In episode 132, Jesse talks to Dr. Lucy Cheke, Psychology Lecturer at Cambridge University, about memory and the negative consequences of obesity.
Types of Memory
Memory is not a monolith — there’s more than one kind of memory.
- Episodic memory is the ability to remember your own personal experience and replay previous events.
- Declarative memory involves remembering facts (for example, “Paris is the capital of France”).
- Procedural memory is the ability to remember how to do a task, like ride a bike.
Each type of memory is stored in different areas of the brain. So, for instance, damage to your hippocampus can affect your episodic or declarative memory, but not your procedural memory.
Dr. Cheke’s most recent research looks at the effects of obesity on episodic memory.
How Obesity Affects Memory
To measure how obesity might affect episodic memory, Dr. Cheke and her colleagues designed a video game in which players where shipwrecked on an island and had to hide specific food in certain spots, across two different days (in the game, not in real life). They then quizzed participants on all three aspects of the game they just played: what was hidden, where it was hidden, and when it was hidden.
People who were obese were less able to accurately identify where they had hidden particular items on a particular day. This indicated to the researchers that obese participants had trouble integrating the disparate components of the video game into a coherent episodic memory.
And it’s not just episodic memory that’s negatively impacted. Other studies have shown that obesity causes problems with working memory (also known as short term memory) and executive function (the ability to switch between tasks, focus, and keep track of information).
Obesity doesn’t exist in a vacuum. There are many related conditions, e.g. high blood pressure, insulin resistance, etc., and each of these conditions affects the brain. One example: in the brain, insulin functions as an important neurotransmitter for learning and memory.
At the same time, it appears that the very fact of having extra body weight does negatively impact the brain, significantly increasing inflammation.
One of Dr. Cheke’s upcoming studies will try to tease apart these contributing factors and figure out which aspect of obesity affects memory. (Spoiler alert: it seems likely that it’s a combination of multiple factors.)
PS: Remember to sign up for our weekly Brain Breakfast email.
Episode Transcript hideshow
— This Week in Neuroscience --
Jesse: Okay, so we're talking about memory in this episode, and appropriately I've got a memory-related thing that actually did just get published recently. It's not quite This Week in Neuroscience, but it's just a couple of months ago in neuroscience. There were findings published by researchers at the University of Waterloo, that when you need to remember something, at least when you need to remember a single-word something, there is, as far as they can tell, no better way to do that than to draw a picture of it. They gave research subjects 40 seconds to either draw a picture of something or to write the word as many times as they wanted to, then they would distract them with classical music for a little bit. They did this several times and then asked people to just recall spontaneously as many words that they either drew or wrote down as they could. By a shockingly large degree, the drawn words were always recalled better; sometimes twice as many drawn words were remembered as the ones that were written down.
They mixed this experiment up, they tried different things to see if anything could stand up to drawing. They had test subjects list the physical characteristics of the objects, they could try to visualize and create mental images, they could view pictures of the objects depicted by the words, and everything was soundly defeated by just drawing a picture of it. Now interestingly, the artistic quality of what was drawn didn't matter. So they were given a word like apple, and some people would draw a perfectly beautiful apple with a texture and a sheen and all that stuff, and others would just draw a stupid little apple icon that any third-grader could've done. But that didn't matter. The artistic quality was totally unrelated to how well it was remembered. So, this is a little memory hack that could be used regardless of your artistic ability.
Now in the original study, they were given 40 seconds to try to get the word into memory, but even when they tightened the time down to just 4 seconds per word, drawing still came out on top. So, this was a pretty solid finding. Drawing slaughtered the field in everything they put against it. The large caveat is that the testing was done on single words only and things that were concrete nouns that you could draw a picture of, not abstract ideas or full sentences or things like that. So, next the researchers are going to be trying to figure out both why drawing is so potent for benefitting memory and how a strategy based around drawing might be applied to other types of information besides concrete nouns.
-- Main Interview --
Dr. Cheke: I first started off being really interested in studying memory, and from the perspective of what kind of memory different animals have, and how memory develops in children. So, that's really what my PhD was focused on, looking at how we can understand what young children and what animals are experiencing when they have a memory. Is that they're reliving the kind of playback videotape DVD in their head like we are as adults, or whether or not it's a much simpler thing that's going on, and how we would be able to experimentally assess that. So that's what I was looking at with my PhD, and then that kind of moved into looking at how we make decisions based on what we remember. Different things can affect how well we remember different events, and that's how I got into looking at how obesity may impact memory and how memory may be important for making decisions as to what you eat and why you might eat certain things at certain times, and how much to eat. So, it was kind of a bit of an indirect path in some ways, because I started off working with birds. Starting off, just being interested in memory generally, then looking at how memory in a nonverbal way, like when you're not communicating your memories, how it affects your behavior, and then from that into how it affects food decisions and basic everyday decisions that we make in our lives.
Jesse: And specifically it sounds like you're dealing mostly with episodic memory, like, "I was here in this place, this thing happened." Not procedural memory, like, "I know how to ride a bicycle."
Dr. Cheke: Yeah, absolutely. So there's many different kinds of memory. So, as you say, there's procedural memory, things like, "I know how to ride a bike, I know how to drive a car," and then there's something called semantic memory, which is more like knowledge and facts about the world. So, I know that the capital of France is Paris, I know that my mother's name is Ruth. And then episodic memory is distinct from those because it really involves a re-experience or a replaying of a previous event. So, not just knowing that the capital of France is Paris, but remembering a time when I went to Paris and what it smelt like and what it sounded like, and what I experienced while I was there. That's really an episodic memory, and that's what I focus on.
Jesse: The difference between semantic and procedural memory is an interesting distinction. The distinction makes perfect sense when you explain it, but I wonder, within the brain, are those sorts of memories stored differently or do we know?
Dr. Cheke: Yeah, they are very different, actually. So, in terms of the kind of theories of memory, there are examples of either declarative memory or non-declarative memory. Declarative memory is a kind of memory that you could state to someone else. So, knowing that the capital of France is Paris, I know that I know that, I can describe it, I can explain everything about that memory. Whereas knowing how to drive a car or ride a bike, a lot of that has to do with what your body knows. What exactly angle do you have to lean when you're turning a bike so that it turns but you don't fall over? You couldn't quite describe that, but your body knows how to do it. And it's the same catching a ball. It's something that you learn how to do and it's something your body remembers, but if you had to describe exactly what you have to do to catch a ballÖ Actually most of the things that we do when we're catching a ball we're not even aware we're doing. So, it's very much whether or not it's something you're consciously able to kind of access and describe vs. something that is a lot more unconscious. What studies have found is damage to certain bits of the brain, so for example the hippocampus, that can damage declarative memory. So, people will be totally amnesiac as far as not being able to recognize their family, not being able to remember what they did five minutes ago, but they're perfectly able to learn tasks that involve procedural memory.
So, the most famous example is a patient known as HM, Henry Molaison, who was kind of the first one that was studied when looking at episodic memory when it was first explored, and he had a bilateral temporal lobectomy, so both of his temporal lobes were completely removed in the hippocampus and a lot of the surrounding areas. And he was totally amnesiac; he wasn't able to form new memories, he wasn't able to remember what happened to him five minutes ago or describe where his house was. But he could learn new skills, and the big one that they tested that with was something called the mirror drawing task. So, in this task you have to trace a shape, which would be very easy if you were just being able to look at your hand. But they block your vision of your own hand and you have to look at your hand in a mirror, so you have to learn to do all the movements backwards. And that's something that, the first time you try it, it's very, very difficult to do, but something that people pick up, they learn quite quickly. The scientist, Brenda Milner, who was studying him, brought him in every day to learn this task. Every day she said, "Have you ever done this before?" He said, "No, I've never done this before, I've never seen this and I don't know who you are." But he was learning the task over time, he was getting better and better and better at it, so that on the third or fourth day, as soon as he came in, he was like, "No, I've never done this before. Oh, it's really easy, isn't it?" because actually he'd been spending four days learning it.
So they're very dissociable in the brain, you can get different people with different legions to different parts that have what we call a double dissociation. So, one kind of legion can impact procedural memory but leave declarative memory intact, whereas another kind of legion can impact declarative memory and leave procedural memory intact. So, we can see that these memories are stored in very different places in the brain.
Jesse: What was it that gave you an inkling that there might be something fruitful to look at there, how a person's body mass index, whether they're obese or not, might affect their memory?
Dr. Cheke: So originally, it was a slightly by-chance reading of a paper. So there's a paper that was published in the ë60s that was looking at decisions made in supermarkets. And you know this kind of standard wisdom, that if you go shopping on an empty stomach, you're likely to buy the entire world's supply of food, whereas if you go shopping after you've just eaten, then you'll kind of buy an apple and a cereal bar and think that's what you can live on for the rest of the week. We're not very good at disengaging from how we currently feel to imagine how we're going to feel in the future, and that's an ability that involves something that's very similar to episodic memory, called episodic foresight. It's basically this idea that when you have to relive something in your brain that uses very similar processes as it does to imagine something that might happen in the future, that basically both of them are kind of constructive processes where we create a scene or an event in our minds.
And so I was looking at the other side of that, the foresight side, and I was interested in looking at to see what literature there was on people being able to disengage from how they currently feel to plan for a future time when they might need something else. And this study on supermarket behavior showed thatóit was very ë60s, so people were weighted as overweight and underweight by the criteria of the scientist standing outside the supermarket and making a decision about whether or not they thought that person could probably due to lose ten pounds, and that was the scientific definition of being overweight at the time. And they found that people who were overweight were not being affected by their current state, they weren't having this problem of buying too much food if they were currently hungry, or buying not enough if they were currently full.
Jesse: They were just defaulting to assume that they'll be hungry later, I would assume?
Dr. Cheke: Yeah, sort of buying an average, large amount. They were defaulting to more of a kind of knowledge-based system, like, "This is what I eat. This is what I always eat at lunchtime, this is what I eat at dinnertime." And so that got me thinking about whether or not people who are overweight generally have a tendency to think about both the past and the future in a more knowledge-based way. Like, here are facts about the world, but less in a kind of individual scenario imagination episodic memory way, if that makes sense.
It's not the clearest story of how I got into it, unfortunately. Essentially, I thought, based on that, well, maybe they're not using their episodic processes. Maybe they're using these rule-based processes. That said to me, okay, well the first question would be is there anything up with their episodic memory? Because it's easier almost to test if people's episodic memories are functioning properly than it is to see if their episodic foresight is functioning properly. Basically it's hard to say, "How well are you imagining that?" whereas it's much easier to say, "How well are you remembering that?" The theory being that they're part of the same process, so if there's something wrong with one, you'd expect there to be something wrong with the other. So that's why I started looking at episodic memory in obesity.
So, what I wanted to look at initially was whether or not there was any evidence that people who were overweight were less likely or less able to be using these kind of vivid re-experiencing, kind of memory processes and more likely to be using kind of fact-based, rule-based memory processes when thinking about specific events that happened in the past. So, coming from kind of the animal cognition literature, it's very much concentrating on how you differentiate, when you're doing a psychological task, between a memory that is, "These are facts I know about the past," and a memory that is, "I am really reliving that past event." And so what I wanted to look at was whether or not, if you got a task that specifically required that kind of reliving of a past event and needed you to integrate all sorts of different information about that specific past event that's not just facts about the world but there's what happened, exactly where did it happen, when did it happen, and be able to integrate those into one memory, would you find that people who are overweight struggled with that more? That's ultimately what the recent study showed.
Jesse: What was the methodology of the study like? That sounds like a complex thing to really probe.
Dr. Cheke: It is, it's a very difficult thing to probe. So I've created this new task and I'm kind of in the process of really getting a grip on making sure that it is definitely testing what we hope it's testing. But basically it's a treasure hunt task video game, and what you do is you have a complex scene, so it might be a picture of a desert with palm trees and an oasis and some plants off to the sideóa complex visual scene. And you're given a little bit of a backstory. You're a person who's been shipwrecked, say, on this desert island, and you've got all these food supplies but you can't carry them all with you, so you need to hide some of them around the scene and you're going to come back at a later point and try to retrieve them. Essentially what they have to do is people move these food items around the screen using the arrow keys and they have to hide them in particular places, and they hide them across a couple of different time periods, and they're just labeled day one and day two, and there's a little night scene in between. So, how it goes is you have four items, you hide them around this desert, and then you have a little night scene, and then you get the same four items again on day two, but you have to put them in different places. And so what you end up having to remember is not only what you hid but where you hid it, but not only where you hid particular items, but that particular item on day one is in a different place than that particular item was on day two. So you have to remember what you hid where and on which day, integrate those together to be able to get the right answer of where you hid a particular item at a particular time.
Jesse: It's almost like that kid's memory card game in reverse.
Dr. Cheke: It is relatively complex, but basically it's a computer game. And so what we found was people who were overweightóand follow-up studies seem to suggest that this is really being driven by people who are obese rather than people who are just in the medium/overweight rangeóthey were less able to accurately identify locations where they'd hidden these particular items on a particular day. But they were also generally less able to remember the different elements individually. So for example, this task was later split down into which of these items did you hide. Doesn't matter where you hid them, doesn't matter when you hid them, just tell us which of these you hid from a selection of items. The overweight people were less able to accurately identify the items that they had hidden. There was another task where they had to recognize locations within the map. Is this where you hid something? Doesn't matter what you hid, doesn't matter when you hid it, just is this somewhere where you hid something? And they were also less able to perform that task. And finally, which of these did you hide first? Doesn't matter where you hid them, just tell us about the order in which you hid things. So, just taking the time element out. They were also less able to complete that task. So, it seems that it wasn't necessarily specific to the ability to integrate, but actually on every single element of this memory task there was a significant reduction in ability and performance in the overweight group.
Jesse: How much of a time delay was there between when people did this hiding, when they were playing the computer game, and when they were tested on it later?
Dr. Cheke: About five minutes, so not long.
Jesse: Okay, so a really short time?
Dr. Cheke: Yeah, yeah. Generally when you're testing episodic memory, you want there to be a long time delay because that's kind of how we experience everyday life, you don't normally have to remember these complex memories five minutes after you've experienced them. But in a lab scenario, the shorter the time delay, the more you're able to actually get a lot tested within a short amount of time. So we left it about five minutes, understanding that people would do a lot better at that than they probably would in everyday life when there's a longer time delay. But actually, you get a lot of individual differences even with that short time delay.
Jesse: Were there a huge number of potential hiding places or potential objects being hidden? Because it seems like in five minutes, to forget whether an object was even part of the group that you hid, that seems like a really short time span for forgetting, unless there's just an inordinately large number of objects in that list.
Dr. Cheke: So there isn't an inordinately large number, but if you think about it, it depends what the objects that you didn't hide are. Because, for example, if I was to give you a selection of items that you hide and one of them, say, was a picture of half of a lemon. If I then show you a picture of a half lime, you might be like, "Oh hang on, was it that?" You see what I mean? Because it's quite a visually similar thing. You have to have quite a specific memory of the original item to be able to be confident that this new thing that looks kind of similar and is conceptually kind of similar isn't something that you hid. So it's not as simple as, "Oh, we're just completely forgetting it." It's more that if you haven't got a terribly vivid memory for the event, you're more likely to be confused by new things and think, "Oh no, yeah, yeah, I hid that," rather than completely forgetting the things you did hide, if that makes sense.
Jesse: Yeah, it does. So, a couple of questions. One of the things I always get tripped up by is the term "significant" in scientific study circles isn't quite what it means in popular speech. Being statistically significant vs. being a large number difference is not always the same thing. What sort of an actual change did you see in the performance?
Dr. Cheke: One of the things that's really important to emphasize is that this is one study. Whilst we talk about statistical significance, this was a small study and it was just a single study. And one of the things that's a big deal in science at the moment is really trying to emphasize that we can't make huge sweeping conclusions about the significance, the everyday meaning of the word, of a particular result when it's just from one study. Because we don't know. There may have been something special or unusual about the 50 people that were in this test that won't generalize. Within this particular study, there's about a 15% difference. So if you split the sample into people who are healthy weight, that's a BMI between 18 and 25, and those who are overweight, so that's 25 and over, we found around a 15% difference in performance between those two groups. You might consider that a relatively significantly large difference in terms of general significance. But again, it is difficult to know from a single study, we'll need several replications to really get a handle on whether or not this is something that we really would expect to be seeing in overweight populations out in the world. Because 50 people in Cambridge, England don't necessarily reflect a different 50 people in, you know, Texas. But it's kind of a starting place.
Jesse: Something else that would actually be interesting, and you might not have enough data with just 50 people in the original study to look at this, but to kind of throw out the middle of the people that would qualify as overweight but not obese, and look at the people that are sort of in the healthy range. So in the US, I think the definition of obese is more than 30 pounds overweight. So if you're between 0 and 30 pounds overweight, you're not obese yet. But to look at what the difference is between truly obese people and people that are not overweight at all.
Dr. Cheke: Yeah, absolutely. So in this study, we had everyone from a BMI of 18, which is at the skinniest end of healthy weight, right up to a BMI of 45, which is kind of into the morbidly obese range. So, we generally say if you've got a BMI between 25 and 30, then you are overweight, and 30 to 40 is obese, and then 40+ is morbidly obese. And what we found in that particular study was generally a linear relationship between BMI and performance, that as you got further from this kind of arguably arbitrary cutoff of 25, then the worse your memory score seemed to be. Follow-up studies that we haven't published yet seem to suggest that it may not be as linear as that, it may be that actually we're really looking at a problem with people who are obese rather than just overweight. It may be that people who are in between, people who are, as you say, not at their perfect, ideal weight but they're not by any means obese, you may be seeing something in between, you may be seeing just a lot of individual differences depending on why people are in that range. Are they someone who eats a lot of fat and sugar but takes a lot of exercise, say? Just as there are a lot of different reasons that someone becomes obese, there are a lot of different reasons that someone might be slightly overweight. One of the challenges is to really pick apart all those different factors that are coming into it to kind of see, right, is it something that's true of the obese population as a whole, or is it just a subset of people who are obese for a particular reason, say?
Jesse: And do you have any speculations on what the actual mechanisms of action underlying this might be? That's not what you were studying in this study, so I'm just asking for fanciful non-scientific speculations. But if you were to do some research to try to find out what was going on, what rocks might you want to look under?
Dr. Cheke: In talking of fanciful unscientific speculations, I have many, many ideas. In terms of what there's suggestive evidence for from other places, there seem actually to be a number of ways that this could be happening. So, when we talk about obesity, obviously the very simplest definition is looking at people with an excess of body fat. But, we all kind of know that that's not the only thing going on generally when someone is obese. Generally, someone who is very obese is more likely to have some comorbid conditions, so conditions that are related to their obesity. So, for example, high blood pressure or insulin dysregulation, or diabetes, like insulin resistanceÖ And there are a lot of other conditions that are related to obesity, and every single one of those has a potential mechanism through which it could be affecting the brain, and there is evidence that it is affecting the brain. So the brain, for example, relies very heavily on having a good supply of blood and a good supply of oxygen in your blood. So, if you have something wrong with your cardiovascular system, if your heart is not healthy, that could have an impact on how healthy your brain is. And when you have problems with oxygen to the brain, that's when you get things like strokes or what's called transient ischemic attack, kind of little strokes that cause little tiny damages to the brain, which may result in the kind of small memory deficits we might be seeing.
So that's one potential avenue. Insulin is a really important factor and one that I've been looking into recently. So we know being obese causes, or certainly is associated with, dysregulation of insulin in the body, and that's why people who are overweight are more likely to become diabetic. But one thing that a lot of people don't know about insulin is that as well as having a very important role in the body for regulating our blood sugar and so on, it also has a very important role in the brain, and it acts as a neurotransmitter, as a messenger within the brain, and is really important for learning in memory. So if you stop insulin working in the brain, you get severe problems with learning new information with memory. Certainly in rodents you do.
Jesse: That's fascinating. I didn't know that.
Dr. Cheke: Yeah, yeah, it's kind of a little known "evening job" that insulin has. So if you have problems with regulating your insulin, it's likely that you will also have problems with all the things that insulin does in the brain as well. And so it's possible that what we're looking at, in terms of these memory deficits that are potentially there and associated with obesity, has nothing to do with body fat per se, that it may be to do with these other issues, like hypertension, or high blood pressure, or insulin dysregulation. However, on the other hand, there is some evidence suggesting that body fat on its own, without any of these things, can cause problems with the brain. So, for example, there was a study done with mice where they basically took all the fat from a fat mouse and put it into a thin mouse, and they found that this was associated with a significant increase in inflammation in the brain of those mice. So, suggesting that just having that fat, even if you haven't done anything to deserve it, if you like, can cause problems with the brain just on its own. So basically there's a number of avenues through which this could be happening, and the likelihood is that in any particular individual you'll have one or two or three of these happening at the same time. So it may be the case that, however you get there, obesity is a problem for the brain, but depending on how you got there and what your particular issues are, it may be a different kind of a problem for the brain, which may result in different kinds of problems with learning in memory. This is one of the things we really want to get down in the nitty gritty and work out what's going on.
Jesse: And to a certain extent, what people should do with this information is the same in any case, is avoid obesity almost at all costs. There's very little advantageous about getting yourself into that situation.
Dr. Cheke: Yeah, I mean the difficulty is that the advice ultimately doesn't change. Eat a healthy, balanced diet and take a lot of exerciseóthat's been the advice all along and it's still the advice. I think the only difference really is for people who are worried specifically about their memory. So, if you're overweight and you need to lose weight for health reasons generally, this is just another of those health reasons. If you're someone who's specifically worried about their memory, then this might be a new thing to try. Eating a healthy diet and taking exercise and so on, and trying to keep within a healthy weight is good for your memory as well as being good for your body and your general health.
Jesse: Did your study look at all at how obesity might affect procedural or semantic memory, or was it purely episodic?
Dr. Cheke: So this particular study was purely episodic. There have been other studies looking at other things. I don't know about procedural memory, but certainly there's been studies looking at what's called executive functions, and that's the kind of work center of the brain, the ability to switch between tasks and the ability to keep track of information whilst you're doing things. The sort of "desktop" of the brain, if you like. And part of that is something called working memory, which is the ability to keep information in your head as you're using it. So, if someone's told you their phone number, the ability to keep that information in your head as you're typing the phone number into the phone. That ability has certainly been shown to have problems associated with obesity, that people who are overweight are more likely to be unable to keep track of that information quite as well during tasks in ongoing decision-making situations. So that would probably link in quite well. Procedural memory, I don't know. I think it's probably likely that it'd be less likely to have an impact on that, because generally the bits of the brain that deal with procedural memory are a little bit less sensitive to stuff going a little bit wrong. The hippocampus is kind of like the canary in the mind, if you like, of the brain. Stuff goes a bit wrong and the hippocampus is one of the first things to start going wrong, and that's the bit that really deals with episodic memory. You're more likely to see problems there when stuff is going generally slightly wrong in the brain than you are in some of the other areas.
Jesse: That's true, isn't it? When the brain takes a hit or when you have too much to drink or something, the laying down of recent long-term memory is always one of the first things to go wrong.
Dr. Cheke: Absolutely, and if you cut off nutrients to the brain, then it's one of the first things to really react to that. So if you cut off oxygen, say, to the brain, something called hypoxia, then the hippocampus is really one of the first things to start getting damaged by that. So, you'd really expect these kinds of episodic memory processes to be kind of the first thing that you'd see going wrong when there's issues with the brain. So, you're more likely to see these kinds of problems than you are to see problems in things like procedural memory in populations where there might be something going wrong with the brain.
Jesse: One line that I didn't exactly understand from one of your papers, it said, "After controlling for age, sex, and years in education, the effective BMI on the individual what/where/when tasks remained, while the WWW dropped below significance." Maybe I was just reading what the WWW stood for wrong, but I thought that that was sort of the combined effect of the what/where/when, and if I was understanding that correctly, I wasn't sure how each of them individually could have significance without the combined lot having significance.
Dr. Cheke: Okay, so there's a couple of points here. So firstly, the individual elements and the combined were tested slightly differently. So, it wasn't just adding up the three what/where/when tests, but the ability to remember each of them as they relate to each other in a particular single memory. So, you could recognize all the things you've hidden perfectly well, recognize all the locations you've hidden perfectly well, but not be able to assign the correct item to the correct location. So, it's this binding of those three information into a single memory rather than just adding up how able you are to remember things I hid in general, places where I hid stuff. It's the ability to remember I hid this exactly there on that particular occasion. So, what we found was all of those were impaired, but when the overweight people were doing the combined integrated what/where/when memory task, the kinds of errors they were making suggested that whilst, yes, they seemed to have some issue in recognizing things that they had hidden and locations that they had hidden, the kind of errors they were making in the combined task were binding errors. So they might put the right object but in the wrong location, or the right object in the right location but for the wrong day. So, it's this ability to integrate them all into a single coherent memory that that particular what/where/when task was testing.
In terms of the analysis that showed that once you controlled for all the other factors, those three remained significant and the what/where/when one didn't. Again, it's coming back to the idea of what significance means in a scientific statistical test. We use something people may be familiar with, called a P value. It's kind of convention that a P value of 0.05 is statistically significant. What that means is that there's a 1 in 20 chance of this result having appeared by chance, and the more you divide the degrees of freedom, so the more variables you shove into an analysis, the less able you are to detect a statistically significant effect with a given effect size. So, effect size is kind of like what you mean by significance in everyday terms, like how big is this effect, how big is the thing you're seeing. Whereas statistical significance is how likely is it to have happened by chance. So you could have something that's really huge, but for various reasons, very likely to have happened by chance. Or something that's really tiny, but is very, very unlikely to have happened by chance. Does that make sense?
Jesse: Yeah, absolutely.
Dr. Cheke: So, the issue with that particular analysis was, with the number of people we had in the sample and the number of variables putting in, you would have to have a very, very, very large effect size to be able to get a statistically significant result. Just because of the number of variables you're putting in, you're kind of chopping up data into smaller and smaller chunks. So, what happened was that the P value once we added these extra values went from being around 0.04 to being around 0.06, both of which are very close to 0.05, it's just that it's moved from slightly to one side to slightly to the other side. So, I'm not saying that it's not a meaningful difference, but it's not a difference that I would take very seriously, because what you're basically saying is that it's gone from havingÖ
Jesse: Just over the line to just under the line, yeah.
Dr. Cheke: Yeah, exactly. So, it's not to be ignored but it's not, kind of, the effect disappeared. The difficulty was that we weren't actually powered to do that analysis, but because of the various ways that the publication goes, the reviewers asked us to control for everything, so we did. So basically that's why that happened. The effect size was pretty large, but with only 50 people, you can't put a million variables into a regression and expect everything to come out significant.
Jesse: You mentioned that you have a couple follow-on pieces of research that you haven't yet published on. Can you give us any hints about those or tell us about any future research that you have planned?
Dr. Cheke: I can't tell you anything about the results at the moment. We've got two studies that should be coming out hopefully relatively soon, one of which is a brain-scanning study doing the same task, but this time we've controlled very closely for the match between theÖ We've only got lean and obese, we haven't got any overweight people. We're just comparing very skinny people and obese people who are very well-matched for their years in education, for their IQ, for their gender, all of that, and putting them in a brain scanner whilst they do these memory tests and then looking at what's the difference in brain activity. So, that's one thing we're doing. The other thing is really trying to pick apart some of these different factors that contribute to obesity. So, someone who is obese may also take less exercise, may also eat a higher fat/higher sugar diet and may also have higher blood pressure and so on. So, the other study is really trying to measure as many of those things as we can and trying to put our finger on what exactly is the thing that has got the relationship with memory. Is it the body fat or is it actually how much exercise people are taking, or what kind of diet? So that's the other study, trying to really pick apart these different factors contributing towards obesity and which of them is important.
Jesse: Yeah, you may well find yourself in a multivariate situation where all of them combine a little bit.
Dr. Cheke: That's my hypothesis, is actually that all of them have a bit of an effect. They have a big effect together, and separately each of them has a little effect, and you get the real problems only when you've got all of them contributing to the problem.
Jesse: I'm curious about your work with the reconstruction of memory and how that is sort of using some of the same mental toolkit as we use to imagine future episodic events, and particularly how that changes from the time when we're infants to when we have a broader conceptual framework for reality.
Dr. Cheke: Yeah, it's a really interesting thing. So, it's been a theory for the last kind of 10 years or so that the way that we remember things and imagine the future is all based on the idea that, as we experience the world and as we experience all sorts of new events, these each give us kind of little building blocks. Things that can happen and events and objects and scenes and experiences that we can either reconstruct in a kind of true way to recreate something that really happened and relive that, or we can reconstruct in a different arrangement to imagine something that could happen, or could have happened, or something completely impossible, like riding an elephant on the moon. My experience of films of the moon vs. my experience of riding an elephant means that I'm able to take those two elements and combine them into an imagined episode. And you know, a very ancient idea about memory is that memory is ultimately not a videotape that's just recorded as is that you just replayed, but it's a reconstructive process where each of the different bits of information that you've experienced are stored throughout the brain, and that when you come to relive that event, that experience, you bring these things together and you rebuild the memory from all the different elements. It's that process that allows false memory to be formed, because you can accidentally include building blocks that weren't actually originally there, because basically your imagination can be like, "Oh, I'm just going to add, also, this other thing into that memory" for various reasons.
So yes, there's this idea that the same process that allows us to take the building blocks of things we've experienced and put them together as something that we've really experienced is the same process that allows us to put them together in something we haven't really experienced, like an imagination, or "pre-experience," if you like, of a potential future episode, and that we're actually doing that all the time. And some people theorize that the reason we have this kind of memory to relive things is because it's kind of a side product of the really important ability to be able to imagine potential scenarios and plan for what we'll do in potential scenarios, and it's that ability that's the real kind of evolutionary adaptation that's really useful. It just so happens that, in developing that ability, we've accidentally ended up with this memory system that allows us to replay past events. So, a lot of the theory is that these two things are completely inherently linked as two sides of the same coin, if you like.
Jesse: You know, it makes a lot of intuitive sense to me that as we get a better conceptual framework for the way that the world works, we're able to kind of create memory shorthands for events that happened because we know that, based on the laws of physics and the laws of human social relationships, things that are likely to happen. Whereas if you're just a baby and you're experiencing things for the first time, you're getting this massive data stream through all your senses but you don't have any context for any of the data yet. I think of it as almost like saving a large raw image file to a computer vs. saving a tightly compact JPEG, because you have the JPEG compression algorithm after you've grown to five years of age or whatever it is. But it's interesting to think of what some of those differences might be. Like, the stage of memoryóobviously we can remember things to get procedural information and probably even semantic information from before we can have declarative memory, which I guess comes online at about four years of age or so.
Dr. Cheke: Yes. That's a really nice analogy, the JPEG analogy, and that sort of theory is at the center of quite a lot of theories of memory developmentÖ As you say, before you can really understand the input properly, then you can't store it as anything useful. So, if you've got this information coming in, it's not that you necessarily can't physically store it, it's that you can't store it in a meaningful and useful way. And what then the issue is is how accessible are those memories later on? Is that information still there? Is that raw image file still there, but we just don't have the kind of filing system to be able to get at it anymore, because our filing system has changed a lot since we were recording these raw image files? Or is it that they just don't get recorded properly because there's no kind of conceptual way of filing them? That's something that we don't know at the moment. It's a really interesting thing.
Jesse: Unfortunately you can't ask a baby anymore than you can ask a lab rat or a parakeet. You're studying crows mostly in your animal memory studies, is that correct?
Dr. Cheke: Yes, in my PhD I was studying crow family birds. And yeah, absolutely, that was one of the big aims of my PhD, was to look at how we can assess this kind of memory in children that haven't got good verbal competence yet, that aren't speaking and telling us about their memories yet, and in animals. That's why the test that I'm using at the moment is a nonverbal test, because having to remember where you put things and go and find them again is something that's much more ecologically relevant to animals and also to little kids. If you're crawling around the floor and you've hidden a toy somewhere, can you remember where it is to go back and get it? And you don't need to be able to describe your inner sensations to be able to perform a task like that, so that's kind of one of the ways that you can get at memory in children.
But there are some really interesting studies being done with babies, looking at how long they look at things and whether or not they seem to recognize. If they look at them for a long time, if they seem surprised by them, or whether or not they're willing to work for them. For example, there's a lot of studies that use little pacifiers, that basically the more you suck on this pacifier, the more a particular image shows rather than another one. You get babies to basically work for something they want by sucking really hard on a particular pacifier. And, for example, they'll work to hear their mother's voice but they won't work to hear a stranger's voice. And people have found that babies remember stories and music they've been played in the womb. They remember them when they're about one month old after they've been born. So, there's certainly some kind of memory going on from a really, really young age, but these questions of what's understood by it and how it's really experienced is a very difficult one.
Jesse: One last question that occurs to me, going back to your study with people with a range of BMIs: did you do anything to try to make sure that people's memories are pure episodic memories and that they weren't using some sort of memory device, like, "I'm going to hide objects that start with the letter R; if it's a raspberry, I'm going to hide it under a rock because they both start with the letter R"? Some sort of mnemonic tool that might make it not purely episodic?
Dr. Cheke: Yeah, this is something that I'm really interested in with this task, because obviously you're hiding things around the place and that gives you an opportunity to do that kind of strategizing. Actually, I have a study using the same task with older adults specifically looking at that kind of strategizing to see if older adults are more likely to strategize to kind of help their own memories. I did do that analysis actually with the obese data and I didn't find any difference at all. So, it's not that they're doing that more or less, but yes, some people do do that. Because you can see where people hid things, you can kind of explore those kinds of strategies. It's difficult because in everyday life these kinds of memories aren't separable. We don't only replay the video and completely leave our knowledge of the world behind. So, it's very difficult to really pick these things apart, so generally it's not somethingówe can't get people to purely use one kind of memory. What we can do is look at what they did and rate it for how much they used one kind than another. So, for example, you can get people to report, "Do you remember that, or do you just know?" And this kind of remember/know distinction, you can look at brain activity for things that people have reported remembering vs. things that people have reported just knowing, and you can see episodic-like memory activity for the things they report remembering vs. more semantic activity for things that they report knowing. So, you can get at it, but generally it's difficult to stop people doing both. You just have to try and get at what you're measuring for any particular trial afterwards.
Jesse: Yeah, I would think thatólike, for example, let's say that somebody had horrible episodic memory but perfectly strong semantic memory. Going through the world, assuming that they're fairly functional, they would probably learn to behave in such a way that takes advantage of their strength and minimizes their weakness. Kind of like if you're a bodybuilder that's strong on your right side and weak on your left side, if you don't make an effort to work out in a way that sort of evens that out, you might just get an even stronger right side and an even weaker left side. I can see something similar happening with the way people use their memories. Is that something that we see at all in literature or in studies?
Dr. Cheke: Absolutely. So, an extreme example are people with absolute episodic amnesia, like their episodic memory just doesn't work at all, and what you find is that they very much use strategies to get themselves through the day. And you've seen some examples. For example, there were some nice studies with amnesic children that had damaged their hippocampus. Because of hypoxia, this lack of oxygen when they were born, they're completely episodic amnesic but they have perfectly find semantic memories. What they found is they couldn't remember their way to school, but they did fine when they got there. They got perfectly good grades, they were perfectly able to understand complex concepts. They just weren't able to remember events, they weren't able to remember what happened to them five minutes ago.
Jesse: That would be even more disorienting, it seems like, to be a kid. You don't think of amnesia in the context of kids very much, but what a strange way to grow up that would be.
Dr. Cheke: Yeah, absolutely. And as far as I know, where they are now, they have jobs and they do function. They're not necessarily able to have jobs that require them to keep track of a hectic day. Because generally, to get something into semantic memory, it has to be a regularity, it has to be something that always happens, it has to be a fact about the world. Keeping track of the goings on of an office day-to-day, that's something that requires remembering specific events and making quick decisions based on that kind of memory.
So, it's amazing actually how many things you realize you wouldn't be able to do very well if you didn't have this kind of memory. It seems pointless to be able to just re-experience something in the past, but actually it's something we need to do on a day-to-day basis all the time. And there's a lot of strategies that you can come up with that generally involve writing stuff downówith Alzheimer's patients, certainly. You may forget you've put an egg on to boil, you need to write down that you've out an egg on to boil and carry that piece of paper with you. It's that kind of thing that actually is a real problem for people with episodic memory deficits, it's that sort of everyday boring, getting on with life, remember that you put an egg on, remember that you put the oven on, remember that you came upstairs for a reason, that kind of thing. That's the kind of memory lapse that we all experience from time to time. "Why did I come into this room?" That's an episodic memory lapse, because you made a decision to do something five minutes ago, but then you've forgotten what that decision was.
Jesse: You might have some really interesting insights on this, it just occurred to me: for episodic memory, it seems like it's one of these things that would be integrally tied up with our own sense of identity. Like, I watch a movie, when a character has amnesia, oftentimes it's like, well, they have amnesia for like the last six months but they remember the six months before that, and so they still kind of have the story of, "I grew up in the city of Philadelphia, my parents do this and this." But if you were a kid and you literally didn't have any series of events telling the story of how you got to be where you are and why you are who you are, what does that do psychologically do these people and their sense of identity?
Dr. Cheke: That's a very interesting question. Certainly they can learn things over time. They learn the town they live in and they learn their parent's names and everything. But what that would do to your sense of identity, I really don't know. I think it's a very interesting question. Certainly you look at what people give as examples when they're asked who they are as a person, and these are always episodic memories. Generally people aren't saying timeless facts about life, they're saying events that made them who they are. So yeah, I think it would have a very profound impact on your sense of who you are. But growing up like that, I really don't know what that would do. It's a very interesting question.
-- Ruthless Listener-Retention Gimmick --
Jesse: So, there's that sort of philosophical saying that we see the world not as the world is, but as we are. And in my case, for the last couple of days increasingly as I am has been hungry. I've been walking in city streets, things like that, sometimes just trying to get a stroll, take my mind off the fact that I'm not having meals when I'm used to eating meals. But of course, that hits you with a barrage of food advertisements and smells wafting from restaurants and all that sort of stuff just reminding you that there's lots and lots of food out there that could be yours if you were willing to break the fast, pull a little money out of your pocket, and just go get it. But the funny thing is, walking down the same city streets, normally I never pay the slightest attention to any of that stuff. The smells barely pass my perceptual threshold, I ignore the advertisements as best I can. But for the last couple of days, everything food-related is bright neon in my mind's eye, whether it is or not in reality. And in that sense, I have something in common with the humble little zebrafish.
We talked about zebrafish before, these guys are beloved of scientists because they're tiny, they breed quickly, they have a relatively well-understood genome, and perhaps best of all, they have clear skin and basically you can look right through them, you can see their nerves, you can see their little mini organs. So they're kind of like a living slide, one of those slides that you would put under a microscope in middle school. They're like that, you can see all the good stuff inside them while they're swimming around. But scientists who did not want to take the time to chase me around this week as I was being hungry instead got a bunch of hungry and non-hungry baby zebrafish and watched to see how their behavior would change and what was actually going on inside them that led to those behavior differences based on whether or not they'd been fed.
Now surprising to no one, hungry things behave differently. If you're starving, you should take more risks. It's not the time to play it safe. If you see something that's remotely food that you think maybe you can struggle into your mouth and get, you better go for it. And zebrafish, like many things in the animal kingdom, they've got both predators and prey, there are things they eat and there are things that they can be eaten by. Not being the smartest fish in the world apparently, you can study their approach or flight behavior by putting a computer screen up against the tank that they're swimming around in and showing them either large or small dots. Typically, zebrafish will flee from the large dots and chase the small dots, thinking that the small dot is something that they're going to be able to eat once they get to it. And true to form, the hungry zebrafish would chase after large medium-sized dots that normally would've been nothing to tussle with if it was a well-satiated zebrafish. So if you are hungry, you are less likely to hide from the big dots and more likely to go after any small dot you see, and even some medium-sized dots, too. But again, all this is conventional wisdom so far, nothing surprising.
But what's cool in this study is that scientists were able to see exactly why this was happening. In the hungry fish, there was inhibition in the HPA axis. That's the hypothalamic-pituitary-adrenal axis, which humans have as well, believe it or not; that is part of the neuroendocrine system. And secondly, the hungry fish had increased activity in serotonergic neurons, those are neurons that respond to the neurotransmitter, serotonin. So as a result of these two effects, additional nerve cells in the fish became responsive to the small and the medium-sized dots, so the representation of the fish's world is actually changing based on the state of the animal's hunger. So when I say something like smells smell stronger or food advertisements look larger and brighter, that might not be all hyperbole, it really does seem like that's what's happening in animal brains and very possibly human brains as well.