Why are Diabetes and Depression Linked? | Sherita Golden, M.D., M.H.S.


So I actually think about we
have a series at Hopkins called Telling Stories About Science. And so, that’s sort of what
I want to do in the next 15 minutes or so, is sort of tell you how a
conversation I had with a mentor early in my career led to
a whole field of research. So, the title of my talk is
why depression and diabetes are associated, connecting
the brain to the pancreas. And so, I sorta hope to take
some of the basic science that you’ve heard this morning and sort of tie that in to some
translational work that we’ve done in the area of
population science. So I like to show this slide. And you may wonder why I’m
showing you a slide with the Hard Rock Cafe and a little baby
that looks somewhat like me. And so the reason is
because right before I started my faculty
position at Hopkins in 2000, I went to the American Diabetes
Association meeting with my mentor, Fred Brancati, and it
was held in San Antonio Texas. And it was memorable for
a variety of reasons. One is that it was the first
time I actually left my son, who was about nine months old,
with my family, and actually left
the state of Maryland. But the other memorable
thing is that Fred and I went to eat at this
Hard Rock Cafe in San Antonio. And it started to storm terribly
while we were in there and neither of us had umbrellas, so
we were forced to sit there for about 2 hours and I remember
during the conversation Fred asked me he said you know
I’ve been preparing all this literature about a link
between diabetes and depression like why would
they be associated? Why do you think
this might happen? And sort of the neuro endocrine
training that I had just had in my endocrine fellowship kind
of got the wheels spinning and that’s what I’m going
to talk to you about. So we probably all recognize
this very famous painting by Edvard Munch,
called The Scream. And I know that many of us have
often felt this way in various scenarios. So if you add the bleach to your color by accident
when you’re doing laundry you woke up at the time you were
supposed to be leaving for work. I don’t know if that’s ever
happened, but you wake up and realize the room is
way to bright for the time it’s suppose to be. And interestingly
it’s kind of fun for me to be at the National Press
Club cuz my son who’s now a junior in high school wants
to be a print journalist. He’s been very frustrated that
many of his writers have not submitted their articles in
time for the pending deadline. Cuz he’s the deputy editor
of the school newspaper. But on a more serious note. Sir William Osler was
sort of the founder of medicine as we know it. Today and the first chairman of
the Department of Medicine at John’s Hopkins University. And he wrote a very famous book
called, Osler Principles and Practices of Medicine, that I
actually got a copy of when I graduated from medical school. And he wrote that book
back in 19, sorry in 1892. And had a chapter
on risk factors for type two diabetes and obesity. And so, what’s interesting is
if you actually look at the left side of the column you recognize
that many of these are what we consider traditional risk
factors for type two diabetes. But what’s interesting
is on the right column, he was a bit visionary,
he recognizes nervous strain or worry as risk factor for
diabetes. And actually before
he recognized it, so did doctor Willis, who actually,
there’s a circle of vessels in the back of the brain named
for him the circle of Willis. He also had recognized this and it was even called
the diabetogenic personality. So, but moving forward to the,
sorta the turn of this century, there had been
a lot of work showing that there was an association between
diabetes and depression. And about a year after I
became interested in the work. There was a meta analysis
showing that individuals with diabetes are about twice as
likely to have depression as individuals without diabetes. And the lifetime prevalence of
major depression is higher in those with diabetes compared
to those without diabetes. This is actually also
a common association in type I diabetes as well, so
it’s relevant to adolescence. So 15% to 20% of adolescents
with type I diabetes will have elevated depressive of symptoms. About 23% have sub clinical
depressive symptoms, and there’s a study called a search
study which is the largest US based epidemiological study
of diabetes in children and adolescents and actually there
are similar rates of depression and type one and type two
diabetes in children and adolescents maybe slightly
higher in type 2. And so, one of the questions is
most of those studies were cross sectional so it sort of begged
the question of the chicken and the egg which one came first. So this is kind of tricky
because if you think about the natural history of
the one set of type one diabetes is typically in late
childhood to early adolescence. And then type 2 diabetes tends
to be a disease of middle age. But then depression’s
interesting. It has its onset between toward
the end of adolescent and young adulthood. And then there’s a second wave
of depression that occurs in mid life as well. So you could see that one might
actually predict the development of the other. And back at the turn
of the century, we were really unclear which
direction was the primary. So one of the first questions
I was interested in and looking at is does diabetes
predict the development of depression. And so you might sort of
hypothesize they would be linked because perhaps is the
psychological burden impose by diabetes that can increase
the risk for depression. So we know that diabetes in
the setting of new complications Particularly if there
are complications that are associated with
functional impairment and in a setting of a lack
of social support and passive coping skills, certainly
can lead to depression. But as an endocrinologist
with a biological background, it turns out that
hypergylcemia itself, which is associated
with diabetes. Tends to actually adversely
effect the parts of the brain that control mood and cognition. So, specifically, the
hippocampus, where there can be atrophy of neurons and
also neuronal apoptosis. So, that’s why it’s not uncommon
in diabetes that we can not only see depressed mood but we can
also see cognitive impairment. So, one study I was interested
in looking at does diabetes predict depression was in
the multi-ethnic study of atherosclerosis, so this is
a very large population based study of cardiovascular disease. But a lot of the population
also has diabetes and these individuals were recruited
from six centers throughout the US and
they were truly multi ethnic, so they were 45 to 85 at
the time of recruitment. They represented African
American, Hispanic Americans, Chinese Americans and whites. And they had no clinical
cardiovascular disease at baseline, so they were fairly
healthy cohort In our turfs. And so, what we were able to do
because Masa had several follow up visits was to actually look
at the presence of diabetes on the first visit. And then subsequently and we defined them according to
their glucose tolerance status. So, at that first visit,
do they have diabetes, do they have impaired
glucose tolerance, or do they have normal
glucose tolerance. And then, we follow them,
and at visit three, there was a depression measure. And we were actually able to
determine whether or not they had depression based on their
answers on a questionnaire and whether or not they were using
medications to treat depression. We also had that information
on them at that first visit. So we were able to exclude
people that had depression at baseline to sort of
remove the confounding. And so, what we found is that
individuals with diabetes at baseline had about a 50% higher
risk of developing depression during followup compared to
those without diabetes and that it was independent
of differences and complications, socioeconomic
status, and obesity. So it looked like that diabetes
did predict depression, but about depression
predicting diabetes? And so this actually will
sort of answer a little bit the question that was asked at
the end of the last session. So here again we had several
plausible biological mechanisms. So one again, the depression and the psychological
burden of the disease my lead through obesity through
a variety of mechanisms. There are also the behavioral
aspects, so typically individuals who are depressed,
don’t like to exercise. They may tend to over eat, maybe
less inherent with the doctors and weight loss recommendations,
that can lead to obesity and insulin resistance. In addition, some of
the treatments that we use for depression can actually
induce obesity. So particularly some of the
atypical antipsychotics are now used as booster therapy along
with the serotonin reuptake inhibitors to induce obesity and
insulin resistance. Finally, what I was
most interested in as an endocrinologist. Was the impact of
chronic stress and depression on
the neuroendocrine system, so specifically elevations
in cortisol, your catecholamines hormones and
inflammatory markers. So we, and again, went back to
the Mesa Study and this time we looked at individuals at
baseline who had depression who did not have diabetes. So we excluded
prevalent diabetes. Defined the same way as they
were in the other half of our analysis. And again we followed them for
two years to look for the development of
incident diabetes. And what’s interesting, individuals with depression
had a lot of risk factors for type II diabetes. They were less
physically active, they consumed more
calories during the day. They were also more likely
to be current smokers. And smoking, even though you may lose weight,
be less obese as a smoker, smoking’s still an independent
risk factor for diabetes. They also had higher
body mass index and higher inflammatory markers. But the question was,
if we control for all of those factors,
do we see an increased risk for type 2 diabetes and
in fact we did. Individuals who had depression
at baseline had about a 21% higher risk of developing
type 2 diabetes. But it wasn’t all explained by
those factors we adjusted for. Which, sort of, made me to think
that there was still a missing link and like what is the
missing link between the brain and the pancreas
that we may make. And that’s what I wanna spend
the last few minutes talking to you about. So this is a body stress system. So Dr. Hussein and
I didn’t really plan to have talks that were somewhat
complimentary, but it actually was very
helpful to set the stage. So in his example, the gazelle was being chased
by a pack of cheetahs. So that does activate our
fight or flight response, and so what typically happens is
that, what happens is that you have any type of stress, here,
activates the hypothalamus and that stimulates the release
of hormones that femulate the pituitary gland–it’s
a little small gland sitting at the base of your brain–to
secrete additional hormones that ultimately stimulate your
adrenal glands over here. Hold on. I think it’s this one. I’m a little IT challenged, so
let me go back to where I was. Sorry about that. [LAUGH] Hold on. I need to stop using the laser
pointer that’s always a thing. So we were right here, so the little triangle
sitting at the bottom, that’s your adrenal gland. That is what ultimately
makes your cortisol. And the other thing that happens
is when that axis, that’s called your HPA Axis, when
that’s activated, it stimulates your sympathetic nervous
system to be activated also. So you have an increase
in cortisol, and you have an increase
in catecholamines, which you may refer
to adrenalin. So if you are that poor gazelle
being chased by the cheetah, this system is activated. But then,
once the gazelle escapes, this should all turn off. So this is a tightly
regulated system. So then the cortisol feeds back
to the pituitary gland and the hypothalamus
just shut this off. Because we don’t want this
system activated all the time. But then what we think happens
in disorders like depression and chronic stress is that you have
this chronic activation of the HPA access. You also, this leads to activation of the
sympathetic nervous system and they co-activate each other, and
that leads to an increase in the levels of cortisol
that I mentioned. And interestingly, when you
activate the sympathetic nervous system, it also activates
your inflammatory system. So you have an increase
interleukin 6 levels. And so interleukin 6 is
sort of the beginning of a very progressive
cytokine cascade. And all of those factors
are associated with insulin resistance. And so that can actually
lead to the development of type two diabetes. And specifically with cortisol
not only does it lead to insulin resistance directly, but it can also lead to
the development of central fat. So we’ve heard about different
kinds of fat this morning, white fat, brown fat But there’s subcutaneous fat
that was discussed earlier. There’s also visceral fat
which is the fat sitting, wrapping around the organs. And it turns out that cortisol,
there are a lot of, cortisol receptors on that
fat and that It motivates and the development of more
fat around the belly. So that kind of central fat
is also associated with insulin resistance. So this is sort of my schematic
of how the hormones might be linked to the development
of diabetes and depression and stress. So again, one of the metabolic
perils of having central fat, so what’s shown on this particular
slide, if you look to the left, you can see that
white on the CT scan. That’s like central visceral fat
that’s all wrapped around your organs. That’s the fat that’s associated
with poor metabolic outcomes. If you look to the right,
you can see that there’s sort of this panel of subcutaneous
fat that’s on the outside. The thought has been that that
fat is not as metabolically active as what’s in the viscera. But I know that there’s some
emerging literature that it probably likely is
contributing as well, but it’s really the visceral fat. So we know that that’s
associated with higher glucose levels and
insulin resistance. It’s associated with
higher blood pressure. More lipid disorders and
also are at-risk for developing type-II diabetes. So what does it mean when
we say insulin resistance, because we’ve used that term
a lot this morning, and so this is just a really
simple schematic. That our insulin receptor,
we have normal insulin action. Our insulin receptor basically
takes the glucose that we break down from food and
brings it into the cells, so our cells can have energy. And what happens in the setting
of insulin resistance or abnormal insulin action is that
there’s a failure of glucose to enter those cells, and so
our cells don’t have energy, and they have to use other
sources for energy. So is there any sort of data
to support the cortisol hypothesis in relation to
depression and diabetes? So this is where it’s kind of
fun to go to the animal models, and I always am jealous of my
basic science colleagues because they can knock out
various things and you can’t really do that
in human subjects research, because the IRB doesn’t really
like that type of thing. [LAUGH] So but this is a very
nice animal model of an enzyme called 11-beta-hydroxysteroid
dehydrogenase. And what this enzyme does is
it inactivates cortisol, so it keeps your body from being
overexposed to cortisol. But what they did in this mouse
is they made a mouse model that overexpresses that enzyme, so this mouse fails to then
activate its cortisol. So you can see in each of these
the wild type mouse is shown on the left, and
then the transgenic mouse is shown on the right, and you can
see clearly that the transgenic mouse that’s making too
much cortisol is heavier. And then it’s also more insulin
resistant, is more obese, and actually has higher levels
of cortisol in it’s liver circulation. So then what evidence
do we have in humans? So we actually have
two sets of evidence. So one is in our
clinical studying and. And so this is sort of the model
I thought about when Dr. asked me what is the link
between depression and diabetes because what
we see in clinical is there’s a disease
called Cushing syndrome. This is a picture of
a woman who has it. You get Cushing syndrome from
making too much cortisol from either a pituitary tumor or
an adrenal gland tumor. And so what happens is these
individuals develop central obesity, so you can sorta
see the central fat they actually develop muscle
wasting, insulin resistance and type two diabetes and
interestingly they can also develop mood disorders
which is very fascinating. So that’s one setting we see it,
we also see the development of diabetes in insulin resistance
if we give patient high does steroids for
inflammatory conditions. So like rheumatoid arthritis or
after they’ve had some type of transplant, we’ll often
give high dose steroids. Those patients are very prone to
developing type II diabetes and insulin resistance. So, the thing that’s been
really interesting for me, and sort of several post-doctoral
fellows that have worked with me is that we’ve actually been
able to look into multi-ethnic study of atherosclerosis, and we have a sub-study called
the Mesa Stress Study. And what we did is that we had
individuals at two Mesa exams collect salivary cortisol. So you can actually measure
cortisol in saliva by having your research participants
chew on cotton swabs. And then you can actually
centrifuge that saliva, and you can measure it. And one of the challenges with
cortisol is that it’s variable throughout the day
like many hormones, so it’s highest in the morning
when you first wake up, and then it drops down
lower later in the day. So we actually had to measure
the cortisol curves over the course of the day, and
we did this in about a thousand people and had them collect 6 to
8 samples throughout the day so we could characterize
the whole curve. And what’s been fascinating to
us is that even independent of depression, we see evidence of
dysfunction of the cortisol excess in the setting of type
two diabetes in obesity. So what we found is that in the
setting of type two diabetes, instead of having that robust
cortisol response when you wake up in the morning,
is actually very blunted and we see that same cortisol
profile in depression. In addition, there’s a loss of variability in
the cortisol throughout the day. So instead of it
reaching a peak, it tends to be higher
at the end of the day. So we see that in diabetes, and
we also see that in depression. So they actually share a very
similar neuroendocrine profile. And then, the one question
that we’ve been able to sort of ask is okay,
if you have diabetes, does it change your
cortisol access over time. Or is it that your cortisol
access predicts a development of diabetes, so this is the same
question we asked earlier but without the depression. And so the jury is still out, so we have done one study showing
that if you have diabetes, it does not predict a change in
your cortisol curve over time. And now we’re actually gonna ask
the other question as soon as we accumulate more diabetes cases. So you may say, so what why
do we care about how our hormones maybe related
to metabolism sort of in thinking about everything
we’ve talked about this morning. And I think it’s because despite
the fact that we have a lot of known effective lifestyle and
other interventions that help to prevent diabetes there is still
a large diabetes epidemic and we’ve really not been
able to fully Implement everything into the environment,
so it really begs the question, are there other biological
targets that we should consider. So I would argue that it may be,
because we see these neuroendocrine abnormalities
In diabetes and obesity, that we may be
able to actually modulate he neuroendocrine system as a
novel way to complement existing strategies for preventing and
treating diabetes. And, in fact, there are sort of glucocorticoid
receptor antagonists, so those are the receptors
that cortisol binds to. There are agents that
antagonize that bonding. And we don’t know yet how
they’re related to obesity and diabetes risk, but I think that’s an important
future area of investigation. And then from a clinical
perspective, because I’m sorta ending my story, I think it also
says though, that in addition to understand the biology, that
it’s important for us to sort of step back and think about how we
care for patients with diabetes. Because it’s not uncommon for the mental health comorbidities
to be a part of that. So we actually are gonna have
to change the way we deliver clear to these patients, so we have more of
a collaborative care approach. One of the very first patients
that I saw when I was a young faculty member called me one
day on a snowy afternoon and said, I have been living with
type 1 diabetes for so long, I just can’t take it anymore and I’ve just given myself 100
units in my insulin pump. This is a true story. I was actually able to call
911 from my house and send, I said don’t come to my house,
go to hers, and intervene. But this is a reality that when
I’m not sort of doing molecules that I actually see in
my clinical practice. So ultimately we wanna move
our patients from feeling like this to looking more like this.>>[LAUGH]
>>[LAUGH] So>>[LAUGH] So I would just like to acknowledge
all of my collaborators over the years as well as
my funding sources. So thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *