distance.
Speaker 1:If you have long-distance flights.
Speaker 2:You are higher in the sky, so you're more exposed to cosmic rays. So this is a retrospective call, so you're looking at the exposure back and trying to see if the disease is. We don't need to wait 20 years or more so that they declare a disease.
Speaker 1:They are, they're already, uh, they have already been exposed, and so we are collecting the, the cause of death for those who are dead, and yeah, and you're in a big metal tube in the sky for most of your working life. So, like it's quite a defined set of confounding factors, you're unlikely to be a flight attendant and a radiological nurse at the same time. Um, so I guess if you, if you are seeing health effects from ionization, potentially you can be pretty sure where that's come from if you're dealing with a flight attendant I think ozone is another one.
Speaker 2:Isn't it also exposed to night shift and uh? Yeah and you know jet lags and we know that jet lags night shift or can also disturb the else, and so they have co-exposure there.
Speaker 1:And ozone. Charles Weschler was saying more ozone up high as well, potentially.
Speaker 2:There is what we call the fume events, some chemical contamination in airplanes. That is not well understood now, but they could also be exposed to some chemicals sometimes. So we have also this occupational cohort. And the third group of people we are looking at are the people who undergo radiotherapy. When you have cancer you have radiotherapy that kills the tumor, but today we are more and more healing from cancer. We are not dying anymore from cancer, so the survival rate is good, it's improving. But then you could have else effects and when you have a breast cancer you could have effects on the heart. Our heart is exposed to ionizing radiation. So that's why we're looking also at side effects of radiotherapy, to better understand these side effects and to try to find the best dose enough to kill the tumours but not too high to save the healthy tissues around the tumours. So we also have these cohorts.
Speaker 1:And are you sitting across all of those studies, Corinne, or do you focus on one particular area?
Speaker 2:No, I'm coordinating. We are a group of 25 people epidemiologists and data managers, statisticians working on all these cohorts.
Speaker 1:Wow, and you mentioned you're doing an interesting one on radon, a very large cohort radon study. Tell us a little about an interesting one on radon, a very large cohort radon study. Tell us a little about that because actually radon isn't something I've spoken about much on the podcast. Actually, I must get James McGrath on to talk about radon, because Ireland struggles with radon quite significantly.
Speaker 2:Yeah, James is a big expert in radon.
Speaker 1:Yeah, tell us about the radon study you're doing. Expert in island radon. Tell us about the radon study you're doing.
Speaker 2:This is called the CORAL project. This is part of the Constance cohort. Constance is a large population cohort in France. 220,000 of people were recruited 10 years ago. We are assessing the exposure, the radon exposure over their life, because all the participants in our cohort provided all the places where they live across their life, so we are able to. When they were kids, we will assess radon concentration based on the cities they were, based on the radon potential in the soil. We cannot do measurement in the dwellings where they were living 30 years ago and so, based on all the dwellings they lived in, we will assess the radon exposure over the life and assess this radon exposure with the diseases that they declared lung cancer also. But the idea is also to look at other types of cancer, because it is established that radon is associated to lung cancer, but it's still not really clear if radon could be associated to leukemia, because the bone marrow could be also exposed to ionizing radiation after radon exposure and there is also suspicion of skin cancer due to radon.
Speaker 2:So other type of cancer and other type of diseases. Neurological diseases could also be associated to radon. We are about to publish a new meta-analysis where collecting all the data on radon and its effect, and that's why we perform these cohort studies to study the association between radon exposure over the course of the life, because the case control studies were mainly done in adulthood, not including childhood, so we want to see if the exposure during childhood may have a contribution.
Speaker 1:So this is ongoing.
Speaker 2:We should have the results in one or two years. We have collected all the data and we are now about to analyse, to treat statistically all this data.
Speaker 1:And is there an assumption that you can be a more vulnerable group with something like radon as well? Like the young, are they likely to be more affected collectively?
Speaker 2:It has not been observed for radon yet, but it's known that the children are more vulnerable to ionizing radiation, more vulnerable to chemicals as well, because they are. They are young, as you say, the tissues and organs are not totally mature and compared to adults and compared to the weight body weight children are more exposed. They inhale more air, they eat more food, they drink more water If we divide by the body weight. Compared to adults, they are more exposed and they are more vulnerable because they are still under development.
Speaker 2:So we know that children are more radio-sensitive than adults, but it has not been showed on radon, because most of the epidemiological studies on radon were made on on adults and just collecting the past exposure, but not going back into the childhood of people yeah, now we spoke about this before kids, kids tend to eat dust and be exposed to all sorts of stuff.
Speaker 1:You know so they. They tend to overexpose themselves to all sorts of things.
Speaker 2:I forgot to mention this, that they have specific behaviour. They are crawling on the floor, they are touching their mouth with their hands, specific behaviour that expose them more to for example to chemical compounds. Semi-volatile organic compounds. We were mentioning phthalates in dust or poly-brominated flame retardants. It has been shown that for some of these compounds the major pathway of exposure is in ingestion for children.
Speaker 1:Yeah.
Speaker 2:And in this cohort study where we will study radon, we try to assess the radiological exposome. You may have discussed what is exposome during the podcast with other people. So exposome is the old exposure across our life, from birth and even in utero, before to death. All type of exposure, environmental exposure, but also social exposure, behaviors. So this is the exposome, as there is a genome describing all our genes. The pending concept is exposome. It was proposed by Christopher Weil in 2005.
Speaker 2:So we are trying to assess the radiological exposome. This is all the exposure to ionizing radiation from birth to adulthood to now. So this is radon, natural radiation. But we also assess the human entropic radiation, so the fallouts from Chernobyl and from nuclear tests. We still have a very low contamination of the environment, but still a contamination. So we are assessing this and assessing exposure through medical. When you are doing a CT scan or a radiology radio. When you are doing a CT scan or a radiology radio, you are exposed to x-rays, ionizing radiations. We are asking to these people all the medical exposure exams they had from childhood to now and for those who were exposed professionally, we also we will know which type of job they had in the past, if they were a nuclear worker or flight attendants or a medical worker.
Speaker 2:So we assess all the routes of exposure to ionizing radiation and, at the end of those, a total cumulative dose from birth to now, and see if there is association with diseases cancer and non-cancer diseases. So this is what we call the radiological exposome I shows you learn something new every day.
Speaker 1:I've not actually heard exposome before. I mean, we've. I've heard the microbiome and things like that, but the the exposure makes some sense, the cumulative exposure and does it tend to talk about one particular pollutant and your total exposure to it? Or could it be a collection of pollutants, all pollutants?
Speaker 2:physical, chemical, biological and also, uh so, your close environment, but also the your social interaction. We know that the social environment is also important for health, so exposome is really including all these risk factors and also the positive ones. If you are exposed to green spaces, green spaces have been shown to be positive for health effects, and if you are living in a city where you can walk, it's positive for your health. So exposome is really the understanding of all what expose people from birth to adulthood and association with else. Yeah, it's a new concept more and more when we were mentioning that we need to break the silos. The objective of exposome is really to make people think larger than only one chemical or only radon or only a virus. It's to make people think that, okay, we're exposed to a large amount of many factors physical factors but yeah, so the exposome is. This is a very new concept, but uh so it's hard to to qualify the exposome of people. That's why the scientists are looking at, uh, radiological exposome or indoor exposome. Indoor exposome is, uh, yeah, what?
Speaker 2:what you're exposed to indoors including the biome and there are people working on urban exposome. Urban exposome is outdoor air pollution in cities, plus noise in cities, plus the fact that it's very dense you cannot walk. So there are studies on children's exposome. You're only looking at what is the exposure of children chemicals and the food you're eating. It was Christopher Wilde who invented this concept in 2005, was the director of the IARC, the International Agency for Research on Cancer. That's interesting.
Speaker 1:yeah, it reminds, me.
Speaker 2:To make people think larger, to be more open-minded, including social exposome, and social sciences, and this is why human biomonitoring is developing, because it's hard to describe all the exposure from birth to now.
Speaker 2:That's why looking at what we have in our blood, in our urine, is more and more used because we imagine that if we were exposed during our life to these pollutants, for those who are persistent they are still in our. Or if we are chronically exposed, they are not persistent but we are chronically exposed to them, such as phthalates. We can measure them in our blood and urine. They are not persistent but we are chronically exposed to them, such as phthalate we. We can measure them in our, in our blood and urine. That's why it's developing a lot in europe, but in also in us.
Speaker 2:It started in us with the name studies where a group of americans were sampled every every couple of years, measuring hundreds of chemicals in in their blood and urine, and we mentioned the low-cost sensors and cheap and expensive techniques. But on the opposite side, we have the non-target analysis that are developing the NTA. I don't know if you have heard about this technique so to now. When we were looking for benzene, we had to know that we want to measure benzene in this building and we were having the analytical sampling and then analysis to detect benzene. We had to know that we want to measure benzene in this building and we were having the analytical sampling and then analysis to detect benzene. So we were very blind because if you don't know that there is a chemical, you are not looking at it, you are not developing the sampling and the analysis we discussed about the Lempost effect.
Speaker 1:Yeah, I was going to say we had a conversation about this. I remember that we had a conversation about this. I remember that we had a conversation about that.
Speaker 2:So the idea of this new NTA, non-target analysis, is that you have no limits. You take everything in your net and you are able to discover quantities of chemicals, and with both gas chromatography and liquid chromatography, when you combine both techniques you are able to so now. With such techniques, it's not 10 VOCs that you measure in the air.
Speaker 2:This is thousands, so it's very expensive. It's very new, so it's not used as routine. So it can be used for environmental samples dust or air and also for urine and blood. So it's a way to assess exposome. If you are looking at this non-target analysis in urine or blood, you discover many, many, many compounds and this was first tested in Europe by Pavel Rotkovsky in Norway. He measured with these techniques settled dust samples and he identified more than 2,000 chemicals in settled dust samples in dwellings. So it was dust. Now it's used at EU level in blood samples, so it's blood exposome. In a way, you have all the markers.
Speaker 2:But so it's challenging because analytically, it's blood exposome in a way you have all the yeah but so it's challenging because analytically it's quite hard and expensive, and then you have a massive amount of data. You have also some statistical issue to be able to, to address and to, to manage all this amount of data. So it's not simple, but it's very promising. I also also to be able to describe our exposome.
Speaker 1:Yeah, because I mean to very briefly explain the lamppost effect. It's a story about somebody looking for their keys under a lamppost and somebody asks them why are you looking for your keys there? And he says because I lost them over there, but it's dark and I can't see. So I'm looking where I can see. I lost them over there, but it's dark and I can't see. So I'm looking where I can see.
Speaker 1:And we talk about that in the context of air quality, because we know from experience that pollutants keep bad company and if we're looking at one pollutant, we could be assuming a health outcome that's toxicologically, I guess, being caused by something else. It just happens to be keeping bad company and we're missing parts of the jigsaw puzzle. So, generally speaking, when we're doing these speciated air quality tests, we're looking at 10 or 20 or maybe even a few more, but there are hundreds and thousands potentially in and around us. But we have to balance that equally with the. As interesting as that is, does it help us get a better outcome? And if we look at studies like Ben's and Max's, if we find that 75% of all the harm is coming from particulate matter, are we better off just focusing in on that?
Speaker 2:or is it the?
Speaker 1:fact? Is it the fact that particular matter is keeping bad company we don't know about? You know, this is the fun thing, I suppose, is trying to unravel that and figure out where this harm is being created and improve our understanding of it exactly.
Speaker 2:Yeah, I don't say that in the future we need to measure everything in with this nta techniques, but this will help to improve knowledge and maybe to focus on some specific compounds that that are associated to human health effects yeah, and if nothing else, it will keep the phd students busy for the next couple of decades for yeah, absolutely, yeah, yeah, I've still got a lot to do in our field For sure.
Speaker 1:So where do you see it kind of heading then over the next five years, corinne, from an epidemiological perspective? Is there a shortage of epidemiologists? Is it a sector that needs to be 10 times as big? Is there advances in science you think are coming? Where does it? Where does your field go over the next kind of five to ten years, do you think?
Speaker 2:yeah, you're mentioning a shortage of epidemiology. This is true that we have more and more difficulties to find students because, uh, students uh quite smart in mathematics or uh, data science, are more going to finance financial issues, studies or to AI. But we are working on a nice object. This is human health, so that's why we can motivate young researchers to join us. Because of this, this makes sense. What we are doing makes sense because we try to understand the human health effects and, as we were discussing, more and more diseases are increasing. Cancer or other neurodegenerative diseases are also increasing.
Speaker 2:We need to understand if there is an environmental cause. If there are environmental causes, so our studies are. With all the new models, the new techniques, with AI or the new way to measure exposure. This will expand, for sure in the future. And we have also societal requests. People are more and more aware of environmental impacts. People are requesting more data. Maybe I'm biased because I'm working in the field, but I have the feeling that there is an associative demand to better understand the influence of the environment or the influence of our work on else, and we need science. More generally, we need science to provide elements for our policymakers. With COVID crisis, we observed that there is high risk that people are taking decision with not science-based and more and more people are propagating fake information. So we need really science and good science and information, but basic I mean factual information, factual data to help policymakers to take their decision. This is important to keep the science at a high level in the future.
Speaker 2:I believe that this is not the only element for decision. There are also, of course, economic issues or societal issues, but this is part of the elements that policymakers need to make their decision Evidence-based knowledge to improve the health of the people.
Speaker 1:Yeah, no, absolutely.
Speaker 2:So I hope epidemiology will help in that sense.
Speaker 1:Corinne, thanks so much for spending a couple of hours talking to me today.
Speaker 2:It's been brilliant really enjoyed it.
Speaker 1:It's been absolutely brilliant. Thank you, thanks for listening. Before you go, can I ask a favor? If you enjoyed the podcast and know someone else who might be interested, please spread the word and let's keep building this community. This podcast was brought to you in partnership with 21 Degrees, lindab, aeco, ultra Protect and Imbiote all great companies who share the vision of the podcast and are not here by accident. Your support of them helps their support of this podcast. Do check them out in the links and at airqualitymattersnet.
