Alison Lee, MD, discusses the health disparities of air pollution in the home.
Alrighty. Great. Good morning, everyone. Good morning, Dr Lee. Well, I'm glad you're here. Thank you very much. I see a lot of pulmonary colleagues online right now. Nice to see hi, Sexy. And I'll turn it over to Jim and he'll introduce us and get it started. All right. Good morning, everyone. Welcome to this installment of our virtual Grand round series this morning. We're very fortunate to have Dr Allison Lee speaking for us. Dr. Lee earned her medical degree from the University of Massachusetts and then went on to complete residency in internal medicine, N Y u, followed by fellowship in pulmonary and critical care medicine at Columbia before joining the faculty at Mount Sinai. She currently serves as an assistant professor in the division of pulmonary critical care and sleep medicine and conducts research examining the influences of environmental and socio economic risk factors in the development of chronic disease, with particular focus on environmental influences in the prenatal and early childhood period and their contributions to development of chronic diseases later in life. Please welcome Dr Allison Lee. Thank you very much. So can everyone see my slides? Great. Perfect. Okay. Um so thank you very much for the introduction and the kind invitation to speak today. I'll be speaking with you about the work that I do, um, in Ghana, specifically looking at an exposure that you may not have heard of, which is household air pollution. I have no disclosures. So just an overview of where we'll be going today. First starting out by talking about poverty and health just offer you a framework of how I approach the problem and from there trying to, um, Relais the importance of early life on the development of chronic disease. So specifically looking at what's known as the Doha had theory or the developmental origins of health and disease. Um, but then really focusing on long development and why early life maybe so important. From there we'll talk about household air pollution, which I believe is a globally important environmental exposure and then moving into work that we're doing with the Ghana randomized Air Pollution Health Study. So it's well described that zip code determines how long you live. This is data showing that these are counties in California, and essentially what you can see here is that in the wealthiest counties such as Marin county average life expectancy is, uh, over 83 years, whereas in the more impoverished communities, average life expectancy is, let's, say, 76 years. Um, and again, this is really just crude data, but demonstrating this strong relationship between poverty, um, here by household income and life expectancy. And of course, if we look worldwide, this is magnified even further. Um, here you can see that throughout sub Saharan Africa, life expectancy is essentially below 65 years of life. Um, in Ghana, where we work, which is outlined here in red, life expectancy is 63 years. So this is compared to an average of 78 years in the United States. And clearly this is not due to genetic differences. Clearly, there are changes in the environment that are a result of poverty and economic inequality that work together to produce health disparities. And these are really the so called social determinants of health. This is by no means an exhaustive list, but I think we can understand that environment such as built environment, access to health care, healthy foods, education, environmental pollution for housing, stress and income all work together to produce health disparities. And it's really important to understand that these health disparities begin to act in early life. And so in the 19 eighties, Barker and his colleagues were trying to understand differences in mortality rates from chronic disease. Um, and they noticed that these while there were reductions in mortality rates across England and Wales, the reductions appeared to depend on socioeconomic status of those communities. And they hypothesized, perhaps that the reason for this was differences, um, in health in early life. And so what they did is they decided to try to understand whether or not there was a relationship between mortality and adulthood and infant mortality at the time when those adults would have been infants, and so they looked at a number of different diseases. Here's looking at ischemic heart disease and essentially looking at the correlation between ischemic heart disease and neonatal mortality by community. And you can see that there's a strong relationship that they noticed and further when they broke this down by socioeconomic status SCS one being sort of the wealthiest communities and SCS five being the most impoverished, you can see that there are stark differences both in infant mortality and also mortality of chronic diseases in adulthood here, just showing chronic bronchitis and emphysema with our standardized mortality rate of 36 versus 188. And this is just that same data described graphically where you can see here the exes are the most impoverished communities. Um, these communities down here are the wealthiest. And so, again, just a strong relationship demonstrated here between ischemic heart disease on the left, bronchitis on the right and neonatal mortality. They went on to see whether or not they could identify sex specific effects. And indeed they did. They found that the relationship appeared stronger than women as compared to men. And so from this data, they concluded that adverse influences in childhood associated with poor living standards increased susceptibility to other influences associated with affluence encountered in later life. They did wonder, though, Could we look at markers in earlier life? So they looked at blood pressure in Children at age 10 and adults at age 36 just for brevity. I'm only showing you the data from 36 but this is the same in Children, and essentially what they did is they wanted to examine whether birth weight as a marker of the intra uterine environment would be associated with blood pressure at these different ages. And what they found is that for men in both men and women, those men and women with the lowest birth weight also had the highest systolic blood pressure. And this was true even when they stratified by current weight. And so I think that these two studies together really start to bring out for important points. One is a vulnerability disease across the life course began in utero. And the work that I do is trying to understand how exposure to adverse environments could be important in that we also start to see that their parallels across organ systems suggesting perhaps an inter relatedness of organ systems and perhaps an inter relatedness of underlying mechanisms. We see, especially with the blood pressure work, that there is tracking of development across the life course. And we also start to see a suggestion of sex specific effects. So now narrowing, narrowing more into long health. Um, hopefully I can explain to you why, as an adult pulmonologist, I'm most interested in studying early childhood. So even before Barker's, uh, doe had theory, this is work by Fletcher and Pido, and this is sort of the seminal work demonstrating FV one trajectories. And what they did is they recruited over 1100 working men from central London, and they followed them with lung function measurements every six months. And they use those repeated measures to construct FV one trajectories. They made the assumption that everyone reached 100% maximum attainable lung function. And then they noted that everyone, uh, even healthy non smokers had a decline in lung function after this time. They also noted that those persons who were smokers and susceptible would have a rapid rate of lung function decline. However, if you stop smoking, your lung function would never return to normal. But the rate of subsequent decline would begin to approximate that of never smokers. And this really was the paradigm of COPD until a number of decades later, when Langdale published this seminal paper in the New England Journal, where they leveraged three cohort studies, the Framingham Offspring cohort, Copenhagen City Heart Study and the lovely smokers cohort that had longitudinal follow up of its participants for more than 20 years with repeated lung function measurements and using this data they categorize people into having normal, um, normal or low F B one a baseline and having COPD or no COPD at study end. And so essentially, they've created four groups. Persons who had normal or low FB one at inception and no COPD at study end versus those who had normal or low FB one inception with COPD at study end. And what they found is that of the 500 people who would go on to develop COPD, a significant proportion here, 62% of those people actually had reduced maximum attainable lung function. And what's more is they discovered that the rate of lung function decline in those persons was actually normal, as compared to those people who had a normal maximum attainable lung function. They, as would have been predicted, we would have a rapid rate of lung function decline, and this was true whether they looked at CCS per year or Z scores. Until really, this starts to bring into question the prior paradigm of COPD, whereas before we thought that the path to COPD was from, you know, 100% maximum attainable lung function with a rapid decline in lung function, Um In fact, now we understand that reduced maximum attainable lung function is a major risk factor for the development of COPD, even with subsequent normal rates of lung function decline. And so that sort of brings into question. Well, what is this low lung function prototype? Is this just the tail of a normal distribution, or could this be a distinct genotype? And if it is, when is that early trajectory set and what factors contribute? And to understand the answer to that question, we have to look at longitudinal cohort studies. This is one of the longest ones, or sorry, the largest ones with over 2400 participants. And in this study they recruited participants at age seven and followed them with repeat spectrometry through age 53. And if we look at these lung function trajectories that were constructed, what you can see is that by age 53 there's three different prototypes that have low lung function. This would be a below average group, a early, below average, with an accelerated decline and a persistently low group. They then went on to look at associations between these finna types and risk for COPD and what they found is that those persons who had below average lung function trajectory had four, almost four times the odds for developing COPD. Those who had persistently low had nine times the odds, and those who had early, below average accelerated decline had 35 times the odds of developing COPD. What's more, when they looked at these three finna types together, they found that these three final types made up over 75% of all COPD cases at 53 years and 100% of the moderate to severe COPD cases. But I think if we examine these trajectories closely, what you'll notice is that they appear to be distinct even at seven years of age. So perhaps we need to look even earlier. And for this there's not many studies. But this is one smaller study from the Tucson Children Respiratory Study, where they measured lung function in infancy at 1 to 2 months of age and followed the cohort for more than 20 years. And what they found is that those infants who had the lowest quartile of lung function had persistently reduced lung function, even after 20 years of follow up, and this was true regardless of the spirit metric measure that they looked at. And so then these data, taken together sort of allow us now to focus in on this early life period, perhaps even extending to the prenatal period to try to understand or examine different factors that may contribute to lung health. So we can think of genetics but obviously also think of environmental pollutants both chemical and non chemical nature. And of course, I've spoken mainly about long health. But I think we can see parallels if we look at cardiovascular development and growth. And we know from other research as well that there is an inter relatedness of organ system health. For example, we know that growth influences cardiovascular health and lung health, lung health influences, cardiovascular health and endocrine and metabolic and is also associated with mortality. So clearly, as described by Barker, we do see inter relatedness of different organs. So moving from here. Hopefully I can now introduce household air pollution and hopefully get you to understand that this really is globally important environmental exposure. Um, that plays a significant role in childhood health. So what is household air pollution? Um, well, 40% of households worldwide are reliant on the burning of solid fuels for their daily cooking and heating needs. And this is disproportionately so in sub Saharan Africa, where more than three quarters of households are reliant on these fuels, and partly due to sort of just the widespread use of these fuels. There's substantial morbidity and mortality on the order of 2.8 million deaths every year and 85.6 million dailies or disability adjusted life years. These are examples of two traditional cook stoves. This is a charcoal stove, and this is what's called a three stone fire, essentially just £3 of mud that are close in close proximity so that the pot can sit on top and then the solid fuel is pushed underneath. And these are combustion and efficient cook stoves, which essentially just means that they produce a lot of smoke. They produce a lot of toxicants at very high levels, and we know that these exposures are occurring over the life course and therefore associate with morbidity and mortality both in childhood and adulthood. These are some examples of solid fuels that are burned, so especially in agricultural communities following the harvest, you'll see crop residuals being burned charcoal as I mentioned to you before, particularly in areas that are near forest, you'll see would being burned areas at higher elevation or areas that have a lot of livestock will also burn, um, animal dung as well, which can be quite toxic. And if we think that, um, tobacco, which essentially is a biomass, um, produces a number of different health outcomes, You know, biomass smoke essentially has the same constituents with the exception of nicotine. And so really, any health outcome that you can think of that would be associated with tobacco smoke. Um, there really is beginning to be evidence that we can see the same health outcomes with household air pollution. I will say most of this work is cross sectional in nature or leveraging questionnaires, and therefore there really is a large literature gap data gap that needs to urgently be filled, particularly looking at exposure response relationships. And of course, when we think about household air pollution, the behavioral context is important. So we know that women are the primary cooks. They continue to cook while pregnant, and therefore we conceptualize that exposure to household air pollution really does begin in utero. We also know that Children frequently will stay, especially young Children will stay by their mother's side, Um, and therefore they also have high exposures. But that's not to say that others are not exposed. If you've ever been to one of these communities during cooking times, what you'll notice is that essentially there is a layer of fog that earlier of smoke. Excuse me that it just lays over the entire community. And so really, there is no one in the community that is unexposed. And so what sort of exposure levels are we talking about? So this is the data that was published almost 10 years ago now using satellite based models to try to understand exposures in New England. So this is sort of equivalent to what you or I may be breathing on a daily basis and, you know, just to point out here that the range is on the order of six to, let's say, 12 micrograms per meter cubed. So keeping that in mind, what sorts of exposures are we talking about with household air pollution? So we don't you know, with our studies, we don't have the luxury of using satellite based models. And so our participants all wear personal exposure monitors, and that means our exposure. With our exposure data, we're able to get real time measures, um, of their exposure. So that's what you're looking at here. So on the X axis is a time series starting at 16 and running through 19 at a little bit before 11. A little bit before 11 o'clock in the morning. And here on the Y axis is the exposure concentration. And I just want to point you out to the scale here. So we're talking in the thousands of micrograms per meter cube, so already just an exponentially higher exposure as compared to what you are. I see every day and you can see that these sort of peaks These are cooking periods, and during the cooking period, the exposures are routinely going up into the thousands of some spikes you know, reaching 6000 or even exceeding 6000. And in this particular example, the median exposure over a three day period was 125 micrograms per meter cubed. This is another example where again you can just see these clusters of exposures again reaching into the thousands and again here. The median exposure is 151.7. We also measured carbon monoxide is another product of combustion. So same graph here, where on the on the X axis is a running time series on the Y axis is our carbon monoxide in parts per million? And you can see that there's just these huge spikes in carbon monoxide exposure that correlate with cooking periods and are well in excess of the EPA guidelines. The degree of exposure is incredibly important, especially when we think about interventions that could be used to reduce exposure. So here's the exposure response curve, and you can see this is for pneumonia, and you can see that the exposure response curve is incredibly steep at lower exposures. But they have biggest plateau as you move into higher exposures. So you can imagine that if you have an intervention that, let's say, has a 50% reduction in exposure, you may still be operating on the flat portion of the curve. Um, and really, the question is, how do we introduce an exposure that's able to move you into the steep portion of the curve and move you below what currently is the W H O interim target for indoor air, which would be below 35 micrograms per meter cubed. And so this is what we've been attempting to do and study in our in our work in Ghana with the Ghana Randomized Air Pollution and Health Study. So in 2013, we recruited over 1400 pregnant women from Contempo, Ghana, and again Ghana's in West Africa, and Contempo is located in the middle belt, and we randomized these pregnant women to receive one of two cook stove interventions. The first thing a gas stove or LPG stove, which is similar to what you are. I'm a cook on at home, the second being an improved biomass stove or violate, which allows women to continue to burn solid fuels but has improved combustion efficiency, properties or control, which is a three stone fire. The inclusion criteria were that the women had to be less than 24 weeks gestation with a singleton fetus. They had to be all non smokers, and they had to be the head cook of a household. In addition, we performed repeated personal exposure assessments. This is both of the pregnant woman and then over the child's first year of life, both of the child and the mother. And we measured carbon oxide and PM 2.5, and we perform this four times prenatally and then three times over the first year of life and the main outcomes for the study, where birth rate and acute lower respiratory infection or pneumonia in the first year of life. But as I'll talk about, we also measured child Anthropocene matrix every three months, and current funding is allowing us to follow this cohort through child age seven years. So these are the stoves again. So this is the three stone fire, the traditional cook stove. This is the bio light stove, so they improved biomass stove. And this is the L. P. G. And it's just important to note that for the study we provided support for the stove. So we fixed any issues that may have arisen, Um, and also particularly for the l P G. R. And we provided an unlimited supply of fuel so that the cost of the L P G gas itself would not be a limitation to adherence to the stove birth outcomes were measured within the 1st 24 hours of life, and this was regardless of whether the births occurred at home or in the clinic. Our pneumonia surveillance or acute lower respiratory infection surveillance involved fieldworkers, who visited each house every week and followed the W H O integrated management of childhood illness. So essentially what this does is it looks for Children that are having difficulty breathing. It looks at their respiratory rate, um, and other symptoms, and essentially categorizes them into having pneumonia or severe pneumonia. We then took any child who was stiffer, unwell and referred them for physician evaluation. And again we provided support. So we provided transportation. All households had insurance through our study. Um, and then the physician evaluation was paid for by the study as well. So we did everything we could to try to promote getting Children who are identified by the field worker is sick to the clinician for a visit, and our primary outcome was physician diagnosed pneumonia or severe pneumonia in the first year of life. So this is the study flow where again we randomised 400 women. They were assigned to our three study arms. These pregnant women. These resulted in 1310 live births. Um, and you can see here, those Children that, uh, you know, unfortunately, did not make it through the first year of follow up. Um, but resulted in over are approximately 55,000 child weeks of field worker follow up. So what do we see with our intention to treat? Well, um, you know, unfortunately, with the intention to treat analyses, we did not see a statistical reduction, um, in birth weight, which was our main birth outcome between arms as compared to control. When also, when we looked at our pneumonia outcome, Um, you know, we similarly did not see a reduction in either physician diagnosed pneumonia or physician diagnosed severe pneumonia. I think this is probably one of those cases where you would look and you would say, Is that one line or three? Um, in the blown up version here we do see a little bit of separation, but the, uh that was not statistically significant. And so why might that have been? Well, you know, our hypothesis is that although rlp grom did have a significant reduction in exposure, um you know, unfortunately, we still did have a lot of overlap in exposure between these arms and further when we look at the average exposure in the L. P G and we see that we did not reach that WHL threshold of 35 micrograms per meter cubed. So perhaps a combination of our exposures were still too high and we had too much overlap between study arms. So what about exposure response analyses? So, as I said, we measured exposure at seven time points for prenatally and three over the Children's first year of life. These are exposure response analyses, looking at the relationship between prenatal and postnatal, household air pollution and pneumonia. So this includes infants who completed one year. Follow up and you can see we have almost 55,000 child weeks of field worker follow up. Um, in total, we had 392 episodes of physician diagnosed pneumonia and 100 and 23 episodes of physician diagnosed severe pneumonia in our cohort. And when we look at the relationship between prenatal and post natal household air pollution, exposure on pneumonia and severe pneumonia, what we see is that it appears that prenatal exposure is associated with an increased risk for the development of pneumonia and severe pneumonia in the first year of life. And when we look at sex stratified analyses, it appears that female infants may be more vulnerable to these effects. And again, these results are all per one ppm increase in carbon monoxide. So these are exposure response analyses. We then wondered, Well, if we see this effect of prenatal exposure on pneumonia, could we identify a sensitive window of exposure? And so when we look at our exposure protocol, what we see is that the exposures were based on gestational age and enrollment. So we had one assessment done prior to the intervention and then three spaced equally between intervention and birth. So these exposure assessments were not done based on gestational age, but rather based on when the women were recruited. And this is just a history Graham showing you sort of the spread of when we performed exposure measurements in the cohort. So here the Y axis is the count of the number of exposures that we had taken at each week. Gestation. As you can see, we have pretty good coverage starting at seven weeks gestation. And so we can use this data in a distributed lag model framework to try to understand the time varying association between increased prenatal household air pollution exposure as index by carbon monoxide and odds for developing severe pneumonia or pneumonia in the first year of life. So here on the X axis, we have gestational age in weeks, the Y axis. We have the odds ratio for developing severe pneumonia, and what you can see here is that when our odds ratio and confidence interval and in the bracket here is a bond Peroni corrected odds ratio or sorry confidence Interval. What you can see is that we identify a sensitive window where those confidence intervals do not include one suggesting perhaps that increased exposure in late gestation is associated with increased risk for pneumonia. Um, and we can see a similar thing that it's hard to see, but this is statistically significant where we see that in late gestation, there appears to be an association between increased exposure to CEO Um, and, uh, this is position diagnosed pneumonia and again, these are per one ppm increase or sorry, this is per one unit log increase in prenatal CEO exposure. So I think we can now add to our results and say, Although we didn't see an i t. T effect, we do see that increased prenatal household air pollution exposures associated with increased risk for pneumonia in the first year of life and that girls may be more vulnerable. We also within a subset of our cohort, performed infant lung function at uh, at one month of age. This was only in 400 mother infant pairs. And of course, you can perform spectrometry and infants. So we did title breathing analyses where essentially you can get an extra Tory flow parameter, which is the time to peak title exploratory flow to exploratory time. You can also understand title volume, respiratory rate and then in ventilation. And what we found is that there was an exposure response association essentially, that increased prenatal exposure was associated with impairment and excretory flow and an increase in respiratory rate and then the ventilation. And again, these are per one ppm increase in average prenatal CEO exposure. And again, when we looked at sex stratified analyses, we see that girls appear to be more vulnerable to these effects. We then wondered, Could these changes in respiratory mechanics influence risk for future pneumonia. And in fact, when we looked, we did see a very small but statistically significant increase in the odds of developing pneumonia again. Here's per one unit increase in respiratory rate so we can now add to these findings and say, um, that increased prenatal household air pollution exposure may also be associated with imperative than lung function. That's measurable at 30 days of life. And again we see that girls may be more vulnerable to these effects. As I mentioned, we measured, um, anthropoid metrics in the Children every three months. Um, so over the first year of life, we have five potential measurements at 0369 and 12 months of age. I will say that data completeness for these measurements was very high. So over 77 or sorry, over 76% of the cohort had all time points measured. Um, and an additional 14% had, um, these anthropology tricks measured at four time points. In these analyses, we did start to see a suggestion that perhaps our LPG arm had effects outside of exposure reduction. So here we see that those women randomized to the L. P G arm. We're essentially all continuing to breastfeed their Children at 12 months of age, whereas in the other two arms that that number was closer to around 80%. So we leverage these repeated anthropology tricks to develop trajectories of growth. And what you see here are the trajectories that we identified for weight length, head circumference and mid upper arm circumference. And I think that's an important observation is that we see that these trajectories appear to be distinct at birth. Although you do note some changes in slope in the first six months of life, we also use W H O Z scores to calculate height for age Z score, wait for HD score and weight for height Z score. And you know, I think the important observation here is really that the cohort is rather a frail cohort. So, you know, in the United States, um, and in other developed countries, for example, you would expect to see that Children are overweight. Um, and have you know, for example, Z scores greater than two? Here we see just an overall shift of Z scores down so that we have you know, a group for each D score that is just persistently below minus two. So, for example, 10% of our population is stunted and another 44% of our population is just persistently at risk for being stunted. And the same is true if we look at underweight and being at risk for underweight and wasted and at risk for wasting so again, just very what I would characterize his overall, just a very frail cohort. So we then used orginal logistic regression to try to understand whether prenatal and postnatal household air pollution exposure were associated with these trajectories. And what we found is that increased prenatal household air pollution exposure, increased risk for lower or smaller height, height for age Z score. And perhaps you know, the factor with the most public health significance is stunting such that a one ppm increase in prenatal household air pollution exposure as index by carbon oxide was associated with an odds ratio of 1.25 When we looked at post natal household air pollution exposure, we saw that increased risk or increased exposure was associated with increased risk for lower or smaller head circumference, mid upper arm circumference and weight for height, Z score or wasting and again here, the CEO is on one ppm increase and the PM 2.5 is per 10 microgram per meter cubed increase. We again looked at sex specific effects, and here the results appear to be to vary based on pollutants. So girls appeared more vulnerable to the effects of PM 2.5 and boys appeared more vulnerable to the effects of CEO. We then asked, Well, if we see an exposure response relationship, might our cook stove intervention be able to ameliorate some of these effects? And here you can see that we don't see any significant findings to the bio light arm. But in the l p g r. We do see that those infants born to mothers randomized to the L. P G R have reduced odds for the development of smaller head circumference and smaller mid upper arm circumference as compared to control. And so we can now add to these findings to say that prenatal and postnatal household air pollution exposure, um, effectively impaired nearly all metrics of growth and increased risk for stunting. And perhaps we're starting to see that the L. P G cook stove can ameliorate some of these effects. So, as I mentioned before, were now funded to follow the cohort through child age seven to look at lung development. So we have lung genotyping visits at age four and seven, and then we perform repeated personal exposure assessments and also continuous community level and central site monitoring. And so I'll share with you some of the results that we have thus far from our age four visit, which has been completed. So for lung function and four year old, clearly you cannot do spectrometry or at least you cannot do. It cannot do it Well, I would say, um so we are employing impulsive Salama Tree, which is an effort independent measure that just requires title breathing to assess lung function of child age four. And so here we're using generalized estimating equations to estimate the associations between prenatal and postnatal household air pollution exposures and lung function as measured by impulses. Salama Tree And what we see is that if we look at all exposures together, increased prenatal and post natal household air pollution exposure is associated with an increase in airway resistance, specifically at our five hertz, 20 hertz and then the area of reactivates. If we look at Prenatal and Post Natal, what we see is that the effect, while present in both perhaps maybe stronger in the prenatal period. So again suggesting an increase in airway resistance, um, measurable at age four years. So we can add to these findings by now saying, um that in addition, early life, household air pollution exposure appears to impair lung function, measurable not only at birth but also measurable at age four years. We then looked at whether poor childhood growth, which again we previously demonstrated, was associated with household air pollution. Exposure is associated with lung function at age four. So we started out by trying to understand whether birth entropy metrics may be associated with lung function. And again we see the same pattern. We see that those Children who have smaller or lower birth weight have higher airway resistance. We see an inverse relationship. Um, we also looked at measurements at age four, specifically, R Z scores, so wait for HD score and height for HD score. And again we see the same pattern. An inverse association where those Children who have poorer or smaller wait for HD score height for HD score have increased airway resistance both at R five and R 20 and then here for the height for age Z score. We're starting to see perhaps an effect on the small airways with an increase in airway resistance, with reductions in height for age. The score we then similar to how we did before we created growth trajectories. So using our measurements at every three months through 12 months of age and also extending this to include the measurement at 48 months and again, we see that our trajectories are distinct and essentially parallel are the Maybe you can say that there's a little bit of catch up growth happening here, but the same pattern is true. We're now we have 13% of our cohort that is persistently wasted and 39% of our cohort that's at risk for wasting. If we look at height for age D score, we see that 9% of our cohort is persistently stunted, and 46% of our cohort is at risk for stunting. So still, this clearly just very frail cohort. So if we look at associations between the Z scores and lung function. So to orient you here, each of these plots is our lung function measurements. So here we have our five x five are 20 and then the difference that are 5 to 20. And here we're looking within sort of each Z scorched trajectory. Uh, metrics. So this is, uh, in reference to a normal trajectory. And then we have our normal low at risk and stunted, and what you can see is the same pattern is emerging here as well. So if we see, uh, that Children who are stunted or at risk for something appear to have an increase in our five and so increase in resistance resistance as compared to those with a normal height for age, the score trajectory and the same is true for our 20. Those Children who are stunted appear to have increased resistance at 20 hertz, as compared to those Children who have a normal height for age Z score. And if we look at trends, these trends are significant as well. We can now add to this list and say not only impair growth in the first year of life. Um, sorry at birth, but also at age four and also from birth through age for appear to be associated with changes in lung function. Specifically, with this pattern of increased airway resistance, we've started to look and this is preliminary data that we used for an R O N application that will be resubmitting in November. But we started to look at the association between early life, household air pollution exposure and blood pressure. So this is just resting systolic blood pressure, so measured after 10 minutes of rest, measured three times an average. And what you can see is that we see a dose response effect where increasing early life PM 2.5 exposure is associated with an increase in systolic blood pressure to the order of 1.37 millimeters mercury per 10 micrograms per meter cubed increase in PM 2.5. What's more, if we try to understand whether or not our cook stove intervention is able to have a lasting effect on blood pressure, we actually see that those infants born to mothers are those Children born to mothers randomized to the L P G arm appear to have over a seven point reduction in systolic blood pressure at age four years as compared to control again. These are just the 1st 80 or so Children. And so we're working on extending these analyses to the entire cohort. Um, and those analyses are underway, and again we see a similar pattern in the larger cohort. So we're now leveraging that pilot data are preliminary data, um, to think about ways that we can better characterize cardiovascular health and Children. And so we're aiming to do this by using ambulatory blood pressure monitoring. So here you can see in a pilot study, we have our Children wearing their ambulatory blood pressure monitors. We also have a study where we're doing at ticket fee. So trying to understand sleep patterns during the day and wake cycles. This is our air pollution monitor. Um, so our PM 2.5 monitor, Um and then the Children are wearing a hex, a skin shirt so we can get heart rate variability. So a lot of equipment, but overall, this was very well tolerated. And actually the siblings of the Children who were wearing this for our pilot study, um, we're actually all wanted to wear the equipment as well, and we're sort of jealous that they couldn't do it. So we have, um you know, high hopes we have a lot of experience deploying, um, you know, exposure devices and Children. So we have high hopes that we can do this successfully in our cohort. And this is just an example of what the data stream looks like. So here's Artie graffiti data. Um, it's just where you can sort of beautifully. See, this is daytime, and this is nighttime data. This is our inventory blood pressure measurements. So you can see the repeated measures of systolic and diastolic blood pressure just over time again. Here, the X axis is time for our measuring period and then in green here at the bottom. Um, these are our exposure assessments. So again, you can see that there's these spikes and exposure that are occurring at dinner time, Um, and also at breakfast time. So I think you know, the final piece of evidence that will say is that perhaps and again, this is something that we're hoping to explore more. But perhaps early life household air pollution exposure. And here again, we're looking at PM 2.5 is associated with increased risk for systolic blood pressure. at age four. So I think, you know, for about trying to summarize and bring all this data together, I think what we can say is that early life, household air pollution exposure, beginning prenatally clearly affects child health. And we see that prenatal exposure influences lung function both measurable at one month of life and the measurable again at four years of life. It appears that this poor lung function then sets the stage perhaps for increased risk for pneumonia in childhood, which we know is a leading killer of Children under the age of five and perhaps may play into a feedback loop, then further impairing lung function, which again we can measure at age four. We also see that prenatal and postnatal household air pollution, um, impairs nearly all metrics of growth. Um, and this is really important if we think about childhood mortality, especially with the fact that we do see a strong association with an increased risk for stunting, which again has public health relevance. And also, you know, we see that these growth metrics then appear to be related to one function where again we see that poor growth translates to increased airway resistance again measurable at age four. And finally, you know, we're starting to see some sort of signal that perhaps cardiovascular health and childhood could also be associated with early life household air pollution exposure. So again, really, just the story of early life health in a related nous of organ system development, really just starting to, uh, set the stage for health disparities over the Life Board. And so, you know, I've I've our work really focuses on household air pollution. But clearly there's many different factors at play here that influence health. And I will say, you know, our intervention are LPG. Intervention was not able to demonstrate a significant improvement in birth way or pneumonia over the first year of life. And I think, you know, I don't think it's that controversial to say that, you know, at this point, I think we can say yes, we need an intervention that's able to reduce exposures more. Yes, we need an intervention that's able to create more of, um uh, more of a gap between interventions between study yards. But I do think that we need to start thinking about how we address the elephant in the room, which is poverty, because I think until we start to think about ways to address poverty, we really will not be successful at trying to alleviate these health disparities. And, you know, my my sort of final thought on this is I think, you know, especially in the age of covid, where we're seeing that this is especially true, you know, in our population that we serve here at Mount Sinai, I think that we have to remember that economic inequality and poverty is an issue not just globally and for people living in Africa, but also for people. Um, here, living with us in New York City. And so with that, I just want to thank all of the fabulous collaborators. Clearly, this is the work of many people and could not be done, especially without the collaboration of our colleagues in Ghana who lied. Just a really superb research team. Um, and who, you know for whom? None of this work would be possible without their support, Um, and collaboration. And then, of course, important to acknowledge our funders. So thank you so much for your attention. Um, and I'm happy to take any questions that you may have great. Thank you, Dr Lee. So, uh, if you have questions or comments, please put them in the chats of discussion. Uh, Alison D c sections question do so Yeah, so? So she's asking. That's a great question. So we I think this is something that, you know we're struggling with right now with all research studies. Um, and especially in Ghana, where there really is no testing capacity. So initially in the country, they were doing testing. But at this point, they've essentially taken the view that they cannot do widespread testing and really in the area that we're working in, they're not testing. So we threw our study. We have deployed every three months. We use the W, H O and MSF covid questionnaire to try to ask you the pain symptoms in the community. Um, and we're also trying to get some additional funding right now to be able to do PCR swabs, um, to test for covid at the start of our age seven assessment, which was slated to begin now but has been pushed back to January. So we're trying to take, you know, different precautions, and obviously also make sure that our staff have the proper PPE, um, to protect themselves because I think especially working in a cohort were testing is limited. Um, it's it's difficult, but yeah, that's a great question. Something we've been grappling with. Has anyone compared curve oxy? Hemoglobin in mothers and newborns are looked at less than 2.5 microns. Particles and placentas. Yeah. So the placenta piece. So So, yes. People have looked at carb, oxy, hemoglobin, and there was an effort a number of years ago to try to use that as a biomarker of exposure. Um, but those those didn't really pan out in terms of placenta. I mean, that data is really interesting. So So some of you may be aware, but essentially, there's, um, data now that came out of the environment cohort, um, in Europe, which essentially found that mothers, when they looked at placental samples, that they found PM 2.5 in placental samples across essentially every single range of maternal exposure. So suggesting that even at very low levels of exposure, PM 2.5 is able to be absorbed into the bloodstream, right and reach the placenta. So that's, uh, we definitely thought of that. Cannot be used as a biomarker cannot be used to sort of look at the impact, right. We know that the placenta is very important and sort of driving fetal growth. So we have placental samples for our entire cohort. And, you know, it's it's something that we're definitely in discussions to try to find some funding to do that because I think that would be, um, really, really interesting in a way to kind of push the field forward. Alison is a couple of questions. Yeah. So Rachel is asking sufficient variation to access what may help with recovery of growth parameters. Yes, we do see variation in post Natal exposures. Um, as I showed in the growth trajectory, we don't really see a growth recovery. I mean, maybe we see a little bit of a change in slope between 12 months to, let's say, 48 months, but we don't see, um, a trajectory that has, like, a rapid rate of improvement, similar to if you remember back to the slide with the large cohort of FV one trajectories where you do see there was one trajectory that had an improvement. We don't see that, you know, like just be limited by sample size. So for our current work, we're following 700 mother infant pairs or mother child pairs. Um, so it may just be that our sample size is not large enough to give us, you know, multiple different, Uh, something the types, but it's a really It's a good question. Um, and something that we should definitely look more into comparative data on PM 2.5 in New York, especially in congested areas. Yeah. So the other piece of my work my K 23 then actually also my current are a one. Um, is working within Doctor Rosslyn writes cohort, which is a birth cohort here in New York City, where we're trying to understand the influence of prenatal and early life ambient air pollution specific PM 2.5 on the same metrics. So my K is actually measuring lung function at age four. Um, and so we are in the process of analyzing those data, but we don't have that specifically yet, but, um, if you look at the work published by her group and also Rachel, you're on the call published by your group as well. I mean, clearly there's a lot of data out there suggesting that prenatal and early childhood PM 2.5 exposures influence child health over the life course. Is there any difference chemically in food cooked on wood stoves versus gas stoves? Yeah. I mean, I think so. The really the exposure piece. Um, I think we have a lot left to explore there. So, as I said, we're measuring carbon monoxide and PM 2.5. Um, you know, we have urine banks so potentially we could look in the urine for their markers, like phs, um, and other market exposure there. Uh, you know, we have filters banked for all of our PM 2.5 exposure assessments. So I think, you know, there still is a lot that we could potentially do with the exposure piece and try to understand how those differences may play a role in health. Obviously, a lot of those are sort of expensive analyses, and we're limited by funding. But I think that's that's a great point that, um, that some of those differences as well and sort of what you eat can play a role, especially if you want to think about childhood growth. Francine, allow people to a nude If you have a question that you'd like to ask. Yeah, Okay. Well, if there's no other questions, really want to thank Dr Lee for a really terrific topic today in great work. Really? Impressive, Alison, to thank you very much for this and thanks everyone for joining us today. We'll see you all next week. Have a good one. Thank you so much. Thanks. Allison. Congratulations. Thank you.
