What's the deal with cortisol?

14:55Emily Oster:Yeah. Oh man, cortisol gets such a bad rap online and it's such a critical hormone. It's so important. So cortisol is a steroid hormone. We've talked about these several times before in wellness, actually. Steroid hormones are really interesting because the site that they act on is in the nucleus, is with the DNA of a cell. So all the steroid hormones, including things like testosterone and estrogen, get into the nucleus and change how DNA is transcribed so it can fundamentally change what cells are doing, which means all these things can do a million different things. It's not like this one protein is responsible for increasing your sugar level or decreasing your sugar level.

15:38Emily Oster:It does a million things. And I think the best way to sort of think about cortisol is that it is there primarily to save your life in the next 10 minutes at the expense of the next 10 years. So cortisol is a hormone that's present to keep you moving and alive after something really bad has happened to you, like you got mauled by a saber-toothed tiger or you are running from something that is trying to eat you.

16:09Perry Wilson:Wait, I'm going to pause you there because I think for most people, the thing they think is driving that is adrenaline. When you tell me you were chasing by a tiger, I think adrenaline.

16:20Emily Oster:Yeah.

16:21Perry Wilson:They're not the same.

16:22Emily Oster:They're not the same. And cortisol is secreted by the adrenal glands, which is where adrenaline comes from too. And they are by all means related and evolutionarily follow kind of similar paths. But adrenaline is much more cardiovascularly active in terms of like flogging your heart and whatnot. Whereas cortisol is more focused on sugar levels. So when you're being chased by a tiger, you want there to be plenty of sugar in your blood for your muscles to use and for your brain to use. And you really don't care about whether there's energy for your gut or your liver or like your bones or for your immune system even, which is a very energetic, energy intensive thing.

17:08Emily Oster:So you're kind of like shutting everything down and focusing all that glucose into muscle and brain. And so when you give someone cortisol or when they're endogenous, when their own cortisol goes up, their blood sugar goes up, Their immune system ramps down. That's why glucocorticoids, which are cortisol derivatives, are immunosuppressants, things like prednisone. Many people may have used hydrocortisone cream like for itches and stuff like that. So hydrocortisone is cortisol, the exact same molecule. There's no difference between hydrocortisone and the cortisol that your adrenal gland secretes.

17:47Emily Oster:I mean, they put it in a cream, obviously, but that's what it is. the problem that people attribute to cortisol and sometimes incorrectly, as we'll go into, is like, what if that stressor lasts longer than 20 minutes, right? You're running away from the tiger. Oh, you made it away from the tiger. Great. You're good. Everything's safe. Everything can go back to normal. What if your whole life is stressor? What if you're constantly feeling like you're being chased by a tiger and your cortisol levels are high all the time? And it tells a really nice story that like, ooh, that would do a lot of bad things.

18:22Emily Oster:But as I said, cortisol gets a bad rap and it might not be as bad as you think.

18:29Perry Wilson:So in terms of just sort of people thinking about like the way that their cortisol moves throughout the day, because cortisol is a, it is a hormone and it's, it's would be a mistake to think of it as just always flat, except when you're chased by a tiger. It is something which is fluctuating through the day, it's highest in the morning. So when you sort of first wake up in the morning, your cortisol tends to be higher and it's lowest sort of in the middle of the night when you are asleep is my sense, like on average for most people. And then it will spike at times during the day. This is something that's quite hard to measure, but in principle it would spike when you are doing either physically or emotionally stressful activities like running from a tiger.

19:12Perry Wilson:Is that true?

19:13Emily Oster:Yeah. Yeah. Yeah. The morning cortisol spike is well described. It is one of the reasons why heart attacks are more common in the early hours of the morning, for example, because your body's sort of revving up to get ready to wake up the day. And we will get to, this is a circadian rhythm, like one of our fundamental circadian rhythms. It's about five times higher in the morning than it is at night. I mean, it's a real shift over time. And if you really dig into the data, you'll find that the level of cortisol might not be as important as the circadian rhythm of cortisol. So one of the things we see in the truly chronically stressed are actually not high cortisol levels.

19:58Emily Oster:It's a loss of that fluctuation. It's flat cortisol levels. Yeah.

20:03Perry Wilson:So I think that's a good pivot into the first piece of what's tough about this entire discussion is going to be really hard about the data, which is like, what are we measuring and when are we measuring? So there are some things we measure in people where I think we can say, maybe you want to do some repeat measurements, but we have a good sense. Something like A1C levels for a diabetic. You test if fasting, not fasting, but we have a good sense. We have a stable sense of like how we test that. And we could use that to say, you know, you're a diabetic, you're not a diabetic. Measurements of cortisol, which are done typically through your blood, although actually they can also do saliva, you can also do hair, you have many different ways to test this.

20:47Perry Wilson:They don't all test the same thing. And because there's this fluctuation, it's actually a really, really hard thing to hit. If you wanted to say, what's the number that's going to tell me I'm chronically stressed in some dangerous way. There's sort of no number like that.

21:01Emily Oster:Yeah. And actually it's a lab test that's pretty bad. So there are some lab tests like sodium level that's just, it's been figured out and we can do it really quickly and it's highly accurate and you test the same blood sample twice and you get the same number and like, it's really nice. And cortisol just isn't one of those things. I mean, one of the major issues is that 95 % of cortisol in your blood is bound to proteins and is inactive. So these are just chaperone proteins that carry it around. And then the free fraction, the 5 % is the thing that does stuff in your cells. But almost all the lab tests out there are measuring the total.

21:37Emily Oster:So is it correlated with the free? Yeah, but not great. And it's hard to know. There's obviously the timing of measurement that's really important. So you really want to get, if you actually needed to measure cortisol, for example, to tell if you of a disease of excessive cortisol production. And these diseases exist. It's called Cushing syndrome or Cushing's disease. You get a test early in the morning when you've got the best chance of capturing things high. The assays that they use for cortisol have a lot of false positives. So there's cross-reactivity with a bunch of other hormones and other substances that people are taking in.

22:15Emily Oster:It's just like, I mean, I'll say right now, like getting your cortisol measured, if you don't have symptoms of a cortisol excess disease is you're unlikely to get any useful information. So I think, okay.

22:30Perry Wilson:So when we talk about the ways to measure cortisol, actually measuring your blood has both the total and free fraction issue. It also has the issue that when you take someone's blood, it raises their cortisol. Oh yeah, yeah, yeah. Because it's like, it's stressful. Like somebody's sticking a needle in your arm and you're like, Jen, it's a tiger. like I better get ready to run away it's not a tiger you asked for that but as a result like saliva is in some sense the sort of best metric of this at a moment because it's not stressful to take someone's saliva but the thing that I wonder is why you know we have continuous glucose monitors why has someone not made something that I stick in my you could put it on a zin pouch something I stick in my mouth and leave in my mouth all the time on like the side of my cheek and then it constantly measures my cortisol.

23:17Perry Wilson:Don't you think Brian Johnson would buy that in like one hot second? Like why is the market not delivering that, Perry? What's the problem?

23:23Emily Oster:No doubt people are pitching that in Silicon Valley right now. Call me.

23:29Perry Wilson:Call me VC funds. We're going to measure cortisol with our mouth sticky. With your mouth sticky.

23:36Emily Oster:I think two things. One, cortisol is complicated to measure. It's It's a hormone. It's a bigger molecule than something like glucose, which is small and easy to measure. And then the second thing is that nobody knows what to do with the answer. I could tell you what your cortisol is.

23:53Perry Wilson:That's not the answer. I'm sorry. That's not the answer to why. They're constantly – people love to measure stuff. There's nothing to do with it. The first answer was right. You should have stopped at that one.

24:00Emily Oster:Okay. We will erase the second. No, it's fine. I can be wrong sometimes.

24:04Perry Wilson:Okay. So this is very hard to measure. and we'll get into like why that probably means you shouldn't measure it. But despite its difficulty in measuring, there is actually a fair amount of research on the relationship between cortisol measured in some way we will get to and various health outcomes. And I will just like, spoiler, I think most of this literature is really bad and it is observational and not causal. And I I've been practicing my complaints. But before we get into them, I do think it's worth talking a little bit more. You alluded to Cushing's disease, but a little bit more about the places where high cortisol can actually be very damaging.

24:52Perry Wilson:We know that this is a problem.

24:54Emily Oster:Yeah. We should talk about Cushing's disease because actually a lot of the influencer work is taking what people see in Cushing's disease where cortisol levels are through the roof and saying, therefore, mildly elevated cortisol levels in normal people also do this thing. And so what Cushing's disease is, is a benign tumor, typically on the adrenal gland that secretes cortisol. It is unregulated. So it's sitting there and it's just secreting cortisol, secreting cortisol, secreting cortisol. And that is quite bad for you. The syndrome of Cushing's syndrome is characterized by pretty dramatic moon faces, what they call it.

25:34Emily Oster:The very round face that you may be familiar with if you see people on very high dose of other corticosteroids like prednisone. Sorry about these terms, but this is what the medical terms are. A buffalo hump, which is like a collection of fat on the upper part of the back at the bottom of the neck. Striae, which are really violacious, like deep purple stretch marks in the abdomen. a lot of accumulation of visceral fat, diabetes, high blood sugar, and so on and so forth. The treatment is removal of the cortisol secreting tumor.

26:08Perry Wilson:And I think it's actually very important to note because some of the things we see in the sort of wellness influencer claim space are things that seem like, well, if you have this Cushing's disease, you will have this. And therefore, if you have a mildly elevated level of cortisol in the normal range, you also get maybe less belly fat, but still some belly fat.

26:31Emily Oster:Right, right. And that has that nice sort of logical step-by-step ring to it. But if you've been listening to Wellness Actually, you know you actually have to prove these things with data. So let's talk about data. I mean, there are clearly cases where high cortisol levels are bad for you. I mean, maybe the most dramatic one is in sepsis and septic shock, where it is fairly well established that if people are coming in with very high cortisol levels in the setting of sepsis, they're more likely to die. It was a nice study in PLOS One in 2019 that took 139 patients with severe sepsis or septic shock.

27:09Emily Oster:And those who had cortisol levels that were high by their definition, had a tenfold higher mortality risk. It's a pretty large effect size. Now, Emily, that's not a causal study. As you said, that's observational. So walk me through like why, yeah, is it the cortisol that's bad or what are you thinking when you hear, okay, high cortisol, tenfold risk of death and septic shock?

27:33Perry Wilson:Yeah. So I do, I think sort of two things in that case, which will come up again later. So one is that, yeah, this is not a, randomized the cortisol. Some people came in with a higher level of cortisol than others. It seems plausible to me that that is measuring something else about how stressful, what else is going on in your life, or even how bad is this sepsis? I realize all sepsis is bad. Sepsis is not good. But could you read this as this is a metric itself of how much is the body struggling in other ways. And this is just one measure of that. I will say in this particular case, the fact that the odds ratio is 10, which means like the mortality risk is 10 times as high, that's actually a quite, that's a very big effect.

28:21Perry Wilson:It is therefore the kinds of biases I have in my mind, like differences in other things that could be going on in people's lives and so on, probably aren't driving a tenfold increase in mortality. So the combination here of the particular outcome, which is so extreme, and how large the effect size is makes me more confident in that result, even though I think it would be better to have a randomization, although this is a case in which I cannot imagine how you would randomize people to different cortisol levels. That's not really an available randomization, which is why we're going to generally need to rely on observational data.

28:58Emily Oster:Yeah, yeah. And for the medical people listening who trained back when I did, there was a time when we gave people cortisol or other glucocorticoids when they had sepsis because we thought, oh, all this inflammation is the bad, like their immune system's going crazy. We need to suppress their immune system a little bit. That was standard of care for a while. We no longer really do that except in rare circumstances. And so it is quite complicated. And I think, you know, it's certainly possible that a large amount of that observed effect size is just due to the fact that really sick people have really high cortisol potentially.

29:30Emily Oster:But most people who are listening to us are not currently having sepsis or septic shock.

29:36Perry Wilson:Yeah. Stop listening. If that's you, stop listening to this.

29:41Emily Oster:Maybe we're the one thing keeping them sane in the hospital bed.

29:45Perry Wilson:We're not.

29:45Emily Oster:All right. We believe in you.

29:47Perry Wilson:Most people are worried about the idea that a chronically elevated or frequently elevated level of cortisol or just having high cortisol, whatever people seem to mean by that, is bad for some health outcome like heart disease being the one we should start with. It's the one that comes up the most. And so this is a place where there is some data and I want you to describe the data and then I'm going to tell you my feelings on it.

30:15Emily Oster:Why it's wrong. I'll set you up in the best way I know how. Okay. I'm going to give you two studies here. One is sort of a prototypical. There's other studies in this space looking at urinary cortisol. Gosh, we didn't even mention that. There's yet another place you can measure cortisol. You can measure it in the urine. And then the risk of cardiovascular mortality. This was from the Journal of Clinical Endocrinology and Metabolism. That's a good journal. 2010, 861 participants above age 65. They got 24-hour urine cortisol levels at baseline, followed them for six years. Over the course of that time, people in the highest tertile, so the highest third of urinary cortisol levels, had five times the risk of dying of cardiovascular disease.

31:00Emily Oster:But there was no difference in deaths from other causes. So there's a big effect size, five-fold from higher cortisol levels. I told you I'd give you one other study just to give you something to talk about. So nope, I'm not. I'm going to wait on the Mendelian randomization because it's going to require some explanation. So let's start with this. My urinary cortisol is in the top third of people's urinary cortisol. I have five-fold the risk of dying from cardiovascular disease in the next six years. Sounds pretty bad. Got to get my cortisol down. Okay.

31:31Perry Wilson:I mean, look, here are some things that raise people's cortisol. being poor, working at night, having few resources, having a stressful family life, having other diseases, things like that. All of those things are independently associated with death from heart disease. We know that like a million different ways. Exercise causes you to be less dying, rate lowers your cortisol overall. All of these things are like going on in the background. And if you look at the people in this study and you look at the terciles, they differ in a bunch of different ways from each other. And it is therefore extremely difficult to imagine that we are attributing this difference in the risk of death to these levels of cortisol, as opposed to attributing them to these other things.

32:30Perry Wilson:Now, this is a case in which actually the whole thing is very complicated because, well, maybe it's these other things raising your cortisol, which is then causing you to, you know, have a higher risk of death. That's an interesting mechanism. But if you ask, like, what's the action item here? What's the, like, thing that would, it's changing these other things. It's not like if you gave someone a shot that magically lowered their cortisol, that would have this positive impact on them. These results are also, this sample is quite small. Not that many people died, which is good. and therefore this sort of hazard ratio of five is like incredibly noisy.

33:06Perry Wilson:It's consistent with an effect that's two. It's consistent with an effect of 13 times, which is definitely not true. This is a statistically noisy, very potentially biased effect. I don't think we learn anything from studies like this. I just think we learn nothing.

33:24Emily Oster:So I completely agree that there's all this stuff that can affect your cortisol that might also independently affect your risk of death. Right? We know in America, at least, being poor dramatically increases your risk of dying from cardiovascular disease. And also dramatically increases your cortisol. So, yeah. So, okay. That's classic confounding. The problem is, we all love a randomized trial, but no one's going to do a randomized trial of supplemental cortisol on people. That's not going to happen. Probably would be considered unethical. So, we're left with some other options. And I think we need to talk about Mendelian randomization.

34:01Emily Oster:Okay. So let's put on our genetic hats for a second. I sort of believe in Mendelian randomization as an approach. And Emily, I think you are more skeptical of it. So I'm going to let me tell you how I view it. And then you tell me why I'm wrong. Okay. Here's the idea. There are certain genes that increase the amount of cortisol you make independent of other things. genes that have to do with hormone synthesis and stuff. And everyone has little variations in their genes and some are just a bit more active than others. And so all else being equal, there are people who genetically live at a higher cortisol level than a similar person who had a different genetic background.

34:43Emily Oster:So, you know, and this is true of everything, not just cortisol, right? Like there's some people just genetically have a higher LDL than other people. And some people have genetically lower LDL. Okay. Because your genetics are assigned at birth, not no, at conception, that exposure happens before you know that you're poor, before you have stress in your life, before anything else. It cannot be related to that. The genetics are not related to those things. Being poor doesn't change your DNA, right? And therefore, we can use those genetic signals and say, okay, here's a person whose genes have them exposed to a higher level of cortisol than they otherwise would have been given everything that's going on in their life.

35:30Emily Oster:And here's a person whose genes have their cortisol lower than they otherwise would have been given everything that's going on in their life. And if the person with the genes that predispose them to high cortisol has a higher risk of cardiovascular disease than the person whose genes predispose them to low cortisol, I conclude, therefore, there's a causal link between cortisol level and cardiovascular disease?

35:53Perry Wilson:The term Mendelian randomization implies that something is being randomized. And so I think when we think about this technique, we want to think about the word random. So what you are suggesting is that the genes are allocated randomly, that when you are given genes, there's a piece of the genetics that are coming randomly to you. That is not true when comparing across people in the population overall. So let me give you a more like direct example of this. Imagine that instead of cortisol, we were thinking about eye color. And you said there's a gene that determines your eye color. In fact, there is a gene that determines your eye color.

36:40Perry Wilson:There's a set of genes. We understand how those work. But it is not, when we look across the population, those genes are not randomly distributed. It's not like randomly when you're conceived, you are given one of the eye color genes. It's random, only conditional on the genes that your parents have. So you have a set of parents and you have, your mom has a set of 23 chromosomes. She has two copies of each. And at conception in the egg, you get one of those. And which one you get is random. But who your mother is, the genetics that she starts with is not random. So when we talk about Mendelian randomization, there's something cool, a cool idea in here.

37:25Perry Wilson:I want you to imagine two children born to the same set of parents. And mom has two copies of each chromosome. And on one of those chromosomes, she's got like a high cortisol gene. And on the other copy, she's got the low cortisol gene. and her children, let's say one of them gets the high cortisol and one of them gets the low cortisol. That's random. And so then we have two siblings who are effectively, everything is, you know, the same except the, in this case, this particular thing is randomly allocated weight. And so you could then get many sibling pairs like that and say conditional on your family, controlling for your parentage, comparing you to your siblings, we're going to see whether the sibling who was randomly allocated the high cortisol gene has worse outcomes than the sibling that was not randomly allocated.

38:21Perry Wilson:That's a really cool idea that uses genetics and randomizes. Do you agree?

38:28Emily Oster:Yeah. All I wanted to say was that in my mind, one of these kids is named Gene and the others named Eugene.

38:35Perry Wilson:I love that so much. Okay. But what, okay. So that technique,

38:41Emily Oster:just trying to throw you off.

38:43Perry Wilson:That technique works like, works sort of in concept the way I've described it. And there are some, actually some issues with using that technique in practice, but at least in principle, there is a real randomization there. Here is the problem. The studies that are being done that are using a technique they are referring to as Mendelian randomization do not actually do what I just described. Instead, what they do is they don't look within sibling pairs. They look just across people in the population. And that is ridiculous because genetics are not randomly allocated across people in the population.

39:21Perry Wilson:I mean, an equivalent thing is if you said, well, I'm interested in the impact of race on something. And I don't think I can just look at people's race because that's confounded with all kinds of other things, but I'm going to look at a gene for skin color and I'm going to relate that to the outcome. Well, that's stupid. That's not, like, doesn't make any sense. That doesn't work. And so this technique is not being, it's not that it's conceptually poor, it's being used in a way that makes no sense.

39:46Emily Oster:Well, so there is a caveat to Mendelian randomization. So to take your skin color example, which says that in order to use this technique, you have to show that the gene is associated with the phenotype, right? So the gene is associated with skin color, let's say, and the gene is associated with the outcome, but the gene is not associated with the outcome through any plausible path outside of the phenotype. So there is like some statistical testing you can do for that. Yeah.

40:18Perry Wilson:That's called the exclusion restriction. And in the cases that you've described where we're comparing, say, two people with different genetics across different families where, for example, one of them is born to a set of parents who have high cortisol and one of them is born to a set of parents with low cortisol. Now your exclusion restriction has failed. And so that just doesn't work. So I think you're right. There is a thing and it's totally failing and these guys don't seem to understand that. I just don't understand. I fundamentally, Perry, this is like drives me crazy because people in medicine use this technique all the time.

40:53Perry Wilson:And I just think they don't understand what they're doing. And I can't understand why they don't understand.

40:57Emily Oster:Let me give you the results of the cortisol Mendelian randomization, because maybe this is a moot point. Okay. So in the observational framework, we've seen that a high level of cortisol increases your risk of cardiovascular death by 500%. Okay. In the Mendelian randomization analysis, genes predisposing to high cortisol increase the risk of cardiovascular death by 8%. Right.

41:21Perry Wilson:I mean, okay. So one thing is, if you think what Mendelian randomization is doing is fixing your causality problem, then I think what you would say, okay, now we've got a better result and it's basically zero. It's basically zero. It basically doesn't matter.

41:37Emily Oster:It does not suggest causality.

41:38Perry Wilson:I will say, I think counter, counter example, as I've noted, I don't think the Mendelian randomization really makes any sense at all. And part of the reason that the effect I think is much smaller is because the genes they're identifying have very little predictive power on actually having high cortisol. And so like just these genes are not have very high penetrance. And so this may be failing and giving you a very small number for reasons that kind of have nothing to do with getting rid of bias. They may just be generating noise. I hate this technique. I understand that I'm being like, and talking about, I hate this technique.

42:20Emily Oster:That's fine. We can't randomize people to high cortisol. We can't supplement them with cortisol or placebo. So we're stuck. We have to make some inferences. But I will say that based on this data, there's not a compelling argument that high cortisol level in and of itself is causal of cardiovascular disease.

42:38Perry Wilson:Yeah, I totally, that I think we totally agree on. I think your question about give me a better technique is an interesting one. I mean, I do think a true Mendelian randomization with a sort of a sibling pair study would be an interesting thing to explore. It could have some other problems, but I think that would be something interesting to explore. I agree this is a very difficult problem. I think one thing you could imagine randomizing if you sort of, at least in a short-term way is, you know, some of the, the kind of like behavioral techniques that might lower people's cortisol, like meditation.

43:12Perry Wilson:You can imagine running a study where you're like one group gets like a lot of meditation support or some other, you know, therapy, something that would lower their cortisol. It probably have other effects. So it's not like a great study, but I, you know, like I, it's something like that to try to affect people's cortisol. I agree. We're not going to find anything, but it doesn't mean it's not interesting to, try.

43:33Emily Oster:To have least luck. So, all right.

43:35Perry Wilson:Sorry I got so exercised, Guy. I feel like people are going to be like, why are you so exercised about this? But it's like, this is my job. This is like a thing. This is like the core of my professional life, other than parenting writing is like statistical methods.

43:50Emily Oster:What you guys don't know is that Emily's first boyfriend was a Mendelian randomization study. So, okay. I want to move rather quickly through the other stuff that people say that high cortisol does if we're going to write off cardiovascular disease. And again, guys, we're not saying that there's no association between higher cortisol levels and these things. We're asking, is it the cortisol, not the life stress and the other stuff? So I I think a big one is diabetes, metabolic disease, right? We certainly know that people with Cushing's syndrome, much higher risk of diabetes. We know that when you put people on steroids like prednisone or hydrocortisone, their blood sugar goes up.

44:37Emily Oster:I was surprised here, Emily. There's a nice meta-analysis of 21 studies. This is appearing in Psychoneuroendocrinology in 2018, 11 ,000 people or almost 12 ,000 people, no significant difference in basal cortisol levels between subjects with and without the metabolic syndrome, which I just wouldn't have... And that's not even a causal study. It's just straight up, there's no link.

45:01Perry Wilson:What's so interesting there is I would have thought that we would see an association there just because of all these other factors. Totally. But it seems like even...

45:12Emily Oster:Even the confounding can't... Even the confounding can't get you there somehow.

45:16Perry Wilson:Maybe that's because it's hard to measure.

45:18Emily Oster:um let's do stubborn belly fat yes so much of it okay how do i mean i have so much stubborn

45:27Perry Wilson:belly fat i don't think you do it's so stubborn i'm always arguing with it it's always making

45:32Emily Oster:these terrible points our producer tamar just rolled her eyes like emily saying she has belly

45:38Perry Wilson:fat how do you describe stubborn well okay anyway i think i i as a middle-aged man stubborn belly

45:45Emily Oster:fat is, I get it. It's stubborn because it's like, I work out, I try to eat right, and it's still there. And that's annoying. So I did look into this for us. Once again, in Cushing syndrome and Cushing's disease, there is a substantial increase in visceral adipose tissue. That's what belly fat is as opposed to subcutaneous fat, which is kind of all over your body. Belly fat is visceral fat, and it's not good for you. Higher amounts of visceral fat are associated and potentially causally with cardiovascular disease. But in a study appearing in clinical endocrinology in 2024, they looked at visceral fat and subcutaneous fat among people, just regular Joes with varying cortisol levels.

46:30Emily Oster:And there was no association between higher cortisol level in the non-pathologic range, in the non-Cushing's disease range, and belly fat. So that's not it, folks.

46:40Perry Wilson:I also think it's worth saying, like, when people are chronically stressed, it actually isn't, we said this at the top, it's worth saying again, it's not that chronic stress means you always have a high cortisol level. It actually means your cortisol is very flat as opposed to moving up and down like it's sort of supposed to typically. And so, like, your cortisol is supposed to be responsive to, like, events and to the day, which, of course, is evolutionary valuable because you want to really be, like, able to respond when the tiger comes. When people are chronically stressed, they can get a sort of flattening of the cortisol, but it's not that it's higher.

47:15Perry Wilson:And so this sort of whole narrative is kind of confused.

47:17Emily Oster:Yeah, it can even burn out. So there's some nice studies in PTSD, in patients with PTSD that actually shows your typical patient with PTSD has lower cortisol levels than controls who don't have trauma. So yeah, this whole idea that like your cortisol is so high and because you're so stressed, if anything, it's blunting the cortisol response or your cortisol might even be low.

47:40Perry Wilson:So you don't need to detox your cortisol. Not that we have any, like people will sell you a lot of stuff, gummies, drinks.

47:48Emily Oster:Well, yeah. I mean, I did look into it in case people want to make their cortisol. You know what? There are some things that'll make your cortisol lower that I think are good for you anyway. So there's a pretty large meta-analysis of randomized controlled trials. You had talked about, you know, randomizing people to interventions to lower cortisol, which of course have off-target effects. This is, once again, psychoneuroendocrinology, 2024, 58 randomized trials, just about 3 ,500 patients looking at different ways to reduce cortisol levels. And mindfulness meditation, relaxation techniques, both reduce cortisol levels by about a third or so, not bad.

48:29Emily Oster:Interestingly, the studies of yoga and Tai chi didn't have a significant effect. And the talk therapies like cognitive behavioral therapy and counseling did not have a significant effect. So maybe mindfulness meditation. Yeah.

48:41Perry Wilson:Exercise is good for your cortisol, lowers your cortisol. We also see that in randomized control trials. Yep.

48:48Emily Oster:Absolutely. Exercise. And we've hit on this several times, but sleep, you guys, especially if you want to maintain that diurnal variation, which does appear to be an important thing for your overall health. Getting adequate amounts of sleep gets the cortisol into the right place at the right time.

49:06Perry Wilson:So is it worth, I think an interesting question here, is it worth measuring your cortisol at all? So we just gave, you know, there are, these are recommendations to, that will reduce your cortisol, but also things which you should probably be doing for general, like health maintenance. Regardless, would you recommend someone measure their cortisol?

49:26Emily Oster:If they had moon facies, a buffalo hump, or a violaceous striae, I would recommend that they have cortisol levels measured. Do you want to know what my cortisol level is?

49:36Perry Wilson:Because I looked it up while we were here because I had it measured in the last time I had a blood test.

49:41Emily Oster:What time of day was it?

49:43Perry Wilson:It was in the morning.

49:46Emily Oster:It was 15 micrograms per deciliter.

49:49Perry Wilson:19. It only got in the sufficient range for my...

49:56Emily Oster:We should say, I can't believe... Yeah, sufficient. You know you die without enough cortisol, right? The flip side of this is that your adrenal glands can stop producing cortisol. This is called Addison's disease, which JFK had. And it causes severe hypotension, sodium wasting, death, you need to supplement. That's bad too. So I think cortisol is just like not something if you're interested in your health and wellness, it's just not worth your time. It's like, don't, you're not going to change your life based on it. And yes, you should de-stress, but not because of cortisol.

50:31Perry Wilson:Look, I think every time we talk about hormones, I am more and more impressed with the endocrine system. I feel like the, you know, the human endocrine system is so like tight and it's so well thought out. And there's so much of the wellness space that's about like, give yourself more of this hormone and more of that hormone, do this. And most of the time, like your endocrine system, like it's thinking about that. Like it thought about when the cortisol should be, when it should be low, like how much you need. It's like estrogen, like we're all kind of moving, moving together. And I just, I don't think it's a system we want to be messing with too much.

51:09Emily Oster:With the exception of hormone replacement therapy and

51:11Perry Wilson:with the exception of, yeah, no, yeah. Okay. There are some things, but like in, like our tendency to want to get inside the endocrine system and mess around with it, I think is over. I think we're doing too much of that outside of, I agree, hormone replacement therapy.

51:27Emily Oster:Yeah, yeah. Cortisol too high is bad. Cortisol too low is bad. It needs to have a specific pattern. You're not going to get this right with like ashwagandha, which does lower your cortisol a little bit, by the way. But like this is one, trust your body a little bit and be zen.

51:41Perry Wilson:Trust your body, but don't trust Mendelian randomization because it's stupid.

51:48Emily Oster:It can be done very well.

51:50Perry Wilson:It can be, but it's not. Okay, enough of that from me. All right, Perry, smash or pass, cortisol?

51:57Emily Oster:I mean, I'm a pass in terms of cortisol as a wellness metric. Obviously, I love the hormone. I die without it. So do you. But I'm a pass. Emily, smash or pass?

52:07Perry Wilson:I am a pass on anything that involves detoxing, retoxing, changing gummies, adapt to this and that. Just trust your endocrine system unless you have a buffalo hump and then get it looked at. All right.

52:23Emily Oster:That's it for Cortisol, your mailbag question of the week after the break.

52:31Perry Wilson:Hi, Emily and Perry. This is Celeste in Detroit. My question is more personal to both of you. Is there anything that you felt certain about like 10 years ago that now you look back on and cringe? change? We all have them and I'm dying to know yours. Thanks. Oh, this is such a good question.

52:50Emily Oster:I'm sure there are many things that I've suppressed and I have convinced myself that I never believed that really I did. But the one I am quite sure I believed strongly and I think I'm going to just cheat a little bit and say this was more like 15 years ago, maybe even pushing 20 years ago. But I was a hardcore fluid resuscitation vigilante when I was in medical training. So when someone came into the ICU with sepsis, it was like a contest. How much IV fluid can I put into this person in the next 12 hours? And back then, we thought this was the right thing to do. This totally makes me cringe now.

53:31Emily Oster:There's been a lot of evidence in the past 20 years that suggests that overly aggressive fluid resuscitation can cause major problems down the road and people get volume overloaded and their lungs don't work anymore and stuff like that. But it used to be a contest among the residents. I gave that guy 15 liters last night and high five. Yeah, that's how you treat sepsis. And I cringe a little bit thinking about that now. How about you, Em?

53:56Perry Wilson:That's amazing. Mine is more personal. It actually vaguely relates to cortisol. So I used to be very sure that you should exercise in the morning before you eat and that you should like not eat anything before you should like exercise fasted. But it turns out that's really stupid. And it's really bad for your cortisol, I think. It raises your cortisol a lot to exercise fasted. And I don't do that anymore. And I think it was part of this space of like, you know, like diet culture that I find now very cringy.

54:35Emily Oster:It's okay. We're all entitled to change.

54:37Perry Wilson:I can't believe you guys were high-fiving. That's like, that's so nerdy and weird. Oh, yeah. It makes me, yeah. Okay. I'm not going to comment. I have feelings. I'm not going to comment. Thanks. All right, that's it for us today. Stick with us next week when we'll ask, what's the deal with pregnancy brain? And we'll have a special guest. Wellness Actually is produced in association with iHeart Media. Our senior producer is Tamar Avishai. Our executive producer at iHeart is Jennifer Bassett. Our theme music is by Eric Deutsch. And our content is for educational purposes only.

55:17Emily Oster:If you like the show, help other people find us. Leave a rating and review on Apple Podcasts or your podcatcher of choice. And help us spread the word about the show. You can follow us on Instagram at wellnessactuallypod. And don't forget, we want to hear from you. Head over to wellnessactually.fm and leave us a question for our mailbag or suggest a topic for a future show.

55:39Perry Wilson:We'll let the influencers have the last word. When your cortisol levels are elevated, you can hang on to a lot of fat around your middle. And this can be really frustrating because it seems like you're doing everything right, but it's not budging. The first thing I want you to do is get rid of the high intensity interval training. The second thing is intermittent fasting. Get rid of it. Eat 30 grams of protein within 30 minutes of waking up. And the third one is the least fun. And that is to stop drinking coffee first thing in the morning on an empty stomach. This third one might piss you off.

56:19Emily Oster:From Taco Night in Tulum to sushi in Tokyo, every bite is rewarding and post-worthy with Amex Gold's four-times membership rewards points at restaurants worldwide. Wherever you dine, points are piling up. So bring your friends along for your next course because it's not all about the posts. It's about the company and the memories. How can gold from Amex sweeten your next food moment? Learn more at americanexpress.com slash explore-gold. Terms and points cap apply.

56:47Perry Wilson:This is an iHeart Podcast. Guaranteed human.