High Cost of Losing Weight With GLP-1s: Researcher Kosali Simon, Ph.D., Explains

(00:00)
[instrumental music plays]

Mark Masselli (00:04)
Many say GLP - 1s is miracle drugs. They're reducing waistlines and have the potential to reduce healthcare
costs, but our guest says we need to carefully examine the data before leaping to conclusions.

Dr. Kosali Simon (00:17)
What we're asking is a very narrow question. We are asking about healthcare costs that can be observed over a
period of time that we can observe. So here are many caveats. If we want to look at five - year follow - ons of
people who have started taking medications, we are going to have to only comment on formulations that were
around a while ago.

Margaret Flinter (00:42)
Joining us is Kosali Simon, who earned her PhD at the University of Maryland. She's a distinguished pr ofessor
and associate vice provost for health sciences at Indiana University.

Dr. Simon (00:53)
We find evidence of small increases in other costs. And now there are a few other studies also that have these
same findings, so we feel like there's a growing consensus that that's what people are seeing.

Margaret (01:10)
This is 'Conversations on Health C are.'

Mark (01:23)
Dr. Simon, welcome to 'Conversations on Health C are.'

Dr. Simon (01:26)
Thanks very much for having me, Mark and Margaret.

Mark (01:29)
Oh, that's great. We want to have a conversation with you about all of your research interests, but let's start with
the economics of GLP - 1 medication. I wonder if you could tell our audience how you conducted the research.

Dr. Simon (01:42)
So, we have done research on what the healthcare cost consequences are for people taking GLP - 1s by looking at
hundreds of thousands of people who have been taking these medications. So, this is going to be very different
from looking at clinical trials. This is in the real world, as people start taking them, what is their usage like? We
look at people after they've started taking them. We compare them and their healthcare use and cost trajectories
to people who haven't started taking them but are kind of similar in tha t we see those people taking them
afterwards, so they're a good control group for looking like people who are similar. What we find is that even
though we know these medications have costs, we're looking for: is there evidence that their other healthcare
c osts will go down. We do not find evidence of that. We find evidence of small increases in other costs, and
now there are a few other studies also that have these same findings, so we feel like there's a growing consensus
that that's what people are seeing .

Margaret (02:58)
Well, this is completely fascinating because, as I'm sure other people have said to you, it flies in the face of what
you would assume to be true. And certainly, in the healthcare community, we know that obesity in diabetics, but
in others as well, is associated with so many adverse health outcomes. I guess one of the questions would be, is
it a case of we're not looking at short - term benefit, or we are looking at short - term benefit versus long - term?
Does your economic model allow for the fact that over time, and we don't have all that many years to date, that
the curve would begin to drift downward again as the consequences of untreated obesity develop maybe a little
bit later in life. Does the model allow for that kind of perspective as well?

Dr. Simon (03:50)
Definitely, and I've got to emphasize, what we're asking is a very narrow question. We are asking about
healthcare costs that can be observed over a period of time that we can observe. So here are many caveats. If we
want to lo ok at five - year follow - ons of people who have started taking medications, we are going to have to
only comment on formulations that were around a while ago. Because newer formulations, we just don't have

yet real - world follow - ups. Even if we're talking abo ut, "Oh, by 2026, we should have data," researchers
typically have access only to data that's about two years old anyway, so we don't get today. So, a lot has to be
said about the narrowness of what we're asking. It in no way questions the benefits that ha ve been seen in
clinical trials. Enormous, enormous benefits in lots of ways.

Margaret (04:44)
And a really important point to make, that this is what we can say about what happens with healthcare costs in
this period of time after being on GLP - 1s, period . Let's talk about future studies.

Mark (04:56)
And when it cures cancer or cures some Alzheimer's disease or something like that, you'll do another study that
will look at the potential impact, right?

Dr. Simon (05:05)
Definitely. We just have to keep t racking because there may be things that are illnesses that their healthcare
consequences would have been seen many years down. And certainly, we're just not going to be able to answer
anything long - term until that long - term happens.

Mark (05:24)
Sure. Ab solutely. Let me push a little on this, because there's another study that found middle - aged adults taking
GLP - 1 drugs for obesity can save over close to $200,000 on lifetime medical costs. Doesn't this contradict your
findings?

Dr. Simon (05:40)
I'm glad you've asked because this is where I get to say, oh, here's a different way we could answer that
question. So that paper that you're referring to is a simulation study, and so you could have an answer off of,
let's actually look at what people are doing. And then another way to answer the question is what might we
anticipate by thinking of a thought process? A thought question is, suppose that everybody who is clinically
appropriate for GLP - 1s gets access to GLP - 1s today. And let's suppose the benefits the y're going to see are
going to be as we see in clinical trials. That is, adherence is not a problem, access is not a problem, all the ways
in which they are going to take it, all as the clinical trials, and that they are going to see these benefits. Then y ou
ask, what might that look like in 20 years, 30 years, 40? You can ask these kind of simulating questions, and
even the studies then say, well, we know there's going to be mortality, longer, healthier lives. And those benefits
we value not just as are th ey reducing medical costs, but if we have somebody living a year longer, regardless of
what's happening to their medical costs, we have a benefit. And so, we can still do a different calculation, which
is not asking what's happening to medical costs, but h ow do we value the benefits of extra life that could be
provided? So thought process simulation versus real - world evidence.

Margaret (07:28)
And I would imagine that our social scientist friends, public health friends, epidemiologist researchers, are
look ing at things like years of life lost, quality of life, ability to work, and all of the other factors that play in.
Tell us how your research is evolving. Is this work leading to other avenues of research within the whole areas
of obesity and drug manageme nt?

Dr. Simon (07:57)
Yes, I think there is so much interest among, of course, society, researchers, no different, trying to answer these
questions. And really, our limit is the data available, right? So, we can talk about how expensive the medications
ar e, and I'll say the data that researchers need to answer these questions also are hard to obtain and are expensive
in what's available. But with what is available to researchers, for the most part, we are starting to ask more
questions as we get more year data. So, we've got, as more people are taking these medications, more ability to
look at subgroups. We've got numbers of people that we can study where otherwise it might be, "Oh, that's a
rare person in my data source." Now we've got a lot more. We can a sk, well, what if they were in these
circumstances? So, I think we'll just continue to see more of subgroup analyses, more of longer - term, more of
cutting into the different formulations, the different stages of medications. What do we see with tirzepatide ,
because we can really only start to look at even semaglutide in very more recent data.

Mark (09:14)
A junior social science question. If you're doing better, GLP - 1 starts to really make an impact, and nothing else
has worked for you, is it possible you're saying to yourself, "Let me check up on this other thing that I hadn't
really gone to before, becaus e I've got something that's really impactful in my life, and I've never had..." We're

hearing a lot of people say, "I've never had this type of results." Maybe they have another problem they want to
get fixed. They're going in for more because actually thi s works.

Dr. Simon (09:45)
Yes, those are rightly what we are catching, is the monitoring that goes on once you have now started taking
GLP - 1s. What we are seeing in our data is you're seeing more monitoring type of visits and costs. And so those
have pot ential to catch other things that, again, in further downstream years, it might be that that is going to
show up as a reduced cost, whereas now what we're seeing are the visits that are incurred. It's all wait to be seen.
We can only say what it is we can see with the data in front of us, and we all know there's huge other benefits
that are probably not just showing up in healthcare costs. And we're trying to value them and observe them and
workforce productivity. Where can we get that data to be able to li nk people? Those are -

Margaret (10:44)
Anyone who stops the conversation at just the tie between cost of drug and associated costs in healthcare and
thought that was the be - all and end - all doesn't belong in research. They need to be someplace else. We've read
that some say the work suggests that we're holding GLP - 1s to a higher standard than we've held other
medications by putting this sort of burden on, given the cost, given the volume of people on it, that we really
have to look closely at the economic t ie. What's your answer to that? I've sort of reflected on that and tried to
think of another similar situation, and I honestly couldn't think of another situation where the economics has
been right up there as close as efficacy, at least in the beginning. So, tell us about that.

Mark (11:36)
How about the HIV drugs?

Margaret (11:37)
HIV, the cost of the drugs wasn't really the issue , in my memory , so much when they first came out. But what
do you think about that?

Dr. Simon (11:45)
Yeah, and here it matt ers a lot whether we think about the medication for the indication. So, when we talk about
GLP - 1s for diabetes, it's a very different story than when we talk about it for weight loss. When we look at
programs that have baked in rules that say medications f or weight loss are not covered. And that is a very
different way of treating a medication and asking, "So what's it doing? Let's now come up with specific cost
analyses for these medications." It's not asked of other medications, where it's just, well, we think there are great
clinical trial outcomes. They show efficacy. Let's go ahead and cover it. And the kind of questions that are asked
of GLP - 1s or the way we see that formularies or insurance coverage programs are approaching GLP - 1s are very
different, and we can point to Medicare and Medicaid to see this. So, for the diabetes indications, all programs
are covering, but for the weight loss indications, it's only about a dozen states' Medicaid programs that cover it,
and some have been dropping it. And th en in Medicaid, as you know, it is only a temporary demonstration that's
covering it - Right ... in the bridge program.

Mark (13:06)
So, think of covering, you cover a lot of territory, and let's go into another area, aspect of your work. You're part
of th e NIH - funded project that examined what happens when healthcare providers have the same demographic
factor as patients, and what happens when those factors are different. And this was part of a, I understand, an
Alzheimer's disease project. Tell us more ab out...

Dr. Simon (13:31)
So yes, there is research that is - we are looking at how the healthcare system functions for Alzheimer's care, for
caregivers, and there's a lot to examine there. The aspect about the providers and the patients, we are still
actua lly in the field on that, so I don't have that much to say on that particular question, except that there's quite
a lot in, it's almost you can draw this common feature. When it comes to many illnesses, we can point to proven
strategies or healthcare that exists as technology, medications. And then when we come to actually studying
how does the human being, how does a patient connect to these care systems? There's a whole lot to study in
how does insurance coverage, how does every factor that is in between the solution and the patient matter. So -

Margaret (14:35)
Well, we certainly -

Mark (14:37)
We know that from -

Margaret (14:38)
We know that well from our experience in -

Dr. Simon (14:39)
And you know well from the work you do .

Margaret (14:42)
...community health, but an area that we've been deeply involved in for decades really is the opioid epidemic
and how to effectively get people into treatment and retain people in treatment. So, sort of along those same
lines, I know you've had research in that area as well. What's the best way, maybe you could say from an
economist's perspective, to help those who are addicted?

Dr. Simon (15:08)
So, another area where we know there is effective medication for opioid use disorder. And these are fairly low
cost compared to the costs on society, employers, individual families, and there's this big puzzle. Why is it that
uptake is low? And that has a lot to do with understanding are there providers nearby who will prescribe? How
does one get connected? How doe s one have a sustainable method of payment? And all of, as you were saying,
all of the community aspects that go into connecting people with the solutions.

Mark (15:47)
We were fascinated with your op - ed in The Wall Street Journal that really focused in o n the Affordable Care
Act and the debate, that the subsidies perhaps were too large. And you believe the system is designed, the way
it's designed is the problem. What was your proposal to fix this problem?

Dr. Simon (16:10)
Yeah, so here's something, I t hink. It's when you start to look at all the formulas baked in, then you start to
think, "Oh, how does the very particular way something is designed actually leave there to be potential ways to
improve?" So, one aspect of the subsidies that may not be very well known among people is the way it sort of
caps how much an individual is exposed to of the price in a way that may lead people who are setting the prices
to say, "Oh, I don't need to really compete on the prices because the person who's going to make that choice
doesn't see the full price variation." So, we pick a percent of my income that is going to be paid at maximum,
and then the rest of the cost is going to be the taxpayer's cost. And so, a system that designs like that, economists
will point out, "Oh, that doesn't necessarily lead to as much price competition. I might compete instead on, oh, I
want to be selected by the features of the program." But it could be, and it's just throwing out there a suggestion.
There are lots of suggestions on how to reform the creation of subsidies, but one aspect to look at would be, is
there a way to get more price competition by the companies that are setting the prices?

Mark (17:37)
And there certainly was no rhyme or reason when Congress came together, the poli tical struggles of how this
was finally shaped, so they certainly weren't drawing on economists, or if they were, we'd love to talk to them.

Dr. Simon (17:53)
I really think that when you design a program, you can't anticipate the so many ways something m ight play out.
And then we just have to be in this rapid cycle of, okay, we see this. We see prices increasing. Let's go rethink.
And this may not be the only, lots of solutions may come up, but we're in this continuous process of learning.

Margaret (18: 13)
Well, I think all of our health strategies, and Medicaid is one of them, we do well to think upfront about what
are the questions we ought to be looking at in terms of impact evaluation and outcome of decisions. And we've
certainly had Medicaid as an i mportant part of our work and really value the Medicaid leaders when they have
thought about having those kinds of studies upfront. One area that you looked at relative to Medicaid was
changes in chronic medication dispensing to children and young adults f ollowing COVID, at a time when
Medicaid was removing many from the program, was during the Medicaid unwinding period. What did you find
in that particular work?

Dr. Simon (19:00)
So, we found that during this time of the unwinding, that is, we know that there were these automatic extensions
given during the COVID era, and that once we went back to the rules as normal, that would mean that a lot of
people would be subject to renewal and likely lose coverage, and we saw many are dropping. And we see that
w hen insurance changes, there is a period of readjustment where there's disruption in coverage. And so, we saw
evidence of some people then having to be on cash, and then cash brings with it the future consequences of, now
we've got somebody on cash, their likelihood of continuously taking medications is not as high as if they had an
insurance source that kept their prices to be less out of pocket. So, we saw those disruptions happen, and it's not
surprising. Other instances of research where there's been in surance churn, insurance changes that happen, show
also that there is disruption that comes.

Margaret (20:10)
And if I could just a follow - up question. When you're conducting research like that, are you looking nationally?
Are you looking by individual Me dicaid programs? Maybe just give our audience a little more insight into the
depth and the nuance that you have there.

Dr. Simon (
Yes, absolutely. Yes. What many economists have done in studying the connection between insurance and how
people take their medications is to typically look for large national data and then look for something that looks
like an experiment. We need to have something that's a shock to insurance, that isn't just happening to
everybody at the same time. And so, when there's Medicai d program changes, we know they're large ones that
happen at the state level. Some states do this, and some states don't, or some states do it in a different year. And
that gives us a comparison. We look at the states that changed compared to the states th at didn't change, and we
say, "Oh, that's more likely like a natural experiment we could learn from."

Mark (21:14)
Margaret, you were sort of asking about how one's conducting research. Doctor, tell us a little bit about your lab
and your colleagues. Are you in an institute? Or I think our listeners would be fascinated to know behind the
scenes.

Dr. Simon (21:27)
Yeah, sure. So, I'm here in Indiana University Bloomington, and across the hall from me here, there are PhD
students, post - doctoral scholars, th ere's other faculty here. And what we do is we do teaching and research. So, I
teach undergraduate students in a healthcare management major. And in addition to that, we do research with
our time that's given by the university for research. We look for que stions that we feel we are able to contribute
to, and we get together as part of the doctoral training that students get. Many of them are coauthors on our
work. We look for in what way can we help give some light? Can we shed some evidence that might have a
bearing? Sometimes it's just descriptive trends. What do we know about how something has been evolving? And
another way we try to approach things is, can we give an answer as if we were doing a randomized control trial?
Except we can't. We just look for when does nature give us an opportunity to say - ... "Oh, this is very likely the
causal effect of insurance."

Margaret (22:46)
I wonder if I could tie your research to one of the national projects that we've been involved in, and that's the
NIH All of Us project. You're probably thinking the same thing, which we were participants in from the earliest
days and continue on through this period of time. Unique , for our audience, if they don't know about it, unique
in that it's been continuous over many years. It's with a stated goal of bringing in people who are like all of us in
this big melting pot of a country, but that the data is made available to scientists for research. And I'm curious,
do you and your teams in health economics research have any interes t in that dataset, given that it is publicly
available and covered so many factors? And to us we just think Dr. Collins really just made a great contribution
to science and health with that work. Tell us a little bit about that.

Dr. Simon (23:46)
It's a g reat dataset, and I'm a participant too in it. And I have been part of a project at the university that's been
trying to help other people also get access and connected to All of Us. And so, there's a lot of promise there. It's
somewhat different, as you k now, from other datasets in that we don't have a sampling frame, so it's not like we
know, oh, this is everybody on Medicare program or such. So, understanding who is a participant and what that
tells us about what we can study. But the unique data there, having all the biomarker data and being able to have
more linked outcomes, that just expands the kinds of questions we can ask. So, it is really a great program, and I
think it's so wonderful. I've heard more recent news about its future outlook that sound s very promising.

Mark (24:45)
We are part of its future. We're excited about it. So, I asked you about everybody else around you, but tell us
about how you got interested in health economic research. What was your trajectory like?

Dr. Simon (24:57)
Yeah, so I think that when we want to talk about specific medications or specific aspects you see in healthcare,
everybody can see in their own lives, their own families, how important healthcare is. And then, I think just
studying economics just seemed to me l ike an area where there's no end to the questions we might have about
how we want to see things become better, how we believe that we can be talking with people about things that
matter to their lives. And so, this has been an area of economics that I foun d very fulfilling in the way that you
can just connect with so many people and interests and think about what we can do, and there's just a lot we can
keep doing. It's always, as you do on your side, on the healthcare and community, this is going to keep u s busy.

Mark (25:57)
So, speaking of economics, there's been perhaps some downsizing of the support that's available for graduate
students and for research. Has that impacted the Indiana University in any way, or your department?

Dr. Simon (26:14)
I think every university is assessing and trying to understand what are the changes, and there are just many
aspects of it. We still try to continue to see, well, right, even with whatever we've got, what's the best way we
can try to find answers? Because we 've just got to keep doing that. But it's not going to change it. I imagine that
lots of generations have had these questions of, how's it going to be two years down the road? What should I do?
And we just try to adapt and be resilient.

Margaret (26:47)
W ell, one of the areas, I know we don't have much time left, that is always a little bit challenging is the role of
partnerships with industry when you're doing research that is affected to a product. And many, I think, of the
pharmaceutical companies have addressed that by setting up foundations or separating out the entity that's doing
the research. The GLP - 1s, obviously, industry is part and parcel of all of it. That's where the drugs came from.
How do you engage with industry in your work as a health eco nomics researcher, using the GLP - 1 work as an
example?

Dr. Simon (27:25)
So certainly, at the Aspen Health event that was this summer, I got to hear and see what you typically in a
university wouldn't get to hear, or an academic conference wouldn't. So, i t's exciting to see what there is from
the commercial side, to see what new innovations are coming up. And I'll also say, as I'm getting ready for a
university networking event on Thursday, where we're trying to get social scientists, along with bench scie ntists
who are involved in the GLP - 1 and related medications and metabolic health area, to come together and say,
"Can we try and understand from each other?" So even within academia, there are lots of these kinds of silos
where we want to learn, and I thi nk that's the same with academia and industry. We want to learn and see what
should we be thinking ahead to our research questions to study.

Margaret (28:27)
And we ask ourselves the same question, I think, within our research institute and within the co mmunity health
center system, which is 32 million patients. Do we have more in common than different when it comes to
looking at these questions? And I think the health economics research maybe has not been as much a part of the
focus and really needs to b e. So, thank you so much for that work.

Mark (28:50)
Well, and thank you, Dr. Simon, for joining us for this interesting conversation, and remember to subscribe to
our videos on YouTube. Find us on Facebook and X. You can also share your thoughts and comm ents about this
program. Take care and be well.

(29:05)
[instrumental music plays]

Margaret (29:08)
This copyrighted program is produced by "Conversations on Healthcare" and cannot be reproduced or
retransmitted in whole or in part without the express wr itten consent from Community Health Center, Inc. The

views expressed by guests are their own, and they do not necessarily reflect the opinion of "Conversations on
Healthcare" or its affiliated entities.

(29:29)
[instrumental music plays out]