This Presentation will explore the patterns of stillbirth across maternal race, ethnicity, and age in recent US data. Gaps in research and opportunities to explore proposed causal linkages will be discussed.
Anne E. Brisendine, DrPH MPH CHES is an Assistant Professor of Public Health Practice in the Department of Health Care Organization and Policy at the University of Alabama at Birmingham (UAB) School of Public Health, and she serves as the Science Director of the Applied Evaluation and Assessment Center. She received her DrPH and MPH degrees from the UAB School of Public Health before completing a Postdoctoral Fellowship at the UAB Civitan-Sparks Clinics and Department of Psychology. She received her BA in American Studies from Barnard College, Columbia University. Dr. Brisendine has a background in interdisciplinary research, strategic management, and publishing, which allows her to be an effective communicator of research and policy to a variety of audiences. Her research and scholarly practice focus on a variety of topics including autism spectrum disorders, developmental monitoring in Early Head Starts, statewide systems development in education and health, early childhood mental health, and perinatal epidemiology.
Dr. Warland has disclosed that she does not have any real or perceived conflicts of interest in making this presentation.
Thanh Dickey: Dr. Anne E. Brisendine is an Assistant Professor of Public Health Practice in the Department of Health Care Organization and Policy at the University of Alabama at Birmingham School of Public Health, and she serves as the Science Director of the Applied Evaluation and Assessment Center. She received her Doctorate of Public Health and masters in public health degrees from the UAB School of Public Health before completing a Postdoctoral Fellowship at the UAB Civitan-Sparks Clinics and Department of Psychology. She received her bachelor of arts in American Studies from Barnard College, Columbia University.
Dr. Brisendine has a background in interdisciplinary research, strategic management, and publishing, which allows her to be an effective communicator of research and policy to a variety of audiences. Her research and scholarly practice focus on a variety of topics including autism spectrum disorders, developmental monitoring in early head starts, statewide systems development in education and health, early childhood mental health, and perinatal epidemiology. Dr. Brisendine’s presentation is titled, Racial Disparities, Maternal Age, and Stillbirth.
Dr. Anne E. Brisendine: Hey, y’all. My name is Anne Brisendine. I’m an Assistant Professor in the Department of Health Care Organization and Policy here at the UAB School of Public Health in Birmingham, Alabama, and I am incredibly grateful for the opportunity to be able to present some of the work I’ve been doing over the past few years on racial disparities, maternal age, and stillbirth.
I think first off, I’d love to set the scene and set the stage by developing some common language, knowing that it is important that we know what we’re talking about when we’re using some of these terms, because when it comes to clinical work and policy or population health-based work sometimes we can be saying the same thing and mean two different things. I always try to work from a foundation. This may be a review for a lot of folks and for some folks it may be entirely new, so bear with me.
We know that a lot of this is jargon and imprecise terms and everything else in between. These are just a couple of things I want to make sure we’ve got a working knowledge of before we get into some of this work.
First, I want to talk about public health. If I had been doing this presentation two, three years ago, there would probably be more folks who would ask if by doing public health I inspect restaurants. But, obviously, with the pandemic, more folks are much more intimately familiar with the work of public health. Right here we’ve got our 10 core functions. We are so much more than epidemiology. We are doing assessment and monitoring, but also communication and building access, and all of this founded on the centrality of equity.
With that, a lot of the conversations we have are around social determinants of health. This is a buzzword that gets thrown around a lot, but basically what it means is that there’s only so much that can be determined with our genetic code. There’s only so much with biological inheritance and environmental exposures. That’s where we’re getting into some of these social determinants of health.
You’ll hear the term; your zip code matters more than your genetic code. That’s some of the things that we’re operating from when it comes to public health. It is those conditions where people are born into, where they grow, they live, they work, and they age. That includes the health system, but it is much more broad than that. We also acknowledge that these circumstances are shaped by the distribution of money, power, and resources at the global, national, and local level. Those are influenced by policy choices on top of all of that.
We talk a lot about health disparities. That is one of those things where we see in the society that we live in, in the United States, that there are differences in health status between different subgroups of the population, primarily between different racial and ethnic groups. Oftentimes we talk about differential access, but it’s so much more than that.
We’re always pushing to move towards a culture of health equity. I love what the Vermont Department of Health puts forth as their thought around health equity. That is that health equity exists when all people have a fair and just opportunity to be healthy, especially those who have experienced socioeconomic disadvantage, historical injustice, and other avoidable systemic inequalities that are often associated with categories of race, gender, ethnicity, social position, sexual orientation, and disability.
We use this graph on the left-hand side of your screen and you’ll see three people reaching up to get some fruit from a tree. Where we are is far to the left. People are coming from different levels of that stepping stone. Even if everyone is the same height, same color, same person, if you’re starting on a lower rung of that ladder, it’s going to be significantly harder if not impossible to reach that fruit. It doesn’t mean you can’t jump, but it’s going to be that much more effort.
We move to the center, we see that we’re then adding boxes to lift folks up to be able to reach that fruit. Some people may need two, some people may need three. If we were just giving everyone one box, it still would be harder for some folks. It’s really tuning that to meet the needs of the populations that we’re working with. Then finally, on the right-hand side of that graphic, you see that everyone is starting from the same ground level. That is our ultimate goal is that everyone has the same opportunity. That it’s not just equally distributed, but it’s equitably distributed.
Then I want to move into talking about the life course perspective. Now, those of us who work in maternal and child health, this is a fairly new conceptual framework, but it’s looking at life, not just as disconnected stages, like the perinatal period or the toddler period, adolescence. Really is an integrated continuum that cuts across generations. We’re recognizing that each stage in life is influenced by all the life stages before it, and that they influence the life stages that follow it.
It’s important to examine exposures during a critical period like pregnancy, but that also doesn’t fully explain why such disparities are existing between these racial and ethnic groups, because that nine months is only a snapshot in time. It is containing just a thin slice of everything that has happened in a woman’s life up until that period. When we look at the life course perspective, we’re trying to understand exposures throughout the life course.
Part of that is the cumulative pathway mechanism. That’s looking at how wear and tear can add up over the lifetime to affect a person’s health and function. We often talk about the allostatic load. That’s going to be the stress hormone burden that people carry, and that can come from physical and social environments. It’s not just a person’s BMI, it’s not just the type of food a person eats, or if they live close to an environmental hazard. It also is part of that social experience. The body’s designed to maintain that balance and stability.
If your body can’t keep that balance or your internal thermostat is ratcheted up too high because you’re living in an experience where there may be higher levels of violence, or there may be experiences of racism that someone is having to overcome, that can put someone’s stress perpetually at a higher level, and you never get to come back down to catch your breath, for example.
That brings us to the weathering hypothesis, which is really the groundwork for a lot of the work that I have been doing over the past years. That comes from Arline Geronimus. She put together this framework to look at the relationship between social inequity and health outcomes in population groups and also how that comes across age.
When we’re looking at age, we’re not just looking at the number of years a person has been alive. We’re not looking at just that measure of development. It’s really a synergistic indicator of that cumulative impact of social inequality. It’s a physical manifestation in deterioration stemming from the effects of racial discrimination, political marginalization, and other biases that leads to that differing risks of exposures across and within population groups.
It’s part of that same group of theories where we’re talking about intersection, for example. Where it’s not just a woman having differential exposures to stresses and risk factors because she’s a woman, but you add to that a woman’s racial or ethnicity that might currently in our system have a lower sense of power. It might be politically marginalized. Those things together essentially biologically and genetically cause a woman to age more quickly than a woman from a different racial or ethnic group.
For example, we think of the life course. With a graph, on your bottom X axis, we’re looking at the life course. Zero, five years, puberty, pregnancy, and then for lack of a better term reproductive potential on your Y axis going up the side. If you look at these two curves, and these are not founded in specific data, but this is a theoretical model. You see that reproductive potential for a white woman, they’ve got more protective factors pushing that potential up higher. That could be socioeconomic status, that could be access to healthcare. That doesn’t mean that there aren’t risk factors, but they’re just fewer. They’re outweighed by those protective factors. Again, on average.
When you look at the African American women’s curve at the bottom, again, on average, we’re seeing more risk factors and that can come from race and racism, that can come from differing access to healthcare and quality of healthcare, and other access to the ability to exercise and other health behaviors that may or may not be available to people depending on where they live, their type of work, all of those sorts of things.
Then finally, I want to talk about administrative data. I deal with administrative data, secondary data, and for the work that I do, I deal a lot with vital statistics. That’s the national vital statistics system, which is coming from the Centers for Disease and Control and Prevention. They call it the fetal death and perinatal mortality files. I, like a lot of folks, prefer to deal on the language of stillbirth, but when we’re looking at the administrative data, it’s fetal death because it is distance, it is objective, it is clinical, it is standardized. We are looking at 88 pages of just how to enter this data into the computer if you’re a clinician gathering some of this information.
It collects a lot of information, but there’s still a lot more information that could be gathered. We know that every situation is unique, stillbirth is common, but it’s still really uncommon compared to a lot of other conditions and outcomes, so we’re limiting our sample size. Then it allows us to speak in averages, but the downside of speaking in averages instead of being able to look at the population as a whole means that we are smoothing out a lot of the nuance. When I am teaching class or dealing with some of our students, I might have an older Black woman say to me, “I’m hearing all of these things. They’re telling me that I shouldn’t have kids, that I was wrong to have gone into graduate school, and that by delaying reproduction and having kids, that I am destined to have a terrible outcome.”
It can look like that from the data a lot of times. I can’t blame anyone from saying that, but again, averages are not destiny. We have to say to folks, “Being in graduate school, living where we do, these sorts of things, that can be some of those protected factors.” That’s what I really hope that folks can keep in mind as we move through some of this data that just because it might say that a woman in her late 30s who is African American has a significantly higher risk of stillbirth, that doesn’t mean that that is the absolute that’s going to happen in every case.
Talking about stillbirth, I think it’s also interesting to note that we’ve got one way of categorizing it at a national level, but still even if we’re looking at countries and states, and all of these different ways that people are categorizing them, in the five years between 2009 and 2014, there were 81 different systems that were either new or modified. Again, that’s global, but some of that’s also within the United States. We’ve got this standardized data, but it is only standardized to a certain point.
You’ll also see as we move through some of this that there’s a lot of other questions that we wish we could ask of this data that we’re not necessarily able to gather or to use in the same way that a clinician with a smaller, more focused study might be able to do. There comes that trade off. We’re able to talk at a national level with this data, but we’re not able to get into some of the more specific cases that we might want to be able to.
This study. A little bit of background. It’s not common that we hear fetal death or stillbirth used as a public health metric. We often hear infant mortality or maternal mortality and morbidity. That being said, we still know that there are vast disparities in the occurrence of stillbirth for non-Hispanic Black mothers compared to non-Hispanic white mothers. We see some similar, those smaller scale findings for Hispanic women. Looking at that though, again, we’re taking small sample sizes and trying to make them even smaller there. As we combine years and as we look at more robust data systems, I wouldn’t be surprised if we found similar baseline information for Hispanic women as well.
Stillbirth disparity is the greatest at the earliest end of the reporting window. That’s going to be that 20 to 23 weeks of gestation. We’re going to see in some data from 10 to 15 years ago, almost a threefold increased risk for non-Hispanic Black women compared to non-Hispanic white women overall on average across all ages.
We know that there are certain risk factors. Advanced maternal age, 35 and up, as I approach that age, it’s hard to see the term geriatric pregnancy and not get a little upset, but advanced maternal age and advancing maternal age is a risk factor. Uncontrolled diabetes, hypertension, obesity. We also know multiple gestations or pregnancy to primiparous women. Alcohol use and smoking during pregnancy is also a risk factor for stillbirth, and as I said, we also know that non-Hispanic Black mothers experience higher rates of stillbirth relative to non-Hispanic white women. There’s a lot of data around this. There’s a lot of research that has been done as you can see right here.
That being said, again, weathering has focused on its application to infant outcomes in a lot of respect, it hasn’t been applied to stillbirth frequently or at all. Given that studies have consistently demonstrated an increased risk of stillbirth with increasing age, as well as racial disparities, we wanted to see how that holds up or doesn’t when we look at the disparity in stillbirth rates between racial and ethnic groups as women age. Looking at those two variables together.
How did we do that? Again, we used these National Center for Health Statistics files. We looked at the live birth and fetal death record for the most recent four years that are available. Again, bringing those years together to be able to have a large enough sample size to really be able to ask some of these questions. We selected for singleton deliveries knowing that when you get into multiple deliveries, there’s just some unique characteristics in there that it’s just harder to measure together. We looked at maternal age. Less than 20 years of age, 20 to 24, 25 to 29, and so on all the way up to 40 and older. Then we looked at maternal race, we did non-Hispanic white and non-Hispanic black. We just simply did not have ability and the sample size to be able to ask some of these questions with enough power of other racial and ethnic groups.
First, we calculated crude stillbirth rates, so not adjusting for anything. For maternal age groups within racial groups, for early, late, and combined early and late. Looking at that 20 to 27 weeks, that early end of the stillbirth spectrum, as well as late, so 28 weeks and later. Then we did a multivariate logistic regression. Calculating those odd ratios for stillbirth within each racial and ethnic group, and then across age categories. We were able to adjust for diabetes, hypertensive disorders, both chronic, so existing before the pregnancy, as well as pregnancy induced hypertension. Getting into some of that. Preeclampsia, and then parity.
The results. When we look at combined stillbirth rates across maternal age groups, the thing that I want you to see from this graph and then from the subsequent graphs is that, first of all, we’re looking at that rate on the Y axis going from zero to 12, and then the green line right there on the top is non-Hispanic Black, and the blue line on the bottom represents non-Hispanic white stillbirth rates.
As we go from the left to the right, less than 20 years of age, all the way up to 40 years and older, we see a slight dip, slightly higher in those younger ages, and then getting less of a risk as we move across both for non-Hispanic black and non-Hispanic white women, but consistent across this entire gradient that green line is higher. Almost a twofold increase in risk for non-Hispanic Black women relative to non-Hispanic white women.
The other thing I want you to see here is that the gap widens with age. You see that in those first two age categories, as those rates drop a little bit, it drops faster for white women, it drops further for white women, and doesn’t come up again as high as it does for non-Hispanic Black women on average. When breaking this apart, we see that the same pattern holds for early stillbirth as well as for late stillbirth.
Moving into our regressions is really where we start to take all of this together and be able to see some of these things. Looking at this table, you’ll see maternal age on your left-hand side, and then those three blocks combined stillbirth, early and late stillbirth. Here, we’re looking at those adjusted odds ratios within racial and ethnic groups across age categories.
Among those bigger chunks, you’ll see the left-hand side in each one is non-Hispanic white women, and on the right-hand side, non-Hispanic Black women. With those pink boxes, what you’ll see are the statistically significant differences. We’re comparing these maternal age categories to those younger women. The women under the age of 20 who are delivering. For non-Hispanic white women throughout combined early and late stillbirth, between the ages of 20 and 35, we’re not seeing much difference in risk. It is not statistically significant. We do see that increase happening around the age of 35 and older. At that point, we do start seeing an increased odds of stillbirth relative to those women under the age of 20.
When we move to the columns for non-Hispanic Black women, however, immediately starting at age 20, we are seeing higher odds of stillbirth relative to those women who were under the age of 20 both combined and late. Then for early stillbirth, we’re seeing that trend start at the ages of 25 and older.
You can see that the age at which some of these age-related stillbirth risks may be coming into play for non-Hispanic Black women are younger than they are for non-Hispanic white women. We’re getting into that conversation about weathering. Again, we’re adjusting here for hypertensive disorders, for diabetes, for parity. We’re holding all of the other stuff constant. Looking at women almost in a one to one category. As they age, we are seeing much different patterns of risk as women age dependent on how they’re classified racially.
Here we are taking within age categories and comparing non-Hispanic Black women to non-Hispanic white women. You’ll note here that there aren’t any white boxes or any pink boxes, I should say, on this graph as we are looking at maternal age and our different categories of stillbirth. That’s because every single number on here is a significantly higher risk for non-Hispanic Black women relative to non-Hispanic white women.
Looking at these odds, you are seeing significantly higher odds in every single category at every single age. You’ll also see that that risk and that those odds get higher all the way up through about ages 30 to 34, and they plateau there a little bit, and then they appear to somewhat be dropping off at ages 40 and above. What we may be seeing there is not a consistent pattern when it comes to weathering.
Conclusions. Wrapping things up. As we said, weathering can be thought of as premature aging related to changes in physiology, biological functioning brought about by prolonged levels of stress leading to poor health outcomes. Arline Geronimus, who I mentioned earlier who has been doing this work on weathering since the early 1990s, this has moved from a theoretical framework for her to more recently when she’s actually using biological samples and genetic sequencing to look at the length of telomeres and genetic age and stress hormones, and all of these things that I as the number cruncher and not as a clinician don’t fully understand, but it holds up.
With this, we’re also seeing that to some extent it holds up among fetal death and stillbirth as well. However, the pattern is not consistent as age increases to 40 years and beyond. There’s that general pattern with an increase in the disparity as age increases, so that final table that I showed, but that really only is up to that age category between the ages of 30, 39. We’re seeing it in 30 to 34, 35 to 39, depending on the category that we’re looking at, but then it begins to get closer again.
As with any study, there are limitations. We can’t definitively point to a causal link for these disparities. As research is getting more sophisticated and more moving beyond the theoretical standpoint, as I mentioned, there’s some of these things we are unable to get some biological and genetic data to feed into these questions. But again, with a common uncommon outcome like stillbirth, it’s really hard to find a large enough sample to be able to truly and thoroughly ask these questions. That’s the hard thing about this is that stillbirth is far too common yet at times it’s not necessarily common enough to be able to ask some of these questions.
Our findings are beginning to lend support for associating outcomes. For these adverse outcomes with race and racism. This is a conversation that has been happening in public health for a long time and is becoming part of the mainstream conversation, but there’s so much more that we need to push for and that we need to be looking at when it comes to collecting data, when it comes to analyzing data, and when it comes to translating some of this theory into some of this applied research and practice. With this study, however, we just simply weren’t able to account for some of these direct measures of biological aging or stress, because that’s just not part of this data.
The other thing is data reporting challenges. There is different state to state reporting requirements as I mentioned earlier. This is the broadest brush that we can get and that common denominator across all of these states. That there are some nuances, things change year to year, and so this is the best that we’ve got and it’s good, but it can always be better.
Again, finally, that lack of information regarding context. There are so many nuances in every instance of stillbirth. When it’s distilled down to this objective distanced data, you’re missing a lot of those information.
Public health implications. These findings raise questions about the medical conditions, health behaviors, and social circumstances of populations that are at high risk of stillbirth. This is beyond what would be expected due to aging alone. That’s the main thing right there. Where do we intervene? How do we prevent this? Some of that is continuing to ask these questions. As we move forward, we want to consider both age and race in the prevention and intervention related to stillbirth that they can’t necessarily be looked at separately. That you really have to bring both of those together to be able to do more specific targeted prevention efforts.
Then future directions, as this moves over to this slide, we want to understand those unique characteristics of those mothers who are over the age of 40.
Is it that at that point aging becomes less important and some of these other race and ethnicity categories become more important or is there something else there? That’s a unique population that we want to look at more. Then again, we just need better data. With better data we can see some of those more nuanced patterns. With those more nuanced patterns, we can then ask better questions, and with better questions, we can get better answers to the question of why. Once we have those answers to those questions, we can truly begin to make stillbirth an uncommon outcome.
With that, again, I want to thank you so much for the opportunity to be able to present to everybody. I am sorry that we cannot all be there in person, but I am hopeful that at our next go round, we’ll be able to be there. Thank you so much again for this opportunity. Please reach out if you have any questions and let me know what you think. If you have any questions, again, I know that there’s a mechanism to submit those, but if that gets confusing, feel free to email me directly. Thank you so much, and hopefully I will see you all soon in person.
This presentation was part of the Stillbirth Summit 2021. This individual lecture will be awarded .75 hours of continuing education credit to include viewing the lecture and completing evaluation and post-test. Once received a certificate will be emailed to the address you provide in the post-test. If you did not register for the Summit WITH Continuing Education, you can purchase the continuing education by clicking here. This purchase will provide you access to all Stillbirth Summit 2021 lectures including continuing education credit. There is no charge for viewing the presentation.
Please feel free to ask questions of the presenter. We will obtain their answers/comments and provide them here as received.