Radiolab - How to Cure What Ails You

Episode Date: May 16, 2025

Now that we have the ability to see inside the brain without opening anyone's skull, we'll be able to map and define brain activity and peg it to behavior and feelings. Right? Well, maybe not, or maybe not just yet. It seems the workings of our brains are rather too complex and diverse across individuals to really say for certain what a brain scan says about a person. But Nobel prize winner Eric Kandel and researcher Cynthia Fu tell us about groundbreaking work in the field of depression that just may help us toward better diagnosis and treatment.Anything that helps us treat a disease better is welcome. Doctors have been led astray before by misunderstanding a disease and what makes it better. Neurologist Robert Sapolsky tells us about the turn of the last century, when doctors discovered that babies who died inexplicably in their sleep had thymus glands that seemed far too large. Blasting them with radiation shrank them effectively, and so was administered to perfectly healthy children to prevent this sudden infant death syndrome...Leadership support for Radiolab’s science programming is provided by the Simons Foundation and the John Templeton Foundation. Foundational support for Radiolab was provided by the Alfred P. Sloan Foundation.

This is Radiolab, I'm Luluh Miller, it's May, which means we are right in the middle of Mental Health Awareness Month. And because of that, I found myself thinking about an episode we did many years ago about what was at the time a brand new way of peeking inside the brain to try to see various mental health conditions like depression and many other things. It's a fascinating piece and while it was recorded 17 years ago, and you'll find some of the language reflects that time, the question at the heart of the story is really timeless, really provocative, really fascinating.

So we are going to air it today. And at the end, I will be back with a short update on where some of the technology has come in the subsequent years. So here we go, buckle up with the episode called How to Cure What Ails You. Wait, you're listening. Okay. Alright.

Okay. Alright. You're listening to Radiolab. Radiolab. Radiolab. From WNYC. Rewind. Hello, I'm Jad Abumran.

And I'm Robert Krollrich. This is Radiolab. This hour we're going to be talking about diagnosis. Diagnosis, the easy kind, we're not going to talk about it. The easy would be you come into my office and I'm a doctor, you have a broken arm, I take a picture, I say, hey, you've got a broken arm, the picture says so. Yeah, because you can see the break right there. But let's suppose you came into my office

and you were sad, you tell me that your sex drive is down, I'm the doctor, so it's just between you and me. Well, right away, my learning tells me that you may be a candidate for depression. But how do I know that you're depressed? What do you mean, how do we know? We talked about it. You just said.

Well, you can't measure sadness or depression. You can't go to a test tube and count anything. Right. It's not hard science. Yeah, because... Until now. What if I put you in one of those fMRI machines that we've talked about so often, I snap a picture of your brain in action and I look at it and from your picture I say you are depressed.

You're gonna tell me I'm depressed just from looking at a picture of my brain? Yes. What? No way. Look, it's now here. Photographic diagnosis of mental illness. This is happening. There is no question. And that, by the way, is Eric Kandel, a professor at Columbia University, who just happens to have won the Nobel Prize for Medicine. Let me give you a little historical background.

Did you get a Nobel? I don't think you did. I still don't believe you. What, just because he's got a Nobel Prize, I'm gonna suddenly turn around and like to- No, no, no, you're not. So let's do this step by step, okay? Step one.

Imagine you're slipping into an FMRI machine, okay? Now, I want you to just look at my face. Why? Is that so difficult? Yeah, I just wonder where this is going. Okay, I'm looking at your face. So now that you're looking at my face, different regions of the brain, they become active.

There are cells in your brain that are saying, I know him. The cells are more active, they need energy. Just like when you run, you have to breathe fast. And to get the energy? Your heart pumps more blood. The body sends a rush of fresh blood to that particular group of cells. And because the blood has iron in it, sends a rush of fresh blood to that particular group of cells.

And because the blood has iron in it, the magnet in the brain scanner can see the iron and therefore see the blood flow and take pictures of it. Many, many pictures in real time. I'll show you a very nice example of this. Eric's now heading off across his ample office with an extraordinary view of the Hudson Valley. And he brought over a picture of a human brain with different colors and different areas.

And he told me when you look at a face. When you image the face, this area lights up. You're pointing to an area of the brain. An area of the brain. A couple of your, on your forehead kind of. That's right, that's right. If you look at a house, some other area lights up, but this area does not light up. You look at another face, this area lights up again.

Every time you see a face, same area? Yep. But you haven't told me anything about emotions yet. That's true. So let's move on to step two. Because we use faces to tell what someone else is thinking or someone else is feeling, looking at faces also triggers... ...an area deep in the brain that is concerned with emotion... ...called the amygdala.

Ahem. Now, very recently... ...a number of people have looked at the amygdala looking at faces, and it's extremely interesting. Step three. Okay. I'm going to take you now to London. Hello?

Hi. Yeah, hi. Yeah, that's good. Hey, who's this? Oh, right, sorry. Who are you? I'm Cynthia Fu. I'm a psychiatrist at the Institute of Psychiatry, King's College, London. And are you like in your 30s or your 40s or your 50s? Is this part of the interview? Rude. No, I wanted to establish that she came into psychiatry. I'm trying to think, when

did I graduate medical school? At a very critical time. That's 1997 when the fMRI machines were first becoming available and so Cynthia was able to do a rather amazing study. She got together a group of people who were clinically depressed and then another group of people who were normal and. Depressed people. And then another group of people who were normal. Healthy people. And she put them in the brain scan machine and showed them. Facial expressions. Faces.

Ranging from more neutral expressions to more sad expressions. So they saw a sad face and then a neutral face and then a sad face. That's right. And what the person in the machine was supposed to do is. To look at these faces and decide whether it was a man or a woman's face.

Eh? What does that have to do with anything? Because while they were doing that... While they're making this decision, the emotion of the face is being processed automatically. The amygdala sees the emotion on the faces at that moment. And the machine... It's like tick, tick, tick, tick, tick, and there were hundreds of pictures. Tenth of a second to tenth of a second to tenth of a second. That's right. Did you see a difference between the people who were depressed and the people who were

normal? Yes. Was it a significant difference or a just barely difference? As a group it was a significant difference. And now she takes the big step. Step four. From the pattern she sees in bunches of people, she feeds all those patterns into a computer. It's called machine learning. She told the program, this is a pattern of brain activity in depressed people. This is a pattern of brain activity in healthy people.

And then. She shows the computer a brain scan of a new person. So this is someone the computer's never met before? Exactly. And she did this a bunch of times. Right, a whole bunch of people. And each time the computer tries to guess, is this new person... Depressed or not.

And what happened? More than 85% of the time, 86% of the time, the algorithm correctly diagnosed whether that person was depressed or healthy. With just a brain scan, a computer and a patient, no doctor needed. Cynthia's computer got the diagnosis right 86% of the time. A computer. When we saw the results, it was like, wow, this is amazing. Wait a second. Has she repeated this?

Well, this is actually the very first time that this has been done with depression and so it's just a pilot study and like you say someone else will have to do it again and again and again but according to Cynthia the potential is fantastic. Psychiatry is going to be absolutely revolutionized by this. I think this method can be applied to any psychiatric disorder. Any? Autism. Yzophenia. Obsessive compulsive disorder. No? Autism. It's a phrenia.

Obsessive compulsive disorder. No way, come on. Why not? Every one of these illnesses ultimately must have an anatomical basis. Every one of these illnesses. So this means that it will soon or one day be possible for a patient to come in and you take a picture of him

in real time or of her and you will have a diagnostic tool that's what you're saying absolutely absolutely you mean to tell me that they're gonna put people in machines and just go no no no no this is not a casual thing you go to the doctor you tell the doctor that you're feeling a certain way, the doctor will talk to you. And then he would come to you and say, well, my learning and the test tells me that you're ill. So that's all that's happening here is now the mental doctor will say, I have a test. There's nothing in this that feels invasive to you? Well, obviously...

It's tunneling into the deep depths of your personhood. No, this is... if you believe that mental illness is a mental illness, it is a structural condition which can be fixed. So it's not the deep inner you, it's the broken you. So it's like the broken arm thing you started off with. So you would put the two side by side? I think I would, and then of course you get to the next.

No, come on Robert, I mean, human beings are way too messy for that. You're too messy for it to be that easy. No, no way. So you think this is out of science's reach really? It's just too... There's a part of me that does think it is out of science's reach. I think it's because you think that they're looking deep inside you. That's what you don't like. I do. I mean don't get me wrong. I find brain scans fascinating when it comes to questions like

where is the soul? What is consciousness, that kind of stuff. But don't kind of get in my head and tell me what's right and what's wrong. What if you're feeling sad and sick? Don't you want to get better? Yes, but I enjoy the comfortable ambiguity that would come from a situation like sitting in a therapist's office and saying,

well, how am I feeling? I'm feeling this way or that way. And in the messiness of trying to describe how you're feeling, there's a vast landscape of things that can happen, choices you can make, therapies you can pursue. Let me just do it this way. Let's say you are sick and you know that you're sick, machine or no, okay? If you are feeling badly, wouldn't it be nice if a machine could help you find the right

kind of help? What do you mean? Well, Eric took me through a little thought experiment. A mind experiment. You've developed the psychotherapy and I've developed the psychotherapy. We each claim it's the best in the world. Now we have an objective way of seeing. This machine allows you to independently of any evaluation see the outcome of treatment. So you can audit the doctor. Audit the doctor and give you evidence that it's working or no. Okay I'm a little bit on board. I can give you 10%

buy-in now. Okay. So how far off is this stuff? Is it gonna come soon? This is very early in the game obviously but I did ask him like how far into the future are we talking about here? Soon or long after you're dead? I'm going to be around a long time. But the question stands, will you make it to see that people actually have a… You know, one can't, in medicine, in all honesty, give a timeline for many of these things. Imaging methodology right now is quite sophisticated, but it's still primitive compared to where it needs to be. You're picking this up in status

and ascend. You've become excited as the thing is beginning to emerge. We see it for the first time on the horizon. So you're saying we got to the story too early. That's what you're saying. Just right. It's not going to be interesting 20 years from now. It'll be obvious. Or in 20 years, it'll be obvious that we were wrong. Okay? That's a real possibility because what we don't know is vast. And I want to tell you a story now about just how wrong people can be.

That's coming up right after this break. Radiolab Lulu, just before the break, Jad and Robert were debating how much science really can know about who you are from peering into one of your organs. And I want to tell you a story now about just how wrong people can be. It begins with a mystery. Sudden Infant Death Syndrome. Perfectly healthy child goes to sleep and dies during the night. It's about the worst thing that can happen to a parent. And each year it does happen

about 7,000 times. Still no one knows why. Oh, and by the way, that was Robert Sapolsky. He's a professor of neuroscience at Stanford University. And Sapolsky tells this story of the moment SIDS was diagnosed for the first time or at least classified in a Terrible mistake that was made around 1900 or so people were beginning to recognize this as a disease entity and nobody knew what was up So people decided let's go dissect SIDS kids meaning when a baby would die They would perform an autopsy exactly you, check the baby's insides. See if there's anything different in them from normal kids. That seems logical.

Absolutely. They measure the size of the baby's lungs. Yep. Normal. They measure the size of the heart. Yep. Nothing strange there. Stomach, kidney, liver.

Yep. Those are all fine. Then they would look in the throat. They look in there and they say, oh my god, these SIDS kids, they have enormous thymus glands. The thymus. The thymus. What is the thymus?

Yeah, what is a thymus? You may wonder. Well, it is a little tiny pink gland that is right here, behind your collarbone, at the base of your throat. And its job is to help you fight disease. It makes one type of cell critical to your immune system. Especially in times of stress. Hmm.

In any case, normally this little organ is about the size of a tiny tube of toothpaste, like the tribal kind. But in these SIDS kids, it was huge. Humongous, enormous, twice the size. Exactly. And since the thymus is dangerously close to the windpipe, doctors came up with a hypothesis. A perfectly reasonable hypothesis. Which was that maybe if you're one of these babies with an enlarged thymus and you're asleep and somehow you roll over wrong, well that gland might press down on your trachea and suffocate you during the night.

So, ding, ding, ding, medical mystery solved. Really? No. They even came up with a name for it. It was called status thymicolymphaticus. Status thymicolymphaticus. It was in all the pediatric textbooks by the 1920s. And you would look in there and there

would be pictures, there would be pictures of the dissected thymuses. Normal size and here on the right. Enlarged. Abnormally large status thymical lymphatic. And in no time at all, doctors came up with a treatment. A perfectly logical therapy. Which is that if we're going to help these babies babies, we've gotta shrink their thymus glands. And to do that, the best solution, obviously, is to

irradiate their throats. Irradiate their throats to shrink their thymus glands. Zap the child's throat with trillions of radioactive particles. Really? Literally? You betcha. And this was considered like something every good loving parent should do?

Absolutely. If you worry about your child being at risk for SIDS, go and get their throats irradiated to shrink the thymus glands. And did it work? Yes, it shrank the thymus glands. But, he says it did have another effect. Decades later, you've killed 20-30,000 people with thyroid cancer. 20-30,000 deaths, that's a real number.

Yeah, that's a fairly big one. So here's my question. How could these doctors have gotten it so, so, so, so, so wrong? Do you know what I mean yeah I do I don't you know what you mean you just answered your own question a minute ago no no I didn't I'm about to answer it right now they're they're playing with radiation you just said that but what they but they didn't know that radiation would hurt you they had no it was a brand new technology but that's not what I was

gonna this was a couple of decades into radiation having been discovered. Isotopes are performing near miracles of diagnosis and discovery. People were just tossing around radiation all over the place. Iodine 131. Radioactive sodium. Radar. Gamma-ray. Neutron.

And this was a period with Madame Curie, like dipping her arm into vats of uranium Radioactivity is harmless and dying soon afterward from cancer People would go into shoe stores and they would have their feet x-rayed Yes, x-ray is a wonderful invention I had that You had that I did

Yeah, you take off your shoes and then you can look at your bones That's exactly what they would do Why would you do that? That's what, that was the thing you could do at the shoe store. It was very cool. Yeah, showing how cutting edge of a shoe store they are. So that's your explanation. No, that may look like the explanation. I mean sure radiation would play a role but if you would have let me say what I

was going to say, I would have told you the real explanation. Yes. Preceded the radiation by like a couple hundred years. Whoa, I have no idea what you're talking about. I'm going to tell you. Back in the 1700s. Okay. Well that far back. That far back before radiation, before your grandpa.

Well before the civil war or the Eiffel Tower or Napoleon. I'm talking when the red coats. Were still wearing red. Yeah. Yeah. This was shortly after the Revolutionary War. I'm talking when the red coats. Yeah. Right about this time, says Cebulski, the first med schools started to pop up in America and a supply and demand issue came into effect. Because with these med schools came med students who needed to learn about anatomy.

And of course, in order to do that... They needed bodies. You know, to dissect. This produced this whole occupation. You could be a resurrectionist. A resurrectionist? Yep. And they would go out and dig up bodies that night. And sell them to the anatomists at the medical schools.

I'll need two more by Thursday. Dig. Dig. Now here's the key point. Since demand was so high, the resurrectionists had to go where the bodies were easiest to get. Which meant, you know, avoiding the fancy graveyards. If you were wealthy, you could have yourself buried in what was called a patent coffin. Hmm? Which was a triple layer coffin which was meant to be resurrectionist proof. But if you were not wealthy, no fancy coffin for you, you'd probably just be buried in a sack in some popper's field just a few inches

under the soil. Very accessible for these resurrectionists. Not surprisingly, that's where they went. Are we still on the same topic? Are you explaining why children died? Yes, yes, yes, bear with me. All right. What I'm trying to make is that the grave robbers targeted the poor.

So much so, that sometimes when they, people would catch these resurrectionists in the act and see like, oh my god that's my dad you're digging up, there'd be riots. Get them! Get them! Stop stealing our bodies! Troops were called out and rioters were shot. Are we talking like hundreds of people torch-bearing?

Yeah, it was townies versus the people who were trying to dissect their dead relatives. Okay, okay, this is a lot of history and I'm very fascinated in quotes. What does this have to do with kids dying of theft? All right, let me bring it home. Not that it hasn't been interesting, but bring it home. Okay, as a result of all of this hubbub over grave robbing, country after country throughout Europe decided, well, let's standardize how science gets its cadavers. Forget all this grave robbing. So they passed laws. Which formalized anyone who died in a poor house,

their body would be turned over to the anatomist. This was like the cadaver version of direct deposit. So grave robbing was gone, but now all the bodies used by medicine, not just some, but nearly all, now came from the poor. Estimates were by the end of that century 99% of the bodies used for anatomy lessons had been derived from poor houses. And that seemed okay. Until...

1936, a guy named Hans Selye showed... That being poor actually warps your body. And now, Robert, now we come back to the case of the mysteriously enlarged thymus. Because if you're poor, you're worried about your job. You're worried about feeding your family. You're worried about the bills. In other words, you are stressed out.

And during chronic stress, your immune system goes down the tubes. And since the thymus is part of the immune system, If you are chronically stressed, the thymus gland shrinks. For 150 years, doctors had been dissecting cadavers, pointing at organs which they thought were normal but which were in fact shrunken from a life of poverty and stress and saying, that's normal. So that when these SIDS babies show up with these gigantic thymuses oh my god in fact that was the first time they'd ever seen a normal one

people had no idea what was normal and what was abnormal and they got it backwards killing about 30,000 people in the process. Now the scary thing, says Sapolsky, is that these doctors were not dumb. No, these were the best, most careful researchers at the time and these were the only logical conclusions that could have been made. And nonetheless, it produced an utter disaster. There's, you know, not the slightest reason to think we're not doing the same thing right now. So it has been 17 years since this broadcast first ran, and I wondered, you know, circling

back to the mental health stuff, the brain imaging stuff that we started the episode with, where are we now? Over the last nearly 20 years, as there have been more and more MRI studies and brain scan over and over, Our machine learning is letting us predict with way more accuracy differences in brains with bipolar, schizophrenia, and depression, but we do not have MRI-based tests for these things in clinical practice.

There just isn't any biomarker that we found that's reliable enough for everyday clinical use. So we're definitely not living in the world that Eric Kandel predicted we might be, you know, where you can just scan your brain and tell you what's wrong with you and how you need to solve it. But we're not quite living in Jad's world either. We're sort of in this in between still. But it's one I have to say I kind of like, where on one hand, the science is getting

better at telling us what's going on with our brains, and yet it's not good enough that doctors can quite use it. It's still forcing doctors to treat us as full individuals, as not just brains, to reckon with our full personhood. And so we walk forward, the in-between, to the slow but steady beat of science. ["Skyfall", by The Bachelorette, by J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J.S. Bach, and J I'm from London, and here are the staff credits. Radio Lab was created by Jad Abumrad and is edited by Sorem Wheeler. Lulu Miller and Latif Nasser are our co-hosts. Dylan Keefe is our director of sound design. Our staff includes Simon Adler, Jeremy Bloom, Bekah Bresler, W. Harry Fortuna, David Gable, Maria Paz Gutierrez,

Sindhu Nianasambandan, Matt Kulthi, Annie McEwan, Alex Neeson, Sara Khari, Sarah Sambach, Anissa Vitsa, Ariane Wack, Pat Walters, and Molly Webster. Our Fact Checkers are Diane Kelly, Emily Krieger, and Natalie Middleton. Hi, my name's Diana, and I'm calling from Madrid, Spain. Leadership support for Radio Labs Science Programming is provided by the Gordon and Buddy Moore Foundation, Science Sandbox Assignments Foundation Initiative, and the John Templeton Foundation. Foundational support for Radiolab was provided by the Alfred P. Sloan Foundation.