Prenatal Programming of Brain Development: The Role of Maternal Stress and Stress Biology

February 14, 2023

Information

YCSC Grand Rounds February 14, 2023

Claudia Buss, PhD, Professor, Institute of Medical Psychology, Charité, Berlin

ID9485

To CiteDCA Citation Guide

00:00So I was able to look at associations
00:04between maternal stress and brain phenotypes.
00:08So this was very exciting,
00:10but the one thing I wasn't so happy
00:13about was that we looked at the
00:16children's brain at 7 years age.
00:18And of course there's a lot happening
00:20in the post Natal period and
00:22there's a lot of interaction and
00:24also continuity in terms of free
00:26Natal stress and post Natal stress.
00:28So we don't really know if we
00:29look at something at 7 years age,
00:31is it something that's really related
00:33to the prenatal environment or is it
00:35really the post Natal environment that
00:37shapes this or at least moderates this?
00:38So I thought.
00:40I actually.
00:43I have to confirm something here. Got
00:46it that it's recorded but I
00:49don't see my cursor. We still
00:51have somebody who can.
00:52Karen, can you help me?
00:55See it's. Oh, I see. See the
00:59cursor is there on mission.
01:03There we go. OK. Thank you
01:07so much. So I thought.
01:14Umm.
01:18Moving forward, no. No. OK.
01:21I just wanted to get you to talk.
01:25Thank you. So I thought if we really wanted
01:28to see what our prenatal influences were,
01:30supposed Natal influences,
01:31what we should do is try to characterize
01:34the brain phenotypes shortly after
01:36birth because at this time point
01:38post Natal influences cannot yet
01:40have exerted their influences.
01:41So this is. This is what we set
01:45up together with my colleagues.
01:46Pathik, what ones on your entringer at
01:48the University of California, Irvine.
01:50We had this pregnancy cohort
01:53that where we did extremely.
01:57Deep phenotyping in terms of
01:59their stress and stress biology,
02:00applying ecological momentary or stress
02:04momentary assessments of stress in their
02:07home environment across four days in
02:10their home environment and we took,
02:12we collected a lot of biological
02:15samples and then we followed up these
02:18children and my specific focus was brain
02:21development based on multimodal MRI
02:24and also cognitive function whereas.
02:27Uh, my colleague Sonia entering ahead
02:29more focus on body composition and um,
02:32cellular aging.
02:33So this is like a pretty well characterized
02:36cohort and I will be mainly talking
02:38about results from this cohort today.
02:40Umm, where we have roughly, yeah,
02:44between, depending on the outcome,
02:46100 and 114 Mother,
02:4814 Mother child diets.
02:51We do have several other cohorts now.
02:54We are part of the ECHO consortium here in
02:57the US and contributed 2 cohorts to that.
03:01And we also have a pregnancy cohort
03:03in Berlin.
03:04And we are trying to harmonize our
03:06data collection in ways that we can
03:09eventually merge these cohorts.
03:10For either for mega analysis or at
03:13least for replication purposes and
03:15this is something that we will be
03:18focusing on to really address the
03:20replication crisis and and see where we
03:24stand with some of these initial findings.
03:28So when I started this work,
03:30there was quite a bit of evidence
03:33from epidemiological studies showing
03:35that there was an association between
03:37maternal stress during pregnancy and
03:39higher risk for neurodevelopmental
03:41disorders and psychiatric disorders,
03:43as well as cognitive impairment.
03:45But anything we knew about really changes
03:48in the brain was based on animal models.
03:51So as Kieran said earlier,
03:53we did publish the first study.
03:56Now 13 years ago,
03:58showing associations between maternal
04:00pregnancy specific anxiety and
04:02reductions in Gray matter volume
04:04in the children at 7 years age.
04:07And as you can see here,
04:09especially these pronounced reductions in
04:11Gray matter volume in the prefrontal cortex,
04:14but here also in the in the temporal cortex.
04:18And this is interesting because these
04:20are brain regions that support some of
04:22these cognitive functions that have
04:24been shown in epidemiological studies.
04:26To be associated with maternal stress.
04:28We then later when the sample was.
04:32Larger also looked at cortical thickness
04:34and whether there were associations,
04:36associations between maternal
04:38depressive symptoms during pregnancy
04:40and cortical thickness.
04:41And as you can see in blue are
04:44several regions in the brain where
04:46the cortex was thinner in children
04:48whose mothers had higher depressive
04:50symptoms during pregnancy.
04:52Again,
04:52very pronounced are the
04:54reductions here in the prefrontal
04:56cortex, and this is also
04:59what mediated an association,
05:01the association between.
05:02Maternal depressive symptoms and
05:04externalizing problems in her children.
05:07So this is something we have also
05:10started looking at in our newborns.
05:12In this other cohort, I told about,
05:14I talked about and here's one example
05:16where we were able to show that
05:19there is an association between
05:20higher perceived stress levels
05:22in the mother during pregnancy
05:24and smaller hippocampal volumes.
05:25And we of course also interested in
05:28whether this has done any kind of
05:30implications for cognitive function
05:32cognitive development later on.
05:33And interestingly,
05:34we didn't see a main effect
05:36of hippocampal volume.
05:38The birth and cognitive function
05:39here in this case at six months age,
05:41but it was an interaction with the
05:43environment and those children who
05:45had a larger hippocampal volume
05:47were better able to benefit from
05:49an enriched environment.
05:50So it really shows this concept
05:52of conditional probability.
05:54So certain phenotypes get established
05:56by certain experiences and then will
05:59determine how future experiences
06:01can shape further development,
06:04which I think is this is a
06:07nice example of that.
06:09And then this is some work I'm I'm
06:12working on with Kieran currently
06:15the the UCI children we have done
06:19gene DNA methylation analysis,
06:22longitudinal DNA methylation analysis
06:24that we are currently analyzing and
06:27something we have started looking
06:29into is we have generated this
06:31wholly epigenetic risk score that is
06:35supposedly based on this paper indicates.
06:38Exposure to glucocorticoids during
06:41fetal development and what we did,
06:45we created this polygenetic risk
06:47score and actually showed that
06:49it was associated with maternal
06:51depressive symptoms during pregnancy
06:53and also that it then predicted
06:55hippocampal volume in the newborn.
06:57So this is just an initial attempt
06:59to try to understand potential some
07:02of the epigenetic underpinnings of
07:04some of the associations we observe.
07:07Umm.
07:10I told you that when I started this work,
07:12Umm,
07:12we were the first who who published on this.
07:16But this has changed dramatically
07:18in the last decade.
07:20There are a lot of studies now in
07:22humans showing associations between
07:24various types of maternal distrust
07:27during pregnancy, depression,
07:29anxiety,
07:30but also perceived stress levels and.
07:34Sorry,
07:35it's too loud.
07:38The various forms of maternal
07:40distress levels and brain outcomes,
07:42and this has been work has been
07:44done in in fetuses and newborns,
07:46infants, children and also
07:48adolescents and young adults.
07:49And so there is accumulating
07:52evidence for this for sure.
07:54But the picture is very heterogeneous
07:56because people are suffering
07:58different types of stress,
07:59different brain outcomes.
08:00They are using different pipelines
08:03for analyzing the MRI data.
08:04So I think we don't have a.
08:07Very good picture in terms of
08:09what replicates and what is like
08:10a true effects and I think we are
08:12getting there and and and I will
08:14be talking about some of the steps
08:16that we need to take but at least I
08:18think what we what we can establish
08:20that there is from different
08:22independent research groups quite a
08:23bit of evidence for an association
08:26between maternal distress during
08:28pregnancy and brain development and.
08:31We are of course very interested
08:33in what is it that the fetus
08:35actually receives in terms of the
08:37the signal of maternal stress.
08:39Because I often get asked what kind of
08:42stress should we be paying attention to.
08:44And I think any kind of stress
08:46we should be paying attention to
08:49because we don't know what in an
08:51individual actually leads to the
08:54translation into a biological signal.
08:57So there might be coping strategies.
08:59There might be like, like.
09:01Certain other resilience factors
09:02that lead to the mother coping with
09:05stress well and not increasing like
09:08different stress biology components,
09:10but others where this might be the case.
09:12And here are some mechanisms that
09:14we think are really important and of
09:16course on the one hand it is cortisol.
09:19We know that maternal cortisol
09:21can pass through the placenta.
09:23There is an enzyme 11 beta HSD two
09:25that converts active cortisol into
09:27inactive cortisone and protects
09:29the fetus from an overexposure.
09:32It's just a partial barrier and
09:34a certain percentage of cortisol
09:35passes through.
09:36And something that Kieran has
09:38actually shown and has very early
09:40work is that this enzyme seems to
09:41be stress sensitive.
09:43So not only is there more cortisol
09:45when the mother is stressed,
09:47but potentially more of this
09:48higher levels can.
09:49Pass through and this will lead
09:52to an increase in cortisol in the
09:55fetal compartment.
09:56And then also there's placenta CRH.
10:00Placenta CRH is identical to the
10:02peptide produced by the hypothalamus,
10:05but there is one very important difference,
10:07and that is that it underlies
10:09a positive feedback loop.
10:10So when cortisol is high,
10:12it produces CRH production in
10:14the placenta and will further
10:16stimulate the maternal HP access,
10:19but also the fetal HP a access.
10:20So under levels of under
10:22conditions of chronic stress,
10:23this can lead to this feed forward.
10:26Cycle of elevated cortisol concentrations.
10:29We are also really interested in
10:32cytokines and inflammatory markers,
10:34not only because of the like very good
10:40evidence for infections during pregnancy,
10:43increasing risk for psychopathology
10:45and we have been of course now worried
10:48during the pandemic also what COVID-19,
10:51how COVID-19 infections during
10:53pregnancy might do to the fetal.
10:56Features and feature development,
10:59but these these these immune
11:02mediators are also stress sensitive.
11:04And This is why we are
11:05very interested in them.
11:06And there are some debate whether
11:08they can actually pass the placenta.
11:10I don't think the evidence
11:12is very convincing.
11:13But what definitely happens
11:16is that there's systemic,
11:18systemic inflammation in the mother.
11:20It will lead to inflammation in the placenta,
11:22and the placenta itself will produce
11:26cytokines into the fetal compartment and
11:29inflammatory mediators will be elevated.
11:33This is what I will be talking
11:34about on the next slides,
11:35variation of maternal cortisol and
11:37interleukin 6 concentrations and how
11:39it affects fetal brain development.
11:41But we've also done work on metabolic
11:45factors like free fatty acids.
11:47We have characterized insulin
11:49and glucose and we have looked
11:51at conditions like maternal pre
11:53pregnancy BMI and we also see that
11:55this has the capability of programming
11:57fetal brain development and there
11:59we have focused specifically on
12:01brain regions that are important.
12:03Our energy homeostasis and see
12:05associations for example between
12:07maternal pre pregnancy BMI and
12:10elevated free fatty acid concentrations
12:12and hypothalamic development and
12:14hypothalamic integrity that then
12:16predicts body composition and
12:19fat gain in the infant.
12:22So this is just another summary
12:24how we're thinking about it,
12:26that various conditions during
12:30early life in the mother when she
12:33is pregnant or even before she
12:35was pregnant can affect maternal
12:38placental fetal stress biology and
12:40thereby affect all the fundamental
12:43processes of brain development.
12:46And um,
12:46then affect cognitive and
12:48affective processes as well,
12:50as well as mental health outcomes.
12:53So Umm,
12:54I want to share some findings
12:56with you on the variation in
13:00maternal cortisol concentrations.
13:01And here in Michelle volume
13:03and seven-year old children,
13:05this was the first cohort again that
13:08I've referred to and what we saw was.
13:10Sorry,
13:10I don't know if you can see my
13:13cursor doesn't work very well that
13:15higher levels of maternal cortisol
13:16concentrations during pregnancy
13:17were associated with larger McKellar
13:19volumes and there was a sex specific effect.
13:22We only saw those in girls and not in boys.
13:24And these larger amygdala volumes
13:27also mediated and association between
13:29maternal cortisol and effective symptoms.
13:31And these seven-year old girls,
13:34when we looked at the newborns,
13:36there was a tendency for exactly the
13:38same effect on larger amygdala volumes.
13:40Only young girls, but not in boys.
13:42It was not quite significant,
13:44but.
13:44And what we did find was that elevated
13:47maternal cortisol was associated
13:49with stronger amygdala connectivity
13:52to brain regions involved in
13:55sensory processing and integration.
13:57And very specifically,
13:58for example,
13:59a stronger connectivity between
14:01amygdala and the anterior insula
14:04and this higher connectivity.
14:09Actually this is yet another another network
14:13where we see a stronger connectivity to
14:16the supramarginal gyrus and mediated the
14:19association between maternal cortisol
14:22concentrations and internalizing problems
14:24when the children were two years old.
14:27So I think this is really important that.
14:30We do see that the variation in
14:33brain phenotypes in the newborn
14:36actually do predict later behavior,
14:39later cognitive function,
14:39and I'll show you some more of that.
14:41So this variation that we see
14:44does seem to be meaningful.
14:47So, Umm, we did find that elevated
14:50material cortisol concentrations
14:51are associated with larger amygdala
14:53volumes and this was associated
14:55with more effective symptoms and
14:57also with increased the mitella
14:59connectivity with cortisol with
15:01cortical structures that were associated
15:04with higher internalizing problems.
15:06And this may support higher vigilance
15:08and offspring of mothers who experience
15:10high stress during pregnancy and
15:12therefore could increase the risk
15:14for effective and anxiety disorders,
15:16although eventually this might have.
15:18Or if an evolutionary purpose to
15:21prepare these children to a potential
15:24stressful extrauterine environment.
15:26Umm,
15:27we did see these interesting sex
15:30specific effects and we are not
15:33really clear why this is and we
15:35cannot say that in general females
15:37are more susceptible than males.
15:40There are a lot of examples where
15:42it seems like for certain exposure
15:44and certain outcomes males seem
15:45to be more susceptible.
15:47But what we continuously see when
15:49we look at variational cortisol,
15:51it seems like females are more
15:54susceptible and some reasons.
15:56Um could be, for example,
15:58that there are sex differences in
16:00the timing of glucocorticoid receptor
16:01expression in the fetal brain,
16:03and there is also sex differences in
16:06placental glucocorticoid receptor
16:07functioning.
16:08And also it has been shown in adults
16:10that chronic stress had different has
16:13different effects in males and females.
16:15So dendritic expansion in females
16:17but retraction in males.
16:19So we don't know yet why this is,
16:22but it's something that we consistently see.
16:25I will now talk about some of
16:28our findings in association with
16:31variation in maternal interleukin
16:346 concentrations and I already want
16:36to say we we looked at.
16:38Whether there is any moderation by
16:40fetal sex as well, and there was not.
16:42So here males and females seem to
16:45be equally affected by higher levels
16:47of interleukin 6 concentrations.
16:49So yeah,
16:50I already said a higher inflammatory
16:52measure is a risk factor for various
16:55neurodevelopmental disorders.
16:56And like various conditions like
17:00obesity and infection,
17:01as well as psychological stress are
17:04associated with higher interleukin
17:056 concentrations. Umm.
17:07It seems like Interleukin six really
17:10plays an important role because in
17:12an animal model, if you block Interleukin 6,
17:17if you give an sorry,
17:18if you give an interleukin 6 antibody,
17:21it blocks the effect of maternal
17:23immune activation.
17:24So it it really does seem to be
17:26to play a very specific role.
17:29As I said earlier,
17:30Interleukin 6,
17:31there's some evidence that it
17:32can pass the placenta,
17:33but there's definitely much more
17:35evidence that it will induce inflammation
17:36in the placenta and the placenta.
17:38Itself produces cytokines.
17:40And then of course there's
17:43really a lot of evidence,
17:45preclinical evidence for maternal
17:47immune activation during pregnancy,
17:49altering fetal brain development.
17:50So all I'm going to show
17:53you now is in the newborns.
17:55And again, we found our larger mikalah
18:00volume in those newborns whose mothers
18:03had higher interleukin 6 concentrations.
18:06And as I said earlier,
18:07there was no sex specific effect here.
18:10Umm. We also looked at a metal icon
18:14activity here and there was a stronger
18:18bilateral amygdala connectivity to brain
18:20regions involved in sensory processing,
18:23like the fusiform,
18:25the somatosensory cortex,
18:26the thalamus.
18:28Also brain areas involved in salience
18:32detection like the anterior insula,
18:35as well as learning and memory like
18:37the cottage and parahippocampal gyrus.
18:40And this is just to show you a scatter
18:45plot of one of these connections.
18:48This is the amygdala anterior
18:51insula connection and how it is
18:54associated with varying levels of
18:56interleukin 6 during pregnancy.
18:57And I should say I'm sorry
18:59I haven't said that yet.
19:01Here we are looking at average
19:03interleukin 6 concentrations.
19:04We collected maternal samples three times
19:07during pregnancy and because interlocken.
19:106 concentrations were so highly
19:13correlated across pregnancy,
19:14we calculated an average.
19:15We felt we are not really in the
19:18position of looking at timing specific
19:21effects because they are so highly
19:23correlated and we only have this
19:25one time measure in the newborn.
19:27So trying to draw any kind of conclusions
19:30of time specific effects I don't
19:33think would be warranted with this
19:35design and what the data looked like.
19:38And Umm,
19:39I have done all this,
19:41all this work with my collaborators
19:45Damien Fair and Alice Graham at back then,
19:48which is you,
19:51Damien Ferris now in Minnesota and.
19:55So they have really let all the efforts
19:57on the resting state analysis and
19:59then anything related to diffusion
20:01tensor imaging that I will be talking
20:04about as well as the brain anatomy
20:06we have done in collaboration with
20:08Martin Steiner and John Gilmore at
20:10the University of North Carolina.
20:11And the next slide I'm going to
20:14show you was really mainly led by by
20:17Damien and here we looked at whole
20:19brain connectivity in association
20:21in the new ones and association
20:23with higher maternal interleukin 6.
20:25Concentrations during pregnancy and
20:27saw associations within networks
20:30again and the salience network,
20:33the dorsal attention network as
20:34well as the visual network,
20:36but also various between network connections,
20:40as you can see here below.
20:43As an additional modality,
20:45we looked at diffusion tensor
20:46imaging and we were specifically
20:48interested in this track,
20:49the unsigned fasciculus,
20:51which is a pathway that connects
20:53the temporal lobe with the inferior
20:55frontal gyrus and has us by the
20:58amygdala and the hippocampus.
21:00And we did analysis along this tract
21:03of different diffusion parameters
21:05as you can see here.
21:07So this is the tract and what we
21:10saw bilaterally was that higher
21:12maternal interleukin.
21:13Six was associated with lower
21:16fracture anisotropy which is a
21:18measure of lower maturation of
21:20this tract and this was
21:22very specifically where around the track
21:24where it passes by the amygdala and I think
21:29this is interesting because it's it's a
21:33bilateral and this is just a scatter plot.
21:37This is here you can see what the tracks
21:39look like and what the results look like.
21:42So it's it's. It's really a pretty nice
21:46linear association and unfortunately only
21:48in a really small subgroup here we had
21:52repeated MRI scans at 12 month age as well.
21:56So we looked at whether there is also an
21:59association between maternal interleukin 6
22:01concentrations and these diffusion measures,
22:03so 12 month age, which was not the case,
22:05it was not significant anymore.
22:06But what had happened is that the
22:09there was accelerated maturation
22:10now over the first year of life,
22:13which I mean it's.
22:14Very small sample and it needs
22:15to be replicated.
22:16But I think it is really interesting
22:18in the if we think about like what
22:21we know about this brain overgrowth,
22:24for example in the context of autism
22:27spectrum disorders that has been shown.
22:29So there might be like an initial delay
22:32and then an overcompensation and maybe
22:35this is something that we see here.
22:39So this is a summary of the various
22:41findings with variation and maternal
22:43interleukin 6 concentrations.
22:45And for all these outcomes that
22:49we have looked at,
22:49we see associations with behavioral
22:52or cognitive function in the
22:54first two years of life.
22:56So this,
22:57it makes a lot of connectivity was
22:59associated with a measure of executive
23:02function response inhibition when
23:03the children were two years old.
23:06Does it make a lot of connectivity
23:08and especially this accelerated
23:09increase also during the first year
23:11of life was associated with cognitive
23:13impaired cognitive development
23:14based on the Bayley scales of infant
23:17development at one year age and the
23:20whole brain functional connectivity
23:22was predictive of working memory
23:24function that two years age.
23:26And because we had these various
23:29cognitive aspects that were altered in
23:33association with maternal interleukin.
23:36Six we wanted to see,
23:38although again it was like a
23:40small sample at four to five years
23:43whether we we have like for a very
23:46general cognitive measure here it's
23:48fluid intelligence and association
23:50between maternal and telekin.
23:526 And this measure of fluid intelligence
23:54which was the case and this was
23:57after adjusting from many variables
24:00that would like be qualified,
24:02would be indicators of the quality of
24:04the post Natal environment like the home.
24:06Environment maternal sensitivity,
24:08for example,
24:09and we also try to see whether we can
24:13identify some structural variation
24:15and in brain structure that might
24:18underlie this association and saw
24:21that potentially specifically again
24:23here in the prefrontal cortex,
24:25the horse triangularis might play
24:28a role in this association.
24:31There is more evidence now also
24:35from other groups,
24:37showing that maternal immune activation,
24:40here also in humans,
24:42is associated with neonatal brain
24:45connectivity here specifically
24:47the the salience network.
24:49This is an interesting study because
24:51it's a it's a real longitudinal
24:54study looking at maternal cytokine
24:56concentrations during pregnancy and brain
24:59circuitry 45 years later in adults.
25:02There is also evidence,
25:05at least in terms of neurodevelopmental
25:08delay from like very impressive
25:11Scandinavian birth records,
25:14and also some interesting.
25:17Studies in nonhuman primates.
25:21So to conclude,
25:22there is evidence for prenatal
25:24conditions like various forms
25:26of stress but also cortisol and
25:28inflamed inflammatory medias like
25:30interleukin 6 to be associated
25:32with fetal brain development.
25:33And showed you evidence for associations
25:35with the size of the hippocampus
25:37and amygdala as well as structural
25:39and functional connectivity of
25:41the amygdala and as well as global
25:44cortical volume and thickness
25:45and the functional connectome.
25:48And it really seems like neural
25:50phenotypes are being programmed.
25:52Increased risk for neurodevelopmental
25:54and psychiatric disorders and
25:56that potentially these stress
25:59sensitive biological mediators,
26:01variation and maternal stress
26:02biology do play a role for
26:05programming in the fetal brain.
26:08I think if we talk about MRI,
26:11we have to also acknowledge this paper.
26:13This is something that has
26:15been published last year and
26:17my collaborators Damian Ferron,
26:19Ellis Graham and Oscar Miranda Dominguez,
26:21who I work with closely,
26:22are all involved in this and.
26:25So I think there is a crisis,
26:29a replication crisis,
26:30something similar that has
26:32affected a genomics a while ago,
26:36where this study really suggests that if
26:38we want to look at brain wide associations,
26:41especially with certain phenotypes,
26:43especially when it's complex phenotypes
26:45like mental health outcomes,
26:47we need very large sample sizes
26:49because effect sizes are small and
26:51probably most studies that have been
26:53published are underpowered and they.
26:55Like,
26:56I think they showed the evidence for
26:58this very impressively in this study.
27:01They also did acknowledge that
27:03there are phenotypes where the
27:06associations are stronger,
27:07like cognitive phenotypes for example.
27:10But in general,
27:11I mean this raises really the question,
27:13what can we do with these smaller
27:15sample sizes and can we still,
27:17are they still worth it,
27:18can we still trust the results?
27:20And Umm.
27:20So we have of course done a lot
27:24of thinking and.
27:25What we what we feel about what
27:27we have published so far and what
27:30we can do going forward.
27:31And so I I still believe that we
27:35have a very good conceptual model
27:38and there's a lot of preclinical
27:41evidence kind of supporting the
27:43kind of analysis we have done
27:45and also like in support of the
27:48specific findings we have.
27:50But in the future we should
27:53still see whether we can do.
27:55Better whether there are
27:58opportunities for replication,
28:00whether we can work together more
28:02closely in terms of the specific
28:04protocols not only for data
28:06collection but especially also
28:08for processing the data they are,
28:11I think there are a lot of
28:14opportunities for collaboration.
28:15And there is this fetal infant
28:17toddler on your imaging group that
28:20that has been founded that really
28:23addresses some of these issues
28:27and were people who work in the
28:30field of Infinera imaging,
28:31come together and share their
28:33experiences and bring their protocols
28:35together. There's a lot of progress
28:38in freely available processing
28:41pipelines and a lot of advances.
28:43Then there are several.
28:45Consortia are trying to bring in,
28:47bring together the various infant samples.
28:49There are like this origin consortium
28:52that is led by Rebecca Nikaya in
28:55Michigan or also the Echo consortium.
28:57And then of course there are larger
29:01representative developmental in
29:02your imaging studies like the
29:04Baby Connectome project and very
29:06importantly coming up the HBCD
29:08study that I think would be very,
29:10very informative.
29:13Something that we have also
29:15think about is how can we huge,
29:17how can we use the larger consortia that
29:21are available right now to potentially
29:24inform the results in our smaller cohorts.
29:26And one of the things that we are trying
29:29to do now is to use the larger consortia
29:32like for example the ABC D study to
29:35calculate polling euro risk scores.
29:37So really in the sample of
29:40several thousand like ABC D.
29:43Um, look at associations with a certain
29:45outcome that we are interested in,
29:47like internalising problems,
29:48and then look at the functional connectivity
29:52that is associated with this outcome
29:54and then apply the weights from this
29:57larger consortium to our smaller samples.
30:00So really a very similar approach to
30:02polygenic risk scores and this is
30:04something we are currently working on and.
30:06We have done this for it makes a lot of
30:09connectivity and internalizing problems and
30:12we're actually able to then use this Poly.
30:16Pulling your risk score to predict
30:19emotional regulation and our infant cohort.
30:21So this is work in progress but this
30:24is and this is led by by Oscar Randos
30:29Dominguez and he he has yeah we I
30:32think we are making good progress to
30:35to see how we can utilize these larger
30:37cohorts and another example is for
30:40example these brain charts for human
30:42for the human lifespan that has been
30:44recently published based on like 100.
30:461000 individuals,
30:47something similar to growth charts so that
30:50you can see where does your data fall,
30:52how representative is it,
30:54and then you get percentiles
30:56based on the larger population,
30:58which I think probably is a good way of
31:01correcting your smaller sample sizes.
31:02So this is something we are
31:05currently working on.
31:07And I'm happy to discuss this further later.
31:10But before I come to the end,
31:12I want to talk about this 4th
31:14area that I wanted to address,
31:16which is maternal preconceptional
31:20stress experiences and specifically.
31:23Now it doesn't work again.
31:30This slide.
31:35Maybe I stay very long.
31:38Thank you so. The UM,
31:42maternal child adverse childhood experiences
31:44and how these might potentially get
31:48transmitted to the next generation.
31:51And so, I mean we're thinking of different
31:56forms of neglect and abuse experiences and.
32:01As we all know, this is a huge problem
32:03because prevalence rates are really,
32:05really high.
32:06I think it's also really important that
32:09people working and and and perinatal
32:12medicine know that this is like such
32:15a high prevalence and that they will
32:17encounter many women who have made
32:20these kind of experiences and that
32:22potentially 1/3 of the women that they
32:25see could have these kind of risk factors.
32:29What we know from like many studies
32:31is that there is an increased risk
32:35in the exposed individual for higher.
32:37For psychiatric disorders as well As
32:40for somatic disorders and metabolic
32:43function like obesity,
32:45and we also understand some of the
32:47mechanisms.
32:47We know that there are alterations
32:50in the endocrine stress system,
32:52but also a very well replicated
32:55finding is increased systemic
32:56inflammation in individuals exposed
32:59to childhood maltreatment.
33:01And what is accumulating more and more
33:03is that also the offspring of these
33:05mothers who themselves have not been.
33:08Victims of abuse also have a higher
33:11risk for neurodevelopmental disorders,
33:13behavioral problems,
33:14but also adverse birth outcomes and.
33:20Also obesity, for example,
33:22and what we have done in in the ongoing
33:27ECHO cohort is because several studies in
33:31smaller or larger samples have addressed
33:34the association between maternal childhood
33:37maltreatment and single health outcomes.
33:39And what these studies do not allow to
33:42address is what is the potential effect
33:45on comorbidity across disorder, so.
33:48We took advantage of this echo cohort
33:51where we had information on up to 4000
33:56Mother child diets and we're about like.
33:59A little more than a third of those
34:03mothers did report that they had been
34:05exposed to childhood maltreatment.
34:07And then we looked at these six outcomes,
34:10internalizing problems, asthma, obesity,
34:13autism spectrum disorders, ADHD analogy.
34:16And as you can see here,
34:19across many of these disorders,
34:22we see a very significant increase
34:24in the risk for these disorders
34:26in children whose mothers had been
34:28exposed to childhood maltreatment.
34:31The highest is for internalizing problems,
34:33but also for autism spectrum disorders.
34:37It's one.
34:38It's a 1.7 fold increase and a more
34:42than twofold increase for ADHD.
34:44And there's also an increase for asthma.
34:47We didn't find any association with allergy,
34:50and the only outcome where we found
34:53an association that was moderated
34:54by sex was obesity.
34:56So only female offspring whose mothers
34:58had been exposed to childhood trauma were.
35:02Had a higher risk for obesity and
35:05what was really interesting.
35:08Is that these mothers, these,
35:12these children clustered into different
35:15groups and there was one group that
35:18you see here who had who had diagnosis
35:21on various of these outcomes,
35:24especially the neurodevelopmental outcomes.
35:28ADHD,
35:28LG,
35:29but also asthma and also internalizing
35:32problems.
35:33And mothers of children in this group
35:35were twice as likely to have been exposed
35:38to childhood maltreatment then in the
35:40other lower risk groups with lower health,
35:43with the lower prevalence of health outcomes.
35:46And then we also did a latent class
35:49analysis to to look at whether different
35:52types of exposures of maternal childhood
35:55maltreatment were associated with
35:56specific outcomes and the child.
35:58But what our data suggested
36:00was that it was rather,
36:02um,
36:02a matter of severity,
36:04because it was those mothers who
36:06had been exposed to more than one
36:08type of abuse and neglect whose
36:11children had the highest risk of
36:13developing these disease outcomes.
36:16So this paper was just accepted for
36:19publication and will be out next week.
36:21Umm, it's not yet out.
36:23We are working on it. And so I want to.
36:29I'll be quick,
36:30I will.
36:31I will only talk like I try to finish
36:33in like 5 minutes and we'll talk a
36:35little bit of the Mecca about the mechanisms,
36:37what might underlie this
36:40intergenerational transmission.
36:41And there has been a lot of focus.
36:43On post Natal factors,
36:45because women exposed to childhood
36:48maltreatment have a higher risk
36:51for postpartum depression,
36:52they have more often bonding difficulties.
36:56They're like impaired maternal,
36:59maternal sensitivity,
37:00which of course are all risk factors
37:03for later pathology and the child.
37:06But the case we wanted to
37:08make is. And that's because of all
37:12the evidence for altered stress
37:15biology in the exposed individual.
37:17After childhood maltreatment,
37:18they will most likely carry
37:20those forward to pregnancy.
37:21It will not stop once they become pregnant.
37:24And I have just shown you that
37:25there's a lot of evidence that like
37:27variation in these biological mediators
37:29can then program the fetal brain.
37:31And this is what we're trying
37:34what we've tried.
37:35To summarize in this review and also
37:38recent another recent review how
37:40the various sequelae of maternal
37:43childhood maltreatment that you see
37:45here will affect the biological
37:47state during pregnancy and can can
37:50affect fetal brain development
37:53and even the maternal behavior.
37:56The post Natal environment that is
37:59being created by that will most
38:01likely be affected by stress biology
38:04during pregnancy and this is.
38:06Something else,
38:07uh Kieran and I are working on
38:09together whether there could be
38:11differences in estrogen sensitivity
38:13potentially that will maybe reduce
38:15estrogen sensitivity in these
38:17mothers who have been exposed to
38:19childhood maltreatment that might
38:21not allow her brain to adapt to
38:23this new situation to prepare for
38:26motherhood as well as in individuals
38:28with higher estrogen sensitivity.
38:30At least this is a working hypothesis
38:33that we are examining right now.
38:37Umm here's an overview of various
38:42associations between maternal
38:43childhood maltreatment and variation
38:46in stress biology during pregnancy.
38:49And so this is what we have
38:52contributed to as well.
38:53And indeed,
38:54there is evidence for higher cortisol
38:57concentrations during pregnancy,
38:58higher inflammation,
39:01steeper increase in this placental
39:02CRH over the course of gestation,
39:05but also other important.
39:06Andrew,
39:07current mediators like thyroid
39:08hormones that are very important
39:10for fetal brain development seem
39:13to be associated with higher
39:15maternal childhood maltreatment.
39:17And to to really make the case that
39:19the transmission already occurs
39:21prenatally and not just pro postnatally.
39:23We wanted to show that already in the
39:26neonatal brain we see associations
39:28with maternal childhood maltreatment
39:29and this was indeed the case.
39:31We saw that neonates newborns whose
39:34mothers were exposed to childhood.
39:37Treatment had actually overall
39:38smaller brain volumes and very
39:41specifically lower Gray matter volumes.
39:43When we looked at whether this was
39:45regional specific or more global effect,
39:46we really saw it was more of a global
39:49effect globally smaller brain.
39:51In these newborns whose mothers had
39:53been exposed to childhood maltreatment,
39:55so really making the point that
39:57it's then probably something like
40:00the post Natal environment that
40:02might also be affected will add
40:04on top of this this early already
40:07prenatally programmed phenotype.
40:09And we were interested in whether
40:12total brain volume and newborns
40:14was associated with cognitive
40:17performance and executive function
40:19and did not find any main effect.
40:22Um in two years and 4 1/2 years.
40:24But again,
40:25we saw a really interesting
40:27moderation by maternal sensitivity
40:29that we observed in a standardized
40:32place situation in a way that.
40:35Infants with larger brain volumes,
40:37we are more able to benefit and
40:40be affected by variation in
40:42maternal sensitivity,
40:43whereas this was reduced
40:46and individuals who are born
40:47with a smaller brain volume,
40:49and this was the case at two
40:51years and also a very similar
40:53pattern at four to five years.
40:57So potentially we want brain
40:59volume could be in Europe phenotype
41:01that indicates differential
41:03susceptibility to the environment.
41:06Umm. And this is something else
41:08we have recently been working on.
41:11This is diffusion tensor imaging
41:14data of our new cohorts that we
41:19recently that we recently established.
41:22And here what we see is that it
41:26seems so we look at diffusion tensor
41:29imaging and we look at this measure
41:33radial diffusivity where higher
41:35scores indicate lower maturity.
41:37And, um, we see that different
41:39forms of depression seem to be
41:42associated with different patterns.
41:44So depression depressed mothers who had
41:46been exposed to childhood maltreatment,
41:49their children at birth seem to have
41:52have a phenotype of delayed maturation,
41:55whereas those newborns whose mothers had
41:58been exposed to childhood maltreatment.
42:01But we're not exposed shelter maltreatment,
42:05sorry.
42:05And we're depressed during pregnancy rather.
42:07Go to pattern of accelerated maturation
42:10and I think this is this is interesting
42:13because we know that both delayed
42:15maturation but also accelerated
42:17maturation might have negative outcomes.
42:19So I think it is really important to
42:22consider both ends of the spectrum
42:24and that's why I thought this was
42:27actually a quite interesting finding.
42:30Umm.
42:32So this is just in terms of
42:34clinical application,
42:35this is something we've developed
42:38with my colleagues Christina,
42:40Hayem and Azania entering at the
42:43charity this kind of the cycle of
42:46biological embedding of child of
42:50adverse childhood experiences and.
42:53And also where potential targets could
42:56be for intervention to break the
42:59cycle after exposure to not even like
43:03to prevent the biological embedding.
43:05But if it has embedded,
43:06if it has already been embedded,
43:08is there a potential for reprogramming
43:10for compensation so that not like
43:13certain phenotypes get established but
43:16then I think what we can do in like
43:20caring natural care is really try to.
43:24Work here.
43:25Work on disrupting this vicious
43:27cycle of the intergenerational
43:29transmission by really trying to
43:32focus or identify women at risk.
43:36And and see how to support women who have
43:39been exposed to childhood maltreatment,
43:42and ideally already during the
43:44preconceptional period or during pregnancy.
43:47But then of course also providing
43:49support in the postpartum period.
43:52But as always,
43:53of course the the earlier the better.
43:57I'd like to just.
44:00Finished with this quote,
44:01it is easier to build strong
44:03children than repair broken men,
44:05so I think it is really important to
44:09understand the very early origins of.
44:13Susceptibility for mental health or
44:17adverse mental health conditions.
44:22Because we can take advantage
44:24of the great plasticity of the
44:27brain during development and yeah,
44:29and deliver targeted interventions to, yeah,
44:32take advantage of this high plasticity.
44:36I would like to close by of course.
44:40Thanking all my collaborators.
44:44Without whom, I couldn't have done this work,
44:47and I'd also like you for your attention.
45:00Thanks so much. Talk to us.
45:01Any questions for Doctor
45:03bus in the audience?
45:07We do have one question already and
45:09from zoom and if Lilia Benoit would
45:11like to to unmute maybe start your
45:14video and always interesting to discuss
45:16individual differences and potential
45:18moderating influences and Lilia, do you
45:20want to pose your question? Yes.
45:23Hi. Can you hear me? Yes, yes.
45:26OK. Thank you very much for for this talk.
45:28It's very, very interesting.
45:30And I'm not an expert at all in methylation.
45:35I'm Shawna Dawson, psychiatrist.
45:37But I'm more interested usually in behavior,
45:42family therapy, communication.
45:45And so it's always very unsettling
45:48for me because I have the
45:50impression that sometime sometimes.
45:52When we measure the outcome of
45:54the adverse childhood experiences,
45:56I do not see how we can measure the
46:00moderating effects of behaviors.
46:03And and the role model effect on the
46:06parent behavior toward the child.
46:09And so even maybe I'm just very biased
46:12because it's an area I don't know much about.
46:14But when I read sometimes obesity,
46:17ADHD, depression, inflammation,
46:18my impression is that all
46:20of these outcomes could be,
46:23might be directly transmitted
46:25just through behaviors like the
46:28maternal behavior towards herself.
46:30I don't know, using drugs.
46:33Being addictions,
46:33feeling depressed or the parental
46:36behavior towards a child,
46:37which is like repeating this
46:39circle of trauma.
46:40And so it's very puzzling
46:43for me because I'm just like,
46:45how I can, how can we measure,
46:46you know, or how could we compare,
46:48should we go back to very biological?
46:53Causation or.
46:54Or could we just say,
46:56oh,
46:56maybe it's just the behavior of being
46:58transmitted and we are actually
46:59measuring something else which is a.
47:02And the body signals of it.
47:06But actually this is not really
47:07the cause of transmission.
47:09I don't know if it's clear.
47:10I
47:10I think it is clear and I
47:12don't think it's an either or.
47:13I think the behavior that is
47:15altered in response to these
47:18adverse experiences will actually.
47:20Moderate or affect how much her
47:23biology is changed in response
47:26to childhood maltreatment.
47:28This is why I had this one figure.
47:30Sorry, I went through it very quickly with
47:33these various sequelae like drug exposure,
47:35like obesity, other risky behavior,
47:38but also her mental health.
47:40And it is indeed the case that
47:42the more risk factors they have.
47:45The higher or the the the the more
47:48pronounced are the differences in
47:50biological markers during pregnancy.
47:52For example,
47:53I don't think I mentioned that,
47:54but in our studies when we have
47:57looked at inflammatory markers
47:58during pregnancy in women who had
48:01adverse childhood experiences,
48:03we see it only elevated in those
48:05women who have also depressive
48:07symptoms during pregnancy.
48:09So I think these like what has
48:12established as a consequence,
48:14sorry, I'm looking at you,
48:15but I should be looking there.
48:17What has been established in terms of
48:20sequella of these adverse experiences
48:22are extremely important and these
48:25include behavioral alterations because
48:27because if certain phenotypes like
48:29depression has been established,
48:31this will affect her behavior
48:33towards her child most likely.
48:35And I think This is why it is so
48:38important that we identify these women
48:40because we cannot change anything
48:42about these experiences that the the
48:44the mothers or the parents made.
48:46But we can of course try to positively
48:49affect some of these sequella,
48:52and I think this will already this
48:54will already be very effective.
48:58Sorry. Thank you.
49:02That was really fascinating.
49:03Thank you so much.
49:04I I was wondering about your comment
49:07about girls only being affected.
49:10With obesity of girls,
49:12of these moms, what, what,
49:13what sense do you make of that?
49:18That that is difficult. I don't think
49:23we really know why this might be.
49:26It is interesting that also in exposed
49:28individuals it seems like that obesity risk
49:31is higher and exposed females than males.
49:34Umm. I really don't know.
49:36We have a good sense of this.
49:38I thought it was really interesting that
49:40this was the only outcome that we saw
49:44differential effects based on offspring sex.
49:48But I actually have to pass.
49:49I don't have a good idea.
49:50I can, I can say that it has been similar
49:54things have been shown in the in the
49:57as I said in the exposed person but.
50:00Yeah, we don't know.
50:01I think this is something to really
50:03something to really look look into.
50:05And I have to say that the
50:08few studies there are,
50:09they haven't always looked at sex
50:12differences and moderation by by sex.
50:14So I think we will have to see
50:17whether this this gets replicated
50:18and try to make sense of this.
50:22Thank you so much. Go catch. One
50:24last quick question.
50:28Yeah, amazing.
50:29Amazing talk and fascinating area.
50:33Quick question about the
50:35studies involving stress.
50:36Were you specifically targeting
50:38populations of parents with high
50:41stress levels due to whatever other
50:44factors that might be causing?
50:45Or were you targeting sort of
50:47more of a general population?
50:49And if if so,
50:51are they levels of stress that we
50:54should be particularly worried about?
50:55Can we quantify
50:56it? Can we know when to intervene
50:59can and when are we dealing
51:00with the natural variation?
51:03I think that's a it's an important
51:05question and very difficult to answer.
51:07So the the first part is very easy.
51:08So the initial studies we did
51:10and the data I presented today
51:12was just a normal population,
51:15so normal variation and stress
51:17our current ongoing work.
51:18We have enriched the cohorts for women
51:21exposed to childhood maltreatment.
51:23So they are slightly higher risk because
51:26we want to investigate this further
51:28and also start looking further into
51:31these the moderating role of these.
51:33So we have enriched.
51:35Our ongoing cohorts.
51:38What are levels that we should
51:39be paying attention to?
51:40I think this is really difficult
51:42because as I said it's not a one-on-one
51:45translation and actually there are
51:47a lot of studies that don't find
51:49associations between variation and
51:51psychological stress and biological
51:53mediators that they measure.
51:55And I think well obviously it doesn't
51:57mean that there is no association,
51:59I don't think it's it's measured correctly.
52:02So something that we have done
52:04for example in this but these
52:06ecological momentary assessments is.
52:08That we see that the intra individual
52:11variation in stress is what is
52:14associated with cortisol concentrations.
52:16Not the inter individual variation,
52:18but how much the individual
52:21varies around her or herself.
52:24Her own mean is what is important.
52:28So I think just based on
52:31questionnaire measures in screening,
52:34it would be hard to say,
52:35oh,
52:35this is what you should pay attention
52:37to and this is what you should not
52:39pay attention to because the how it
52:41gets translated into signals for the
52:43features might be very different based on,
52:46as I said,
52:47various resilience factors.
52:49But of course I think in general
52:51we should be paying attention
52:53to mental health and stress,
52:55especially in the enduring prenatal care.
52:58Because even if maybe it's not high
53:01enough to affect the fetus in a negative way,
53:04I mean you can do something for the mother.
53:06And I think in general it's still
53:08a problem that mental health issues
53:10are not very much the focus or
53:12at least are paying attention,
53:14being paid attention to enough during
53:17prenatal care and similarly the
53:20these adverse childhood experiences.
53:24I I always say I I really had a
53:27very difficult time to establish,
53:29um,
53:29the study and in Germany because
53:31I got a lot of feedback saying you
53:33cannot ask pregnant women about
53:35adverse childhood experiences.
53:37You will read,
53:38traumatize them and you should
53:39not do that to a pregnant woman.
53:41And I really had to argue that
53:43I think we're doing something
53:45good for the women because these
53:48thoughts will get illicit anyway,
53:50because now they are becoming parents.
53:52They will think about their own childhood.
53:54And it is important to do this in a very
53:56secure environment and then offer them help.
53:58And they are actually papers out
54:01there on exactly this topic saying
54:03that women would hope to be asked
54:06about these experiences,
54:08especially,
54:08for example,
54:09if it has been sexual abuse,
54:10because this can really affect
54:12a vaginal delivery and it's
54:13very important that these women can
54:15address these kind of concerns.
54:17Early on our studies, for example,
54:20we've had these cases and we went
54:22to the delivery room with them and.
54:24Some of them, it really helped them
54:26to prepare for that and sometimes I
54:28still have to do a cesarean section
54:30and others were then able to really
54:31try it and and go through with it.
54:33I think it's it's in general I would
54:36say we should start paying attention to
54:39stress and mental health issues in general.
54:43And even if even if it's not like
54:45a extremely higher or toxic level
54:47that it would affect the fetus,
54:49we can do something good for the
54:51for the pregnant woman and that
54:53will have a positive impact.
54:55But I cannot say, well,
54:57this is a cut off that you should pay
54:59attention to and others you you shouldn't.
55:01It's very difficult of course I
55:03think you're rallying cry for.
55:06The support for pregnant individuals
55:07is a great way to end this.
55:08Thank you again, Doctor Bush.