Latest / The Biological Psychiatry Podcast / Biological Fingerprints of Adversity
Transcript
- Elvisha Dhamala: From the Biological Psychiatry family of journals, this is the Biological Psychiatry podcast, ⁓ where we discuss the latest research on psychiatric disorders and their treatments. ⁓ Thank you for joining us. I'm your host and Social Media Editor, Elvisha Dhamala. ⁓ What if different types of childhood adversity leave fundamentally different biological fingerprints, ones that are already detectable before kids even hit their teenage years? We know that early life stress shapes long-term health outcomes, but whether growing up in a neglectful home versus being exposed to trauma produces distinct patterns of biological wear and tear, and whether that distinction matters for mental health has remained an open question. A new study in biological psychiatry, Global Open Science, followed children across six years, tracking aging markers including cellular health, brain aging, and pubertal timing. They found that deprivation and threat each carved out a fundamentally different biological aging profile, and only one predicted psychiatric problems years later. Joining me today is Chase Antonacci from Stanford University, first and corresponding author of a recent paper in biological psychiatry global open science. welcome to the Biological Psychiatry Podcast. Chase: Thanks for having me. Elvisha Dhamala: so let's get started. When we talk about early life stress becoming embedded biologically, what does that actually mean and why is the field increasingly looking at biological aging as a framework to understand these effects? Chase: Yeah. So, you know, when I think about early adversity and we think about biological aging and kind of this idea of embedding, I think it's helpful to think about, you know, what is early adversity and what are kind of the downstream consequences. So we know that early adversity, things like exposure to, you know, abuse or trauma and also things like growing up in a poverty or with kind of chronic resource scarcity. These sorts of experiences in childhood are associated with increased risk for mental health problems in adolescence and across the lifespan, and also higher risk for disease outcomes as well. And so when we think about early embedding, we're really asking how do these early experiences essentially get converted into lasting change and how the body is really built and regulated? so really this kind of change in a physiological set point. So across things like hormone regulation and cellular metabolism and developmental tempo. And you know, to your question of why is the field thinking more about Biological aging as kind of this lens. So we really think that adversity doesn't just hit one system, but it really hits many at once. And so biological aging kind of gives us a common currency to compare across all these systems. So things like cells or hormones or brain, you know, brain development and metabolism. And we can really across all of these systems ask whether a child's biology is running ahead of or kind of behind their, their typical chronological age. And then the last thing I'll say too is that, you know, we know these aging markers are not just correlates of adversity, but they also seem to predict disease, risk, and psychopathology. So we mentioned this big UK biobank study in the paper that recruited over 400,000 adults. And they found that faster biological aging actually predicts the onset of things like depression and anxiety. So it really does seem like biological aging may be a pathway from early stress to later illness. Elvisha Dhamala: A key distinction that you make in your paper is between threat-based and deprivation-based adversity. Can you walk us through that distinction and why it matters for understanding how stress affects our brain throughout the course of development? Chase: Yeah, so this is really interesting. And this is kind of the literature around taxonomies of adversity is something that I've in our group has been interested in for several years now. So if you think about it, there's this huge diversity of experiences that kids can face as they grow up, different environments, different types of, you know, households, you know, just hugely different experiences. So this threat versus deprivation idea Has come from a number of folks. I think the most prominent are Margaret Sheridan and Katie McLaughlin's work, where they've laid out the dimensional model of adversity and psychopathology or DMAP. And within DMAP, you have these two kind of qualitatively different types of experiences where they argue that the entire kind of exposome of exposures can be boiled down to threat and deprivation. So threat is really the persistent, you know, kind of presence of harm. or the threat of harm to one's physical integrity. So you can think of things like abuse or violence or trauma exposure. So kind of too much of something dangerous. And then in contrast, you have deprivation. Or sort of the absence of expected environmental inputs. And that would include things like neglect or poverty, or even low parental support, or kind of, you know, lack of typical cognitive or social understimulation when kids are young. And so kind of the too little of something that's needed. And there's been some work showing that these different types of experiences shape different neural circuits. So threat seems to tune fear learning circuits and Whereas deprivation seems to affect circuits for cognition and executive functioning. So kind of the goal of and part of the idea of this paper was to really ask whether that specificity extends beyond the brain to these kind of multi-system biological aging profiles. Elvisha Dhamala: In this study, 225 kids were followed across four waves over about six years and you collected a comprehensive battery of biological measures. Can you give us a sense of who these participants were and what was being measured at each visit? Chase: these kids in the study, there were 225 kids and they were all local from the San Francisco Bay Area. And they were recruited back in, I think starting in 2012 when they were around ages nine to 13. And then they were community samples. So everybody was healthy at the time and nobody had any mental health problems. And then we followed those kids across four waves that were each spaced roughly two years apart. So about six years in total, all the way into late adolescence. And at the baseline visit, when they came in, we administered 16 different measures of lifetime stress exposure that span kind of this broad global range. So we looked at things like threat and trauma exposure, a number of measures of parenting quality, perceived social support, and neighborhood disadvantage. So even things like neighborhood poverty or pollution burden. And then at each of these four waves, we also measured six different markers of biological aging across four key systems. So those were cellular aging, where we looked at mitochondrial DNA copy number and telomere length, which are kind of classic biological aging markers. We also looked at neuroendocrine aging by looking at salivary cortisol, or kind of this the classic stress hormone. We looked at metabolic and somatic aging with BMI or body mass index and self-reported pubertal stage. And then we also looked at neurological aging by calculating brain age from structural MRI. And then at the final wave in late adolescence, we had the kids complete the youth self-report, which is a broadband measure of psychopathology. So we looked at internalizing symptoms like anxiety and depression, externalizing like conduct or behavioral problems, and then total problems as well. Elvisha Dhamala: in terms of your analyses, you used a technique called canonical correlation analysis or CCA. this approach allow you to do that more standard statistical approaches might not? And why was it the right tool for this question? Chase: Yeah, so CCA is interesting. you know, if you think about most standard statistical or modeling approaches, you usually have one or a set of predictors and then one outcome at a time. So you might ask, for instance, does cortisol relate to trauma or does BMI relate to poverty? So you kind of can get a separate pile of correlations or statistical associations, and that might miss how, for instance, these markers move together. And we know, of course, that biology doesn't work as kind of one variable at a time. Instead, really adversity, we also know, seems to shift a coordinated pattern across these systems. And so CCA is a nice type of modeling because it's built exactly for that. It finds kind of a combination of stress dimensions and a combination of biomarkers that you can think of it as almost the they go together the best. It's all kind of like a whole system handshake. And so concretely what we did is we use PCA to reduce these 16 measures of stress exposure into three independent dimensions. So those were cumulative stress and trauma, low parental support, and neighborhood disadvantage. And then what CCA allows us to do is link those three dimensions to six aging biomarkers. And then instead of asking, for instance, this Does trauma correlate with cortisol? You get these emergent multi-system phenotypes by using CCA that are really difficult to see when using univariate tests. Elvisha Dhamala: let's get into your findings. You identified two distinct biological aging phenotypes. One was the deprivation phenotype and the other was the threat phenotype. Can you describe what each of these looked like in terms of the stress dimensions that were implicated and also the biological markers that they were associated with? Chase: Yeah, so what we found was kind of interesting. We found these two core multisystem stress aging phenotypes that emerge. The first one, as you mentioned, is that we called it a deprivation deprivation phenotype. So on the stress side, that was characterized by low parental support and high neighborhood disadvantage. And then on the biology side, we saw higher BMI, older brain age, and lower mitochondrial DNA copy number. And so in the paper, we read this as kind of a cellular metabolic attrition or almost a wear and tear under chronic resource scarcity. And we think that this phenotype really maps nicely onto historical models of allostatic load, where you have sort of sustained energetic insufficiency that really grinds down metabolic, neural and cellular systems. And then in contrast we had this other phenotype or the threat phenotype, where on the stress side it was characterized almost purely by trauma and cumulative stress. And then on the biology side, by earlier pubertal stage, higher BMI again, actually higher mitochondrial DNA and higher cortisol. And so we read this as almost a coordinated somatic acceleration and kind of neuroendocrine mobilization. So really the body speeding up its developmental clock and almost revving up stress physiology to enable, you know, for instance, a child to respond in environments characterized by persistent danger. Elvisha Dhamala: So one thing that you mentioned that's really interesting is that the mitochondrial DNA copy number was showed opposite relationships with the two phenotypes. it was reduced under deprivation, but elevated under threat. What might that tell us about the different biological processes that are happening in response to stressors? Chase: Yeah, this was a really interesting finding from the paper where we saw that so mitochondrial DNA copy number is roughly how many mitochondrial genome copies exist per cell. And people think that that's roughly a proxy for cellular energy capacity and mitochondrial stress. So in the aging literature, typically high higher mitochondrial DNA copy number is seen as adaptive and is associated with longer health span, longer lifespan. In our study, what we found in these kids is that deprivation was associated with reduced mitochondrial DNA copy number. And that's consistent with this idea of mitochondrial attrition or chronic kind of stress and energetic scarcity wearing mitochondria down, or this almost like a depletion phenotype. But then interestingly, with threat, we saw something different. it was associated with elevated mitochondrial DNA, which we think. We propose in the paper might be consistent with more of a mitochondrial biogenesis or the cell building more mitochondria to meet the heightened energetic demand of this kind of sustained kind of somatic neuroendocrine activation in the context of persistent danger when a child might need to respond quickly, in response to their environment. So The same marker might actually mean opposite things depending on the adversity context. So if you studied mitochondrial DNA in isolation, for instance, by just pooling all stressed kids together, these two signals may very well cancel out. And you might conclude that stress actually does nothing to mitochondria, but only when using this multivariate dimension-specific approach, it beals that there's actually these two kind of dual pathways. Elvisha Dhamala: you also found that BMI was the one biomarker that showed up across essentially all dimensions of stress. Why do you think metabolic dysregulation emerged as such a convergent signature? And how should we think about elevated BMI throughout the course of development? Chase: Yeah, it's a really good question. So in our sample, BMI was the one marker that was independently predicted by all three stress dimensions. So by cumulative stress and trauma, by lower parental support, and by neighborhood disadvantage. And each of these seemed to push BMI up on their own. you know, for the question of why does metabolism seem to be this convergent signal? So we know that metabolism and and BMI By proxy really seems sits downstream of almost every other stress pathway. So cortisol, for instance, cortisol and HPA, axis dysregulation, disrupted sleep. You can think about the food environment and food insecurity, reduced physical activity. A lot of these are, we know, well-established consequences of exposure to early adversity, and they all funnel into metabolic regulation. So we think it might be kind of this final common pathway. At the same point, we should note that. This is, we we, you know, reviewers brought this up and we mentioned it in the paper, is that this BMI finding really is a population level statistical association. It's nothing close, you know, to a diagnosis. So an individual child with high BMI is not necessarily a child who was exposed to adversity. And, vice versa, adversity doesn't doom any given child to high BMI. at these ages, we think that elevated BMI might not even be read necessarily as biological aging per se, but really understood as an early indicator of metabolic dysregulation that increases risk for cardiovascular and metabolic risk in adulthood. Elvisha Dhamala: Another thing that was quite interesting is that you found that the stress biomarker associations were strongest at the first assessment and actually got weaker over time. What does this mean for how we should think about timing during development? Chase: Yeah, so this is a really interesting point. What we found is that for both threat and deprivation, their links to biological aging were really the tightest and the strongest at baseline. So when these kids were between ages nine to 13 years old, and they seemed to you know, slightly fade or attenuate across following waves. So for instance, low parental support was associated with mitochondrial DNA. But that correlation dropped over 75% from the first to the last visit, or as well, low parental support and BMI also dropped by 36%. so the point that we make in the paper is that it seems like the biological mark of early adversity seems, at least in our sample, to be laid down relatively early. And then these systems more or less stabilize. And that's why we use this term of imprinting. At the same time, normal adolescent development, we know puberty, growth, the new kind of s social stressors that adolescents face, those also add a lot of variance across time, which may kind of swamp the signal of you know early early adversity. So I th I think those two things are not mutually exclusive, but the idea from the paper seems to be. And for our findings seem to be that early stress is really most strongly associated with these, you know, biological aging markers early on. And then it doesn't seem to drive continued dysregulation across adolescence so for better or worse, it seems that early stress seems to be kind of set in these early biological aging markers and doesn't drive continued dysregulation across adolescence. Elvisha Dhamala: when you looked at whether early stress predicted the rate of change in these biomarkers over adolescents, you found that there were no relationships essentially. Was that surprising? And how does that fit with your imprinting interpretation that you just described? Chase: Yeah, so we thought what we might find is that different types of stress may be related to, for instance, kind of early biological aging or these set points by the time a kid enters the study. And we might see other types of stress, maybe more proximal or distal stressors that are associated with change across adolescence and these biomarkers. in fact, we saw that no stressors, at least that we asso that we surveyed were associated with change. In any of these biological markers across time or change in the trajectories. you know, we think that the trajectory estimation is actually a harder question. So with four time points per kid and person-specific slopes being noisier than intercepts, and a modest sample. We think that it's possible that there are other types of stressors. When these kids get older, they're facing new types of peer stressors that may shape the trajectories of biological aging. But at least with our sample, everything up to you know the age, you know, nine to thirteen seems to be largely consolidated by that that time point. so this may really point to the need to focus on the pre-adolescent period as a window for surveying stress and also understanding kind of like biological integration. or biological imprinting of stressors and also may signal to the need for focusing on the pre-adolescent period as a window for prevention. Elvisha Dhamala: Another thing you were interested in were the mental health outcomes. And you found that only the threat phenotype predicted psychopathology six years later, but it did so across both internalizing and externalizing problems. Can you tell us a bit more about that and what might be driving this transdiagnostic risk? Chase: Yeah, so we found that the threat phenotype predicted all three outcomes six years later. So, like you mentioned, total problems, externalizing problems, and internalizing problems. And in contrast, the deprivation phenotype predicted none. So it really was this almost transdiagnostic signal where exposure to early threat raised the risk broadly for many types of problems rather than any one specific disorder. And we think that this may point to almost a general mechanism rather than symptom-specific ones. So that may be associated with we, you know, there's a fair amount of literature suggesting that exposure to early threat is associated with heightened arousal or emotional reactivity and really seems to tune the brain's fear and salient systems. also early threat is associated with earlier pubertal timing, which itself is a robust and well-replicated prospective risk factor for adolescent psychopathology. And then we also in the threat phenotype see this kind of chronic HPA or cortisol upregulation, which we know impairs regulation of negative affect and behavioral control. And so what we're seeing is that the findings with biological aging seem to largely parallel what has been posited in the developmental psychopathology literature and neuroimaging literature for some time now. The threat, but not deprivation, really seems to shape. This threat reactive circuitry that we know is associated with broad socio-emotional problems and psychopathology risk. Elvisha Dhamala: So one thing that was really interesting is that in your regression models predicting psychopathology, the stress component of that threat phenotype was a significant predictor, but the biomarker component of that phenotype was not. So can you tell us a bit more what that actually means and how that might be shaping the relationship between threat and psychopathology? Chase: So with CCA, you basically for each canonical variant, you get two latent scores on each side. So for instance, for threat, you might you get a threat, a stress related threat latent factor, and then you also get its corresponding biomarker related latent factor. And what we found is that in joint models entering both the stress component And the biomarker component into the same model, the stress component was a significant predictor, whereas the biomarker component was not significant after accounting for stress exposure. So we don't think that this necessarily means that biology itself is irrelevant. And indeed, on its own, the biomarker component was significantly associated with psychopathology. We think what's actually probably going on is more related to just kind of underlying math behind how CCA works. So CCA builds the two components to share the maximum amount of variance by construction. So basically finding the group of stressors that is most strongly and specifically related to a group of biomarkers. And so when you put both in one regression, the stress side essentially soaks up a lot of the shared variance and leaves the biomarker side with only a small, you know, unique remainder of variance. So it's a big statistical ask to have the stress and the biomarker sides each predicting independently above and beyond themselves because statistically they are, you know, they're they're coupled by construction. so we think the design of the study can't necessarily tell us whether you know, for instance, that biology biomarkers might mediate the stress to psychopathology link if it's a correlated marker of a shared underlying process, we think settling that will require more kind of prospective causal designs. you know, we think that we saw deprivation, for instance, is really associated with this pretty striking allostatic load type pattern. So older brain age, lower mitochondrial DNA and higher BMI, really this kind of biological wear and tear pattern that we know is associated with things like cardiometabolic risk and cognitive problems in adulthood. So, deprivation is almost certainly not benign. but instead the outcomes that we looked at were self-reported emotional and behavioral problems. And we know deprivation effects may surface instead in things like cognition and executive functioning or physical health, or even perhaps later onset depression. But at least in this study, what we were what we were able to capture was mostly psychopathology and late adolescence. So and even, classic dimensional models would suggest that that deprivation tends to be associated with cognitive and executive functioning problems in late adolescence. In this study, we didn't assess cognitive and executive functioning in late adolescence. We were mainly focused on psychopathology. So I think it's very plausible that deprivation and its effects on biological aging are associated with maladaptive or you know difficulties in behavior. outside of socioemotional problems like psychopathology and with domains like cognitive and executive functioning. Elvisha Dhamala: thinking about the clinical and translational implications of your work, if these biological aging profiles are already set by late childhood, what does that mean about when and how we should be intervening? So does this change how we should be thinking about prevention? Chase: Yeah, this is a great question. I think that you know the point that we make in the paper is that it's probably important to think about moving upstream and moving earlier. if the biological mark is largely in place by age 11, which is what we see pretty strongly in our sample, then that might suggest that interventions that are arriving in adolescents may be working after the biological set point is already in place. and so We argue in the paper that the leverage in being able to change these trajectories that these kids are on and and prevent mental health problems may actually lie earlier in childhood. Elvisha Dhamala: my last question is if there's one key message you'd want clinicians, researchers, and families to take away from your work, what would that be? Chase: I think the point that we would make is that it really matters what specifically children are exposed to and what kind of adversities they face, not just how much. And this is the argument that's been made in the kind of stress taxonomy world for a long time, that the type of adversity matters and not just the amount. We saw that threat and deprivation seem to leave different biological fingerprints and each carry different types of risk. So if we want to understand and eventually interrupt how stress gets under the skin, we have to move beyond just counting adversities and start really asking what specifically children face. At the same time, we also see this very clear signal that across all of these biomarkers and this really broad global picture of early stress exposure, that early experience seems to shape Biology and become largely embedded by late childhood or early adolescence. And we would make the argument that, you know, as scientists and researchers and clinicians, that it's really important to understand and more deeply characterize stress exposure and biological aging development within the pre-adolescent early childhood period. Elvisha Dhamala: That's all for today's episode. Thank you to Chase Antonacci for sharing his work on the effects of childhood adversity on developmental biology. For our listeners, you can find the full paper in Biological Psychiatry, Global Open Science, and we'll include the link in our episode description. To learn more about our journals, please visit biologicalpsychiatryjournal.com. You can find The Biological Psychiatry Podcast wherever you get your podcasts. Subscribe now so you don't miss our next episode and follow us on our social media channels for highlights and updates. Thanks for listening. ⁓ This podcast is for informational purposes only and does not constitute medical or professional advice. ⁓ The views expressed are those of the participants and do not necessarily reflect those of the Society of Biological Psychiatry, ⁓ its family of journals, or its editors ⁓ All rights reserved, including those for text and data mining or use in AI systems. ⁓