Yale Psychiatry Grand Rounds: May 20, 2022

May 20, 2022

Lustman Awards

Tanner Bommersbach, MD, MPH; Dan Tylee, MD, PhD; Terrell Holloway, MD; Zach Harvanek, MD, PhD; and Nientara Anderson, MD, MHS

Information

ID: 7864
To Cite: DCA Citation Guide

Download Transcript

00:00Thank you Trisha and and welcome everybody.
00:04The and welcome to the
00:07Lussman Award celebration.
00:09The Lussman Awards are just a wonderful.
00:13Opportunity to acknowledge the incredible
00:16work done by young scientists in
00:19the Department of Psychiatry and the
00:21Residency and Fellowship programs.
00:23It's also a great time to celebrate
00:25the the career and accomplishments
00:28of Seymour Lassman and and Doctor
00:30Pittenger will describe his
00:32background in just a little bit.
00:35But the Lussman family plays a
00:37special role in the hearts of
00:39the Department of Psychiatry.
00:41It's a family that's been involved,
00:43connected to the department
00:44across 3 generations,
00:46and I particularly want to add
00:48my thanks to the the things that
00:50I know that that Chris and and.
00:53And perhaps others will express
00:56to Susan and Jonathan Katz,
00:59who've been really incredibly supportive
01:02of our ability to continue the lesbian
01:06awards over the last many years.
01:09The last man awards have really emerged
01:11as one of the nation's leading awards
01:14for research done in the residency.
01:18And if you go back over the list of honorees,
01:21you'll see many people who
01:23have emerged as leaders in the
01:25Department of Symmetry here at Yale.
01:27But really,
01:28all across the country and and people
01:32who've had whose impact has really.
01:38Been measurable on the overall effort
01:40to relieve the the public health burden
01:44of psychiatric disorders and addictions,
01:47so this is really an incredibly
01:49exciting and special day and I
01:51look forward to the presentations.
01:53I just wanted to add one
01:55thing before I close.
01:56There's no grand rounds next week due to
01:59more because of the Memorial Day holiday,
02:01but the following Friday,
02:03June 3rd will have the state of the
02:06the state of the department address.
02:08So I hope that we'll see everybody there
02:11for the state of the department address.
02:17Which is also another great
02:18opportunity to reflect on the
02:20year that we've all been through.
02:22So without further ado,
02:24let me introduce Doctor Chris
02:26Bittinger who will get us underway.
02:42Chris, you're muted.
02:56There we go. You think that after two
02:58years I have this figured out. That's OK.
03:04So now I just need your
03:05screen back slides back.
03:19And multitude of slide decks,
03:22it's all well. There we go.
03:26Good alright, let me try again.
03:30Welcome everyone.
03:31It's a pleasure to have you all join
03:33us for this year's last minute awards.
03:35As John said,
03:35this is really a special day in
03:37the life of the department and one
03:39of my favorite grand rounds of the
03:41year as we celebrate our commitment
03:43to training and the spectacular
03:45work of some of our of some of our
03:48trainees as well as our commitment
03:50to mentorship and the participation
03:52of mentors in the success of
03:54these trainees and the history.
03:57Of of Seymour Lessman,
03:58who was a giant in in our department
04:01and then the Child Study center years
04:03ago and whose family has generously
04:05supported this award since 1973.
04:08As an in honor and continuation
04:11of his legacy.
04:13So, Seymour Lessman,
04:14for those of you who aren't
04:16familiar with him from previous from
04:18previous lesson award celebrations,
04:20had a remarkable career.
04:22He served in the Army in World War Two.
04:24Then he obtained his PhD in psychology
04:26from the University of Chicago in
04:28his MD at the University of Illinois.
04:31And while at the University of
04:33Illinois he became interested in the
04:35Physiology and behavior of newborn infants.
04:37This was a time when the field was rife
04:39with arguments about nature versus nurture,
04:42which I think we've long since transcended.
04:43Because of course,
04:44all interesting things.
04:45Or both, but at that point this was a very,
04:48you know, active point of of
04:51disagreement and conflict in the field.
04:53And that became the focus of
04:56doctor Glassman's work.
04:57He came to Yale in 1955,
04:59where he completed psychiatry,
05:01residency and Child fellowship,
05:03and he joined the psychiatry faculty in 1962.
05:06And then two years later,
05:08he was promoted to the rank
05:10of full professor.
05:11I can clearly remember when I was
05:12a trainee and John would make that
05:14statement that he was promoted to full
05:16professor just two years later and
05:17John Pier out, all of us and and say,
05:20let that be a model for you, but.
05:23But truly remarkable remarkable
05:26achievement on the faculty.
05:28He was a dedicated teacher,
05:29gifted clinician,
05:30and a careful and creative scientist.
05:32He became increasingly interested in
05:35problems of impulse control and adolescence,
05:37and among numerous other contributions.
05:39He was a key member of a group of
05:41scholars together with Al Solnit,
05:42Anna Freud and Joe Goldstein,
05:44who wrote a landmark text in the
05:46field of child maltreatment beyond
05:48the best interests of the child.
05:51See,
05:51we're lussman died tragically at
05:53a young age in a boating accident
05:55in the early 1970s and one of the
05:57leading lights in our in our field was
06:00taken from us and shortly thereafter in 1973.
06:03His family created this award,
06:05and it's been given continuously
06:07every year since 1973 with I,
06:08I think,
06:09puts us at 49 years.
06:12As John said,
06:13if you look at the the
06:15plaque of past winners,
06:16you'll see quite a few
06:18leading lights of psychic.
06:19If you're both at our institution
06:20and across the country,
06:21and so are our awardees today are
06:25joining an illustrious company.
06:28As I said, we're not only
06:30celebrating the awardees and the
06:32accomplishments of our trainings here
06:34in the Yale Department of Psychiatry,
06:36but also of their mentors.
06:38None of us succeed without the
06:39support of our of our mentors.
06:41They're modeling.
06:42You know financial support, their guidance,
06:45and so we're going to have each of our
06:47awardees introduced by by a key mentor and.
06:52And while the focus today is on
06:54the trainees and their work,
06:56we we applaud the mentors as well.
06:59So with no, I'm sorry I should just be.
07:02I've already thanked the the Lussman family.
07:03I also want to thank the
07:05Lussman Awards Review Committee,
07:07group of faculty who joined us,
07:09YOUNGSTON show Marina Picciotto.
07:12Wow, if I just rattle them off
07:13the top of my head,
07:14I'm going to leave someone
07:15out and that would be upset.
07:16But just I appreciate the Group of the
07:19Group of faculty who helped us review
07:21these for the last several years.
07:23So without further ado,
07:25I will invite Bob Rosenheck guns to
07:28introduce our first winner of us been
07:31awarded first prize winner for 2022.
07:34And that's the wrong one. All right?
07:40So while we go ahead, go ahead.
07:42I'm working out the technical.
07:44Yeah well, while we fiddled with the slides,
07:47I very happy to introduce Tanner
07:49boomers back from Grand Forks, ND.
07:54And of course, he has been a superstar
07:57who published 15 papers as a resident.
08:01And worked very closely with me and
08:04a group that was Co led by Greg Lee,
08:07who's a young investigator.
08:08All of you should get to know.
08:13Tanner, I thought his magnificent
08:17productivity was a product of coming
08:20to Yale, but in fact he was the superstar
08:24before he came to Yale, graduating
08:26summa *** laude from college and getting an
08:30MPH from Hopkins and working with a number
08:33of leading lights there.
08:36His interests are broad.
08:38His curiosity of is boundless,
08:40and what I particularly would emphasize
08:42is the not just the number of papers,
08:46but the fact that he has emerged
08:48as a very talented writer.
08:50And I think we often don't recognize
08:53how important it is to cultivate the
08:57skills of clear scientific writing.
08:59And he's become just a pleasure to
09:02to read his work as it's developed.
09:06And so without.
09:08More ado, let me hand it over to Tanner,
09:10who will present his paper that
09:12appeared in JAMA Psychiatry.
09:16Well, thank you so much Doctor
09:19Rosenak and I just want to thank Bob.
09:22You know, despite what he
09:23says coming into residency,
09:25I knew very very little about about
09:27research and and most of what I've learned
09:30can be attributed to Doctor Rosenak,
09:32who's who's just been such an amazing
09:35mentor to me and I really could
09:37not have asked for a better mentor
09:39when I came into to residency.
09:41I also want to thank Gregory as.
09:44Thompson, our senior author on this
09:47paper who's been a friend and a mentor
09:50to me over the last couple years,
09:52especially.
09:52And I also just want to thank
09:54the the Lossman Foundation.
09:55It was fun to hear more about about
09:58Doctor Lossman for for their support
10:00of of resident research and giving
10:02us really the opportunity to share
10:04our research with the department.
10:06So thank you.
10:08Turning to the the title of our paper
10:10over the next 15 or 20 minutes or so,
10:12I'm going to be talking about recent
10:14trends in the national suicide attempt
10:16rate and specifically looking at.
10:18The mental health utilization
10:20of individuals who attempted
10:22suicide from 2008 to 2019.
10:25Suicide prevention has been an interest
10:28of mine since since medical school,
10:31and I think it's one of not only the greatest
10:34challenges facing the field of psychiatry,
10:36but also one of our greatest
10:37current public health challenges.
10:38And so I think it's a really important
10:40topic that we need to continue to
10:42invest research in next slide, please.
10:47And the reason that it's such a challenge
10:48for us currently is over the last two years,
10:50we've seen a pretty precipitous rise
10:52in deaths by suicide in our country.
10:55Our national suicide attempt,
10:56our national suicide rate over the
10:58last two decades, has risen by 35%,
11:01and that's in the context of declines
11:04that we saw in the 1980s and the 1990s.
11:07It's also very concerning because
11:09over the last two decades,
11:11other developed countries across the world,
11:13especially countries in Europe,
11:14have actually seen a decline
11:16in their suicide rate.
11:17While we've continued to see
11:19a pretty steady increase.
11:22Next slide,
11:24please.
11:25And when we're thinking about
11:26suicide prevention,
11:27it's not only important to
11:28look at deaths by suicide,
11:30but it's also important to look at
11:32trends in all types of suicidal
11:35behavior across the suicide continuum.
11:37So this was data that was just
11:39released by Samsa for last year.
11:40For year 2021.
11:41You can see at the bottom of the pyramid
11:44here at 12.2 million adults are reported
11:47serious thoughts of suicide last year,
11:49and that corresponds to almost
11:515% of the total US population,
11:54one in 20.
11:56Individuals 1.2% of the total US
11:59population reported making suicide
12:01plans in the last year .5% about 1.2
12:04million reported suicide attempts
12:06and a much smaller fraction of those
12:09individuals died by suicide, about
12:1146,000 so still way too high of a number.
12:15Next slide, please.
12:19And so while the majority of
12:21individuals who attempt suicide
12:22don't go on to die by suicide,
12:24we know that suicide attempt
12:26continues to remain the strongest
12:28predictor of future suicide and is
12:30an important group for us to study.
12:32Longitudinal studies indicate that
12:33anywhere from 7 to 13% of individuals
12:36who attempt suicide will go on to
12:39die by suicide in their lifetime,
12:41and the majority of these deaths
12:43estimated anywhere between 70
12:45to 80% tend to occur in the 12
12:47months after a suicide attempt.
12:48And so we know this year after a
12:51suicide attempt is a really critical
12:53period of time for us to intervene
12:55and prevention strategies really
12:57rely on individuals receiving some
12:59sort of mental health intervention
13:01after a suicide attempt.
13:03Yet we know that a fairly large
13:06amount of individuals don't actually
13:08receive services in the year after
13:11a suicide as a suicide attempt,
13:13and we don't know this percentage very well.
13:16We don't actually have a
13:17great understanding of what.
13:19Percent of individuals actually
13:21receive services in proximity to their
13:25suicide attempt because most of the
13:27studies that have been done on this
13:29topic have been done either in fully
13:31insured samples or in individuals who
13:32receive care after a suicide after a
13:35suicide attempt because they present
13:36to emergency departments or they
13:38get admitted to an inpatient unit.
13:40Several studies that have been done
13:42in a large HMO population of a fully
13:45insured sample show that about 40%.
13:49Suicide Attempters had a healthcare
13:51visit within the week before
13:53their attempt and 95% had a visit
13:55within the preceding year.
13:56Yeah,
13:57like I said,
13:57we know that a large percentage of
13:59individuals actually don't receive
14:01care in proximity to their attempt
14:03and and looking at veterans and
14:05the Veterans Health Administration
14:06is a good example of this.
14:08So in 2018 it was estimated that 63%
14:11of veterans who died by suicide didn't
14:14have an encounter within the VA in
14:16the 12 months prior to their death.
14:19And so.
14:20An alternative way,
14:21and perhaps a more accurate
14:23approach to really understanding
14:25how often individuals who attempt
14:27suicide are receiving services
14:28in proximity to their death,
14:30maybe to use population based survey
14:32data of individuals who are insured and
14:35individuals who aren't individuals who
14:36got care after their suicide attempt,
14:39and individuals who have gotten no care,
14:41and so that was the approach that
14:43we wanted to take in our study.
14:45So, next slide, please.
14:49So this was the title of our paper that
14:51was published this spring and we were
14:53interested in looking at 4 main issues.
14:56The first is what are recent trends in
14:58past year suicide attempts from years
15:002008 to 2019 among adults in the US.
15:04We also wanted to know what risk
15:06factors were associated with increased
15:08suicide attempts over this time period.
15:11Third, we wanted to understand that
15:13amongst individuals who attempted suicide,
15:15what were their self reported
15:17trends in using mental health?
15:18Services during the year of their
15:20attempt and finally of individuals
15:22who didn't receive services.
15:25We wanted to know what were their self
15:28reported barriers to receiving that care.
15:30Next please.
15:32So just a quick note on our method.
15:34So we used a population based nationally
15:36representative survey data from the
15:38National Survey on Drug Use and Health,
15:40which is a population based survey that's
15:43administered annually by Samsa and it
15:46administers computer questionnaires to
15:48individuals at their household addresses,
15:51or delivers in person questionnaires
15:54by a trained interviewer,
15:56and so our sample included adults
15:58over the age of 18 from 2008 to 2019.
16:01And the Nysda assesses suicide attempts
16:04with a relatively simple question.
16:07It asks in the last 12 months,
16:09did you try to kill yourself?
16:10So in this way it really tries to get
16:12at suicide attempts with fatal intent
16:14and it tries to screen out other types
16:17of suicidal behavior such as non
16:18suicidal self injury and this resulted
16:20in us having a pretty large sample
16:23of close to half a million adults.
16:26Next slide please.
16:28And so for our first question are
16:30what are recent trends in past year
16:33suicide attempts from 2008 to 2019?
16:35The bottom line here is we found that
16:37the suicide attempt rate increased
16:40from 481 per 100,000 individuals to
16:43564 per 100,000 individuals which
16:45corresponded to about a 23% increase
16:48when we adjusted for other changes in
16:50the population over this time period.
16:53And you can see the pretty precipitous
16:55rise in the graph here from 2008 2009.
16:58All the way to 2018-2019.
17:02Next slide, please.
17:04Secondly,
17:04as I said,
17:05we were interested in what risk
17:07factors were associated with this
17:09increase in suicide attempts and
17:10we found a significant increases
17:12in several different subgroups,
17:14but especially among the unemployed,
17:16the unemployed,
17:17their suicide attempt rate more than
17:19doubled over the course of the study.
17:21We also saw a pretty large increases
17:23among young adults aged 18 to 25
17:26individuals with substance use
17:28disorders and individuals who
17:30were not currently married,
17:32and after we tried to control for all these.
17:34Factors and put them all
17:35into a multivariate model.
17:37We found that the time trend of increasing
17:39suicide attempts over the course of
17:41our study still remains significant,
17:43which means that we weren't able
17:45to identify all of the factors
17:48that were responsible for this
17:49rise in suicide attempt,
17:51and there's likely other factors
17:52that we didn't have access to
17:55that were likely contributing
17:56to this increase in suicide
17:58attempts. Next slide, please.
18:01Now, turning to mental health
18:02service use in the last year.
18:04We wanted to know what were the
18:06trends among individuals who attempted
18:08suicide in their use of past year,
18:10mental health services and what we found
18:12was that there was no significant change
18:14in any of the services we looked at.
18:16We looked at outpatient inpatient
18:18and psychotropic medication use
18:20from 2008 to 2019 and if you can
18:22see the the dark blue line here,
18:25this was prescription medication use.
18:26So this was the most common type of
18:28mental health service received by
18:30individuals who attempted suicide.
18:31Anywhere from 40 to 50% of individuals
18:34who attempted suicide in our study
18:36reported receiving some sort of
18:38psychotropic medication in the last year.
18:40Outpatient services were then the
18:42next most common and right around
18:4440% of individuals reported
18:46receiving outpatient services.
18:48A smaller amount received inpatient
18:50services and a much smaller amount.
18:52Received substance use services.
18:54Next slide, please.
18:57And so finally we wanted to know what were
18:59the percent of individuals who reported
19:01needing services but did not get services.
19:03And so in this most recent year in 2019,
19:06we found that 45% almost half of
19:09individuals who attempted suicide reported
19:11needing services but not receiving them.
19:14And when we looked at the
19:16most common reported reasons
19:17that they didn't receive care,
19:18about half of individuals said
19:20they didn't receive care because
19:21they couldn't afford the cost,
19:23about 1/4 of individuals said they simply
19:25didn't know where to go for treatment and 16.
19:27Percent were either concerned about
19:28the opinions of others or were
19:30concerned about confidentiality,
19:31which both are related to mental
19:34health stigma related concerns
19:36about seeking care for.
19:38For suicidality.
19:40Next slide.
19:40So when we look at a summary of our findings,
19:45the first point is there's been a
19:46substantial increase in suicide attempts,
19:48about a 23% rise in suicide
19:50attempts from 08 to 2019 without any
19:53corresponding increase in service use.
19:55We found that while mental illness was
19:57highly associated with suicide attempts,
19:59the greatest increase has actually occurred
20:00among individuals without mental illness.
20:02It's occurred among individuals who
20:04are unemployed and young adults.
20:06Individuals with substance use disorders,
20:08or individuals who are never married.
20:11We also found that only 40% of
20:13individuals who attempted suicide.
20:14At any sort of outpatient mental
20:16health visit in the past year and
20:18nearly half of individuals reported
20:20an unmet need for treatment,
20:22which was largely due to cost barriers or
20:25not knowing where to go to obtain care.
20:28So turning now to the implications,
20:30the first implication of our study
20:34next bullet.
20:37Was that service use in our study was
20:39significantly less than in other studies
20:41that have looked at fully insured samples.
20:44So like I said in our study,
20:45only about 40% reported an outpatient visit,
20:49where in the previously studied insured
20:52samples upwards of 95% reported an
20:55outpatient mental health visit,
20:56and so when we're thinking about research,
20:59it's important to to think and to
21:01remember that that these insured
21:03samples are fully treatment,
21:05seeking samples of individuals who receive.
21:07Care and emergency departments after
21:09their attempt may actually overestimate
21:11the percent of individuals who are
21:13receiving care after a suicide attempt.
21:15But perhaps more importantly,
21:16I think it's concerning.
21:17Given that most of our current
21:20suicide prevention strategies
21:21really rely on healthcare,
21:22contact, our most common prevention
21:24strategies right now.
21:26Include things like a
21:28routine suicide screening,
21:30making sure individuals experiencing
21:32suicidality receive evidence based treatment,
21:34and making sure individuals receive
21:36safety planning before they leave.
21:37Training patient units or our
21:39emergency departments,
21:40but it's impossible to really do any of
21:43these evidence based interventions unless
21:44people are walking through the doors,
21:47and unless we're seeing them
21:48in our healthcare settings.
21:49And So what I'd like to talk
21:51about next is really,
21:52I think,
21:53a potential need to diversify our
21:55suicide prevention interventions
21:56to not only focus on high risk
21:59individuals who frequently have contact
22:01with us in the healthcare system,
22:03but to also look at more public,
22:04health oriented interventions that
22:06begin to address the social conditions.
22:09Which people live?
22:11Next point,
22:12our findings also just demonstrate a
22:14very large unmet need for treatment
22:16during the most high risk period for
22:18fatal reattempts, and so we know,
22:21based on prior literature,
22:22that the most high risk time for
22:24dying after a suicide attempt is in
22:27the year after that suicide attempt.
22:29Yet our study found that nearly half
22:31of individuals are reporting an unmet
22:33need for treatment during this time,
22:35which really speaks to.
22:38Just how big of an issue
22:39access continues to be,
22:41and I know that this is
22:42something that all of you know,
22:43especially those of you who who
22:45work in our emergency departments
22:46or in inpatient units and try
22:48to get care in the community for
22:50individuals after they're discharging.
22:52That access continues to remain a
22:54significant issue and specifically
22:56cost was cited as the number one
22:59barrier to care in our study.
23:01It's also important to note that
23:02most of the data in our study was
23:04collected after implementation of the
23:06Affordable Care Act when we had a lot of.
23:08Different policies and insurance
23:10expansions that were really
23:12targeted at lowering the cost
23:14of care and improving access,
23:16but it is discouraging to see that
23:19cost still remains such a barrier
23:21to care for folks next slide.
23:24And so when we're looking at potential
23:27interventions to to work on in the future,
23:30I think it's important that there's
23:31really kind of two potential
23:32areas that we can look at.
23:34The first is interventions within
23:36the formal healthcare system.
23:38Next slide. And so there are a
23:42number of evidence based strategies
23:43within the formal healthcare system
23:44that we can continue to focus on.
23:46Like I said, routine suicide screening,
23:48not only in psychiatric settings
23:50but also in medical settings.
23:52There's been a fair amount of evidence
23:54to show that brief interventions in
23:56the emergency department and follow
23:58up contact through postcards or phone
24:00calls after patients are discharged
24:02from our emergency departments,
24:04can lower suicide attempts and
24:06suicide rates after discharge,
24:08and we know that suicide safety
24:09planning is a.
24:10And it's based intervention,
24:11so I think we need to continue to make
24:14sure that our emergency departments
24:16and healthcare systems are equipped to
24:18really implement these interventions.
24:20But I think we also need next slide,
24:24public health oriented interventions
24:26outside of the healthcare system.
24:29Interventions that really begin
24:30to address the social conditions
24:32in which people are living and the
24:34social conditions that we know
24:35precede suicidal behavior.
24:36And so in our study,
24:37for example,
24:38individuals who were unemployed
24:40showed the greatest.
24:41Increase in suicide attempts over
24:44the study period.
24:45There's been a fair amount of
24:47evidence to suggest that states
24:49with more generous unemployment
24:51benefits actually so reduced suicide
24:54rate at the state level after those
24:56more generous unemployment benefits
24:59are implemented next.
25:02The same can be true for anti poverty.
25:04Income supports things like temporary
25:06assistance for needy families
25:07or TANF benefits.
25:09There's a fair amount of evidence
25:11to support that states with more
25:13generous TANF benefits actually
25:15see a reduction in their suicide
25:17rate at the state level.
25:19One particularly interesting modeling
25:21study from economist estimated
25:23that we may be able to reduce 3000
25:26suicide deaths in our country.
25:27That's one in nine suicide deaths each year.
25:29If we were to improve.
25:31Increased tanaff benefits by
25:34a relatively marginal amount.
25:36Next slide,
25:37I think we need to continue to
25:38invest in Community based crisis
25:40response which is receiving a lot
25:41of attention right now to make sure
25:43that we're reaching people during
25:45times of crisis and referring them
25:47to an appropriate level of care.
25:50Next,
25:50we need to continue to expand
25:52gatekeeper trainings to make sure that
25:54lay individuals in schools and in churches,
25:56people that frequently come into
25:58contact with suicidal individuals,
26:00feel comfortable doing a basic suicide
26:03risk assessment and referring individuals
26:05to an appropriate level of care next.
26:09We need to continue to look at means
26:11control when we look at systematic
26:13reviews and meta analysis of effective
26:15suicide interventions means control
26:17consistently demonstrates the most
26:19amount of evidence at being able to
26:22reduce our suicide rates and and so
26:24this is mainly focused on firearm
26:27restriction for higher high risk
26:29individuals and also focusing on
26:31safe firearm storage for individuals
26:34who possess firearms.
26:35And finally,
26:36this would be a talk on epidemiology
26:38or health services.
26:40If we didn't focus on data and really
26:42improving our our surveillance systems.
26:43Right now, the CDC is only.
26:47Tracking and monitoring deaths
26:48by suicide and we don't really
26:51have any monitoring systems or surveillance
26:53data on other types of suicidal behavior.
26:56Things like non suicidal self
26:58injury or suicide attempts.
27:00And so my last slide.
27:02I just like to talk a little bit
27:04about these surveillance systems.
27:06And so this is a map of the current
27:09CDC state level surveillance,
27:10and so you can see on this map that
27:13CDC basically tracks which states
27:15have the highest rates of suicides,
27:18and in this map it's the
27:20states with dark blue.
27:22But this sort of map isn't
27:23really all that actionable.
27:25It doesn't really tell us what might
27:27be occurring at the state level.
27:29What might be the high risk populations
27:31in these States and so one alternative
27:33approach would be to really expand
27:35our ability to do county level.
27:37Surveillance next slide.
27:40And so an example of of county
27:42level surveillance may be
27:43something that looks like this.
27:44This was a a recent geographic analysis
27:46that we did trying to identify high
27:49risk counties in the US for suicide.
27:51So all the counties in red are actually
27:54the 50 counties in the US whose suicide
27:56rates are over triple the national
27:58average and whose rates continue to
28:00increase at a rate much at a rate
28:02much higher than the national average.
28:04And so this type of data,
28:06I think,
28:06is much more actionable.
28:07It would allow us to really
28:09target our interventions.
28:10Or grant funding to these particular
28:12counties to begin to understand what
28:14is happening at these counties.
28:16What are the high risk populations and
28:18what might we target to begin to reduce
28:21the suicide rates in these counties?
28:23Next slide.
28:25And so finally, I just want to acknowledge
28:27my my mentors on this project.
28:29Again.
28:30Doctor rosenak.
28:32Doctor Reed and again give special
28:33thanks to the to the Seymour Lossman
28:36Foundation for sponsoring this award.
28:38So thank you so much.
28:45Thank you Tanner. That was great.
28:47Such an important topic and
28:49really very clearly presented.
28:51We do have because in addition to his
28:53other manifold talents that you've
28:55heard about Tanners proven himself
28:57to be a master of time control,
28:59and so we have a minute or
29:00two here for questions.
29:01If anyone has questions or comments
29:04before we go on to our next top.
29:07Power. How much of a factor
29:10do you think that since cost
29:12seems to be so much of an issue?
29:15How big an issue and how much
29:17of a role does the fact that
29:20psychiatrists frequently don't even
29:21don't take insurance must much less?
29:23Medic, Medicare or Medicaid?
29:29Yeah, I think I think that's a really,
29:30really important issue and I think
29:33most of us you know who who take care
29:36of patients on this call know that
29:37it's not necessarily our patients with
29:38public insurance that we have a hard
29:40time finding out patient care for,
29:41but it's our patients with private
29:43insurance and our patients who are
29:44uninsured and that actually in this
29:46particular study we found that suicide
29:48attempt rates have been increasing
29:49amongst those two populations.
29:51Individuals with private insurance
29:53and individuals who are uninsured
29:55and and so the access problem,
29:57I think, really is.
29:59Was focused on on those two two
30:01groups of individuals.
30:08And are you noted that your strongest
30:10predictors of the OR correlates
30:11of the increased suicide rates,
30:13weren't mental health diagnosed?
30:14They were age and unemployment and
30:16the other factors that you listed.
30:18But did you look at mental health?
30:20Did mental health diagnosis
30:22increase in parallel?
30:24And if you look at them as a mediator,
30:26maybe unemployment leads to
30:27an increase in depression,
30:28which mediates the increase.
30:30Made of suicide.
30:33Good question, good question.
30:34We didn't look at it as a mediator and
30:37one of the limitations of the the NIS.
30:40That data set is it? It doesn't.
30:42It's focused on substance.
30:43Use mainly doesn't focus on mental health
30:45and so we only know kind of in the data
30:48set of major depressive episodes as
30:50well as serious psychological distress
30:52which is a nonspecific kind of marker
30:55or indicator of psychological distress.
30:57In the last year.
30:58So it doesn't give us a lot of
31:00information about mental illness.
31:02We did find that.
31:03Individuals with depression and and
31:05serious psychological distress or highly
31:08associated with attempting suicide.
31:09But we didn't actually see
31:11increases in the suicide attempt
31:14rate amongst these populations.
31:16That was more focused on
31:17individuals who are unemployed.
31:18Individuals with substance use disorders.
31:20Like I said,
31:21or individuals who were never married.
31:24Right? Thank you and congratulations again
31:27Tanner for your for your first place.
31:30Last minute award you will be receiving
31:33both a certificate and and a cache award.
31:35Can't give it to you in person today,
31:37but that will be on its way to you.
31:40All right, I'd now like to invite Renato
31:44Pallanti to introduce the first of our
31:47Co 2nd place awardees Daniel Tiley.
31:51Good morning everyone. It's
31:53a pleasure to introduce Doctor
31:55Daniel Kelly. He's a third year
31:57resident and role in the NRP program.
32:01After completing his MDPC training at
32:04the SUNY Upstate Medical University,
32:07under the mentorship of Doctor
32:09Steven and Steven Paul. Then
32:12join my group in December 2019 and since
32:14then it became a key member of my team,
32:18leading several analysis and
32:19contributed to many projects.
32:21Going to happen.
32:23He did all of these also having
32:25a very busy clinical schedule.
32:28Because of his accomplishment this
32:30year, then received the Chairs Choice Award
32:33from the Society of Biological
32:35Psychiatry additionally into
32:372020 also received the D3 Award
32:39from our department to work on
32:42a study to extract high quality
32:44depression digital phenotypes
32:46from the Yale New Haven Health System.
32:48Electronical triggers today
32:51is going to present his work
32:53recently published in JAMA Psychiatry,
32:55where we showed the complex
32:56dynamics linking. Psychopathology
32:59and immune function using genetic data. So
33:02without further ado please then.
33:07Thank you Renato for that
33:09really generous introduction.
33:11I want to make sure y'all can see my
33:14screen that I think I've done it wrong.
33:22So if I do this, can you see my slides?
33:26Presenter View presenter view. How about now.
33:30Hang on one SEC there we go perfect. OK,
33:34so the work I'm going to be
33:36presenting to you today is in the
33:39area of genetic epidemiology and
33:42it's specifically cross disorder,
33:44genetic epidemiology.
33:45Looking at relationships between psychiatric
33:47disorders and some related phenotypes,
33:50not quite disorders, personality traits,
33:53and basically representative.
33:56Sample available sample of autoimmune
33:59diseases, allergic conditions
34:01and inflammatory conditions.
34:03So kind of like running the gamut.
34:06Two things that are important to
34:08know about this is that this is
34:10basically leveraging GWAS data.
34:12So genome Wide Association study
34:14data which looks at the effects of
34:17changes in the letter, the base,
34:19the letter base pair at different positions,
34:22positions in the genome in
34:24association with the disorder.
34:25So these are basically data
34:27that were generated elsewhere,
34:29usually by large consortia,
34:30and I'm using the data,
34:32repurposing it and putting it
34:34together for some novel analysis.
34:40This is the the sort of the citation.
34:44If you're interested in
34:44learning more of the details,
34:45we won't be able to get into many of
34:47the details today with the limited time.
34:50But for some context,
34:51if you search the literature,
34:54you'll find no shortage of
34:56reviews and meta analysis.
34:57Looking at epidemiologic associations
35:00between different immune related conditions
35:03and different psychiatric conditions.
35:06What what I think is not clear yet
35:08is whether there are specificity in
35:10these relationships across disorders,
35:12or whether the patterns are really
35:14more or less the same across disorders.
35:16It's sort of an open question
35:18and something that we're we're
35:20using genetic data to look at.
35:22So epidemiologists are interested
35:24in explaining these associations,
35:26and there's different possible explanations.
35:29One thing that might be happening is there
35:31could be a causal effect of 1 disorder,
35:34increasing the risk for another.
35:36Of course,
35:37it's really hard to demonstrate causal
35:39effects on a population scale, really.
35:41You need good experiments to
35:43to demonstrate causal effects,
35:45and that's just not possible here.
35:48In general, you know,
35:49so one hypothesis is that immune disorders
35:51are causing psychiatric disorders,
35:52and we have some basis to to say
35:55that that might be possible.
35:58So we know from clinical observations that
36:00immune diseases that infiltrate the CNS,
36:03so things like multiple sclerosis,
36:05lupus vasculitis and also auto
36:09antibody mediated encephalitis,
36:12we see those with autoimmune disease,
36:13but also after infections like
36:16COVID and paraneoplastic syndrome,
36:17so these can cause.
36:19Neuropsychiatric syndromes,
36:20but it's important to know that
36:23they're relatively quite rare,
36:25and also usually there are other
36:29neurological findings changes in mental
36:31changes in overall mental status,
36:33fluctuating level of consciousness.
36:35People are ill when they have these.
36:37Oftentimes it's very,
36:38very rare for them to present
36:40with only psychiatric symptoms.
36:44And then we know from experimental
36:46data in humans and in animals that if
36:50you administer an immune disturbance,
36:52if you administer cytokines,
36:53or if you administer bacterial endotoxin,
36:56basically you can induce behaviors
36:58in people and in animals that
37:01that are sickness behavior.
37:03And these are also accompanied by
37:06emotional and cognitive changes,
37:07some of which seem to recapitulate features
37:11of depression depending on who you ask.
37:14And then from animal
37:15models of adult exposure,
37:17but also gestational exposure to
37:19different types of immune disturbance,
37:21we know that there are
37:23neurodevelopmental effects and there
37:24are social behavioral effects.
37:26The figures that I'm showing here
37:29so that the top right is a pet FDG
37:33study looking at it was actually a
37:36a pet dopamine binding study looking
37:39at changes before and after the
37:41administration of of interferon alpha.
37:44In the context of treating hepatitis
37:46CI believe and the bottom is looking
37:48at newborn neurons in a mouse model.
37:53I another hypothesis is that psychiatric
37:56disorders are somehow causing immune
37:58disorders or contributing causally.
38:04Clinical observation points to the idea
38:07that stress might be important here.
38:10We know that acute psychological stress,
38:13a laboratory stressor like the tree
38:15or social stressor produces autonomic
38:18changes and neuroendocrine changes,
38:20and also measurable changes
38:22in peripheral immune markers.
38:24So when you stress a healthy participant,
38:25there are transient increases
38:27in I-1 beta IL 6 TNF alpha.
38:30What's really?
38:31I think interesting.
38:32That individual differences in the
38:35subjective experience of that stressful
38:37experience actually are correlated
38:41with different different aspects
38:43of the changes in the peripheral
38:45immune milieu and and the anti
38:47inflammatory cortisol and milieu.
38:51We also know from human studies that
38:54self reported stress or objective
38:57measures of deprivation or poverty are
39:00predictive of symptom severity in humans.
39:03For a number of immune related disorders,
39:05and we know from animal models of
39:07immune related disorders that if you
39:09if you subject those animals to stress,
39:11it worsens the histological
39:13progression of disease.
39:15And that's probably best understood for
39:17asthmatic conditions and and inflammatory.
39:20Health conditions in
39:21intermetallic conditions.
39:24And there's also the possibility
39:27of bidirectional effects.
39:29That's actually suggested by
39:30some of the longitudinal studies
39:32that look at the temporal.
39:34You know the temporal patterning
39:35of these associations,
39:37and then I wanted to mention
39:39here that there's a large
39:41body of literature looking at.
39:43Psychiatric samples people as
39:45seen for psychiatric disorders,
39:47and they show group differences in a
39:51variety of peripheral immune markers,
39:54so it's not really clear what
39:55the significance of that is.
39:56We're so we're sort of left with
39:57a chicken or egg problem there.
40:00There are other explanations too.
40:03Another possibility is that there's
40:05some shared underlying biology,
40:06so maybe there's a shared
40:08pathological mechanism,
40:09or there could also be
40:11correlated genetic mechanisms.
40:12So so different mechanisms,
40:13but they travel together on the genome
40:15because they're so close to each other.
40:17It's also possible that there's
40:19some external causal factors.
40:21Some third variable effects going on.
40:24If things that come to mind
40:25are environmental exposures,
40:26social determinants of health.
40:29And a related hypothesis is that.
40:33There is actually a true causal factor
40:35that's just correlated with one of
40:38your phenotypes and causal for the
40:40second or mediation is possible,
40:42so one phenotype may cause may cause.
40:47As the actual causal phenotype
40:49leading to the outcome,
40:51and so when I think about this possibility,
40:53I think about things like
40:55health related behaviors.
40:57So substance use aspects of BMI,
41:01exercise, diet, sleep,
41:02duration,
41:03but also social connectedness and then
41:06the social determinants of health.
41:08And then individual differences in
41:10the stress response.
41:11Also possible mediator here.
41:16For completion, it's also possible that
41:19mediating or confounding effects are
41:21happening in the other direction too.
41:23So these are relevant for
41:25understanding the next.
41:27The next slide here.
41:29So what we did was we assessed genetic
41:33correlations between psychiatric
41:34and immune related disorders,
41:36and we did this using the LD
41:38score regression method.
41:39So this method basically looks at all
41:41of the positions on the genome and
41:44looks at the association for disorder
41:46A&amp;B and and looks to see if they're
41:49significant and if they're in the same
41:51or opposite directions of effect.
41:52And it kind of.
41:54That summarizes that data
41:56across the whole genome.
41:57To describe the proportion of shared
42:01heritability between the disorders,
42:03I know that's a lot.
42:05I'm going to kind of walk you
42:07through it because the slide is busy,
42:09so we found a predominance of
42:12positive genetic correlations.
42:14They were modest in strength.
42:16The correlation coefficients were between,
42:18you know, point.
42:20Basically,
42:21.8 to point .20.
42:24.08 I'm sorry to .2 and we saw kind
42:29of a clustering of correlations
42:31involving the inflammatory bowel
42:33disorders but also one of the
42:36biliary disorders and lupus,
42:38and those seem to be positively
42:40related to schizophrenia and
42:42mood and anxiety disorders.
42:44We also saw predominance of positive
42:46relationships for asthma and hypothyroidism,
42:48and those included a slightly
42:50different set of psychiatric
42:52phenotypes for some immune phenotypes.
42:55Like celiac disease and another
42:57of the biliary diseases we saw
42:59sort of a mixed pattern across
43:01disorders with some positive and
43:04some negative correlation across
43:06different psychiatric traits.
43:07And then some of the immune
43:09related disorders seem to have a
43:12predominantly negative correlation
43:13across across the psychiatric traits.
43:15At least the ones that were
43:17available to measure.
43:19And.
43:20Another thing that we did here
43:22is we included genetic data for
43:24these additional risk factors that
43:26I showed you on the last slide.
43:28So the health related behaviors,
43:29the sleeping behaviors,
43:30and what we found was that many of
43:35these potentially confounding or
43:37mediating phenotypes show significant
43:40genetic correlations with both
43:43immune and psychiatric disorders.
43:45The effect sizes were stronger for
43:47the psychiatric disorders on average.
43:52So the next thing we did was we wanted
43:55to use genetic data to try to answer
43:58the question of whether whether a is
44:01causing B or B is causing A and the
44:03way we did that was with a method
44:05called Mendelian randomization,
44:07and it's a complex method.
44:09I can't explain all the details of it, but.
44:12The brief version is that if we assume
44:15that the causal loci for phenotype,
44:19if we assume that we know the
44:20cause of LOCI for phenotype A,
44:22we can then look at phenotype B.
44:24Those same positions on the genome,
44:26and we can.
44:27We can ask, do those also show risk
44:30for free and type B in those positions?
44:33And if that's true,
44:34there might be a causal effect
44:35in that direction,
44:36and we can ask the opposite question.
44:38The causal loci for phenotype B?
44:40Do they show any evidence
44:42of effects in phenotype a?
44:43So it's a it's a more directional analysis.
44:46So for all of the 44 correlations
44:48I showed you in the last slide,
44:50we basically performed this
44:53analysis bidirectionally.
44:54So testing both directions of
44:55effect and we found a total of
44:58nine significant effects here,
44:59and again I'll walk you through
45:01this slide because it's a bit busy.
45:03So eight of these effects were
45:06effects of psychiatric.
45:08Patrick phenotypes on immune phenotypes
45:10and seven of them were positive,
45:12so the solid Red Arrows that you
45:15see on the figure are all pointing
45:19from various psychiatric phenotypes
45:22to various immune phenotypes.
45:26Specifically,
45:26we saw effects involving major depression,
45:29schizophrenia,
45:30and also this process disorder phenotype,
45:33which is composed of I think,
45:358 different psychiatric disorders.
45:37But but depression and schizophrenia
45:39figure heavily in those because
45:40they're some of the largest samples,
45:42so the effects are modest.
45:44The average odds ratio was 1.1.
45:47The largest effect that we saw was
45:49an effective depression on asthma
45:52and then the smallest effect was
45:54the cross disorder affect on.
45:56Asthma,
45:56so the way to understand these
45:59is the genetic liability for the
46:02psychiatric trait increases increases
46:04the genetic liability for the immune
46:06related trade and the window when
46:08you'd like to do it.
46:11One of these, one of these
46:13relationships was negative,
46:14so risk tolerance was associated
46:17with decreased genetic liability
46:19to allergic rhinitis for hay fever.
46:25And we had some effects that did not
46:28survive the sensitivity testing and
46:31then the multivariable adjustment.
46:33So you can see that sort of
46:36dashed blue line on the bottom.
46:38It's dashed because the relationship
46:40was interrupted and the things that
46:42interrupted it were all of the Gray
46:43phenotypes on the bottom that have
46:45little lines pointing toward it.
46:46So body mass index,
46:48cognitive processing all of those
46:49seem to attenuate the relationship,
46:51suggesting that the relationship could
46:53potentially be mediated by other factors.
46:56Right?
46:57It's up to him. Can I ask people to
47:01mute please? Other than that, thanks,
47:03yeah we had time.
47:06We we saw one we saw one effect
47:10of hypothyroidism on MDD.
47:11It was it was not significant
47:14after multivariable adjustment so.
47:16And so we didn't make.
47:17We didn't consider it a robust basically.
47:21So interpretation discussion.
47:24Basically, these genetic analysis
47:26find stronger support for the
47:28idea that psychiatric disorders
47:30might be contributing causally
47:32to immune related disorders.
47:34We found relatively little
47:35evidence in the opposite direction.
47:39There are a number of limitations
47:41that I don't necessarily think
47:43we need to get into here,
47:44but in context, so there is.
47:47There is epidemiologic
47:48support for MDD and asthma.
47:50There is less consistent epidemiologic
47:52support for schizophrenia and the
47:54inflammatory bowel disorders and there's
47:57really no literature linking risks
47:59risk taking behavior to hay fever.
48:01And we do replicate the results of some
48:04prior Mendelian randomization studies,
48:06which is encouraging.
48:09Our results do need to be reconciled with.
48:12A lot of epidemiologic studies that find
48:15that a prior immune disorder increases the
48:17risk for a subsequence like yatrik disorder,
48:20so the temporal patterning in these
48:22studies is opposite what we what.
48:24What our data suggests so.
48:28I was trying to think about
48:29ways to explain that,
48:30and one of the the line of thinking
48:33that I followed is that you know
48:35on the whole these immune,
48:37these immune related disorders,
48:39particularly like the inflammatory
48:40bowel disorders.
48:41They're rare, whereas psychiatric
48:43symptoms are relatively common.
48:46And it's possible that people who are
48:49being ascertained for an immune disorder
48:52have some amount of subclinical,
48:54undiagnosed psychiatric illness
48:56ahead beforehand that that is
48:59likely just based on the how common
49:02psychiatric symptoms are.
49:06We also have to reconcile our work
49:08with some some studies that look
49:11at cytokine concentrations and
49:13psychiatric disorders that seem
49:14to suggest the cytokines are are
49:16causal for psychiatric disorders.
49:20And then future studies in this
49:23in this area are going to need
49:26to include these third variable
49:28confounding or mediating phenotypes.
49:30And because they exert outsize
49:33effects on the data in the future,
49:35I would love to integrate markers of HP,
49:38a access functioning inflammatory markers,
49:41immune cell counts,
49:42there's GWAS data for these sorts of things.
49:44And then of course, it's possible that
49:47larger data sets may actually yield.
49:50May actually yield bidirectional effects or
49:53effects of immune on psychiatric disorders.
49:56So that.
49:57It's my study and thank you for listening.
50:01Thanks for the department and the
50:04Lesbian Foundation and family and.
50:07The award the the panel who
50:10made the decisions.
50:11Yeah and thanks Renato and
50:12my lab mates for getting you
50:14know making this possible.
50:17Thank you Dan.
50:18Really interesting work with some
50:21some provocative and counterintuitive.
50:23Things to think about in
50:26the interest of time,
50:26we're going to move straight on.
50:27Can I ask you into?
50:31To close your smile. Yeah.
50:37And we will move on.
50:42So our other Co Co 2nd place
50:45winners are a pair of of of cranes,
50:49Zach Harvey Neck and Terrell Holloway,
50:52and they've asked me to
50:54briefly introduce them.
50:55Terrell and Zach are both senior
50:58residents in the Neuroscience Research
51:00Training program and have really been
51:02wonderful members of the program,
51:03not only with the the
51:05science that you'll hear,
51:06but also as clinical leaders and also
51:09as just wonderful citizens and and and
51:12and leaders in the residency as a whole.
51:14In the NTP in particular.
51:16Terrell did his bachelor's at
51:18Brown and his MD at Mount Sinai
51:21before and during his MD.
51:22He did some really first
51:24rate basic science work.
51:25Looking at neurotransmitter
51:27mechanisms with with relevance
51:28to some of our recent interests
51:31in psychedelics interesting.
51:33And then when he came here,
51:34he shifted his focus to to what
51:36you're going to hear about today,
51:38which is the focus of social
51:40determinants of health and in particular
51:42race and racism on psychiatric
51:44outcomes and this combination.
51:46Of a focus on social determinants
51:48of health and a deep understanding
51:50of underlying molecular mechanisms.
51:52Puts Terrell in a really special
51:54position to be able to do syncratic
51:56synergistic work with the type
51:58you're going to hear about today.
52:00So I Carbanak did his bachelor's at Duke
52:03in biomedical engineering and biology,
52:04and then he did his PhD in at Michigan,
52:08where he studied the aging in the
52:10lowly fruit fly chrysocolla and when
52:12he came to the NTP here at Yale,
52:14he he kept that focus on aging.
52:17But switched to something we like to
52:19think of as a little bit less slowly
52:21that the human power human mechanisms
52:23of aging and how they're influenced
52:25by stress and psychiatric phenotypes.
52:29I love that the two of them
52:31Co submit this paper.
52:32I can't remember that happening
52:34in the previous last minute.
52:35Last minute words,
52:36but it it speaks to this this strength
52:39not only of their science but also
52:41of their their collaborative spirit,
52:44and it's great to see to see their
52:47work synergizing like it does here.
52:49I know they're going to acknowledge
52:50their mentors during the talk,
52:52but I want to give a shout out
52:53to Derrick Gordon.
52:54Would you just Sinha and could shoot?
52:57Metric this one.
52:58Terrell is that take you awhile.
53:01Thank you very much for the introduction,
53:03Chris. And thank you all for
53:06taking part in this presentation.
53:07Zach and I will be alternating the
53:10slides throughout and so you know just
53:12as a means of just giving you a sense of
53:14how this is going to go back and forth.
53:16Just presenting that,
53:16could you go to the next slide, please?
53:20We have no disclosures to to share in
53:22regards to any conflicts of interest and
53:25and just as it means of like how big
53:28this slide this this presentation was,
53:30we had to curtail it a little
53:32bit just to accommodate time.
53:33And so to start off.
53:35As we all know,
53:36there's been a lot of epidemiologic
53:38data that has demonstrated this
53:40long standing mortality gap
53:41between black and white Americans.
53:43As we can see on the slide on
53:45the the graph below to the left,
53:47we see that though that there's
53:49been a decreasing.
53:50Uh, decreasing age and mortality,
53:52or decreasing deaths per 100,000 over
53:54the years that have gone on since 2014.
53:58But we still see is a persistent
53:59gap between the life expectancy
54:01of black and white Americans.
54:03Similarly looking to their right
54:05and the CDC as reported in 2020,
54:07this health life expectancy
54:10gap has actually persisted.
54:11And also in light of the the
54:14pandemic has actually increased.
54:15And so.
54:17Given the fact that there have also been
54:20external factors such as differential
54:22access to care and institutional bias,
54:24that may inform some of these,
54:26these disparities growing
54:28literature has shown that increase,
54:30increasingly,
54:31that biomarkers of health also are different,
54:34and these two different populations,
54:36just to name a few as.
54:39Doctor Tylee House was able to
54:42introduce earlier sciatic kinds of
54:43health and signatory cytokines.
54:44In particular have been shown to be
54:46increased in African American populations.
54:48In addition to.
54:51Increases and changes in.
54:55Methylation and also in telomere length
54:58in regards to how short and telling
55:01your life like in African American
55:03populations and so in trying to understand
55:06what it is that's driving this,
55:08this mortality gap.
55:09One of the ways we're that we were
55:11looking at what we were interested in
55:13looking to see is how epigenetically
55:14if this will also pan out to
55:16show these differences and may
55:18inform this mortality gap.
55:20Next slide please.
55:23Estrel so one way we can conceptualize
55:26these health disparities is is
55:29a accelerated biological aging.
55:31Because black individuals in
55:32America might be dying earlier
55:34because they're aging faster.
55:36Prior work has shown that that
55:38life experiences like trauma and
55:40life lifetime adversity can lead to
55:42accelerated biological aging and and
55:44we know that black Americans face
55:46higher levels of stress and trauma.
55:48Recent development developments and
55:50epigenetics have led to epigenetic clocks,
55:53one of which is called brimmage,
55:54that we'll be using here,
55:56which are sets of epigenetic markers
55:58that have really been trained to
56:00predict aging related outcomes and
56:02allow us to estimate a biological
56:04age even in healthy populations.
56:06Next slide.
56:09So taking both of those things
56:11together and we were interested
56:12in in a Community example here in New Haven,
56:14in analyzing to see whether or not there
56:17was any difference in this genetic clock
56:19between black and white Americans,
56:21and also what potential factors may
56:24inform this difference in aging rate.
56:26And so if you could go to the next slide,
56:28please we actually through
56:30the yellow stress Center,
56:32had recruited a cohort of about 400 healthy
56:37volunteers and essentially had assessed.
56:40Various psychological managers through
56:42the cumulative adversity inventory which
56:44is 140 interview survey that assessed
56:47traumatic life offense chronic stress,
56:50major life events and recent life events.
56:53In addition, that's not shown here.
56:54We had assessed health through self
56:57reported Cornell Medical Index survey,
56:59which is another 195 battery question
57:02assessing their health in addition to.
57:04Collecting biological samples,
57:06whole blood,
57:07from which we we we extracted
57:10DNA and also cortisol,
57:12ACTH and other biomarkers of health.
57:17Next slide, please.
57:19And so, like Trell was saying,
57:22we wanted to do this in
57:23a in a healthy cohort,
57:24we this population has no chronic diseases,
57:27no substance use outside of nicotine or no
57:31substance use disorders outside of nicotine.
57:33They're not on any prescription medications.
57:35We did limit our analysis to only
57:38individuals who identified as black or white
57:40through the sample size considerations,
57:42and these groups did have some
57:44notable differences in in BMI,
57:46alcohol use, and years of education,
57:48but they were similar in gender.
57:50Age, income and current health
57:52status and I do want to emphasize
57:54that health status aspect.
57:56Uh, you know.
57:56The Cornell Medical Index is is a
57:58very comprehensive questionnaire of
58:00medical and psychiatric symptoms,
58:01and we saw no significant
58:03difference between black and white
58:05participants on these measures.
58:07Next slide.
58:09So when we look at the psychological
58:11measures and compare the two,
58:12what we find is that black participants
58:14report more stress and more trauma in
58:17comparison to their white counterparts.
58:19Next slide, please.
58:20And when we then correlate
58:22that with the biological data,
58:24what we see in the left side is that
58:27what we know that that grim age is
58:29correlated with advanced health in
58:31both black and white populations.
58:33But when we are actually doing
58:35the linear correlation of this,
58:37we find that actually race
58:39accounts for about a one point.
58:42UH-75 year different or one point
58:44scuse me my side is acting up about
58:48a 1.73 year difference between
58:50black and white participants.
58:51Again echoing Zach's point of
58:53this being a pretty much like a
58:5630 year old healthy population.
58:58This is a significant difference
58:59between the two groups.
59:01Next slide please.
59:02And so similarly when we look at trauma,
59:05what we find is a similar thing of that.
59:07Essentially black Americans
59:09also report more trauma.
59:11And when we subcategorize the
59:13trauma by type next slide, please.
59:15What we find is that there is an increase on.
59:18There's an increase in all four
59:19of these different categories.
59:20Assault of violence being 1D for
59:24the second one being other injury or
59:26shocking event warning of traumas,
59:28or death of a relative or loved one.
59:30We wanted to separate these by
59:32subtype only just to get a nature of
59:34the trauma and to see whether or not
59:36there are any differences in regards
59:38to the nature of the trauma and how
59:40it informed this type of aging.
59:43Next slide,
59:43please.
59:47And since we're seeing these differences
59:49in lifetime stress and and trauma
59:51and epigenetic aging between our
59:52our black and white participants,
59:54we we next wanted to ask whether
59:56stress overall on the left and
59:58traumatic events in particular on
01:00:00the right predicted a grim image in
01:00:03both black and white participants.
01:00:05And the answer is is yes,
01:00:06we see that a higher stress and
01:00:09and trauma predict higher grim age.
01:00:12But as these plots emphasize,
01:00:13the black participants had higher stress,
01:00:15higher trauma.
01:00:16And grim age when compared to
01:00:20white participants next slide.
01:00:22So next we used mediation analysis
01:00:24to ask whether the higher levels of
01:00:26stress and trauma were responsible,
01:00:28at least in part for the differences
01:00:30in grim age between black and
01:00:33white participants.
01:00:33And to give a brief
01:00:35orientation to these diagrams,
01:00:36the indirect effect explains
01:00:38how much of the relationship,
01:00:40basically how many of those years
01:00:42are explained of the relationship
01:00:44between race and grammage is
01:00:46explained by the mediator.
01:00:48The numbers by the arrows are the
01:00:50coefficients in the model and the
01:00:51direct effect is how much of the
01:00:53relationship is unexplained by the mediator.
01:00:55Now,
01:00:55while both are are highly significant,
01:00:57you'll notice that when comparing
01:00:59the magnitude of the indirect effect,
01:01:02the trauma subscale is essentially.
01:01:04Equal to that of the total
01:01:07cumulative adversity inventory,
01:01:08and it's notable that these
01:01:10results hold hold true when after
01:01:13we account for differences in BMI
01:01:15alcohol use in years of education,
01:01:18we went on to do some further analysis
01:01:20that we don't have time to show you today,
01:01:22but those really demonstrate
01:01:23that assaultive trauma,
01:01:24in particular when when Terrell is
01:01:26breaking down those categories,
01:01:27was a particularly powerful mediator of
01:01:30this relationship between race and age.
01:01:32Acceleration next slide.
01:01:35So to to summarize our study,
01:01:37we've shown you today that black
01:01:39participants have higher grim age
01:01:41and face more cumulative stress,
01:01:43and in particular traumatic stress
01:01:45when compared to white participants.
01:01:47And these higher levels of trauma
01:01:50partially explained the increased
01:01:52scrimmage we observe in black participants.
01:01:55It's worth noting that these
01:01:56findings are in a relatively young
01:01:58and remarkably healthy population,
01:02:00suggesting that the biological
01:02:01embedding of stress and trauma in the
01:02:03epigenome is occurring prior to the onset of.
01:02:06Of illness,
01:02:07and it's also worth noting that there's
01:02:09specificity of these findings to trauma.
01:02:12We didn't see differences between
01:02:13black and white participants in
01:02:15recent life events or or sort
01:02:18of the chronic daily stressors.
01:02:19And while black participants had a
01:02:21higher prevalence in all categories
01:02:23of traumatic events,
01:02:24assaultive trauma was a particularly
01:02:26potent mediator of the relationship between
01:02:29race and increased scrimmage acceleration.
01:02:31Next slide.
01:02:33And so the the implications of
01:02:35these findings are are very broad.
01:02:37And first of all,
01:02:38given our limited sample size,
01:02:40we're only able to compare
01:02:42black versus white Americans.
01:02:43And so and also as described
01:02:45previously mentioned,
01:02:46this is one of the first to actually compare
01:02:48this specific clock grim age in a young,
01:02:51healthy cohort sample.
01:02:53These health disparities, though,
01:02:55that they're so observed later in life
01:02:57this this study has emphasized the fact
01:03:00that essentially that these epigenetic
01:03:02changes happen a lot sooner than.
01:03:04When we start seeing these burdens
01:03:06of disease start to take hold in
01:03:08late adulthood and then furthermore,
01:03:10I think kind of point to adolescence as a
01:03:14means of an intervention point in trying to.
01:03:18To try to like start to slow down
01:03:20some of these insults from taking
01:03:22hold in regards to epigenetic age,
01:03:23in particular in regards to the second point,
01:03:27these are particularly relevant
01:03:28and amongst African Americans who
01:03:31are 22% more likely experience of
01:03:32violent crime and more than twice the
01:03:34likely to experience a violent and
01:03:36lethal encounter with law enforcement
01:03:38and in particular to the events of
01:03:40the of what happened last weekend.
01:03:42Understanding how these affect
01:03:44personal health would be one of
01:03:46the things that is paramount.
01:03:48Understanding how this life expectancy gap
01:03:50persists over such a long period of time.
01:03:53Furthermore,
01:03:54as as we have kind of alluded to,
01:03:56given the fact that these there's
01:03:59still a persistent direct effect
01:04:01of of race on accelerated aging,
01:04:04it opens the door for us understanding
01:04:06and house how some of these more complex
01:04:08sociological interactions like racial
01:04:10trauma and experience discrimination,
01:04:12also impact health.
01:04:13And since in this study we weren't
01:04:16able to actually assess or ask about.
01:04:18People's experiences as a turn
01:04:20as it related to discrimination.
01:04:21I think one of the things of which
01:04:23that would be really interesting to
01:04:25look at is how bees unique experiences
01:04:27also affect the genetic age and age.
01:04:31Next slide, please.
01:04:33And so with that.
01:04:34As we've rushed through
01:04:36this this very dense paper,
01:04:39we would like to thank our mentors,
01:04:41Derek Gordon, Michael Black also
01:04:43helped support us in the in this.
01:04:46In this sax mentors.
01:04:48If you.
01:04:51And also Chris said and young
01:04:53son also have have been.
01:04:55Instrumental in regards to helping
01:04:57guide us in regards to kind of
01:04:59developing this project and and what
01:05:00steps to do after this project is
01:05:02complete and finally we would like
01:05:04to thank the the lesson family and
01:05:06foundation for this opportunity and
01:05:08both of our financial grants in
01:05:10regards to supporting this work.
01:05:17Thank you both for a great presentation.
01:05:19And again, I applaud not only the
01:05:21the science, but the collaboration.
01:05:23Well done, we are in the interest of
01:05:26time going to go straight on to our
01:05:29last presentation on our honorable
01:05:30mention to the loss of an award.
01:05:32And I want to invite broad right
01:05:35to introduce the entire Anderson.
01:05:40Thank you so much
01:05:41and good morning.
01:05:42My name is Darwin Boatwright and
01:05:44I and assistant professor in the
01:05:46Department of Emergency Medicine,
01:05:47and I've studied bias and
01:05:48discrimination and medical education.
01:05:50Today I have the distinct honor of
01:05:52introducing doctrine and Tara Anderson,
01:05:53who will present some of our work
01:05:54on the prevalence and influence of
01:05:56microaggressions in medical school.
01:05:58Doctor Anderson has been a long
01:06:00standing member of the old community,
01:06:02having graduated from Yale University,
01:06:03Yale School of Medicine,
01:06:05and is now a member of the illustrious
01:06:07Department of Psychiatry at Yale.
01:06:08I thought the honor to collaborate.
01:06:10And learn from Doctor Anderson.
01:06:11Since we first met in 2015 and we
01:06:13began discussing the possibility of
01:06:15science and medical literature examining
01:06:17racism and medicine
01:06:18and brainstorming ways we could
01:06:20address this gap in knowledge.
01:06:21These discussions produce some
01:06:22of the work Doctor Anderson
01:06:23will share with you today.
01:06:25Examining the prevalence of
01:06:27microaggressions and their
01:06:28association with medical
01:06:29student mental health and also
01:06:31medical student satisfaction.
01:06:32Doctor Anderson.
01:06:36Hello everyone, we're happy to be here and
01:06:38we'll get started on the presentation.
01:06:40It's going to be a quick one
01:06:42so next slide please Chris.
01:06:45So I want to start with
01:06:47psychiatrist Chester Pierce,
01:06:48who originally defined microaggressions
01:06:50directly quoting from him.
01:06:51He defined them as the subtle, stunning,
01:06:54often automatic and nonverbal exchanges
01:06:56which are put downs of blacks by offenders.
01:06:59The offensive mechanisms used
01:07:01against blacks are often innocuous.
01:07:03The cumulative weight of their
01:07:05never ending burden is the major
01:07:06ingredient in black white interactions.
01:07:08Today migrations are regarded as casual,
01:07:11verbal or nonverbal slides,
01:07:13whether intentional or unintentional,
01:07:14which communicate.
01:07:15Rogatory messages based on the targets
01:07:18modules, marginalized group membership,
01:07:19and here are some examples of
01:07:22different types of microaggressions,
01:07:24illustrated from the I2 hashtag
01:07:27I2M Harvard campaign.
01:07:28So you can see them here.
01:07:31So our study objectives.
01:07:33Next slide please.
01:07:35We sought to characterize the experiences of
01:07:37microaggressions for US medical students.
01:07:39How frequently did they occur,
01:07:41and who was most likely to experience them?
01:07:43Next slide.
01:07:46We wanted to know if microagressions
01:07:47we want to test the association
01:07:50between microaggressions and medical
01:07:51student well being and mental health.
01:07:54Next slide.
01:07:55We also wanted to assess the
01:07:58association between microaggressions
01:07:59and medical school satisfaction.
01:08:01Next slide,
01:08:02we distributed a cross sectional
01:08:04Internet based anonymous survey
01:08:06to medical students in the US
01:08:08between 2016 and 2017.
01:08:09Questions on microaggressions were
01:08:11adapted from the racial ethnic
01:08:13microaggression scale and we asked
01:08:15if respondents had experienced
01:08:16various types of microaggressions
01:08:17how frequently they occurred and the
01:08:19reasons they believe they were targeted.
01:08:21These questions did not directly mention
01:08:24racism, prejudice, or demographics.
01:08:26And participants were instructed to
01:08:28include microaggressions committed by
01:08:30fellow students, residents and faculty,
01:08:31and so here's an example of one of
01:08:34those questions. Next slide, please.
01:08:38Questions on medical school satisfaction,
01:08:40like this one, were adapted from the
01:08:43Institutional Betrayal Questionnaire
01:08:44and then afterwards respondents filled
01:08:46out the PHQ 2 PHQ 2 depression screen,
01:08:49which is 2 item depression screen
01:08:51and their demographic information.
01:08:53Next slide. Here are the demographics
01:08:55of our final sample.
01:08:57We had 759 respondents from
01:08:59over 100 medical schools,
01:09:01so it was a substantial sample.
01:09:03Next slide.
01:09:05What we found the experience of
01:09:08microaggressions was incredibly common.
01:09:1098.7 participants reported
01:09:12having experienced at least one
01:09:13microaggression in medical school,
01:09:1733.9% reported having experienced
01:09:18a microaggression almost daily
01:09:20in medical school.
01:09:22Next slide.
01:09:24The most common attributions of
01:09:25the reasons people felt they were
01:09:27experiencing microaggressions were gender,
01:09:29race, ethnicity, and age.
01:09:31As you can see here next slide.
01:09:35So respondents also described the
01:09:37microaggressions they had experienced,
01:09:38so here's 1 the respondents said
01:09:40I had an older male physician
01:09:42as a first year mentor.
01:09:44I shattered him every other week in his
01:09:45clinic for my intro to clinical medicine.
01:09:47Required course he would regularly
01:09:49introduce me to patients as a pretty
01:09:51face to talk to while you wait.
01:09:53Next slide,
01:09:54I choose to wear my hair and it's
01:09:56natural state sometimes and one
01:09:57of my professors made a comment.
01:09:59Did you get electrocuted after that
01:10:01comment was made him and another
01:10:03professor preceded to laugh.
01:10:04Next slide.
01:10:09Next slide, please.
01:10:12So here are the result and more on
01:10:15the experiences of microaggressions.
01:10:17About half the respondents experienced
01:10:19at least one microaggression a week.
01:10:21So we divided the cohort into two groups,
01:10:23a higher exposure group which experienced
01:10:25microaggressions at least once a week and
01:10:28a lower exposure group which experienced
01:10:29microaggressions less frequently.
01:10:31As you can see,
01:10:32students who identified as black,
01:10:34Asian, multiracial and being a
01:10:35female sex were the most likely to
01:10:37be in the higher exposure group.
01:10:39We also performed an intersectional.
01:10:40Analysis of race and sex assigned
01:10:42at birth which showed that white
01:10:44males reported the lowest average
01:10:46microaggression frequency scores and
01:10:48black females reported the highest
01:10:50mean microaggression frequency scores
01:10:52within all racial or ethnic groups.
01:10:56Mean microaggression scores were
01:10:57lower for meals than for females.
01:11:00Next slide please.
01:11:02So here also this is the correlation
01:11:04between the the PHQ 2 depression screen
01:11:07and the frequency of microaggression.
01:11:09So we divided the microaggression
01:11:11frequency score into 4 quartiles,
01:11:12and we found that as the frequency of
01:11:16experiencing microaggressions increased,
01:11:17the likelihood of respondent having a
01:11:19positive depression screen increased
01:11:21quite substantially in a dose response
01:11:23relationship even after adjusting
01:11:25for demographic factors such as race,
01:11:27ethnicity, sex, SES,
01:11:29urine, medical school,
01:11:30clinical experience, etcetera.
01:11:32Next slide,
01:11:33please.
01:11:35Compared to the lower exposure group,
01:11:37respondents with higher microaggression
01:11:38exposure were less likely to recommend
01:11:41their medical school to friends
01:11:42less likely to donate to their
01:11:44medical school after graduation,
01:11:45and less likely to consider staying
01:11:47at their institution for residency.
01:11:49Next slide, please.
01:11:51In addition,
01:11:52higher exposure respondents were over
01:11:53three times more likely to have this.
01:11:56This is past behavior three times more
01:11:58likely to have missed class because
01:12:01the environment was unwelcoming.
01:12:02They were also four times,
01:12:04nearly four times.
01:12:05More likely to consider medical school
01:12:07transfer and medical school withdrawal
01:12:09compared to the lower exposure residents.
01:12:12Next slide please.
01:12:14So implications increase frequency
01:12:17of experiencing microaggressions
01:12:19may impact medical students
01:12:21health mental health negatively.
01:12:23They are common experience and
01:12:24bipac and female medical students
01:12:26experience them more frequently and
01:12:29students with multiple marginalized
01:12:30identities such as black women
01:12:32are impacted the most next slide.
01:12:36As our data on avoiding class or
01:12:38considering withdrawal from medical
01:12:39school suggests the experience of
01:12:40microaggressions may increase the
01:12:42attrition of diverse medical trainees,
01:12:44which would imperil efforts to
01:12:45diversify the physician workforce,
01:12:47possible interventions to decrease
01:12:49microaggressions include education,
01:12:51improved reporting and remediation,
01:12:53and incorporating racial attitudes
01:12:55assessments into admissions and hiring.
01:12:58And finally,
01:12:58we want to propose reevaluating
01:13:00the use of the word microaggression
01:13:02and replacing it with specific
01:13:04terms such as racism or sexism.
01:13:06Our findings show serious associations
01:13:09between microaggressions and
01:13:10outcomes for medical student,
01:13:12and we hope our research helps
01:13:13dispel the ideas that these
01:13:15microaggressions are a lower priority
01:13:17or or an inconsequential issue in
01:13:19medical education and training.
01:13:21Next slide you can click
01:13:23through the references.
01:13:25There's about 3-4 pages,
01:13:26so I want to thank thank my mentors,
01:13:28down Boatwright and Anna Reisman,
01:13:30the Lussman Family Foundation
01:13:32and award committee.
01:13:34Of course, our illustrious coauthors.
01:13:37Especially the brilliant doctor let
01:13:39and you can go to the last slide.
01:13:42And also of course,
01:13:43thank you to my friends and family,
01:13:44especially my mom and dad who
01:13:46are actually tuning in from Sri
01:13:48Lanka and who are pictured here.
01:13:49So and thank you to you all for listening.
01:14:01Thank you so much the Antara rounding
01:14:05out our presentations. And I see your.
01:14:08I see your parents there welcome.