THINKWELL

AGEWELL

When the Brain Gets Better: Rethinking Cognitive Decline

Dr. Heather Sandison, ND (00:01.314)
Welcome back to the ThinkWell, AgeWell podcast. I am your host, Dr. Heather Sandison. And today I’m joined by Dr. Stephen Chamberlain, a truly interdisciplinary physician scientist and assistant professor of neurology at Oregon Health and Science University. He brings together a rare mix of naturopathic medicine, acupuncture, computational biology, and more than two decades of data science experience across multiple industries.

At OHSU, he completed advanced training in bioinformatics and complementary medicine research. He’s contributed to studies on botanicals for cognitive resilience and serves as the geneticist at the Oregon Alzheimer’s Disease Research Center. He’s also the principal investigator on a five-year NCCIH grant studying salutogenesis, how health is restored, using multimodal non-pharmaceutical interventions for age-related cognitive decline.

He teaches AI and machine learning, continues a clinical acupuncture practice, and is working to build whole person models that help us understand why some people decline and others age well. So Dr. Stephen Chamberlain, welcome to the show.

Steve Chamberlin (01:14.202)
Thank you. Thank you.

Dr. Heather Sandison, ND (01:16.256)
I want to dive in to first, know, how you got into this. How did these two paths of data science and naturopathic medicine and I mean, it’s more than two paths, right? Acupuncture and traditional Chinese medicine. How did all of this come together for you?

Steve Chamberlin (01:31.496)
So that goes really way back to my dad who was a physician and a psychiatrist. And he worked at a hospital called Minniger Foundation in Kansas, of all places. But he was really early. He was involved in the American Holistic Medical Society. So he was a very early holistic sort of MD in the late 60s and early 70s. they were doing.

really interesting things there like early psychedelic therapy and things like that. And so he was, I always think of him as like a pioneer in that area and he was my inspiration. so I wanted, and there was a lot of study of consciousness and things like that. so that was my inspiration. I went away to college thinking, I want to be a doctor. I want to be like that. And I got to school and it’s like, well, nobody’s teaching this. So I got…

I just ended up in research and was kind of drawn to stats and programming. You know, after I did, I was a biology major and then that sort of led to a career just in data science in multiple places. And I always thought, I’m going to get back, you know, I wish there was a way to get back to what I saw, you know, then. then I was here and I started developing an interest in trying to go to acupuncture school, trying to switch my career. I was in, you know, lots of big.

corporations for years and then I moved to Oregon and I went to this conference at NUNM, the naturopathic school here and I knew I shouldn’t go because it would make me want to go to the school and so I did, but I did and it was transforming the mind, was called transforming the mind and they talked about, they had a biostatistician and he was applying the sort of physics ideas to consciousness and to study natural medicine.

And, and that hooked me. And then I volunteered at their research, center at the health got, for a couple of years to decide if I wanted to go. And then I decided, and I decided, I mean, I, I love natural medicine. I like naturopathy and Chinese medicine because I liked the integration of the old and the new, but I also liked it’s also like you study every major health system on the planet, you know, with those two degrees put together. so for research, it’s really good too. And I, you know,

Steve Chamberlin (03:48.84)
from the beginning of my decision to go back to school with naturopathy, I wanted to integrate, I wanted to do research, I wanted to do clinical work too, but I really wanted to do research so I could draw on all my computational experience. And talked to Heather’s Wiki there quite a bit in the beginning, and then, and she was a big guide for me. And then,

You know, and then I took six years off and went to school, got both of the degrees full time. And then, and then I came and then I wasn’t sure exactly which way, how the research would open up. And I came to OHSU, to their department of medical informatics here and did a fellowship in, computational biology and bioinformatics or medical informatics. was a national library of medicine. So that was, I had been talking to them too, for a long time. And, and,

And I was able to get in and they accepted a naturopath there. It’s pretty, I think it’s fairly open here at the school. and then, and from there, just, know, first it was cancer, natural products and cancer was my interest. And then I got interested in, saluted genesis. heard a talk, you know, given by Helene Langevin at the NCCI in 2021, where she gave a really nice sort of easy to understand description of.

and was actually a grand challenge. wanted people to study that. So that’s kind of how I switched from sort of natural products and cancer to, to salutogenesis and then, and, know, switch my focus to, cognitive impairment, know, and.

Dr. Heather Sandison, ND (05:28.674)
Let’s dive deeper into salutogenesis. This is health restoration. We’re both naturopaths. So in our language, this is like the vis medecatrix nature, right? Like this is the healing power of nature. This fundamental belief system, right? That the body has the ability to heal when we use a whole person model. And I think that was the belief, you know, was living in that truth with a capital T that made me open to Dr. Bredesen’s

Steve Chamberlin (05:58.324)
Mm-hmm.

Dr. Heather Sandison, ND (05:58.379)
approach that maybe think, maybe this might work because I fundamentally believe that the body can heal itself, the brain not being an exception to that, right? So I’m curious how, you know, can maybe take us back to that talk where you were turned on to salutogenesis. How do you define it in a way that makes sense for patients and clinicians? How did she explain it that day where you had that aha moment?

Steve Chamberlin (06:21.684)
She used a lot of pretty pictures, like simple graphics, she showed, one of the things she showed was that health is a continuum, and it has two directions. Instead of a binary sick or healthy, so there’s a continuum, and you can get less healthy along that continuum, and then eventually end up in a disease state, and that if you know where you are in that continuum, it’s easier to reverse the health, and that…

and that there are, you know, it goes in both directions too. And so it comes from, the term was originally coined by a psychosocial researcher in Israel in the seventies named Aaron Antonovsky. And he actually studied, he studied survivors of concentration camps from World War II. And so he was really interested in why are there people that are resilient in that horrible.

you know, that really unimaginably horrible situation and some people aren’t. So his focus was more on what are the factors that created that resilience. And then, one of things I like is oftentimes people compare it to the pathogenic approach. So there’s the salutogenic orientation and the pathogenic orientation and they’re complimentary. They’re not one is right and one’s wrong, but the pathogenic orientation is all about like what are disease processes and then the salutogenic is what are health processes.

And then I like there are tables, know, what is, you know, comparing the two, but one of the things I like is that in pathogenic, you know, in the pathogenic orientation, there are risk factors and you’re mostly trying to avoid those risk factors. You know, you’re like living. So, and in the, the pseudogenic orientation, risk factors aren’t necessarily bad. And in fact, they might make you stronger, you know? So there’s that, that idea that a risk, so,

So they can actually be salutogenic. So what I like about that is that you’re not like hiding from the world. know, it’s like you’re building your own health resilience, you know, so it’s exactly. Yeah. Like that. Yep. Yep. That’s the way I think of it. So that’s the one. And the other is, know, I’d already mentioned that it’s a, it’s a continuum, you know, it looks at health as a continuum. So I’m really interested in, so I got really interested in, you know, computationally that continuum. What is it like?

Dr. Heather Sandison, ND (08:26.936)
like the hormetic effect essentially.

Steve Chamberlin (08:45.486)
A lot of people think that there are tipping points along that continuum, like you’re sort of in a stable dynamic state and then there’s a tipping point into a disease state. You can detect the tipping points and things are more reversible for a tipping point. So, and I’m also interested in models that look at the whole person along that trajectory.

Dr. Heather Sandison, ND (09:05.774)
You know, it’s interesting as you say that, I am thinking back to probably a I saw somewhere or something I read in someone’s email newsletter, but it was this aging continuum and how at 44 and 77, it seems like people kind of fall off a cliff for some reason. And you mentioned this in the context of a disease state, and we certainly see that like around a MOCA score of 16 or 18 gets a lot harder to bring someone back.

Steve Chamberlin (09:20.786)
Yeah, just like that.

Dr. Heather Sandison, ND (09:33.326)
And so I’m curious, how do you apply this whole person model, this salutogenesis health restoration model to aging, and to studying brain aging and neurodegeneration?

Steve Chamberlin (09:46.612)
So I know, so I’m working my current research is with a lab model. So I’m actually working with mice and they have cognitive decline too. so, so I will say that my, one of the ways that I work with this is I find, I either find or create populations that have, have the health restored already. So that’s, that’s the way I’m working with it right now. So I’m working with intervention. So in a saluted genesis model,

research model, aren’t, I’m not really studying whether this thing restores health or not. It has to already have been shown to restore health. I’m studying what that trajectory looks like, you know, so I have to, I have to create the, create that trajectory. So that, you know, in the, in the lab, I can, I can set that up and I can create that. can, you know, mice, there are lots of things that have already been shown to restore cognition in mice. And so I’m doing that. And then I have, you know,

the aging model, know, so the side-by-side aging models. So one of the things I do a lot of molecular omics, you know, with different tissues. And so I’m looking at how does, you know, what, what all is changing along that trajectory of that health trajectory and, the trajectory of getting less healthy. And one of the questions the NCCIH has is, are you reversing things, whatever you’re looking at, are you just reversing the decline or is it a different process altogether to restore health?

And once your health is restored, are you healthier? Is it stronger than it was before your health declined? So those are the questions. What’s that? Oh, don’t know yet. I just finished my experiments and I’m working with the data right now. But the other place to do that is with data repositories and real life data, clinical data that people have, which is one reason I reached out to you. And so I’m kind of looking for people.

Dr. Heather Sandison, ND (11:20.792)
What’s the answer? What’s the answer?

Steve Chamberlin (11:42.93)
you know, people that are experiencing this, seeing this in clinic. There are a lot of, have a repository here, you know, for Alzheimer’s people, just clinical and tests and, know, all these different things. And there’s some big national repositories. And so that would be another way. So I’m just looking at longitudinal data, but the main thing is that, you know, I might have a primary outcome of cognition, you know, how either that’s getting better, that’s getting worse, whatever it’s doing, but I want to track.

you know, the whole person, you know, what is changing along that, you know, so it’s not defined by all of the other systems necessarily, you know, it’s defined by the condition of interest, which is cognition, but, but like, how do all those other things change? And I do a lot of network, you know, analysis. So how do they, how do they change together? How are they changing together? And, and then multi-scale, like, you know, if I can get omics data, I can, I’ll use that, you know, along with.

you know, any data that I can get, but really trying to represent the whole person is the idea.

Dr. Heather Sandison, ND (12:42.958)
Okay, so you are studying these trajectories of cognitive health, really whole person health, that both improve and decline, they go in both directions, for mice and humans, right? And so what have you learned so far? I think some of these big questions that you don’t have answers for, but what has this taught you so far about aging and the possibility for improvement?

Steve Chamberlin (12:46.28)
Yep. Right. Right. Right. Right. Yeah. Right.

Steve Chamberlin (12:59.017)
Yeah.

Steve Chamberlin (13:02.888)
Well, so I know, so yeah, so I’m new still to this and so I haven’t analyzed the data from my experiment, but I know one of the things, so I started off, so cognitive non neurodegenerative cognitive decline is more acceptable that you can reverse that. And as you know, it’s not really believed that you can reverse Alzheimer’s cognitive decline as much. What’s that? Yeah, yeah, I think, yeah, it seems, I mean, and that’s what’s drawn me in.

Dr. Heather Sandison, ND (13:25.198)
That’s changed. It’s changed quite a bit in last couple of years.

Steve Chamberlin (13:32.756)
The interesting thing is that all of these clinical repositories with human data, they have reversion rates. So they have certain rates like the ADNI, the Alzheimer’s disease neuroimaging repositories, a really big one used for research on national repository. And it has like a 15 % reversion rate. And I’m not sure from what stage to what stage. The reversion is just that their cognition is improving.

Dr. Heather Sandison, ND (13:56.057)
So can we define reversion versus.

Steve Chamberlin (14:01.972)
And they get, so I’ve been, you know, I’ve been looking at the research and most of the research is interested in project. I mean, there’s a lot of interest in predicting these trajectories, you know, especially for Alzheimer’s, but the interest is when are they going to hit dementia, you know, or when are they going to, when are they going to decline worse? And, and they always throw out the people that get better because not always, but I mean, the story, the ones that I’ve looked at, they, they exclude them. And I know that that’s not the focus.

There’s a lot of question people don’t know why that is. I mean, because this is observational data mostly, and so we don’t know what people are doing, but they think that it will, you know, it’s not what they’re interested in. So, and we have that here in our data. So the OADRC that I work in has a data repository. It’s part of the national network and we have data for Alzheimer’s people that can be used for research. And we have them here too. And everyone knows about it. We even have names for them. Nicknames are called the bouncers or the wobblers or the…

because they go back and forth, you know, with their cognition and, and there is a need to study those people feel like there’s a need. And that’s how I got started. So I, to study, solutogenesis, like I’d said before, you have to have, you have to have a population that gets better and a population gets worse. And so that’s easy to do in the lab with, you know, with age related cognitive impairment with mice and, and, and, and, know, and even age related cognitive decline in people.

You know, as I guess more acceptable to be restored. But, but I started off thinking, well, I started off wanting to study, you know, apply solute genesis to Alzheimer’s and like, well, there’s no pop, there’s not supposed to be any population of anybody like getting better. And, and that’s what led me to looking into these databases and finding, and then you, and I know, you know, Dean Ornish and Dr. Bredesen and then Dr. Isaacson and you know, Florida. And so, so there’s, there.

That’s one of reasons I’ve reached out to you. There are people that are publishing these results, and then there are these repositories have people that also get better that no one really studies.

Dr. Heather Sandison, ND (16:09.734)
Even the Pointer trial that was out recently also showed that. What patterns or hypotheses are emerging from these cases where people on the path to dementia are improving?

Steve Chamberlin (16:12.104)
Yeah, yeah, yeah.

Steve Chamberlin (16:23.774)
So most people, when I talk to people, and I haven’t dug into the data yet, the human data, but most people think it’s an error. it’s, know, that’s mostly, or it’s like they, you know, they just didn’t do a good job at, you know, administering the tests that day or something. Right, and people do go up and down. Yeah, so nobody really knows, but they kind of, it’s ignored. That’s pretty certain, you know, that they ignore that group.

Dr. Heather Sandison, ND (16:42.294)
days and bad days, that kind of.

Steve Chamberlin (16:53.524)
So that’s the niche I want to work with. But that is required to study pseudogenesis. But there are people, mean, like we just said, are people like you that are reversing cognitive impairment and Alzheimer’s.

Dr. Heather Sandison, ND (17:06.574)
Well, and before we hit record, you mentioned a personal friend of yours who you know had recovered some cognitive function. So what do you think is driving an ability to recover function? Because you agree, right? You’re part of the reason we’re having this conversation. You agree this is possible. It’s possible to regain cognitive function even in an Alzheimer’s patient. Maybe not all of the way after a certain point. It’s not guaranteed, certainly.

Steve Chamberlin (17:11.464)
Yeah. Yeah. Yeah.

Steve Chamberlin (17:20.329)
Yeah.

Steve Chamberlin (17:24.818)
Yeah. Yeah. Yeah.

Dr. Heather Sandison, ND (17:33.848)
But what do you think drives that potential ability to recover function?

Steve Chamberlin (17:38.516)
I think it’s the multi, you know, there seem to be variations, I’m not an expert on it, but these multimodal treatments or multi-domain treatments seem to be the real key. And I thought I’d heard like the finger people are starting to do that combined with the new medications. I don’t know, I’m not an anti-medication person, but I don’t know. So I think that’s really key, but I think there’s…

you you would know more than me. think there’s a belief that those multi-domain approaches are not just working on the, you know, neurology, they’re working on multiple systems that restore cognition. And I’m really interested in, you know, I’ve mentioned this before, tipping points in neurodegeneration and resilience and all of that. And I, there are papers about that that I’m reading right now. And, but, so,

Dr. Heather Sandison, ND (18:32.322)
And then would you venture to guess what early signs or biomarkers seem most promising for understanding where that inflection point would be?

Steve Chamberlin (18:36.168)
Well, well.

So that’s my area research. That is the core of my research. So that requires, I don’t know yet. So I mean, know people say, know, cardio metabolic issues are going on too. And, but I’m new at it. But my interest is using, you know, complex, you know, network models to find those points. And then those points, you know, not only can you find the inflection points,

Dr. Heather Sandison, ND (18:41.952)
Okay. And what do you think? You can speculate.

Steve Chamberlin (19:07.22)
but they should be able to guide treatment. Like if you can find out what it is that’s changing at that time and you can detect it. There are methods, there are network methods, dynamic network biomarkers I’m really interested in that use, they measure multiple things and how they’re correlating together. then, so I’m not a, I’m definitely a network person and not a single biomarker person, but I’m interested in there are, those have been shown in some experiments with other conditions that they can predict.

proximity to a tipping point, know, so they’re really good for, they’re not only predicting proximity, but they can kind of detect what’s going on because they’re usually, you know, based on omics data. So you can, so they can guide treatment and guide their experiments, which show that things, know, conditions are more reversible before the tipping points, which is just sort of common sense. they, but that’s, so that’s my interest. So it’s not exactly what it is as much as it is like,

characterizing that trajectory from a whole person perspective and in a sort of complex multi-scale, multi-omics, maybeomics, but a whole person. So there’s something called network physiology, which this intersects with, which is kind of new. then NCCIH has just funded another project at Stanford, started in August to define the physio, which has a model.

Dr. Heather Sandison, ND (20:21.614)
Yeah.

Steve Chamberlin (20:34.356)
of the entire physiological system in a person and how it interacts. it’s a dynamic model. so that intersects. So my thought is, I mean, I’m doing a, I call it a poor person’s version of that, just doing correlations. But the idea is like, how does this thing change over time? And can you detect it in light? Yeah.

Dr. Heather Sandison, ND (20:55.052)
Yeah, with multiple inputs and just, I mean, innumerable variables. We do this at a, I mean, it’s costly, it’s expensive, right? But we do this at a one-on-one level where we get tons of data about people. And then we look at it at multiple points over time. We’re talking toxin data and inflammatory data, nutrient data, infectious burden, full immune function.

Steve Chamberlin (21:13.886)
Yeah.

Dr. Heather Sandison, ND (21:24.01)
understanding, you know, like all of these things that we’ve, I’ve listed these things in the book and also on this podcast before. But then as a provider, I’m to put all that in my brain and kind of make sense of it and then create a treatment plan. As you’re describing this, it sounds like, you know, sort of leveraging that sort of idea, this complex system science approach with lots of data, but

Steve Chamberlin (21:33.694)
Yep. Yep.

Steve Chamberlin (21:45.758)
Mm-hmm.

Dr. Heather Sandison, ND (21:49.007)
essentially plugging it into like an AI model. I have to imagine that this all gets a lot easier, a lot less expensive to use this whole person multi-scale model to transform how, not only how we study cognitive decline, but then how we test interventions certainly, and then how we deploy them at scale so they can really have an impact.

Steve Chamberlin (22:07.081)
Yeah.

Yeah, yeah, I mean, that was one reason I reached out to you because I don’t have clinical experience, you know, with this and I have computational, I have my friend, you know, who I’ve seen do a lot of this and have a miraculous result. But I wondered and I know I’ve read your book and I know I’m familiar, you know, with all of the things you look at with the approach you use. And, and yeah, and my question was, you know, is there a need for a complex model or is there, do you do well, you know, just kind of.

evaluating the data, know, clinically. And then of course there’s, you know, the way I see science is, you know, you have the luxury of building these complex models that of course can’t, aren’t practical in clinic. And then, then, you know, you have to figure out how to translate those, you know, which usually means finding a subset, you know, of those things that are easy to get, you know, in clinic and are collecting clinic. And, but yeah, that was a question. I mean, that, was one of the things I wanted an opinion.

from you who does this kind of thing, is there a need for a complicated model which puts all these things together and looks for tipping points? Or are you already doing that without the model?

Dr. Heather Sandison, ND (23:18.722)
Well, I think the need and we’re in some ways trying to respond to this, but the need is to make this accessible because it’s a thousand dollars to see me as a doctor, right? It doesn’t, and my, can only see eight people a day and that’s that’s a full day for me because it takes 90 minutes to sit down and learn about someone’s health. And it takes another 90 minutes to go through all this data with them and describe it in a way that makes sense so that it’s motivational to make these changes.

Steve Chamberlin (23:33.396)
Mm-hmm.

Dr. Heather Sandison, ND (23:48.205)
And that is not accessible to everyone. So how do we create a tool, and how do we leverage the technology so that this can be accessible to anybody who is Medicare eligible? Because it’s so expensive. It’s such a horrific disease. It’s so costly to care for people. I know very well how much memory care costs. I know how much it costs to deliver, and I know how much it costs for patients.

Steve Chamberlin (24:01.939)
Yeah.

Steve Chamberlin (24:14.536)
Yeah.

Dr. Heather Sandison, ND (24:16.984)
But that’s the financial cost. Now the burden to caregivers, the cost that doesn’t get counted as often, the burden to caregivers, the absence of that grandparent in the lives of the grandchildren, the absence of the mom in the lives of her children because she’s in that sandwich generation having to care for her mother or her father. There are so many costs to this disease that yes, I think any effort to scale

Steve Chamberlin (24:35.368)
Yeah.

Dr. Heather Sandison, ND (24:44.546)
What Dr. Bredesen has created to make it less expensive, more accessible, easier to deploy is a very worthy endeavor.

Steve Chamberlin (24:55.368)
Yeah, that helps.

Dr. Heather Sandison, ND (24:56.846)
Yeah, we are starting to offer the guide. Guide helps a little bit. Guide is basically some care management, some through a, it doesn’t have to be through a clinic, but we’ll have a nurse practitioner joining us who does bill Medicare and she’ll be deploying the Bredesen approach. Plus we’ll have this additional care team, basically clinical coordination happening to support with respite care and Meals on Wheels. And, you know, it’s so challenging.

Steve Chamberlin (25:13.214)
Mm-hmm.

Dr. Heather Sandison, ND (25:26.37)
to have dementia, to be elderly, and then to navigate the healthcare system, even if these resources are out there for you, just getting them, getting access to them, doing the paperwork, getting through that process can be, it can feel insurmountable, especially when, you know, your partner has got diabetes and is in pain or can’t hear, you know, they might have full cognitive capacity, but they’ve got these other issues that are going on. So it can be really challenging. So I do think that we desperately need

Steve Chamberlin (25:32.756)
Yeah, yeah, yeah.

Steve Chamberlin (25:48.616)
Yeah.

Dr. Heather Sandison, ND (25:56.303)
solutions that leverage technology and computational analysis. Tell me how, tell me when.

Steve Chamberlin (26:03.38)
One of the things I’m doing now, which I mean, of course, stuff I’m doing with animals is, you know, is not, that is not translated yet. I mean, some of the things that we’re working with, the biomarkers with the mice are translatable and clinic like CRP and I work with 8-OH-DG and the plasma too. But I’m also trying to do, I’m trying to get some money now to do a…

work with our data repository here, our people data, I don’t like the word human, like it sounds like people, I don’t know, people repository. so, and that’s to work with to do untargeted, either proteomics or metabolomics in their plasma, know, which is clinically available and maybe not too expensive, you know, to do and might capture a broad range of things, you know, so that might be a translatable thing, I’m not sure.

I mean, I know what we pay. I mean, it’s $200 or $300, know, something like that. And it can be pretty holistic. So that’s kind of what I’m working with right now with the data, our data repository.

Dr. Heather Sandison, ND (27:16.696)
You teach machine learning and how to interface with AI. How do you see these tools helping us to understand these complex interventions, complex system science recovery patterns, the trajectories and these tipping points? not all these things are linear.

Steve Chamberlin (27:30.44)
Yeah. Yeah.

No, they’re not. They’re not linear at all. that’s, yeah. so the, that’s why, that’s why Dr. Longavan put that grand challenge into 2021 because of AI. She’s like this, you know, it’s always been, and I, you know, I, I did research when I was in school, which was 09, you know, to 2015, HealthCut Research Institute. And it was always an issue.

Like, well, our medicine has multiple interventions and everybody’s like, you can’t study that with a conventional, you know, randomized controlled trial. Yeah, yeah, no, I’ve heard you guys. Yeah, yeah, I have, yeah.

Dr. Heather Sandison, ND (28:07.756)
That’s exactly why it took so long for Dr. Bredesen’s first trial to happen, is because they rejected, yeah, you’ve heard that story. It’s such a breath of fresh air, right, that the trend has really shifted, that everyone is accepting now, no, we actually don’t want to limit the variables, we want the synergistic effect of multiple interventions.

Steve Chamberlin (28:20.457)
Yeah.

Steve Chamberlin (28:26.098)
Yeah. Yeah. And that’s what I have my, my ex the work I’m doing right now with the animals is, is, that’s one aspect of it is it has multiple interventions. So I’m looking, I mean, I’m trying to restore health and study that, but I’m also doing it with multiple interventions and look at how they interact. So we had, so anyway, the AI is just the compute power is, you know, is there now and they, and the ability to handle lot, you know, very large quantities of input variables. that, that’s why.

You know, she put that grand challenge forth and we got, my department actually got a grant on that grand challenge, which is different from my grant, which is a data generation. So she just was asking people to collect all of the data they can think of on, this is for people with diabetes, collect that data on them for a year, just everything, lifestyle, lab tests, imaging data, wearables, you know, all of that, you know, and now.

AI can handle it. mean, there are methods out there, deep learning and all that. So AI means a lot of things and, know, chat GPT is kind of the AI people think of now. I think, I don’t know, I don’t know what people think, but yeah.

Dr. Heather Sandison, ND (29:37.77)
Yes, that’s certainly my experience. You say AI and I think of my relationship with ChatGPT.

Steve Chamberlin (29:42.546)
Yeah, yeah. So that’s generative AI. So that is, that’s the kind of AI that will answer you, answer questions and things like that. But the other kind of, so anything that’s machine learning and predictive is technically AI, artificial intelligence. but there are machine learning has been around for a long time. There are simple, like logistic regression, sort of simple methods that been around support vector machines. All of that is AI and then neural nets, neural networks.

They’ve been around a long time too, but they didn’t work very well. then, then when the compute power got better around 2010 or 11, they were able to build much bigger neural nets with lots more, a lot more connections. And that’s when deep learning came and deep learning is where everything changed. And so deep learning are just really gigantic neural networks with lots of layers and lots of connections in the model. And they can pick up all these nonlinear, you know, nonlinear relationships.

And then deep learning, deep learning can be predictive and it is in healthcare a lot of times. So you can feed in image data where a variable is a pixel, you know, like in an image and, and, or, you know, I work a lot with RNA-seq, which has 15 or 16,000 genes expression level. And so you can feed all of that in that’s, know, 15,000 variables. so, and then there’s multimodal where you put the imaging and the RNA-seq and the clinical data, you know, from the notes. And then there’s mining of clinical data, you know, with AI to get

concepts, know, out of complex. but the AI, even like chat GPT and that’s considered natural language processing, even chat GPT, those are large language models. Those are deep learning neural nets. They’re all deep learning neural nets. And so that came, you know, that was, like I said, compute power got to where it could handle creating larger neural networks. And that’s when everything changed.

Dr. Heather Sandison, ND (31:37.934)
So do you see that playing a role in like what stage, like early detection, personalized protocols so that we can be more precise? Do you see that in helping us predict who’s most likely to respond to different interventions, whether it’s a medication or a lifestyle intervention? Like how do you all of that and more, I’m sure.

Steve Chamberlin (31:52.616)
Yeah, all of that. Yeah, all of that. I think definitely early detection. So one thing that does really well is it takes in more data. So you can take in a lot more data and get more accurate at detecting. I know in Alzheimer’s, I’ve been interested in trajectories, and those are really long, as you know. people are really interested in project.

predicting what trajectory you’re gonna go down or what path from a point for years out in the future. so they’re good at that. think so there’s, there are levels of models, there are predictive models, which risk is a really common prediction in healthcare, but you don’t really know what to do with that. So that’s one of the things that I teach in my class. I’m really interested in decision science. Like how do you…

How do you make a decision from a model? Okay, it’s great. You can predict all this stuff. people get all caught up and, I can predict this and that, but it doesn’t ever solve a problem. And there’s an actual problem in healthcare with people building models and they don’t get used because people are so fascinated with just predicting something. then there’s beyond that is a prescriptive model. And a prescriptive model tells you what to do, know, recommends a treatment. And I did all of that. I came from banking actually before this. I was one of my industries and we did all that.

kind of at the time we were considered a head of healthcare as far as our sophistication. we did all of that. We would predict what’s the net income from all these different options and things like that. But that’s going on now. But I mean, so I think clearly a model in my mind would identify early detection, but beyond that, where are you on that trajectory? Are you near a tipping point? Are you about to tip? what are the things that…

you need to reverse that proximity. And so the model should be able to, if you’re doing omics and things or even proteomics from plasma, should give you, you can, there are ways to like see what groups of proteins are sort of changing together and what function that might belong to. And then that might guide what treatments would be appropriate. I mean, I know, it’s just the multi-domain as though it’s like, well, everybody should exercise in ether.

Dr. Heather Sandison, ND (34:08.289)
yeah.

Dr. Heather Sandison, ND (34:14.734)
Well, this is my next question. Based on everything you’ve learned from your acupuncture background, your naturopathic background, all of the computational data analysis that you’ve done, what are the lifestyle factors that seem most impactful for maintaining, optimizing, restoring cognitive health? What are the top things that you would suggest someone do?

Steve Chamberlin (34:16.562)
Yeah. Yeah.

Steve Chamberlin (34:37.212)
Yeah, I still, I’m still kind of like I said, new to that. I know I’m doing exercise and a botanical with the mice. so yeah, it’s, yes, it does. There’s a huge body of evidence that it does work for both of those work for the mice. Nobody’s ever tested the two together. So I’m still in the process of seeing what my data said, but that’s why I.

Dr. Heather Sandison, ND (34:47.478)
Does it work for the mice?

Steve Chamberlin (35:04.798)
chose those two because there’s already a huge bite. We do here, we do a lot of Centella research here and Ashwagandha and very much so. Even if there’s an Alzheimer’s mouse model called 5XFAD, which we use a lot here. And it does so with them too. It reverses their cognitive.

Dr. Heather Sandison, ND (35:25.528)
So exercise and herbs, you heard it here first.

Steve Chamberlin (35:27.688)
Yeah. Yeah. Yeah. So, so well, it had, they haven’t been tested together as much. mean, they have, mean, like my friend did, my friend who you had mentioned who has miraculously reversed her cognitive decline and held it stable as a huge exerciser. And she continues, she’s very motivated to exercise. I think that’s got to be one of the number ones, but again, that’s not,

I don’t have any answers yet for my own work. But what I think is that it’s the combination of things. That’s always what I put in my grants and that’s what I… I’ve heard that from our vice chair of research here that they’re very much open. People here in our neurology department are very much…

you know, into the multi-domain and very much on board with that. But they don’t study it and they don’t know how because it’s kind of, you know, it needs new methods and it’s not. So, but he’d mentioned at one point that the problem with that is you see these effects with multi-domain and then everybody goes out and studies the single intervention and then they don’t work, you know, when they look at them individually. So I’m really big on synergy. I’ve always been a synergy. When I first came here,

I was, I did cancer research and I was, I didn’t just study, was trying to predict synergies between natural products and cancer drugs. I wasn’t just trying to find a new drug, you know, from a natural source. I was really interested in synergies because there were, there was evidence in cancer that certain, certain like curcumin when it’s taken with a cancer drug makes the drug more effective in the cancer tissue and protects the healthy tissue at the same time. So there’s amazing things in, I mean, that’s all another topic, amazing things in.

plants, found, you know, I did find, I studied the effect on neurons of Centella and in vitro and, and, and, you know, looks like Centella is anti-inflammatory and also works on, reducing oxidative stress. but Centella is also, it’s also used in derm applications. And when I, when I, when I treated these neurons, their collagen turned on like crazy and no one’s, you know, looked at that as far as does that affect in cognitive impairment, but.

Steve Chamberlin (37:51.144)
But we have seen that Centella increases the density of dendrites, you know, and so it’s actually changing the shapes of the neurons. So it seems like, you know, the neuron might wanna restructure the extracellular matrix. But so I actually did a test.

Dr. Heather Sandison, ND (38:07.246)
Our listeners might know centella asiatica as go to cola. And this is an often used herb in Ayurvedic medicine and traditional Chinese medicine throughout Asia. And we know that it’s very helpful for wound healing, skin health, antioxidant, anti-inflammatory. It helps with like microvasculature and potentially vascular support.

Steve Chamberlin (38:11.954)
Yeah.

Dr. Heather Sandison, ND (38:31.65)
with mood and then cognition because it’s supporting BDNF is potentially one. mean, certainly if you have better circulation and less anxiety and less inflammation and less oxidative stress, you might get downstream effects that include cognitive support, but it also directly impacts BDNF.

Steve Chamberlin (38:50.292)
Yeah. One of the things I was interested, so I really focused on the interactions between the compounds in the plant because it’s got hundreds of compounds in it. Yeah, the whole plant. Yeah. And we did both. like, there’s a few, like a few triterpenes, which they think are the active ingredient in Centella and there’s Caffeola quinic acids. And so we looked at those separately. We combined just those and then we looked at the whole plant. And so we looked, kind of got a, kind of an idea of some of the interactions. But one of the things I did was a

Dr. Heather Sandison, ND (38:59.17)
the whole plant rather than an extract.

Steve Chamberlin (39:20.144)
It seemed to me like, because I did RNA-seq, I looked at sort of untargeted, like all of the gene expression going on. And then I mapped that to pathways and kind of figured out what are some of the functions that might be, might be activated. And I, it looked to me like some of those functions were working together, you know? So I thought, well, how, you know, is there, is, is there a wisdom of the plant? Is it like making, and so I did statistical tests of randomness to see like, are all these things.

You know, that are being turned on by this plant, just random, you know, randomly happening. And they weren’t, you know, they, they were not, they would not, the chance that that those are all happened by random chance was like less than 5%, something like that. so there, um, so that was another, that was another thing that that was cool. But, um, so, um, yeah, so I, you know, at least at this point, my, my whole thing is like the interaction, you know, the synergies between things is what really works. And that’s why I think, you know, it.

doing multi-domain lifestyle stuff with a medication might be okay or might actually make, I’m all for like, if somebody has to take a medication, can you make it work better and have less side effects? Yeah, yeah, yeah.

Dr. Heather Sandison, ND (40:28.91)
Of course, yeah, same. What gives you the most hope right now in the Alzheimer’s research landscape?

Steve Chamberlin (40:37.784)
the multi-domain treatments by far, yeah, by far. Yeah, I’m not, I mean, I’m aware of the drug things. Nobody seems happy with the drug development, actually, I mean, that I know. mean, no matter where you come from, but maybe, you know, if you combine some of those with a multi-domain, you know, they might like work really well.

Dr. Heather Sandison, ND (40:58.482)
I really think, you know, the amyloid, I mean, at least my framework is that amyloid tau, these are there to protect the brain, right? They’re antimicrobial, they engage metals in interesting ways. They are there to protect the brain. And so if we just take them out without addressing the toxic exposure, the oxidative stress, the infectious burden, then we just end up with a less well-protected brain.

Steve Chamberlin (41:06.535)
Yeah.

Dr. Heather Sandison, ND (41:24.226)
But if we can first kind of do our work, clean up the messes, stop the exposures and the burden, then what we can do is we can turn off that microglial activation, right? And then if we get the amyloid out and there’s not as much to protect us from, fantastic. Maybe we’re in a much better place. I think that there could be wonderful synergy between those things. But if we just go in and rip the amyloid out, that doesn’t seem to help most people.

Steve Chamberlin (41:42.034)
Yeah. Yeah.

Steve Chamberlin (41:49.364)
Yeah, yeah, I agree. I agree. Yeah, no, yeah, I know, I hear that. And I agree with you. They seem so complimentary and like the neurogenic effect of exercise and I hope like if you, know, getting them out is great, but then repairing everything, you know, it seems like the two would go together really well.

Dr. Heather Sandison, ND (41:53.068)
Madonna can actually be really harmful.

Dr. Heather Sandison, ND (42:08.878)
Is there a diet that you see work best in mice?

Steve Chamberlin (42:13.484)
that’s a good question. I haven’t, if I do more mice work, I might do the diet. I haven’t looked into that as much. I am interested in that. I know that there are some people that are doing the multi-domain are using very different diet approaches. Like I know the approaching is an ornish. And so that’s a really good, yeah, I am really interested in that. And diet is definitely something that’s not too hard to do in the lab, you know, in an experiment like that.

Dr. Heather Sandison, ND (42:28.738)
Yeah, yeah, like Ornish is vegan. Yeah.

Steve Chamberlin (42:43.034)
Exercise has turned out to be much trickier than I thought in a lot of ways. I don’t know, because it affects a lot of things. definitely, so we put the centella in their drinking water and it affects how much they drink. the ones that are for some reason, the mice that exercise tend to drink a lot less of the water, you know, the centella water. you know, we do, from memory, we do something called the Morris water maze, which is a swimming test.

And so, so the mice, you know, the exercise mice are in a lot better shape, you know, theoretically, you know, for that swim and, they’re natural swimmers and they float. That’s like, it’s not a risk to them, you know, to, swim, but, but they, but so, there’s, you know, exercise is not a linear dose effect either from what I can see in the literature with mice. Like there’s a sweet spot and older mice, if you exercise them too much, it actually hurts, you know? And so,

Dr. Heather Sandison, ND (43:40.536)
just like the people.

Steve Chamberlin (43:42.162)
Yeah. Yeah. And then, and then it can, the neurogenic effect can take, be delayed, you know? So when do you test? So I’ve run into some issues with like exactly how do you, when do you test them and how does it, I just learned it’s interact, you know, studying interactions, even in the lab is very complicated, much more complicated than I thought. And, and, because you just have to know like how do those mechanisms interact with each other and when do you test and what’s the length of time? I think Centella, I mean, we’ve done other.

dose testing with Centella and it’s more linear, know, like the more they drink, the better their memory. But it’s also interesting. This is a tipping point. We’ve done tests where we’ve given Centella at different ages, you know, throughout the age of the mice and it doesn’t do anything before about 12 months. Like when they’re before middle age, it doesn’t make a regular, a younger mouse smarter. You know, it only does it once they’ve started to decline. And that’s, I’ve seen that. I mean, I’ve seen other researchers that have seen that with other natural.

products they’re using for memory with the mice.

Dr. Heather Sandison, ND (44:42.944)
interesting like it’s almost the antioxidant it’s the anti-inflammatory process that’s actually beneficial and if you don’t have a bunch of inflammation or oxidative stress then it doesn’t have anything to like work on there’s no Delta.

Steve Chamberlin (44:50.248)
Right. Yeah. And that’s a total tipping point. So that to me, that’s like, you know, that’s a tipping point. that would tell, you know, if you can, and I’m sure you can pick that up somehow, you know, what, what is it that’s telling you, okay, Centella wouldn’t be worth it to give right now, you know, or maybe it would be.

Dr. Heather Sandison, ND (44:58.862)
Interesting.

Dr. Heather Sandison, ND (45:09.826)
So Steve, I ask everyone this, but what brings you joy at this stage in your life?

Steve Chamberlin (45:14.892)
all of what I do. I am 67 and just kind of at the beginning of a career. It’s taken a while. Like I went, you know, I’ve been in school and doing post-docs and stuff for like 15 years, maybe now, but I always, you know, I just got married. I got into a relationship with the late in life. I love meditation. You know, I love nature. You know, I’m here in Oregon. It’s unbelievable. And I go, you know, we

We go all over, we travel, we camp, we hike a lot. And I do all the multi-domain things, know, I box, I like martial arts and boxing, and this isn’t a boxing accident. No, I try not to get hit in head too much. Yeah, so I…

Dr. Heather Sandison, ND (45:55.456)
Okay, you’re not getting traumatic brain injuries, are you up there?

not that kind of boxing. What’s your routine? What do you do for your brain health based on what you’ve learned?

Steve Chamberlin (46:06.996)
No, I do. So I take supplements. actually just took so I just took the supplement designed by Dr. Bredesen for the last month. I’ve taken different things. The neuro Q. Yeah. And it really, I really, you know, we had our top like Centella researcher here did a grand rounds and she was talking about, she was doing just talking about the literature with Centella and Ginkgo and then coffee fruit and, and, and

Dr. Heather Sandison, ND (46:15.31)
the narrow code.

Steve Chamberlin (46:35.252)
curcumin. She’s talking about the research and at the end she showed that supplement. said she kind of picked it out as what she thought might be the best you know supplement from her and so I tried it and it really made a difference and so I do do supplementation but that’s not I’m not all I’m not a fan of just doing supplementation with everything else. I do exercise, I box, I hike, I lift weights, I keep my mind active you know I really try to study I love to learn I try to keep you know keep

studying and learning new things, which my job is really good for. And I get up in the morning, I meditate, I sometimes do pretty intensive mindfulness retreats, which are very good for that. I love music. I do. So I get up and I exercise and I do meditation, spiritual connection, aside from religion, but just like that connection to everything. Nature is one place I really like.

I live on the edge of a forest in Portland. can live downtown and be on the edge of a forest. So, and, yeah, exercise though, I have to say, you know, I I’m pretty good with my diet. I’ve learned not to be too strict. I mean, that’s part of it. I’ve been like, when I first got into naturopathic school, I drove myself crazy about being strict, but I do, you know, try to do smoothies, try to get, you know, a bunch of veggies in every day as a core and protein. And I do do better with low carbohydrate, you know,

diets leaning towards the diets. Yeah, yeah, definitely.

Dr. Heather Sandison, ND (48:01.889)
Keto-ish. Steve, it’s such a privilege to have you here. And I’m just so grateful that you brought such clarity and curiosity and scientific depth to this conversation around AI and the data and what it means for cognitive impairment as people age. it’s really fun to talk to someone whose work sits at this intersection of data science, natural medicine, neurology, human resilience, aging.

And really, I think that what you’ve illustrated is that there’s this potential for understanding cognitive health in a really new way right now. Your focus on pseudogenesis and this whole person modeling, and this, you know, again, this possibility of cognitive improvement rather than inevitable decline, it’s so refreshing to hear and to know that this is being represented in the research world.

Steve Chamberlin (48:37.992)
Mm-hmm.

Dr. Heather Sandison, ND (48:56.16)
And there’s this reminder, I think throughout our discussion here that the brain is dynamic, it’s interconnected, it’s far more complex than we might realize. I have a reverence for it certainly every day, but also far more capable of repair than we’ve been taught to believe. And so I hope that everyone who’s listening, this encourages you to think differently about how you’re aging and to question the assumption that decline is unavoidable.

Steve Chamberlin (49:07.22)
Mm.

Dr. Heather Sandison, ND (49:21.878)
and really to explore what the possibilities are when we think about the entire system and we do exercise and we think about these herbs and how we can leverage them to really optimize as we grow older so we can think well and age well. So thank you, thank you, thank you for joining me today, Steve. Is there anything that you wanna leave our listeners with or any way that they can learn more about your research or about you?

Steve Chamberlin (49:37.332)
Mm.

Thank

Steve Chamberlin (49:49.202)
I think if there’s any way, I don’t know how many people have, I’ve had several people with dementia that I’ve, know, in one way I had to care for and I know about the resources. If there’s any way, you know, to support resources for people that do have cognitive impairment and might not have anybody, you know, like family, don’t have family around or they, family, you know, doesn’t want to deal with them or something, you know, do that. Like there’s a need for that, like to really support people that don’t have anybody, you know.

impairment. That would be my urging for people.

Dr. Heather Sandison, ND (50:26.358)
Yeah, what a compassionate answer. Thank you for the work that you’re doing. And thank you for helping us to just illustrate and illuminate that there’s a future where cognitive resilience isn’t just the exception, but the expectation for everyone. So thank you, thank you, thank you, Steve. Such a privilege having you here.

Steve Chamberlin (50:40.68)
Yeah, that’s Yeah.

Yeah, thank you. Thanks.