The Peter Attia Drive - #167 - Gary Taubes： Bad science and challenging the conventional wisdom of obesity

00:00:00.000 Hey, everyone. Welcome to the drive podcast. I'm your host, Peter Atiyah. This podcast,

00:00:15.480 my website, and my weekly newsletter all focus on the goal of translating the science of longevity

00:00:19.800 into something accessible for everyone. Our goal is to provide the best content in health

00:00:24.600 and wellness, full stop. And we've assembled a great team of analysts to make this happen.

00:00:28.880 If you enjoy this podcast, we've created a membership program that brings you far more

00:00:33.280 in-depth content. If you want to take your knowledge of this space to the next level,

00:00:36.820 at the end of this episode, I'll explain what those benefits are. Or if you want to learn more

00:00:41.320 now, head over to peteratiyahmd.com forward slash subscribe. Now, without further delay,

00:00:47.740 here's today's episode. My guest this week is Gary Tobbs. Gary's an investigative science and health

00:00:54.460 journalist. He's the author of multiple books, including the case for keto, the case against

00:00:58.320 sugar, why we get fat and what to do about it, good calories, bad calories, bad science,

00:01:04.160 and Nobel dreams. Most of you who recognize Gary's name will undoubtedly recognize him

00:01:08.860 from his work in nutrition. And that's covered obviously in the most recent of his books.

00:01:13.940 However, we spend the first part of this discussion really talking about his background,

00:01:17.940 how he got into journalism as a physics major, and how he spent the first really decade and a half

00:01:25.020 of his career writing about science and how he was particularly drawn to bad science or what would

00:01:30.620 be called pathologic science. Gary almost stumbled into health sciences in the late 90s,

00:01:36.480 and really he's never looked back since. In this podcast, we probably spend about two-thirds of it

00:01:42.740 covering his career up until that point and about the last third getting into this. Gary provides a

00:01:49.140 great history of the overview of obesity research and provides his explanation for why the conventional

00:01:57.720 wisdom today is incorrect. Whether or not you come away from this episode believing that Gary is right

00:02:03.620 or wrong, I hope you do come away from this episode at least acknowledging that there perhaps remains

00:02:10.040 some uncertainty with respect to our understanding of obesity and its causes. So without further delay,

00:02:16.240 please enjoy my conversation with Gary Tobbs. Hey Gary, how you doing? I'm good, Peter. How are you?

00:02:26.740 I'm good as well. I feel like it's been a very long time. I don't know the last time I actually saw you.

00:02:31.420 It's been how many years? Probably three or four. San Francisco, we had Virta offices. Yeah,

00:02:37.880 God, that's maybe four or five years ago now. Who knows? We were certainly younger. Speak for

00:02:43.760 yourself. Come on. Oh yeah, that's true. You may be aging the other way. I forget.

00:02:50.740 There's so much to talk about. I want to start with some of the stuff that probably a lot of people

00:02:54.760 forget about you, which is long before you were writing about nutrition, you were writing about

00:02:59.240 other disciplines of science. And I'm curious how you got interested in this whole thing. So

00:03:03.620 remind me what you studied in college. You went to Harvard and did you study physics? Was that your

00:03:08.840 undergrad? Yeah, I was a physics major. I actually started off as astrophysics major and then got a

00:03:15.460 C- in quantum physics. And my advisor suggested I change my major. So we switched to applied physics and

00:03:23.580 I spent my senior year taking writing courses for the most part. But yeah. What drew you to physics in

00:03:29.680 the first place? Were you a good student in high school? Yeah, well, you got to be a good student in

00:03:33.360 high school to get into Harvard. I read science fiction books growing up. This was the 60s. You

00:03:37.680 know, we all wanted to be astronauts. Now our kids want to be billionaires. Back then they wanted to be

00:03:43.140 astronauts. So I read science fiction books. I studied astrophysics. I have an older brother, as you know,

00:03:49.680 who's very smart and was studying physics and I was competing with him. So that's what I did.

00:03:55.980 Problem is my brain couldn't wrap itself around the concept of Hamiltonian. It was beyond my ability

00:04:03.040 to comprehend. So that was the end of that. So the whole time you're studying physics,

00:04:09.360 you're thinking, I want to now apply this. And I think you went to Stanford thereafter to do a

00:04:15.200 master's in aerospace, if my memory serves me correctly. I did. And I was still thinking

00:04:19.980 astronaut, weirdly enough. At Stanford, I actually lived with a few naval pilots in the same housing,

00:04:28.340 you know, apartment building. And I realized, one, that there really wasn't any need for a six foot

00:04:33.380 two, 220 pound astronaut. When you could send up a five foot, 950 pound astronaut, it was in better

00:04:39.860 shape and more fuel efficient. And B, that I wouldn't survive very well. Any kind of military hierarchy

00:04:46.260 that required blind acknowledgement to superiors. So started aiming towards journalism from there.

00:04:54.720 Where was that seed planted of, call it insatiable skepticism and a refusal to bend to authority?

00:05:02.100 On one level, I once had a editor at Discover Magazine who asked me sarcastically or facetiously

00:05:09.320 whether my brother was as, quote, well-adjusted as I was. And my response was, I didn't have to grow up

00:05:15.120 in his shadow. My youngest son now is 12 years old and he's a basketball obsessed and he plays with

00:05:20.660 the local AAU basketball team. And I noticed that the best players are the ones who have older

00:05:25.340 brothers who play. They're the scrappiest or the toughest. They refuse to give up a rebound.

00:05:30.860 And I think because they've been playing with their older brothers their whole life, I mean,

00:05:34.860 the ones that can compete quit and go into another sport and the ones that can sort of refuse to give in.

00:05:41.420 And so on some level, that probably started my skepticism. But then when I was, you know,

00:05:47.680 so I go to journalism school, I get, I want to be an investigative journalist.

00:05:52.000 I had read All the President's Men by Woodward and Bernstein. I was fascinated by it.

00:05:57.180 The only job I could get out of journalism school was I had two semi-offers. One was in the Dallas

00:06:02.840 Morning News and in 1979, I wasn't willing to move to Dallas. The other was at CNN in Atlanta where they

00:06:12.880 didn't allow cigarette smoking in the newsroom and I was a smoker. So I stayed in New York and took a

00:06:18.820 job with Discover Magazine. And three or four years in, I did, maybe two or three years in, I did a piece,

00:06:25.980 reported an article on the Shroud of Turin.

00:06:29.080 So at the time, researchers from Los Alamos were taking this high-tech imaging equipment

00:06:34.120 to Turin to study the shroud. And they had concluded, so this is supposedly the burial

00:06:41.160 shroud of Christ that it appeared in the historical record right around the 11th or 12th century,

00:06:47.200 right when there's a big market in sort of fake religious artifacts. So there's a strong prior

00:06:54.220 belief that this is a fake religious artifact, but the Los Alamos researchers take their high-tech

00:06:59.260 imaging equipment to Turin and they say, and they get permission from the church to do it,

00:07:04.420 and they say they can't understand how this was made, even though it's also been carbon dated to

00:07:09.180 the 11th or 12th century. I remember a Sunday evening, I was living in my studio apartment in

00:07:14.560 New York. I was reading their reports that they had published, and I forget what it was that struck me

00:07:20.580 as profoundly unsupported by the data. But I actually called the researcher in Los Alamos at

00:07:27.500 home to say, how could you possibly interpret this from what you did? And it was just the step one in

00:07:35.400 realizing, and maybe because I had gone to Harvard and I'd gone to Stanford and I knew a lot of people

00:07:40.040 who went into science personally, there's nothing fundamentally different about somebody who goes

00:07:44.760 into science and somebody who goes into journalism, other than we have sort of different mechanisms of

00:07:49.860 wanting to understand what truth is. What was his response? I don't remember. I don't remember.

00:07:55.220 But if he certainly wasn't able to satisfactorily defend the interpretation, it was also an interesting

00:08:02.800 conclusion about the observation about negative results, which comes up all the time. When somebody

00:08:08.780 gets a negative result, when they're unable to determine how the shroud was created, does that mean that

00:08:14.900 the shroud was created by supernatural means that are beyond the ability of the equipment to detect it

00:08:20.880 doesn't mean that the equipment that they used are simply inadequate for doing the job. And in effect,

00:08:27.000 almost 40 years later, we're having the same debate online this week about the cause of obesity.

00:08:33.560 Are you doing the right experiments? Have you refuted the hypothesis? Or are the experiments flawed?

00:08:39.720 And that's always a fundamental issue in science. Did you know the harm of cigarette smoking when

00:08:45.600 you were a smoker? It's a good question. My mother died of lung cancer and she was a smoker,

00:08:49.940 but that was about 10 years after I started. And it was interesting. So I started smoking in 1978

00:08:55.920 when I went out to Stanford for graduate school. I was depressed. The sunshine in Palo Alto wasn't

00:09:01.700 helping you. You missed those dark Boston winters. You know, it's interesting. I mean, it's getting

00:09:10.220 deeply into my psychopathology here. I've suffered mild depressions my whole life, as you know.

00:09:16.240 You go from Harvard where I play football. I'm well-liked. I built a sort of four-year career at the

00:09:24.260 school. You think of yourself in a particular way, and then you go someplace completely new where nobody

00:09:28.860 knows you. I didn't handle that well. So I started smoking to, I don't know, maybe to spite the kids

00:09:35.220 who seemed so happy and complacent in the Stanford sunshine when I was definitely not happy and

00:09:41.120 complacent. And then it took me 20 years to quit. I don't know if children, and we discussed this when

00:09:46.880 we were doing NUSI, I don't know if kids ever think in terms of the things they do being dangerous.

00:09:52.000 We both did some very dangerous things when we were younger. We were both boxers.

00:09:55.900 I want to actually talk about your boxing history because the article you wrote in Playboy is such a

00:10:01.060 great one. But I don't know if I ever told you this story, but the entire time I was boxing, I never

00:10:05.820 thought it was dangerous. I had this true mental block that was, I don't get hit, so it's not dangerous.

00:10:13.400 Like, nothing's happening to me. I'm evading, and I'm hitting, and da-da-da-da-da. Which, of course,

00:10:18.700 was categorically untrue because my style of boxing was actually to get hit, tire out opponents, and

00:10:26.220 eventually overcome them due to superior conditioning. But it never seemed to hurt. And it

00:10:31.380 wasn't until I suffered a very, very severe concussion when I was 20. I still remember it

00:10:38.380 very well. I was actually hospitalized for two days, had a CT scan, had significant bruising,

00:10:43.260 had significant cerebral contusions, and had a headache that lasted for three months. This was

00:10:48.660 the hardest I've ever had my bell rung. And it was only at that point that I thought, oh my god,

00:10:54.440 you actually get hit in this sport. And by that point, I'd already decided I didn't want to be a

00:10:58.840 professional boxer. That was the irony of it. I was already in college, and I was now only boxing as

00:11:04.180 something fun to do. But it literally took that event for me to go, this sport is crazy. And from that

00:11:11.340 point on, I would look at it very differently. But while I was doing it, I never thought that.

00:11:16.080 I only got into it competitively after my best friend got killed boxing. Yeah, I mean,

00:11:20.780 there's a lot of things. I was a football player in high school and college. And when you're a football

00:11:24.100 player in high school and college, back then, getting your bell rung, it just was what it was.

00:11:28.720 I mean, I clearly had at least one concussion in high school football. My best friend in college,

00:11:33.560 who was actually a lot of kids' best friends, a very charming, wonderful young man of Puerto Rican

00:11:40.560 descent from New York, was an amateur boxer in New York, went back to boxing after his football

00:11:45.660 career came to an end due to knee injuries and got our senior year. He had his first fight and he got

00:11:50.960 killed in the ring. He was knocked out. His head hit the mat. We were in Lowell, Massachusetts,

00:11:55.820 watching the fight at the time. And, you know, he never woke up. My memory is of unplugging him a week

00:12:03.940 later. It did not stop me from then moving to New York when school was over, actually three years

00:12:09.920 later. And when I got to New York starting to box because I was bored. Did you think about him?

00:12:14.440 When I did it? Yeah, sure. But I never thought that what happened to him could happen to me until

00:12:18.780 I got knocked out in my, in the golden gloves. At which point there's this awareness when you get

00:12:25.560 knocked unconscious, when you wake up, that some people never wake up from that, that that might

00:12:30.000 have been the moment to which your candle was blown out in effect and there's no relighting it.

00:12:35.800 Talking about this is interesting because a lot of what we do in life and talk about,

00:12:39.560 when you're talking about prevention of chronic illnesses and prevention of disease and

00:12:43.900 prevention of addiction, it's getting young men and women who don't think and don't process risk

00:12:51.600 the same way we do as we get older. I always thought it was fascinating that the less life we

00:12:56.160 have less to live, the more risk adverse we have. If something bad happens, you're actually losing less

00:13:01.900 of your life than you are when you're younger. And I'm sure the behavioral psychologists like Kahneman

00:13:07.440 and Tversky could say things or do say things about that.

00:13:11.080 Tell me about the Playboy article that you wrote after that golden gloves tournament. I know you

00:13:15.780 still have a, at least the last time I was at your house, maybe an eight years ago, you actually had a

00:13:19.600 framed photo of you lying prostrate in the ring.

00:13:24.280 You know, I still do. It's in our coat entryway. I often thought of putting it up here instead of this

00:13:31.500 picture because I refer to it as my hubris protection. So while I was working at Discover

00:13:37.400 magazine and I was boxing, so this was 1983, 84.

00:13:41.500 By the way, not too long after Duck Koo Kim was killed by Ray Mancini. I mean, the early 80s was a

00:13:47.020 very difficult period in boxing. There were a number of high profile deaths in the ring.

00:13:51.960 Yeah, he was killed while I was boxing. And we actually did a story on it at Discover about what

00:13:57.460 happens to the brain. A good friend of mine, a New York Times reporter now named Denise Grady was 1.00

00:14:02.660 writing that story. And while I was boxing and getting ready to box in the golden gloves and she

00:14:07.240 would come and then slip like articles under my door about whatever we were calling the traumatic

00:14:14.040 event back then. Yeah, I was just a young man who was, I was boxing. I had a friend, a good friend who

00:14:20.080 was a Norman Maller's nephew. And then Norman Maller had a group of young men, not even that young,

00:14:27.200 who met at a gym that's no longer there on 14th Street in Manhattan back when that was a very

00:14:32.780 questionable part of town. And we would go down, I think it was every Saturday morning. We would

00:14:37.700 box for a few hours and then we would go out to lunch. And yeah, from that, I got the idea,

00:14:44.620 why don't I box in the golden gloves and write about it? So I originally was doing it for Sports

00:14:49.260 Illustrated, but then I realized that I was technically too old to do it, that I missed the cutoff by about

00:14:55.180 two or three months and that if I wanted to box, I would have to revise my birth certificate. And so

00:15:01.900 when I realized that, I talked to the editors at Sports Illustrated and I said, you folks, you do

00:15:06.120 investigative journalism. We were in the same building, Discover and Sports Illustrated were

00:15:10.100 both owned by Time Incorporated at the time. I told the editors, you do, you know, hard-hitting

00:15:15.760 investigative journalism. If I want to continue to do this article, I'm going to have to fake my age.

00:15:21.040 And I think that's a problem. And the editor has said, well, it is now that you told us. So we pulled

00:15:26.420 it from Sports Illustrated and one of the editors I worked with had a good friend at Playboy. So I pitched

00:15:31.580 it to Playboy and my sister-in-law got angry at me for writing for a magazine that doesn't portray women 1.00

00:15:38.160 in the most ideal light, depending on how you look at it. And she had a point, but I said some of the best

00:15:43.520 writers in America, write for Playboy. And maybe later I'll have the platform by which I can be too

00:15:50.600 good for this approach. Anyway, we did it. I won my first fight against a policeman from Staten Island

00:15:57.080 who beat the crap out of me in the first round. And then I realized that if I punched back, that might

00:16:02.900 help. And I knocked him out in the second round. And then the second fight was a week later and I got

00:16:08.780 knocked out in I think a minute and 37 seconds by the fellow who ended up winning our division,

00:16:14.540 who was not particularly good, but I had never really had enough time to work on defense. You

00:16:19.360 had been boxing for many years competitively. I had been doing this for about four months at any level

00:16:25.760 other than a Saturday morning hobby. But it was a reminder that anytime I think I'm, some of us grow up

00:16:31.400 with this kind of James Bond complex. I think it's fair to say between the watches and the fast cars

00:16:37.920 and the archery and things like that, you might have it also that we think we're so cool, we can

00:16:42.580 do anything. And that was my reminder that I'm certainly not. And that there are things I should

00:16:48.700 stay away from if I want to have a long and healthy life, among them being boxing. And I haven't gone

00:16:54.100 back and never went back into the ring after that. What's the process like of learning to write? I was

00:16:59.780 rereading an article yesterday that you had recently written. And don't get too flattered by this,

00:17:05.120 but I forgot how well you write. I think I write reasonably well. I've learned to write better

00:17:09.880 over the years. And I think a lot of that came from how often I used to write, how often you used

00:17:15.420 to rip what I used to write apart. Back when we were at NUSI, I was doing so much of the writing

00:17:19.600 and you were doing so much of the editing and revising of internal stuff that I would write.

00:17:24.540 And that process, of course, I think made me much better. And I can now see it when I look at the

00:17:28.600 writing of other people. But your writing today, I think is really excellent. At least it's a style that I find

00:17:34.100 really good. You can write about science in a way that's very readable. How does one learn that

00:17:39.060 craft? When you show up at Columbia, which ostensibly is the best journalism school in the

00:17:43.780 country, I assume you're coming in incredibly raw. Did you believe you had some talent, some hidden

00:17:49.860 talent for writing? Or is that not even a part of what they're selecting at a top journalism school?

00:17:55.760 Is it more the thought process they're selecting for and the writing is the easy thing to teach on top?

00:18:00.740 Well, I thought I could write because I used to write letters to friends and girlfriends that I

00:18:07.080 thought were clever and they liked. So there was always a sense that somewhere inside me, I was a

00:18:11.520 writer. When I first started taking writing classes, it was actually at night school at Harvard. After I

00:18:17.180 got my master's degree from Stanford, I went back to Boston and I don't even remember what I was doing

00:18:22.080 for a living. But over the course of a semester, I took two writing courses, one taught by a science

00:18:28.280 writer from the Boston Herald, who I later worked with at Discover Magazine, and one taught by a

00:18:34.500 fiction editor from the Atlantic named Michael Curtis, who had been a roommate of Thomas Pinchon

00:18:40.300 and Dick Farina at Cornell in the early 1960s and was already a famous fiction editor at the Atlantic.

00:18:46.620 And the very first article I wrote for him could probably remember the subject if I had to. But what I do

00:18:52.600 remember is it came back with the words puerile written in one inch high letters on the front

00:18:58.180 underlined twice. And I had to look up puerile, which meant childish and amateurish and thoughtless.

00:19:05.860 Over the course of a semester, Michael Curtis started to hammer me into being a functional writer by

00:19:13.820 constant and relentless criticism of what I did. And I appreciate it. In fact, the Boston Herald

00:19:21.100 science editor was so nice. He would read people's work and he believed you should never say anything

00:19:26.900 negative about their work. And I gained absolutely nothing from being in his class. He was a lovely

00:19:31.800 man, but I don't think anyone else did either, except maybe an unjustified sense of their own talent.

00:19:38.940 Curtis hammered into me. And then when I got to journalism school, it was just about basically

00:19:42.640 teaching you to write in a report. And so it's a steady learning process. And then as a journalist,

00:19:47.660 when I started at Discover, I started as a reporter. So Time Inc. back then had reporters who did the

00:19:52.840 information gathering and then we would write files and the files would be given to the writers and the

00:19:57.380 writers were true craftsmen. And I started off, I don't know, the first story I remember doing was

00:20:03.260 about some type of particle accelerator at Michigan State. I had the writers there, friends who were

00:20:10.960 writers would read this draft and then relentlessly critique it until I ended up with something that was worth

00:20:16.860 submitting to the editors. And then the editors would do the same thing before it would be published.

00:20:21.640 And as I became a writer and got older, I did that same thing for other reporters who came in under me.

00:20:26.820 And there are a few very well-respected journalists who I think benefited from me being relentlessly

00:20:32.600 critical. And then the actual process of writing is just, you know, I told you this years ago,

00:20:37.680 Calvin Trillin, the famous New Yorker writer, had a phrase he called the first draft the vomit out,

00:20:42.600 where you just get everything down on paper and it kind of relieves the pressure of a first draft.

00:20:48.680 So you get everything down on paper, the computer, and then you just keep rewriting it and rewriting

00:20:53.840 it and rewriting it until it reads like something I would have wanted to read if it hadn't been written

00:20:58.800 by me. And because I read a lot and I read a lot of very talented authors, I know what good writing

00:21:05.160 looks like and feels like. And you just continue to revise and edit until it gets there.

00:21:10.340 Who's the best living science writer, in your opinion?

00:21:14.340 I don't think I could answer that question. There are a lot of people who have a lot of

00:21:19.120 different skills. It's like saying, who's the best living football player or the best? My

00:21:23.180 12-year-old is always asking me who I think the five best basketball players, you know,

00:21:28.220 was Will Chamberlain better than Steph Curry? Well, they were different. There are science

00:21:35.640 journalists. There's a young man at the Atlantic now whose name I'm going to forget and read somebody

00:21:43.480 who writes a thoughtful, well-reported article on a different subject every three or four days.

00:21:51.280 And I often read his stuff and I think, could I have done that when I was his age and, you know,

00:21:57.500 worked 70, 80 hours a week? Possibly, but I don't think so. And I don't think I could have done it

00:22:03.260 as well. When I was growing up, the best out there was Jim Glick, who wrote, remember, Chaos

00:22:08.260 and the Feynman biography, the name of which I forget at the moment. There were a whole series of science

00:22:15.240 journalists, a whole collection of science journalists who came of age in the 1980s when

00:22:20.280 science magazines had first started. There was a magazine called Science, well, Science 80,

00:22:25.620 then it was Science 81, Science 82, there was Discover. Some of those like Charles Mann and Steve

00:22:32.660 Hall, John Tierney. Chaos in the Blood. Chaos in the Blood, yeah. Just exquisitely beautiful,

00:22:39.960 thoughtful writers. In fact, I can't, Charles Mann, he's a friend and Steve's a friend and I think I

00:22:47.380 call, Charles Goes by Cam and I can't read Cam's books because they're so painfully, the writing is so

00:22:52.960 painfully beautiful that I feel inadequate when I read them. It's very difficult. The same goes for

00:22:58.120 Steve. I think what they do, they do much better than I do, but we're now the senior figures in the

00:23:03.140 field. What is it about their work though? You once told me that great science writing isn't really

00:23:07.840 about the science. What is it about Steve, for example, that you read and think he's doing that

00:23:12.880 so well? It's a combination of comprehensive, rigorous reporting. So you know you have the story,

00:23:18.680 although you can never really judge if somebody else has a story right, unless they're writing

00:23:22.160 about your field. The prose, style, the ability to tell a story around a complicated subject.

00:23:30.080 Yeah, I was going to say, I just read The Ghost Map by Stephen Levy. Have you ever read that?

00:23:34.120 No. It's an extraordinary book. I almost, you know, I hope that's the best book he ever wrote

00:23:40.080 because if he's written better books, then I'm in the wrong field. And I know many people say that

00:23:47.020 anyway, but on one level, just the prose. So the ability to write a sentence and a paragraph that

00:23:52.640 other people want to read, but then the story that you tell and the digressions to the story and

00:23:57.080 keeping it always interesting and making it both bigger than it other people might see and yet

00:24:02.680 simultaneously getting the details of the lives and the times. And I'm sure every genre has its own

00:24:09.620 challenges, but science has this challenge that you're inherently writing about a subject that's very

00:24:14.660 complex. It's hard for even the people who are scientists themselves to wrap their heads around.

00:24:20.140 So I'm still going to have to think about who the best science, who I think the best science writer

00:24:24.800 is today, if there is one. And there's so many out there today. The question you always ask is what

00:24:30.280 books out there, like The Ghost Map. And again, on one level, it's about the John Snow and the cholera

00:24:36.860 epidemic in England in whatever, 1864 in London, and how John Snow and his compatriots figured out

00:24:44.520 that this was cholera was not bad air, but an infectious agent in the water supply. But then

00:24:51.340 it's about the birth of modern cities, about urbanization, about the problems that cities

00:24:57.060 faced coming together, about cities into the future, and then fundamentally about the scientific

00:25:03.820 process, the challenge of an outsider doing science, which might be one reason why it resonated so much

00:25:10.360 with me. You start with a small story, and you turn it into something large and intricate and

00:25:15.460 meaningful, extraordinary book. So at what point during your time at Discover did you become interested

00:25:21.420 with the subject matter that would eventually become your first book? Yeah, my background was in

00:25:25.720 physics, so it was natural for me to write about physics. And as Harvard physicist named Carlo Rubia,

00:25:30.900 working at the European Center for Nuclear Research, CERN in Geneva, had his research

00:25:37.820 collaboration of 150 people. A physicist known as UA1 had discovered two fundamental particles that were

00:25:46.140 sort of the two of the three last remaining, well, two of the four last remaining particles of the standard

00:25:52.640 model of physics. So physics, the electricity and magnetism had been united with the weak force, the

00:26:00.040 electroweak theory, and that was a great step forward in the 1960s and 70s. And by 1973, that theory had

00:26:08.620 mostly been confirmed, and it predicted the existence of two particles that transmit, or particles that

00:26:16.420 carry the electroweak force. This was the WNZ boson. And Rubia's group, Rubia, who was this very

00:26:22.720 controversial physicist who taught at Harvard, had convinced the Europe to, in effect, spend the money

00:26:30.000 to build an accelerator that had the power to detect the WNZ particles. And when they announced the

00:26:36.880 discovery, we did an article on it at Discover. I forget who, I must have written that. But then

00:26:42.720 Discover was part of Time magazine, and Time has their Man of the Year cover, right? So Discover had this

00:26:49.200 concept of the scientist of the year. And because of this great discovery, and everyone knew that Rubia

00:26:55.120 was going to get the Nobel Prize for it, Discover made Rubia the scientist of the year. So I get to

00:27:01.760 fly to Geneva. Actually, I flew to Belgium, rented a BMW, drove through Paris, went to Geneva. Time had a

00:27:09.140 very good expense account back then. My life has been downhill ever since. Anyway, I spent time with

00:27:14.820 Rubia in Geneva. I interviewed his friends. I wrote him up as a profile for the scientist of the year.

00:27:20.580 Time Inc. flew him to New York at the Explorers Club. He was given the award.

00:27:25.340 And this was before Stockholm or after Stockholm?

00:27:28.060 This was before Stockholm.

00:27:29.640 Okay. He was awarded the Nobel Prize wedding. Like, it's always the fall. So was it the fall of 84?

00:27:34.280 85, yeah.

00:27:35.740 It was 85. Okay. They published the two bosons in like January or February of 83. Is that right? Or was it

00:27:42.360 earlier, 82? Must have been probably 83. Okay.

00:27:46.240 So anyway, the point is, not only does Rubia get this award, he gets a nice gold Rolex from the

00:27:50.580 Explorers Club. So he's very grateful to Gary Taubes for, he thinks somehow, and I probably was

00:27:57.560 the one who suggested him for scientist of the year. So that April, which would be April of 84,

00:28:03.300 he comes to Washington. There's the big physics meeting every year is in Washington. And Rubia gives

00:28:08.120 a presentation announcing that he's now on to the next step, which is physics beyond the standard

00:28:14.060 model. And he shows what they've got, and he shows a few events, and we could talk about that.

00:28:19.260 And the message is that all he has, it's the biggest discovery in physics in 40 years. And all he has to

00:28:25.820 do is turn on the accelerator, which will happen that autumn, and gather more data and nail it down.

00:28:31.680 And it's Nobel Prize number two. And so I was at that talk. So I came up to him afterwards. I said,

00:28:37.820 look, it's, you know, this is terrifically exciting. And it's rare that anyone predicts a

00:28:41.780 great discovery in advance. So can I write a book about this? Can I maybe I can come to CERN,

00:28:48.080 live with you guys and document this as it happens. And Rubia liked me, he thought of me as his

00:28:55.040 personal scribe. So he said yes. And I put together, well, first, I put together a proposal

00:29:01.200 for the Atlantic that they rejected because Charles Mann and his co-author at the time,

00:29:07.220 Bob Kreese, had just put their own proposal in to do a piece on high energy physics. So when I was

00:29:13.800 always following behind Mann and Kreese, if you're going to follow behind two writers,

00:29:18.180 they're two very good writers to lose out to. Anyway, so I pitched it as a book deal instead.

00:29:23.720 Actually, I was approached by a publisher who was publishing a book by Shelley Glashow,

00:29:30.220 who's a Nobel laureate at Harvard, who I knew when they had asked Shelley who the good science

00:29:34.720 writers were. And Shelley said, you should get Taubes. He's probably the best. So they approached

00:29:41.080 me. I had this proposal from the Atlantic that they rejected. I sent it to the publisher and they

00:29:46.760 said they'd be willing to pay me to do the book. So I left Discover Magazine and flew to Geneva

00:29:52.460 and got a room at the hostel on the lab and was, today we'd say I was embedded with the physicist

00:29:57.400 for the next nine months. And thought I was going to write about a great discovery and pretty quickly

00:30:02.320 realized that this was far less compelling. The evidence were far less compelling than Rubia had

00:30:08.980 claimed in Washington and that there was a very good story here and it was probably not the story

00:30:13.620 of a great discovery. You know, when you ask what informed my approach to science and science

00:30:19.860 journalism, that was the learning experience. So there's 150 physicists on the experiment.

00:30:26.480 So the gist of it is you've got, well, an atom smasher, a collider. So you're colliding

00:30:31.580 subatomic particles together at speeds as close as you can get them to the speed of light.

00:30:37.600 The faster the particles go, the greater the energy in the collisions. And then you're looking in the

00:30:42.480 collisions for the signs. For particles that might be made that your theory of physics,

00:30:48.160 the standard model does not predict. So in that sense, you're testing the standard model.

00:30:53.560 And as the accelerators are able to put more and more energy into these collisions, you're able to

00:30:58.820 test your predictions of the standard model further and further out. And then you've got to

00:31:03.120 understand the predictions, which is part of what we're dealing with here. But so you've got this

00:31:06.900 huge particle accelerator. If I remember correctly, it was maybe four miles in circumference. It straddles

00:31:11.960 the French-Swiss border. At points on the accelerator where you're colliding the two beams

00:31:18.860 together, you build a detector. And a detector, these things are the size of a mansion, three,

00:31:24.620 four stories high. That's where the money, that's, you know, they're as expensive. So the accelerator

00:31:30.360 itself might cost $50 million or $100 million. Today they cost billions. Back then it was hundreds

00:31:37.240 of millions. And the detector, which I always thought of as kind of a camera that's created

00:31:42.860 to photograph, to be able to detect the passage or the emittance of all the various particles that you

00:31:51.900 will come out of these collisions. So you've got photon detectors to detect photons that come out.

00:31:58.880 You've got muon detectors to detect muons that come out. And so you collide it. And of the billions

00:32:05.040 of collisions you're making, I don't know, every day, you're looking for one collision that creates

00:32:12.240 one particle that your standard model doesn't predict. And just to put that in context, as you

00:32:19.320 said, you could have billions of collisions per day. But I think if I recall, I mean, Rubia would say,

00:32:24.460 look, you might get 20 a year that are relevant. And so you just think about the signal to noise ratio

00:32:30.720 there. And again, I think just for folks to understand, you're trying to look at, so you'd

00:32:35.380 have a prediction of the mass effect of that collision. So you have to have some conservation

00:32:39.720 of mass. You have to have some conservation of charge. And you're trying to sort of coalesce all

00:32:45.720 of these different things to look for discrepancies that are occurring at almost an infinitesimal fraction

00:32:52.880 of the number of collisions. It's very hard to wrap your mind around.

00:32:56.420 That are predicted to be occurring or not occurring.

00:32:59.960 That's right. They may not even occur. Yeah.

00:33:02.680 You've got the particles colliding. Everything that comes out of that collision, you have to

00:33:06.720 detect. Because if you miss something, it's beginning to sound a lot like calorimetry,

00:33:11.780 which is a subject we should get to later. If you miss something, then that collision is worthless to

00:33:18.140 you. And you're likely to be fooled because you're likely to think some particle was created that wasn't

00:33:22.660 created because, you know, you do have to have a conservation of energy. So if a certain amount

00:33:28.020 of energy goes off in one direction, you better get the same amount of energy going off in the

00:33:32.160 other direction. And ideally, with every particle you detect, and you detect the energy in the

00:33:37.360 particles, that's part of what the detector does, and you recreate this collision, you know that

00:33:42.660 everything's balanced. If it's not balanced, does that mean that your detector failed? Or that a

00:33:47.360 particle was created that your detector couldn't see because it's beyond the standard model?

00:33:51.920 Now you have to know your detector better than parts per billion issues because you have to be

00:33:59.240 able to estimate whether or not your detector failed. So on some level, this idea that science

00:34:04.340 is about signal-to-noise, you have to be able to predict the noise almost perfectly, to understand

00:34:10.140 the noise almost perfectly. And this is a sort of repeating theme in everything I've written about,

00:34:15.340 even if I don't phrase it like that. In NUSI, we used to talk about signal-to-noise problems all

00:34:19.900 the time. If you don't understand your background, which in high energy physics is the standard model

00:34:26.920 predictions of what you would expect to see if there's nothing new out there, and what you would

00:34:33.520 expect to see from the understood flaws of the equipment. So that's the issue, the problem we're

00:34:40.620 confronting. So you've got parts per billion effects that you're looking for, or parts per trillion

00:34:45.300 effects. And you have to understand the flaws in your detector to better than that. And so as,

00:34:51.400 what's his name, Rumsfeld would have said, you've got your known unknowns and your unknown unknowns,

00:34:55.620 and you're constantly trying to figure out your unknown unknowns, and you never fully figure out

00:35:00.620 your unknown unknowns. So physicists, we read about this, there's a talk recently of a new particle

00:35:08.380 possibly being discovered. And physicists say, well, they haven't discovered it yet because it's only a

00:35:13.380 three-sigma effect. And a three-sigma effect is what, you know this better than I do, a 0.000...

00:35:20.080 Probably 0.00015 or something like that. But our threshold would be six-sigma to really say...

00:35:28.320 So the reason for the six-sigma, or actually five-sigma, is because you want to make a lot of

00:35:34.480 room for the unknown unknowns. You want to say, this is how I perceive it. We can calculate that the

00:35:39.740 sigma is a calculation of the... Standard deviation, yeah, the difference from the mean.

00:35:44.420 Yeah, but there's going to be stuff you don't know. There's going to be stuff that's going to

00:35:48.840 fool you that you don't know. So you want to have an event that's so absolutely, fundamentally

00:35:53.880 frigging outside the range of what you think you know, that you can confidently assume that what

00:36:00.200 you don't know, the unknown unknowns, can't explain it either, and it's new physics.

00:36:04.420 Before you go further, there's one thing I want you to explain to listeners, because it's a beautiful

00:36:10.480 part of physics, but it doesn't exist in every other field of science, which is the relationship

00:36:15.300 between the theoretical and the experimental physicists. I think it's worth pointing out here

00:36:20.880 how that is a collaborative nature in physics.

00:36:24.020 Yeah, physicists divides up into theory and experiment, okay? It wasn't always quite so clear,

00:36:29.780 but even if you think of, you know, Einstein was a theorist, Eddington, who did the seminal test of

00:36:36.580 the theory of relativity, was an experimentalist. So as physics evolves, you have people who wrestle

00:36:42.640 with the theory, and then they have to come up with a theory that makes predictions that can be

00:36:46.460 experimentally tested, and then you have experts who are experimentalists, whose job is to do the

00:36:51.780 experiment, and then there's got to be crosstalk between the theorists and the experimentalists,

00:36:56.220 because the experimentalists have to know precisely what they're looking for, and they need the help

00:37:00.420 of the theorists to tell them how this might manifest itself. And the experimentalists have

00:37:06.080 to understand their detectors, their experiments, to this sort of superhuman ability. That's one of

00:37:12.100 the defining characteristics of an experimentalist, is the awareness of an exquisite physicist, is the

00:37:19.300 awareness of how their equipment could fool them. And so this is how physics has always developed,

00:37:23.380 and then you get to the point that theorists' job is to come up with a good theory and make a prediction,

00:37:29.580 and the experimentalists' job is to test that. In fields in biology and medicine, you don't have that

00:37:34.680 diversion, which I think is a problem, and we talked about this through our years. You've got the theorists

00:37:40.260 and the experimentalists are merged into one person. Everyone's a theorist. Everyone should be capable of doing

00:37:45.880 the experiment, and I think they're two entirely different skill sets. And they both require such a

00:37:53.120 fine level of detail and thinking and rigor that I'm not sure that, you know, after a certain point,

00:38:02.300 like, say, from 1900 onward, I don't think one person could encompass both aspects of, you know,

00:38:09.220 they're almost different endeavors. So to get back to CERN, because this is where it's relevant,

00:38:14.660 when I get to CERN, I move into the hostel, I'm hanging out with the physicists. Once Rubia published

00:38:20.740 the W and Z discoveries, and everyone knew this was the place to be in physics, even before that,

00:38:25.740 he collected a group of very ambitious young, mostly men, physicists from some of the best physics schools

00:38:32.320 in the world, who came to work with him to analyze the data.

00:38:36.200 Has UA1 and UA2 collapsed at this point?

00:38:38.840 It's always both. You always need both detectors, because if you see something at one, you still need

00:38:43.560 independent replication. Yeah. So if you don't see it in both, you got a problem. That's why even now,

00:38:48.860 the Large Hadron Collider at CERN has four, I think, four or five different detectors.

00:38:55.700 It's not enough to see something in one detector, you always need independent replication. And again,

00:39:00.980 that's true of every science. So you have this young men who are doing the analysis, and they're

00:39:05.200 sort of the, they've become, they've moved in, they've come in late, they've become sort of the

00:39:09.980 chosen quartiers of Rubia. They hang with him, they eat with him, they meet with him,

00:39:15.940 they tell him what he wants to hear, which is what's happening with the data. And then you've

00:39:20.400 got the physicists who actually built the damn detector, and the technicians who built the detector. 0.85

00:39:26.280 And these are, you know, these, I used to think of them as they'd come from, they don't come from 0.94

00:39:30.400 Harvard and Stanford and Oxford and Cambridge. They work at like the red brick universities in the 1.00

00:39:35.960 Midland and the UK, and they've been there from the beginning. They built this mansion-sized piece

00:39:41.940 of $30 million equipment with, by, with their hands. There's no mass produced. They know how it

00:39:49.740 can fool them, because they built it. So as soon as I get there, I've got these young physicists who

00:39:54.200 are new to the experiment, who are all excited about this great discovery. And I'm also hanging with

00:39:59.260 these, the old hands who built the equipment by hand, and they're basically poo-pooing everything and 0.53

00:40:04.320 saying, this is crazy. We cannot understand the flaws in our detectors well enough to make the

00:40:09.040 kind of claims that Rubia has been making. Now, are they voicing that? Well, they're trying to,

00:40:14.660 but the problem is you need a group leader who's willing to hear that. So once Rubia started talking

00:40:20.840 that he had made a discovery in Washington in 1984, what he wants to hear, he's gone out on a limb,

00:40:27.320 right? I thought he only predicted it. I didn't think he actually said it had happened.

00:40:31.840 You know, you've probably read my book more recently than I have. Even if you're claiming

00:40:37.700 that it will happen, if you think of, there's this term that the physicists use, pathological

00:40:42.540 science, which is the science of things that aren't so. And my whole career on some level has

00:40:48.400 been writing this. Right or wrong now, I think I understand pathological science, and I've studied

00:40:53.320 it as much as any human being alive. One of the defining characteristics of pathological science is

00:40:58.620 people commit themselves publicly to a result based on premature evidence. It's not ironclad.

00:41:07.320 They haven't locked it down. There's still a chance they could be wrong, and they don't understand

00:41:11.480 the likelihood of that chance. No one ever does until you get very good at this when you realize

00:41:16.580 that chance is enormous. Science is supposed to be hypothesis and tests, right? The idea is you're

00:41:21.620 supposed to rigorously test your hypotheses, and ideally you're trying to prove that you're wrong.

00:41:26.760 Richard Feynman, first principle of science, said you must not fool yourself, and you're the easiest

00:41:34.000 person to fool. So you get evidence that you've discovered something new. Your assumption should

00:41:38.940 be that you're wrong, that somehow your equipment is fooling yourself, and now you try and go out to

00:41:43.880 demonstrate that's true, and you do everything you can to prove it. And if you fail to find out how

00:41:49.540 you're wrong, now you write a paper that says evidence for the observation of particle X question

00:41:56.620 mark, and you put together a presentation, and you go around, you start in your department at whatever

00:42:02.240 university you're at, and you give the presentation to your colleagues, and basically you're asking them 0.91

00:42:06.840 to explain to you how you fucked up, if you'll pardon the language, because surely you did. And if they 0.93

00:42:12.300 can't explain it to you, you march around the country of the world giving this symposium, you still

00:42:16.960 haven't published the paper, the evidence for the observation of particle X question mark, because

00:42:22.100 you're still working under the assumption that you screwed up, and you just don't understand either

00:42:25.900 your equipment. And as you're traveling around the country, people should be saying to you, well,

00:42:29.740 what about this? Did you think of this? Did you know that the, you know, the muon detectors, actually

00:42:34.400 there's a glitch in the detectors that are made in Belgium versus the ones made in Switzerland, and the

00:42:39.400 ones made in Belgium are going to spark once every seven and a half months. Have you checked that?

00:42:44.760 And if you checked it, great. If you didn't, you go back to CERN or wherever you are, and you check it.

00:42:49.420 And so you keep giving this lecture. Finally, at the end of, I don't know, 20 seminars, nobody can

00:42:54.720 explain to you how you fucked up. Now you publish the paper with the question mark. And what you're 0.97

00:42:58.680 asking them to do is tell you how you screwed up, because if you assume that you did, and then they

00:43:04.080 tell you, you don't resist it. But if you claim that you didn't, and you publish the paper first,

00:43:09.480 now what you're doing, instead of that exercise of trying to find out how you assuredly screwed up,

00:43:14.260 you're trying to collect evidence for how you got the right answer. So you go from paying attention

00:43:18.680 to the negative evidence to paying attention to the positive evidence. Again, everyone does this.

00:43:23.420 I'm accused of that all the time. It's like a study comes out that confirms what I believe. I don't

00:43:28.540 question it the same way I will question a study that comes out that appears to challenge it. So once

00:43:35.040 Rubia had gotten to the point where he was predicting a discovery, all he wanted to hear from his

00:43:42.440 physicists was the evidence supporting that. That's the first mistake along the way. So the

00:43:48.540 physicists who wanted to tell him about the problems with the detector and all the ways

00:43:52.840 the detector could fool them and what they weren't thinking about and what they were missing, if those

00:43:56.720 people did their job, they were sort of seen as downers. So after a while, what happens is you get

00:44:02.300 the sort of polarization in the experiment where the principal investigator tends to selectively

00:44:07.600 listen to the people who support what he wants to believe and selectively resist, subtly resist the

00:44:14.820 people who argue against it. And then as the experiment goes on, the more the people who tell

00:44:20.480 him what they want to believe confirm his beliefs, then that confirms his belief that he should not

00:44:26.000 listen to the critics, the skeptics. And this happens in all of science is what we've seen in an extreme

00:44:31.440 way in COVID. It goes on in climate change. It goes on in nutrition and chronic disease. Once you

00:44:38.240 decide you know what the truth is, you tend to stop listening to the people who disagree with you,

00:44:44.040 and then you surround yourself with the people. This is also classic groupthink theory.

00:44:48.860 At what point did Rubia realize you were no longer his scribe in residence during this 10-month period?

00:44:55.280 Probably when he read a draft of the book.

00:44:57.200 You know, Carlo was interesting because he had enormous self-confidence. So you have to picture

00:45:03.380 he's larger than life. He's Italian. He's overweight. He's got this exuberance. 1.00

00:45:08.400 He's a Nobel laureate.

00:45:09.520 He's a Nobel laureate. I went to Stockholm with him. You know, I was fitted for my tuxedo with Carlo. I

00:45:15.640 mean, that's how close we were. Once I started realizing there were serious issues with this

00:45:21.560 experiment, I would challenge him on this. So I didn't hide my skepticism. I would say,

00:45:27.780 well, you're saying this, but so-and-so from Edinburgh or Collège de France or University of

00:45:33.540 Pisa is telling me this. I wouldn't out them like that, but this is what I'm hearing from your

00:45:38.180 experimenters. And Carlo, my take was that he was something of a pathological liar. So he defined 0.98

00:45:45.060 truth as what was convenient for him that he could get people to believe. So I would challenge him on an

00:45:50.660 issue and say, well, you say this, but that's not what I'm hearing from your researchers. And he

00:45:56.080 would say, yeah, well, that's a good point. And then he would take a step back and tell me something

00:45:59.640 else, hoping I would believe it. And if I challenge that, he would acknowledge it and then tell me

00:46:04.740 something else. And we just keep going. It never got to the point where I felt from him any awareness

00:46:10.560 that what I was going to write was not the story he wanted to tell. And I think this is a classic

00:46:15.340 problem with, you know, people of his sort of quality and strength of ego, which is he couldn't

00:46:21.080 imagine that I could see the universe differently than he did. I don't think he could imagine that

00:46:25.220 anyone could really see the universe differently than he did. When I wrote the book, it was called

00:46:29.420 Nobel Dreams, and described what had happened. He had an executive committee at the experiment made up

00:46:35.020 of, you know, a dozen or so physicists from all these collaborating institutions. And he asked them,

00:46:40.280 he wanted to sue me. And he asked them to join in the suit so that would have more power. They

00:46:47.960 refused. And what they didn't tell him is that most of them had read my book and critiqued it in draft

00:46:55.060 as a favor to me. He didn't say that in the book, because that I said that in the book,

00:47:00.140 I would have created problems for them professionally. The book was accurate. So, but he didn't see the

00:47:05.280 universe that way. Did you like the book itself? The writing makes me cringe.

00:47:10.280 I can see the book it could have been. You know, it came out when I was still 29 years old,

00:47:17.220 I guess. The story in it is interesting, but I can't read it. My 29-year-old prose still makes me

00:47:23.500 cringe, so. It came out in what year, 87-ish? Maybe it was February 87, yeah. So I would have been not

00:47:31.160 29, but 30. Did you go back to Discover at that point, or did you immediately jump into

00:47:36.800 the investigative work that would ultimately become the book, Bad Science?

00:47:41.680 No, I went back to Discover. I was freelancing for Discover. And actually, it's interesting. When the

00:47:46.720 book came out, it was February 1987. I was living in the Upper West Side of Manhattan. My publisher's

00:47:53.680 idea of getting me publicity was to get page six in the New York Post, which is the gossip column. So I

00:47:59.340 heard that page six was covering it. And it was, I remember it was a dark, snowy day in New York and

00:48:04.860 dreary. And I went across the street and went down Columbus Avenue and there was scaffolding and there

00:48:10.560 was a newspaper stand. And I bought the New York Post. The headline of page six is egghead squabble

00:48:16.600 over Nobel Prize. And in it, this Nobel laureate, Carlo Rubio, is calling me a 30-year-old writer in

00:48:23.120 New York, an asshole. And I'm thinking that my career in, yeah, I'm never going to eat lunch in this 0.98

00:48:28.360 town again. And my career in journalism is over. And as I continue to report stories, I start talking

00:48:34.780 to researchers about my book and Rubia and they want to know about Rubia. And they'll say to me,

00:48:40.540 well, if you think Rubia's bad, you should write about so-and-so. And it turns out that there are

00:48:45.140 characters like him in every area of science. So we can talk about this pathological science. People

00:48:50.140 tend to think of fraud, right? And when you're committing fraud in science, you're knowingly

00:48:55.600 trying to fool other people by manipulating the evidence. But there are two ways you can manipulate

00:49:01.640 the evidence. You could create a signal or you could do an inadequate job of studying your

00:49:07.440 background. Remember, it's a signal to noise problem. So you could fudge a signal or create a

00:49:12.540 signal or move a signal, or you could ignore the background. And if you ignore the background and do

00:49:17.200 a poor job in the background, you'll still end up with a big signal to noise ratio. You haven't

00:49:22.000 technically committed fraud. You're not actively trying to fool anyone else. You're fooling yourself.

00:49:26.900 And that's the kind of science Rubia did. He was ambitious. He was looking for discoveries.

00:49:33.420 Doing a background analysis is the hard, relentless, rigorous grunt work of science. It's endless and

00:49:39.940 thankless. Because if you do it right, all you'll do is prove that you were wrong all along.

00:49:44.480 Publicizing a signal and getting a Nobel Prize is the good part, right?

00:49:47.620 How do you reconcile that dichotomy, which is, this is a guy that's not committing fraud,

00:49:53.340 of course. But in your words, he's doing bad science because he's omitting to do the background

00:49:59.240 check. But at the same time, he is the guy through sheer personality and force of will

00:50:06.000 that discovers these two bosons that up until the discovery of the Higgs boson were the last two

00:50:12.860 pieces of this puzzle. And I don't think anybody's disputing the work that went into that Nobel Prize.

00:50:17.460 It's an unnerving revelation, right? It suggests that even at the uppermost echelon of science,

00:50:23.260 you can have people that are committing these omissions.

00:50:26.300 One of the messages in my book, I didn't go into this in great depth, but the papers that

00:50:30.260 won him the Nobel Prize, and again, I'm remembering from almost 40 years ago now,

00:50:35.080 those are probably wrong too.

00:50:37.020 He rushed it, I think it would be generous. Yeah.

00:50:39.380 He rushed them. Yeah. Because he was competing, but the particles had to exist.

00:50:44.000 That discovery had already been made. The neutral currents in 1973 or 74, I think it was Fermilab.

00:50:50.840 So those particles had to exist, and they had to exist with the characteristics that he

00:50:55.120 was looking for. So when he published the discovery based on, I think it was seven

00:50:59.980 events that had these characteristics, no one asked, were these really W particles?

00:51:06.320 In retrospect, most to all of them probably were detector flaws that they didn't understand yet,

00:51:13.180 that they still had to work out over the next year when they're trying to discover particles

00:51:17.340 beyond the standard model. This was the issue with Rubia. It was, the reason he got the award

00:51:22.400 immediately is there was a political issue behind all this also, which is since World War II,

00:51:28.000 all the major discoveries in physics had been made in the U.S. The U.S. had spent the big money

00:51:32.920 on particle accelerators. So this was CERN's opportunity, Europe's opportunity to achieve

00:51:38.360 relevance again. And the award was given almost instantaneously because this discovery made Europe

00:51:46.540 basically redistributed the center of gravity of the physics world back to, you know, if nothing

00:51:52.540 else, the center of the Atlantic. When I was reporting the book, one of the things I did was travel around

00:51:56.580 the country and actually around Europe talking to other physicists about their work, their discoveries,

00:52:01.880 and about Rubia and trying to rationalize what they told me with who he was and who I thought he was.

00:52:11.000 In defense of Rubia, none of this gets done without him. No, I'm going to rephrase that. None of it gets

00:52:16.140 done as quickly without him. It all gets done better without him. There are people who, from the experiment,

00:52:24.580 who Rubia ground up and chewed up and spit out, whose careers basically came to an end because of him.

00:52:31.640 And I said this in, I remember the introduction of the book. I said there were sort of three types

00:52:35.460 of physicists in the world. There were those who were at Rubia's level and told me he was brilliant.

00:52:41.720 There were those who knew Rubia well and said he was a very smart guy, but they wouldn't want to

00:52:47.260 work for him. And then there were those who worked for him. And there were points in the course of the

00:52:53.040 experiment I would sit in on group meetings and watch some young French physicist be reduced to tears

00:52:59.620 by Rubia's bullying. And bullying can be very beneficial in science because you're really

00:53:04.260 trying to force someone to think critically and force any sloppiness out of them. But there were

00:53:09.660 times I considered like, look, just ask Rubia to step inside. Don't write the book. Punch him out. 0.88

00:53:15.960 Will physics be the better for it? I decided physics might be, but I couldn't afford it. 0.53

00:53:22.240 So, you know, that's what you wrestle with as a young writer. The reviewers who didn't like my book,

00:53:26.460 which included Christopher Lehman Hount at the New York Times and Jeremy Bernstein,

00:53:32.060 who was a New Yorker physicist writer writing, if I remember correctly, for the Boston Globe.

00:53:37.640 They thought I had no right to write the book I wrote, to take on a physicist of this level to,

00:53:43.780 I think it was Bernstein who said this sort of sociology and character flaws of physicists at

00:53:49.880 this level don't interest him in the least. The only thing that's important is this glorious

00:53:53.720 cathedral of knowledge. It wasn't my interest.

00:53:56.920 So when did you happen upon Fleischmann and Pons?

00:53:59.560 So I went back to Discover as a freelancer. I had a terrific contract with Discover. I loved

00:54:04.420 living in Paris where I wrote Nobel Dreams. So I moved back to Paris writing for Discover

00:54:09.260 and then Time Inc. sold Discover. Science magazines have trouble making money because it turns out the

00:54:14.980 only people who really read them are high school boys, or at least back then were high school boys,

00:54:19.880 and they don't spend a lot of money. So they don't support advertising revenue. So Time Inc. sold

00:54:24.500 Discover to a low budget publisher and I lost my freelancing gig. And I had a choice of living in

00:54:31.540 Paris and getting by on scrapping to do freelance articles or moving to the States. I had friends in

00:54:39.180 L.A. who were writing screenplays and making enormous sums of money. And they said, as have been

00:54:44.720 said to screenwriters in the past, come to L.A., write screenplays, get rich, do what you want.

00:54:50.420 So I moved to Los Angeles, started writing screenplays. And in March 1989, Cold Fusion broke.

00:54:58.020 And my publisher called me up and said, do you want to write a book about this? And I needed to

00:55:04.160 supplement my income. I thought I could spend a year doing the book and make enough money to write

00:55:09.100 screenplays for two years after that. And then, as is my want, I got obsessed and ended up spending

00:55:14.780 three years on the book and owed my father $40,000 when I was done. I've bonded with many

00:55:20.320 writers over the years, young writers, about how much money we owed our parents by the time we were

00:55:25.260 done with our first book. You know, it was another fascinating story. I actually read the so-called

00:55:30.680 fusion to chemist Stan Pons from the University of Utah and his mentor Martin Fleischman from Southampton

00:55:38.180 University in the UK announced that they've basically created nuclear fusion in a test tube.

00:55:44.200 March 23rd, 1989. It's the front page of the Wall Street Journal. Within a day, it's front page news

00:55:50.580 all around the country. The very first article I read in the Los Angeles Times mentioned there was

00:55:55.520 a competing group at Brigham Young University. And as soon as I saw that there was a competing group,

00:56:00.720 I assumed they were wrong. Because as you mentioned, with UA1, UA2, you have two experiments competing

00:56:07.080 with each other. So the motivation is to establish reliable knowledge beyond reasonable doubt. That's

00:56:13.500 what you're trying to do in science. But if your motivation is also to beat the other people to the

00:56:19.340 reliable knowledge, you're going to allow one motivation to compromise the other. And now the

00:56:26.260 beyond reliable doubt part is going to be tossed out in an effort to win. That was one of the fundamental

00:56:31.420 problems at CERN with the work UA1 was doing. UA2 got everything right. Because once they came in

00:56:36.940 second, they were perfectly happy to do the good job I think they were doing all along.

00:56:41.280 On first principles, did you just think it was ridiculous based on your knowledge of physics?

00:56:45.900 No, I didn't know enough to think that. I just smelled bad science. 0.84

00:56:50.720 It's been a while since I read bad science, actually. I think I read that actually the first one I went back

00:56:55.400 to read. But did they claim it was net energy positive? Yeah, the idea is how do you make

00:57:00.980 nuclear fusion? Well, you put heavy water in a test tube with an electrode made of palladium and you

00:57:07.600 plug it into the wall and the deuterium gets sucked into the palladium. And the idea is somehow it gets

00:57:14.160 compressed so much that it fuses and generates ideally neutrons, which are one sign of gamma rays or

00:57:22.380 some radioactivity, which would tell you that there was a nuclear reaction going on, if not at least

00:57:27.480 generating a lot of heat. So Pons and Fleischmann had had an explosion in their laboratory with one

00:57:33.240 of these cold fusion cells. It was being run by Pons's son. It might have been their garage. I forget

00:57:38.140 the details. And when the cell exploded, right, if you're looking for trying to create nuclear fusion

00:57:43.260 in a test tube and your cell explodes, eureka. But even more important to that, it wasn't just that

00:57:48.920 there was a competition with the people, a physicist at Brigham Young, 30 miles, 40 miles down the

00:57:54.620 highway from the University of Utah in Salt Lake City. But the physicist at Brigham Young had studied

00:57:59.980 something called muon catalyzed fusion, which was a type of cold fusion of a sort and had been shown

00:58:07.680 Pons and Fleischmann's proposal and had started working on this after reading Pons and Fleischmann's

00:58:13.060 proposal. So Pons and Fleischmann thought their discovery was being stolen from them by a physicist

00:58:18.040 who should know enough to know whether it's right or not. And as I discuss this at the very end of

00:58:24.020 my research, when I spent about eight hours with the then president of the University of Utah, a very

00:58:29.260 thoughtful, wonderful man named Chase Peterson. And he was explaining to me, you know, we were talking

00:58:34.780 about their thought process and going public with something that became this huge fiasco. He said,

00:58:40.420 you know, yeah, you know, somebody's nothing makes your girlfriend look more attractive than your 0.61

00:58:44.320 best friend trying to steal her from you. And so once they decided that these people at Brigham Young

00:58:49.700 were trying to steal it from them, they decided it must be right. They had to throw a press release.

00:58:53.940 They threw a press release and this and this enormous phenomenon kicks in where everyone in the world

00:58:58.340 tries to replicate it because it seems it holds the promise of infinite free energy. It's the wealth

00:59:04.600 of OPEC and anyone can study it. It costs like three thousand dollars worth of materials to put

00:59:09.580 together a cold fusion cell. So you have basically created a bad science generator.

00:59:16.400 When did it become abundantly clear to you what was going on? How long into your investigation?

00:59:22.100 This is a very different kind of story because this is immediately front page news. So when I jump

00:59:27.300 into it, there are meetings, there are conferences, there are other people discussing this. I, you know,

00:59:33.200 I probably flew out to Utah within, you know, a week of having that conversation with my editor. And

00:59:38.800 Utah is, of course, surrounded by, you know, journalists who want to know if infinite free

00:59:43.140 energy has been created. Funny coincidence. The Wall Street Journal ran it on the front page. The

00:59:49.420 author was a science writer for the Wall Street Journal named Jeremy Bishop. And he sort of took,

00:59:55.280 so I write a book explaining why Jeremy Bishop was sort of not thinking things through clearly when he put

01:00:02.260 this on the front page. I moved back to New York in 1993 or so. I move on to a block on the Upper West Side

01:00:10.000 of Manhattan where Jerry Bishop and I are looking into each other's windows from across the street. Anyway, the

01:00:17.760 world was finding out that this was wrong. I'm following all the attempts to replicate it. In a sense, this was how

01:00:23.560 science is supposed to work, right? Somebody makes a claim of a discovery and they publish the data necessary to

01:00:29.960 replicate that discovery by independent laboratories. And so independent laboratories go out and try to

01:00:35.640 do that. And around the world, chemists and physicists tried to do that. And most people got the right

01:00:41.780 answer, which is cold fusion doesn't exist. But a half dozen laboratories around the world, for reasons

01:00:47.800 that tended to be different than each lab, got the wrong answer. Usually it's, we can talk about that. I mean,

01:00:55.080 there would be a lot of interesting reasons why. So the phenomenon keeps going because of the signal

01:01:00.120 being generated from these six labs while again, all these other labs in the world are working out

01:01:04.680 the background. And some of the best scientists in the world are involved with, this is a major

01:01:10.360 discovery if it's true. So one of the things that attracted was some of the very best physicists and

01:01:14.700 nuclear physicists in the world. And then I had the end chemists. And I had the opportunity to work

01:01:19.420 closely with these people, basically to embed myself with them, while they tried to replicate

01:01:25.180 the experiment and did in effect the background analysis that Pons and Fleischmann and Steve Jones

01:01:30.380 of Brigham Young never had the time to do. So now you get to find out all the ways your equipment could

01:01:35.260 have fooled you and assuredly did. And that's the science working the way it should work. And so by

01:01:41.260 June of that year, it was clear that this was wrong, that they had Pons and Fleischmann,

01:01:45.500 that whole thing was wrong. And then I continued reporting it for about nine months to tie up all

01:01:51.180 the loose ends, to explain some of the other positive results. But it was clear to the world

01:01:56.140 by within two to three months that this was just wrong. And then it faded from the papers.

01:02:01.820 And then you wrote a book about this, of course. So the book was really less about letting everybody

01:02:06.940 know this was wrong. The purpose of the book was to explain how it became wrong, right? How did this

01:02:11.420 happen? Yeah, I became obsessed with how this happened. It's still what fascinates me. So,

01:02:16.940 you know, when I'm writing about nutrition and chronic disease, obesity, diabetes, the question

01:02:21.100 is, how do people establish conventional wisdom, dogma, the ruling theory in any science and on what

01:02:26.940 evidence? And in this case, cold fusion, I wrote it as kind of, I saw it as a sort of case study in

01:02:32.940 bad science, something in my fantasy, something that every graduate student who goes into science would

01:02:38.140 be required to read so they could see the price, all the different ways you're going to fool yourself

01:02:43.420 and the price of doing so, of claiming something to be true that then later turns out not to be.

01:02:49.020 What are the hallmarks of pathological science? Again, there was a great paper written about this

01:02:53.820 that wasn't it brought back to light after the cold fusion debacle?

01:02:58.380 Dr. Irving Langmuir, Nobel laureate chemist, gave a talk at IBM in 1957 on pathological. He

01:03:05.740 called it pathological science, which is the science of things that aren't so. These are

01:03:10.060 things he said. It's not about fraud or manipulation of the data. It's about the scientists not realizing

01:03:15.180 how easily they can be fooled. And there are certain signs and symptoms. The effects tend to

01:03:19.980 be at the very level of the equipment's ability to detect them. There were four criteria that he

01:03:27.980 listed. I should remember them, but I can't.

01:03:30.620 Yeah, there was a sensitivity analysis in there as well. Small perturbations.

01:03:35.100 But let me give you an example sort of how this plays out. So you've got these test tubes, right?

01:03:40.780 And this is a common thing. And they're supposedly generating excess heat. But the physicists are

01:03:45.100 saying if you've created nuclear fusion, you should be generating gamma rays and neutrons. It should be

01:03:50.460 radioactive. Chemists at the Georgia Tech Research Institute, which is attached to Georgia Tech

01:03:55.820 University. They put together cold fusion cells and they borrow neutron detectors. In fact, you could

01:04:02.940 think of them as Geiger counters from the physics department. And they've got these cold fusion

01:04:08.300 cells and they plug them in. The cells are bubbling away and they hold the detectors over the cold fusion

01:04:14.140 cells and the detectors go off. And then they move the detector away and the detector stops rattling.

01:04:20.780 And then they bring it back over the cold fusion cell and it starts. They're detecting neutrons,

01:04:25.660 clearly. So they write up a paper. They have a press conference. And at the press conference,

01:04:31.980 one of the local physicists from Georgia Tech says to them, do you realize that these neutron detectors

01:04:38.860 are humidity sensitive? Now, they're chemists. They've never worked with neutron detectors before.

01:04:43.660 So this is at the press conference. And they go, tell us more about that. And they say, well,

01:04:49.020 when you held it over the cell, the cell was bubbling. It's humid. That might have been why

01:04:55.180 it went off, right? They're doing what scientists are supposed to do ideally before publication.

01:05:00.460 So they go back to the lab. And now they do what they should have done all along is they do a control.

01:05:06.460 So control is same experiment without the deuterium, just do it with water. So now you're generating,

01:05:11.820 yeah, now, now they hold the detector over the control and it goes off and they hold it over.

01:05:16.540 And they realize within a day of their press conference, probably 12 hours of their press

01:05:20.620 conference that they've screwed up. And the reason they screwed up is because they in effect chose the

01:05:25.740 wrong control. They thought the control was just not holding it over the cell, but they weren't

01:05:30.860 controlling for the humidity. Fits back to understanding your background. What are all the other possible

01:05:36.140 causes of what we're seeing? And we have to go through methodically. And this is what takes time,

01:05:41.500 time, time. And you need the help of every smart person, you know, to point out all the different

01:05:47.100 things that you might've thought for that you now have to design an experiment to control

01:05:51.660 for. So everywhere in the country, somebody did an experiment and claimed that they had seen signs

01:05:57.260 of cold fusion. Ultimately, you could explain it away as some aspect of the background that they didn't

01:06:03.020 understand. And it was always, so it was chemists trying to do physics. It was physicists trying to

01:06:07.980 do chemistry. And often if these cells are really creating nuclear fusion, they got to be generating

01:06:13.420 more heat than they're taking in from the wall. You know, they're plugged into the wall. So they have

01:06:18.460 to generate more heat than they, expenditure has to be greater than intake. It's a Chico thing here. 0.53

01:06:23.740 And the science of calorimetry is very complicated. So there are people out there who spend their

01:06:28.860 whole lives doing nothing but calorimetry. They were called calorimetrists in the chemical world,

01:06:33.980 and they were capable of doing these experiments right. When they did the experiments, they saw no

01:06:39.420 excess heat. When somebody who wasn't a professional calorimetrist did the experiment,

01:06:44.940 they inevitably saw a little bit of excess heat, often because they didn't realize that there's some

01:06:48.860 feedback going down the cord. So you're going to get energy coming back into the cell from the power

01:06:53.740 source. If you haven't spent your whole life doing this, you don't know the details of the background.

01:06:58.700 How long did it take for Pons and Fleischmann to come to grips with their errors?

01:07:04.060 Well, I don't think they ever did.

01:07:05.580 One of them's not even alive anymore, right? So he went to his grave believing...

01:07:09.580 I think they both might have passed away by now. Fleischmann would have been at least 50 in 1989.

01:07:15.820 So, you know, it's funny. There's a section of the Twittersphere that likes to point out that I was

01:07:20.620 asked once at a debate in England whether or not I would ever change my mind about the uselessness

01:07:26.860 of the energy balance equation for obesity. And I said, no, I doubt it. So constantly,

01:07:31.980 I get these tweets directed at me where people say, look, why are you even discussing with Taubes?

01:07:36.300 He said he would never change his mind anyway. And the reason I said that is because you have

01:07:39.500 phenomena like Pons and Fleischmann. I've witnessed people live through this. It takes a kind of

01:07:44.940 superhuman intellect to be able to say something I believe beyond reasonable doubt and have staked my

01:07:51.900 reputation on is actually wrong. It's always easier for somebody to believe personally that

01:07:57.900 all the other experiments screwed up. People just didn't do the experiment right. And I used to joke

01:08:02.060 that even if cold fusion was real, someday, back when I made this joke, I was probably thinking around

01:08:07.420 2021, there will be the Stan Pons Memorial cold fusion nuclear plan, you know, on the coast between

01:08:14.300 San Diego and LA pumping out enough energy from three beakers of heavy water to power the entire

01:08:20.780 southern empire. And I'm going to be crawling around the outside of that perimeter fence with, 1.00

01:08:26.220 you know, a beard and pieces of Dunkin' Donuts in the beard and a tattered copy of Bad Science looking 0.99

01:08:32.780 for where the damn thing is plugged in. Because I know it's a con, right? I am never going to be able to 0.99

01:08:37.580 change my mind and accept. But then on the other hand, my argument to the cold fusion people was,

01:08:42.940 if you want us to believe it, create cold fusion cells, make billions of dollars. The day that my

01:08:48.380 neighbor buys a cold fusion power car, I'm going to have to start believing it or I'm in trouble.

01:08:52.940 Until then, I think an enormous amount of skepticism, if not closing my mind is perhaps the correct thing

01:08:59.420 to do. Well, I do want to come back to the point about your flexibility or lack thereof with respect to

01:09:06.140 current belief systems around nutrition. But let's now really pivot from a career that spans a

01:09:12.620 decade and a half writing about physics and other means of science to the pursuit of public health.

01:09:19.340 How did your curiosity get peaked there? What took you into a new avenue of science?

01:09:24.780 The cold fusion book, Bad Science, I interviewed around 300 people for this book. Horace Freeland

01:09:30.380 Judson was one of the great science writers of the late 20th century. He wrote a book about the history

01:09:36.300 of the molecular biology revolution. Anyway, he said I had done more research for the stupidest scientific 1.00

01:09:43.420 subject than any human being alive. I was obsessed with how this had happened, how people made the 0.98

01:09:48.780 mistakes, how the sociology worked. Again, I became obsessed with pathological science as a concept.

01:09:54.700 And when I was done with the book, I had a lot of friends in the physics community who really respected

01:09:59.180 and appreciated what I did. And some of my friends in the physics community were involved in this debate

01:10:04.220 about whether or not electromagnetic fields from power lines can cause cancer. And that involved,

01:10:10.300 so on one level physics, because what we know about physics is they can cause cancer because their

01:10:15.580 wavelengths are too big to interact with something as small as cells. And on the other side, you have

01:10:20.620 epidemiology. You have the science of disease, how diseases spread, and researchers saying we can

01:10:27.020 measure the field strength around power lines and somehow make an association between that and the

01:10:33.020 levels of cancer and people living near the power lines. And so my physicist friends were horrified at

01:10:38.940 the level of science they were seeing. And when my book came out, they said, well, look, if you're interested in

01:10:43.100 bad science, maybe you should look at this stuff in public health because it's terrible. And so the very first

01:10:49.580 article I did, I did an article for the Atlantic on the power line cancer connection. And it was dependent on

01:10:57.020 this field of epidemiology, which I had never paid any attention to, never had any reason to. But I had spent the

01:11:03.180 previous six, seven years learning about how hard science is to get right and being schooled and tutored and

01:11:10.140 lectured into believing that you have to be extraordinarily rigorous and methodical and

01:11:15.900 relentless or you'll get the wrong answer. And here was a field where basically people just said,

01:11:21.340 we're going to look at this, we're going to look at that, we're going to get an association,

01:11:24.060 we're going to assume it's causal. They considered doing the kind of rigorous testing of hypotheses to

01:11:29.820 be a luxury that they couldn't afford. So I wrote the Atlantic piece and then I wrote a piece on

01:11:36.060 epidemiology for the journal Science that was infamous. It used to be that if you wanted to

01:11:44.620 publish critics of the field, first you published Alvin Feinstein, who was at Yale,

01:11:49.900 and then it was Feinstein and Taubes and now it's Ioannidis. And Feinstein and Taubes have kind of been

01:11:56.220 dwarfed by John Ioannidis' contribution. And Feinstein was a controversial character.

01:12:02.140 Both great pieces, by the way, which we'll be linking to in the show notes.

01:12:05.740 Thank you. So Feinstein's piece, you've got a field of science where you can't do the

01:12:10.780 experiments. So that's the first problem. That's a problem with every field where there's an active

01:12:15.420 controversy. If you can do the experiments easily, like cold fusion, I think I've discovered infinite

01:12:21.660 free energy. Well, tell me how you did it. Let me see if I can do the same thing. And if I can't,

01:12:26.860 we got to talk. And if he can't, and she can't, and they can't, then you probably screwed up, dude. 0.85

01:12:34.380 If you can't do the experiment, then depending how interesting it is, you then develop a field

01:12:40.940 of science around basically a hypothesis. And often as more and more people see the same phenomenon,

01:12:48.860 but without using the rigorous experimental techniques of an experimental science,

01:12:54.220 science, they start to believe that this hypothesis is true. And you, again, what I described with

01:13:00.520 Rubia, with this sort of group thing phenomenon, where you collect the people who see the signal,

01:13:05.660 and you ignore the people who work on the background. In epidemiology, there was never

01:13:10.520 any significant work on the background. So you could think of epidemiology where you have a cohort of

01:13:16.380 people, say the Nurses' Health Study, the most famous in the U.S., 110,000 nurses,

01:13:21.800 and you give them questionnaires, food frequency questionnaires, and you ask them what they're

01:13:28.300 eating, and they tell you what they're eating, and then you follow them, and you see who gets sick

01:13:32.240 and who doesn't. And then you look at what the people who get sick tended to be eating versus what

01:13:38.280 the people who didn't get sick tend to be eating. And you have an association between diet and disease,

01:13:45.440 and there's no causal information in that association. But now you hypothesize that the

01:13:52.620 association is causal. Whatever they're eating or not eating causes disease, and there's no way to

01:13:57.820 rigorously test it. So the process of science breaks down.

01:14:01.540 Let's stop for a second, Gary, and explain to people why the intuitive is not so. Because I think it's worth

01:14:08.720 spending a second on the fact that all of this stuff that we talk about in terms of the rigor of

01:14:14.240 scientific thinking, it's a relatively recent phenomenon. It's less than 400 years old.

01:14:19.780 All of the things you're describing are less than 400 years old, which means from an evolutionary

01:14:24.580 perspective, they're non-existent. It's not wired into our DNA.

01:14:28.640 Just might add that everything I'm describing is in... So Francis Bacon, 401 years ago,

01:14:34.700 publishes Nova Morganum, which is sort of the beginning text in the scientific method.

01:14:39.720 And everything's in there. I had to go back around 2001, 2002. I read Bacon because I was

01:14:45.320 thinking maybe I'm just crazy. But if you think back to 4,000 years ago,

01:14:49.560 40,000 years ago, right? If you think back into the roots of our genes, the ability to pattern

01:14:56.180 recognize and make an inference that by definition is presumed to be causal from an association must have

01:15:03.260 been a vital trait for our success. I saw Larry doing X. Larry has more mating opportunities than

01:15:11.840 me. Doing X must be the reason that Larry has more mating opportunities than me.

01:15:17.420 That's true. So you make the association and then you test it. I'm going to do X and see if I get more

01:15:24.040 mating. And evolution is going to test it, right? Because if John now does X and it ends up getting

01:15:30.840 him killed. But we didn't codify it. I think that's the difference. We didn't live long enough

01:15:35.300 to see the result because the arc of the information coming back is too long. So you're right. Evolution

01:15:42.140 did follow the experiment, but at the individual level, we never got to learn from it. So come along

01:15:48.460 401 years ago, and of course, many steps along the way with the introduction of true randomization,

01:15:54.580 statistics, which became an important tool, et cetera. You have this concept that is very foreign to

01:15:59.380 people. And while anybody can say very glibly, well, correlation doesn't equal causation. I think

01:16:06.080 anybody has heard that and understands what it means. I don't think it's entirely clear to people

01:16:11.740 why control and randomization matter and how ubiquitous bias is. So why is it that people living

01:16:20.000 next to telephone poles who have a higher likelihood of getting cancer are not getting it because of the

01:16:26.420 telephone poles or the power lines, rather? Let's understand our background, right? So now this

01:16:30.780 is the crew. So the background is all the ways you can be fooled. Okay. So we see more cancers close to

01:16:35.740 the power lines than we see far away. Can we explain that as something other than the power lines cause

01:16:42.640 cancer? And so what else could explain this? That's the issue. And now you've opened up a door where we

01:16:49.560 have an infinite number of possibilities. That's why right now, the probability that your hypothesis

01:16:54.320 correct has gone down enormously as soon as you accept the reality that there are an infinite

01:16:59.720 number of other possibilities that you have to rule out. Some are very likely. You'd refer to,

01:17:03.860 you used to refer to their first order factors and second order. And as you go down, they become less

01:17:09.300 and less likely. And this, when we talked about the physicists needing a five sigma effect, that's

01:17:14.300 because they're going to say, well, we can rule out all the major factors and we want to leave a lot of

01:17:18.380 room for the fourth, fifth, sixth order variables because we can't get to everything. So maybe people

01:17:24.940 live closer to power lines are poorer than people who live farther away from power lines. They might

01:17:31.000 be a slightly lower socioeconomic status. Okay. Nobody wants to live next to power lines. They're

01:17:36.440 unsightly. They make noises. So if you can afford not to, you don't. So maybe what you think is being

01:17:41.820 caused by the cancer is actually a result of socioeconomic status. I think a lot of what the

01:17:47.660 epidemiologists publish are false positives that could be explained by socioeconomic status.

01:17:53.300 Now the epidemiologist will tell you that they correct for that. Don't they look at the household

01:17:58.360 incomes? If they look at the household incomes, there are major studies. The Nurses Health Study,

01:18:04.220 the most famous study in America, never looked at household incomes. The point is many people don't.

01:18:09.220 So let's say you had to get the health records of people, right, to know how much cancers they had.

01:18:15.420 So how'd you get their health records? Well, we called the people up and we talked to them and

01:18:23.240 asked them if they wanted to be involved in the study and can we get their health records. Well,

01:18:27.500 the people we called, we got to consider we're calling them during the day. So maybe the people

01:18:32.180 who answered the phone are the people who aren't working, right? So maybe there's a socioeconomic

01:18:39.120 status. Maybe there's a bias there because the more affluent people, the people who actually have jobs,

01:18:44.160 are not home to answer the phone during the day, right? Maybe. So there's a whole world of things

01:18:50.900 like that. Or maybe the people more likely to participate in the study are the ones that have

01:18:55.860 worse outcomes worth reporting. Want to participate in the study because God knows, you know, little

01:19:02.320 Jimmy had a brain cancer and they never understood why. And now you're giving them a reason. So they want

01:19:07.560 to be a part of the study. And, you know, this is the process that Rubia didn't like to do of doing

01:19:12.760 the background analysis. It's very hard to do because you have to, each one of these factors

01:19:18.640 is a hypothesis in and of itself. So you have to be able to test the hypothesis. Maybe the biases in

01:19:25.140 the phone sampling process caused what we see. How do we test that? You know, if you can't do an

01:19:30.500 experiment, it's very hard to test. And then there's this factor. We were talking about Nobel

01:19:34.700 Dreams, my first book. So Rubia and company are spending the fall and the winter of 1985-86

01:19:43.640 trying to understand their background. They understand more and more of their background.

01:19:48.440 They finally come to the revelation that they're probably not seeing anything. And other people are

01:19:52.440 doing the same thing also. And there's a meeting in the Aosta Valley in Northern Italy where they're

01:19:58.380 going to discuss all of this. And I went to that meeting. And in the meeting, an elderly Italian 0.84

01:20:04.020 physicist, I forget his first name. His last name was Altarelli, a very cool guy. Altarelli stands up

01:20:11.480 in the meeting. Rubia and everybody is there. And Altarelli begins to explain all the different ways

01:20:17.340 that the equipment could have fooled Rubia into thinking every possible other explanation for the

01:20:23.080 event. So maybe in this case, the muon comes off and it gets lodged in the infinitesimal dividing line

01:20:31.200 between the two detectors and you don't see it. Maybe a neutrino comes off on this one and it

01:20:37.660 bounces off the side of the tau detector and you don't see it. And by the time he's done,

01:20:43.520 and the chapter in the book was called the Altarelli cocktail, and you come up with all these

01:20:48.140 different ways, one of this, one of that, one of these, are all the unknown unknowns. You can now

01:20:53.300 imagine them because around the world physicists have spent months trying to. And it's clear that if you

01:20:59.200 begin to understand the background and all the unknown unknowns, you could explain everything

01:21:02.680 Rubia and company has seen. It's not a five sigma result. It's not even a one sigma result.

01:21:08.000 That's where the discovery died, was in that meeting in Aosta. Again, this is programming my

01:21:12.880 thinking. In cold fusion, all these different labs that had published positive results, I told you what

01:21:17.940 happened at the Georgia Tech Research Institute, they all had a different explanation. It was another

01:21:22.000 version of the Altarelli cocktail, but the point was whenever they saw something that confirmed their

01:21:26.920 preconceptions, they accepted it. Okay, so there's a selection bias in what you pay attention to,

01:21:33.940 and you're always going to pay less attention to the things you don't want to see. So in epidemiology,

01:21:40.600 people might look for one explanation. Say, I believe socioeconomic status probably explains a good

01:21:46.000 deal of these phenomena, and people just don't measure it or they don't measure it right. But even if it

01:21:50.280 can only explain a quarter of the phenomena or a tenth, I can guarantee you that there are probably nine

01:21:55.240 other things that could cover the other nine. The difference is in epidemiology, right, you're

01:21:59.900 looking for 95% confidence level. You're willing to claim a causal effect based on, there is no causal

01:22:06.920 information, and you're willing to claim a causality based on two sigma, not five, not three, but a 95%

01:22:14.520 confidence level. When if you actually try to think about it, if you say, let's treat this like the

01:22:20.320 physicists treat their science, we can go through and we can come up with so many possible alternative

01:22:25.680 explanations that there's a very high likelihood that what you're claiming is a discovery is a false

01:22:32.000 positive. In the case of that example, what would be the alternative experiment? How would we, if we're

01:22:40.480 trying to answer the public health question, which is, it's inevitable that we have power lines, we have to

01:22:46.260 make an infrastructure decision, which is, can they traverse the areas where people live? One would

01:22:55.020 argue that the length of time it would take to answer that question, which would be to randomize

01:23:00.420 people independent of every variable imaginable, assuming you could get people willing to comport

01:23:06.340 to such an experiment for a long enough period of time to see a result. I mean, you can quickly convince

01:23:11.280 yourself that it's simply impossible to do a true experiment here. And yet an answer to that

01:23:18.080 question should be known. So what is good enough? Well, and that's the debate. This is what the field

01:23:24.140 of chronic disease epidemiology has been going through. So infectious disease epidemiology, like

01:23:29.440 with COVID, is a very different issue. You can, in effect, do randomized controlled trials like you do

01:23:34.460 with vaccines. Here, as you just explained, they're effectively impossible. Certainly with something like

01:23:39.960 power lines, because you're never going to be able to randomize people to live near power lines or away

01:23:44.320 from power lines. You know, I think they're just beyond the limit of science. So the question is, how do you

01:23:49.720 deal with that issue? The conventional thinking has been to embrace this idea of prudent avoidance. So, you know,

01:23:56.500 we do that when we wear masks for COVID. But prudent avoidance says you create regulations to try and

01:24:04.300 minimize whatever fields are being emitted from these power lines. So the power lines pay more.

01:24:09.920 They're not avoiding a potential harm doesn't come without its own harms, as we know, in the COVID

01:24:16.040 debate about what happens with shutdowns. So you can say, well, we can regulate the power lines better.

01:24:21.040 But that means the power companies are going to raise their rates. And now people are going to have

01:24:25.440 to pay more for power, which means their socioeconomic status is going to be lower per dollar earned.

01:24:30.840 There are positives and negatives. And these have become social value judgments more than anything.

01:24:36.820 The question is also, what's the incumbent on the researchers? You know, we're both every fan of

01:24:43.560 science tends to be a Richard Feynman science fan, if not a Richard Feynman as an individual has become

01:24:51.480 apparently far more conflicting. Feynman says, and I quote this in the beginning of good calories,

01:24:57.920 bad calories, a fundamental, what you need fundamentally in good science is this sort of

01:25:01.720 bending over backwards, to be honest about what you know and what you don't know. So as soon as you

01:25:06.120 start doing, and he says this in his 1974 Caltech commencement talk, where he talks about the

01:25:10.940 difference between advertising, good science, good science, the it's incumbent upon the investigators to

01:25:16.320 talk about all the limitations in their research. And I think in an ideal article, you know, published

01:25:22.360 journal report, the limitation section would be longer than the results section. Because what we want to know

01:25:27.860 is all the ways they could have fooled themselves, realizing that they very likely did. And then to

01:25:32.800 discuss it with complete and utter honesty, rigor. And I have critics who would say I don't do that. I

01:25:38.440 personally, I try to do that the best I possibly can, because I think that's the ethical moral

01:25:43.660 obligation of doing science or writing about science is not over exaggerating it. If the people with the

01:25:50.120 scientists covering the power line controversy said, we think we know this, but these are all the

01:25:57.080 things we don't know. Nobody acts. That's the problem. So if they're right, like we think

01:26:03.040 cigarettes cause lung cancer, but we never did a randomized control trial. Now the lung cancer

01:26:08.860 effect is huge. That's like a six sigma effect, even if there's no causal information. So we could

01:26:14.180 kind of believe that's causal because we can't think of any alternative explanation for why smokers have

01:26:19.740 such a hugely increased risk. If you're honest about what you see, that's fulfilling the obligation of the

01:26:25.460 scientists. But now nobody changes their behavior. So if you're right, that say power lines cause

01:26:31.180 cancer, people are going to get cancer because you haven't forcefully enough made your argument. And

01:26:35.560 that's the conflict of public health science. How do we get people to change their behavior when

01:26:41.120 we really don't know if they should?

01:26:43.420 And this is an enormous problem. It's the dietary salt, dietary fat. It's where we're going to be

01:26:49.700 going in the next couple of minutes here. So you've struggled through power lines and cancer.

01:26:53.920 You've come to wrestle with the pitfalls of epidemiology. What brings you into the den of

01:26:58.940 nutrition?

01:26:59.640 I'm living in LA, Venice near the beach, doing freelance science journalism and working on

01:27:06.660 screenplays, probably still not making any money. So my friends were wrong about at least my ability

01:27:12.500 to get rich writing screenplays. I call up my editor at science and I say, look, I got to pay the rent

01:27:18.880 next month. I need a story I can turn over quickly. Do you have anything I could write that'll

01:27:23.760 generate a paycheck? And so researchers have just reported, they're reporting on the first results

01:27:29.560 on the DASH diet. This dietary approach is to stop hypertension. It's a low fat, lots of fruits

01:27:35.380 and vegetables, dietary approach. And the paper's coming out in the New England Journal of Medicine

01:27:40.540 and they have a preprint. They have a pre-release copy. It's embargoed. So I should write about this.

01:27:46.740 And you've never written about nutrition and health science?

01:27:49.260 Never written about nutrition. No.

01:27:51.440 Okay. So he's pretty desperate if he's giving you this, because this is a little outside of

01:27:56.100 your wheelhouse, right?

01:27:58.140 No, it's one page in the magazine. It's not a lot going on. So the way you do a one page story,

01:28:04.400 right? You call up the principal investigator. If the article hasn't been published yet,

01:28:08.500 you ask the PI, you interview him, you ask the PI for the names of a couple of other people who know

01:28:13.140 about the research, you could comment. Three interviews for one page is doing your job. So

01:28:18.700 it takes a morning to do the interviews, an afternoon to write the article. It's a day and a

01:28:23.100 half. I get my rent money. What I didn't know about this article is it had been leaked to science

01:28:28.660 in advance. And it had been leaked with a list of researchers who I could interview.

01:28:34.080 This is what year?

01:28:34.720 This was 1998 or nine. My editor gives me the list. He doesn't tell me it was leaked or if he did,

01:28:42.260 I didn't pay attention. So I get the article. I call up the principal investigator, Larry Appel

01:28:47.400 at Johns Hopkins. I interview him. I ask him for the names of people I could talk to. I call up one

01:28:53.820 of the people on the list. This is a former president of the American Heart Association at the University

01:28:58.980 of Alabama, Birmingham. And this person tells me that they can't talk about the paper. I thought,

01:29:04.600 I lose their funding. And I say, we're talking about a diet trial. The New England Journal of

01:29:09.960 Medicine. Nobody loses their funding for talking about that. It's not going to come out until after

01:29:15.560 the embargo. The woman, she refuses to talk to me. She won't tell me anything. I said, look, if there's

01:29:20.920 something wrong with this paper, let's go off the record. Not for attribution, like Woodward and

01:29:26.620 Bernstein in the garage with Deep Throat. Tell me what's wrong with this paper, what the issue is,

01:29:32.180 because if you don't tell me, I'll never know and I'll report it incorrectly. She refuses even to do

01:29:36.920 that. So then I get off the phone with her. I call up one of the people that Larry Appel at Johns

01:29:42.180 Hopkins has given me. And he's a researcher who starts yelling at me. This guy is the grand old

01:29:47.840 man of the field. I don't know this. And he starts yelling at me over the phone that there's no

01:29:51.640 controversy over salt and blood pressure. And I say, I'm not calling about salt and blood pressure,

01:29:56.620 professor. I'm calling about this diet trial that lowered blood pressure in the New England

01:30:00.500 Journal of Medicine. And he continues to berate me that there's no controversy. So I get off the

01:30:05.400 phone with him and I call up my editor at Science. I said, I had an American Heart Association

01:30:10.040 former president refused to talk to me because she would lose her funding, she said, if she did.

01:30:15.920 And then I had this other guy yelling at me that there's no controversy over salt and blood pressure

01:30:20.460 when I'm not writing about salt and blood pressure. There must be a controversy about

01:30:26.580 salt and blood pressure that I know nothing about. So I'm going to write up this article,

01:30:31.440 get my paycheck. And then if you don't mind, I'm going to look into this assumed salt blood 0.96

01:30:35.520 pressure controversy and see what we're missing. And I spent the next nine months reporting that.

01:30:40.780 I interviewed about 85 people for one magazine article. Turns out that it is one of the most vitriolic

01:30:45.840 controversies in the history of medicine. And even though already by 1998, we'd all been eating low

01:30:51.620 salt diets in America for 15 to 20 years, it was clear that the randomized control evidence never

01:30:59.000 really supported the intervention and that it was backed up by a lot of bad epidemiology. And research

01:31:05.480 is assuming that associations were causal that weren't, and that were even questionable associations

01:31:10.180 to begin with. And while I was doing that story, this fellow who was yelling at me,

01:31:15.840 I like to joke, he sounded exactly like Walter Matthau over the telephone. By the way, he's still alive.

01:31:21.260 He's about 101 years old. So while I'm running him down as a scientist, that's evidence that maybe

01:31:27.120 he understands nutrition and diet far better than I do. Anyway, while I was interviewing him,

01:31:33.020 this Walter Matthau character, it was clearly the, you know, I had spent 10 years of my life studying

01:31:38.440 bad science. It was clear this guy was one of the worst scientists I'd ever interviewed in my life,

01:31:42.540 in the bottom five, at least. And I thought I had interviewed the worst. He took credit not just

01:31:49.080 for getting Americans to eat less salt, but eat less fat as well for the low fat diet we had all

01:31:54.980 been on since 1984. So at one point, I called up my editor and I said, look, this guy was involved in

01:32:00.880 this controversy, the fat controversy in any way, the fat story, there's got to be a story there.

01:32:06.060 The message from Nobel dreams and bad science was that bad scientists never get the right answer.

01:32:12.720 You know, it's just, it's too hard to get the right answer for you to go in being sloppy and

01:32:18.020 slipshod and lazy and ambitious and get it right. Nature isn't that kind. So I said, you know,

01:32:25.000 when I'm done writing about salt, I'm going to write about fat. I have no idea what the story is.

01:32:28.880 I've been eating, I've been living in LA, eating my egg whites and probably a 15% fat diet.

01:32:35.240 Dean Ornish would have been proud of me. But if the dogma was based in any substantive way on

01:32:40.120 this fellow's work, there's a story there. And I spent a year writing that piece. I interviewed

01:32:45.440 145 researchers and administrators for one magazine article.

01:32:51.100 The Big Fat Lie?

01:32:52.200 No, this was the soft science of dietary fat. So this was, the salt story was called the

01:32:58.360 political science of salt, political in parentheses. And the fat story was called the soft science of

01:33:04.420 dietary fat, soft in parentheses. They both won National Association of Science Writers,

01:33:09.600 Science and Society Awards.

01:33:11.940 After you wrote these two articles, one on the soft science of fat and the other on the

01:33:17.280 political science of salt, what made you decide to go even further and write what would become

01:33:23.200 perhaps the biggest and most controversial piece you wrote, at least in the newspaper?

01:33:27.160 I believe it was for the Times Magazine, wasn't it?

01:33:29.660 Yeah, New York Times Magazine. I wanted to write a book when I was done with the two science articles

01:33:34.840 on whatever was happening in medical science that could lead to these kinds of mistakes. Remember,

01:33:40.600 my obsession is pathological science. The nutrition aspect of it is just an interesting vehicle through

01:33:46.440 which to explore it. I realized that if I did a book then, I would go broke. I was married,

01:33:51.760 I had responsibilities. Remember, I had come out $40,000 in debt just doing the Cold Fusion book.

01:33:58.540 It was clear I didn't work fast and I didn't want to work fast and that I would not be able to get

01:34:03.760 an advance large enough to cover the time it would take to do the book. So I was living in New York.

01:34:09.180 I was having lunch once a month with an editor from the New York Times Magazine because,

01:34:14.960 among other things, we shared an affinity for the same local French cafe in the village.

01:34:19.500 And we would talk about story ideas and we decided it might be a good idea to see about,

01:34:25.340 to ask the question, what caused the obesity epidemic? And I said to this guy, when I was

01:34:29.340 reporting the dietary fat story for science, I had met up with an administrator from the NIH who said,

01:34:35.400 you know, it's interesting. When we told people to go on low-fat diets in 1984, we assumed we really

01:34:41.460 didn't have the evidence to support the heart disease connection. And the message of my story is

01:34:46.600 they never got that evidence. But we thought if nothing else, we'd be telling people to avoid

01:34:50.320 the densest calories in the diet. So if they avoided fat, they'd lose weight and that would

01:34:55.000 take care of the obesity and overweight are the greatest risk factors for heart disease.

01:34:59.940 And he said, lo and behold, now we have an obesity epidemic. And apparently people stopped

01:35:05.580 eating fat and eating more carbohydrates and that got them fatter. So I always had this two

01:35:10.240 hypotheses for what caused the obesity epidemic, which you can see in the data begins somewhere

01:35:15.780 between 1978 and 1991. And it coincides with two fundamental changes in the American diet. One was

01:35:22.660 the embracing of the idea that a low-fat diet is a heart-healthy diet. So carbohydrates in general go

01:35:28.660 from being considered inherently fattening, which is sort of the conventional wisdom up till the 1960s,

01:35:34.420 foods. And then they get transformed into heart-healthy diet foods. And you may be too

01:35:39.400 young to remember this, but we all stopped eating butter and started having pasta and bagels every

01:35:44.740 day. And lo and behold, everybody starts getting fatter. The other thing is high fructose corn syrup

01:35:49.760 came in. 1977, 78, high fructose corn syrup 55 comes in, which can replace sugar in Coca-Cola and Pepsi. And

01:35:59.140 by 1984, it saturates the beverage industry. And this coincides with the beginning of the epidemic.

01:36:04.980 And people like Michael Pollan had suggested that high fructose corn syrup was the cause of the epidemic.

01:36:10.540 So we decided I would write an article about what might have caused the obesity epidemic. It was new

01:36:17.260 enough then that people cared, and this would be an important story. And in the course of writing that

01:36:22.900 article, I came upon five studies that had been finished but not yet published, that had all been

01:36:29.280 discussed in conferences, so I could discuss them in the article, which had compared the Atkins diet,

01:36:34.720 which is a low-carbohydrate, high-fat, eat-as-much-as-you-want diet, to the kind of low-fat,

01:36:41.680 calorie-restricted diet the American Heart Association was pushing. And then all five trials,

01:36:47.120 the Atkins diet, not only did people lose more weight, but their heart disease risk factors

01:36:54.120 improved. So remember, I'm coming at this kind of from, I'm programmed to think from a scientific

01:37:00.240 perspective. I won't say that I'm coming to it as a scientist, because I know people don't like to

01:37:05.980 hear that from a journalist. There are two fundamental, there are two hypotheses out there.

01:37:10.480 One is that people get fat because they eat too much, and the other is that people get heart disease

01:37:14.800 because they eat high-fat foods. And now you run a diet trial where you compare a high-fat,

01:37:22.560 eat-as-much-as-you-want diet to a low-fat, calorie-restricted diet, and your two hypotheses

01:37:29.300 would predict that the high-fat diet, the Atkins diet, those patients would get fatter because they

01:37:35.380 can eat as much as they want, and clearly they got fat to begin with because they eat too much,

01:37:40.040 and that they would have worse heart disease risk factors. And in both cases, a hypothesis failed

01:37:46.220 to pan out. So from a scientific perspective, the first five clinical trials on low-carb,

01:37:52.540 high-fat diets, actually it was the second through sixth, there was one earlier one that saw the same

01:37:57.420 thing that no one discussed, refuted two of the fundamental hypotheses of modern nutrition science.

01:38:03.740 How did you find these? I mean, I know one of them in there is the Minnesota Heart Study, which

01:38:08.520 wasn't published until 1989, right? Yeah, no, this is Eric Westman's first trial at Duke. There was a

01:38:16.260 group at the VA hospital in Pennsylvania that was doing a trial. There was some researchers in Long

01:38:22.900 Island that did a trial. Why weren't they published? They just hadn't been published yet. They had been

01:38:28.940 finished, and the research we're talking about. Then once I did my article, it made it both more

01:38:35.560 important for them to get published and simultaneously harder for them to get published because nobody

01:38:40.240 wanted to hear. They wanted to see what these people really had, but not actually have to publish what

01:38:46.660 these people really had. But over the next two or three years, they all came out in the journals.

01:38:50.980 How did you sort of come to grips with this? This is now challenging your own beliefs. Presumably,

01:38:56.760 you had believed that this low-fat diet was healthier. Had you ever struggled with your

01:39:01.100 weight? Yeah. I was a college football player, so in college, I tried to be as heavy as humanly

01:39:06.580 possible and eating constantly. I don't know if you did this in college, but we would go out at

01:39:12.000 11 o'clock at night to Elsie's, a diner across the street from the dorm, and eat 1,000 calories or

01:39:19.020 1,500 calories worth of the biggest sandwiches they had, and then go home and go to bed. I mean,

01:39:23.780 that's what you, you know. Anyway, I could get up to 240. My boxing weight was 212. And once I turned

01:39:31.660 30 and simultaneously had moved to California, I just started drifting upward. While I was reporting

01:39:36.760 the science article, I was simultaneously doing a piece for Discover Magazine on the mathematics of

01:39:42.780 the stock market. And so I was up at MIT interviewing an economist who ran the Laboratory for Financial

01:39:48.300 Engineering up there about his research, you know, trying to establish whether people like Warren

01:39:52.840 Buffett are brilliant or just lucky, which is an interesting question because you're looking for

01:39:58.200 the signal of talent over the signal of luck, over the background of luck. And you don't, of course,

01:40:04.260 fully understand your background. So we got to talking about good science and bad science. I told

01:40:08.800 him about the dietary fat article I was writing, and he said, oh, well, if you're doing a story on fat,

01:40:13.800 you got to try Atkins. He said his collaborator, Warden's father lost 200 pounds on Atkins. And

01:40:19.960 this fellow at MIT is a Asian American. And he said he gave up, basically gave up white rice and lost 40

01:40:26.160 pounds. So I went back to LA where I was living. At that point, I was unmarried. My parents had passed

01:40:31.640 away. I had no children. If I killed myself on Atkins, the only one who might care were my cats.

01:40:37.480 And so I did Atkins as an experiment and lost 25 pounds in six weeks. And then like everyone else,

01:40:45.820 I sort of drifted off the diet by the time I started this New York Times Magazine piece.

01:40:50.220 If you go back to read the New York Times Magazine article, there's a line in it that says,

01:40:53.880 overweight, of course, is caused by taking in more calories than you expend, which today I think is

01:40:59.340 a meaningless statement and both wrong and meaningless. There were a lot of ways I had already,

01:41:04.560 because of my research on the dietary fat story, I could already accept the idea that these diets

01:41:10.120 would not raise the risk of heart disease. It was clear to me the dietary fat story was

01:41:15.280 uncompelling at best. There was no compelling evidence to avoid saturated fats. The question

01:41:22.440 about the fundamental cause of obesity and how best to regulate weight was something I didn't really

01:41:26.920 understand when I did that article. But that exceedingly controversial article on the front page of the

01:41:31.280 New York Times Magazine will indeed get you a large book advance. If the subject's one of

01:41:35.720 natural interest to readers, it did get me a large book advance. People like to say that that's the

01:41:41.060 only reason I came to the controversial conclusions I did, because it would sell more books.

01:41:46.160 They don't realize that a large advance in New York was enough to live for four years and the book

01:41:51.000 took me five years. That's when I started the research for good calories, bad calories,

01:41:56.800 and got obsessed with that story. And how did your thinking of the nuance around this topic

01:42:03.280 evolve during the writing of Good Calories, Bad Calories? I mean, it's still evolving. That's

01:42:06.900 what's so bizarre about this, is I still wake up at three in the morning thinking, why didn't I say

01:42:11.740 that 13 years ago? So there's these two issues, right? Which, I mean, I don't think we'll have time

01:42:15.980 to go into both of them. The one being the saturated fat issue, which is that dietary saturated fat

01:42:21.520 drives atherosclerosis. And the other being the cause of obesity, being calorie imbalance.

01:42:29.740 And you kind of go after both of these in parallel, but they're really different. They're not mutually

01:42:36.920 exclusive. Right. But they're related. They're very related. Yes. And that's the issue. You know,

01:42:41.700 the dietary heart story was we get heart disease because of the saturated fat content of the diet,

01:42:47.420 elevating LDL cholesterol. And here a guy like Atkins comes along and says, well, look,

01:42:51.240 don't eat carbohydrates. So if you want to minimize your risk of heart disease by the dietary fat

01:42:56.760 store, you avoid fats, saturated fats. You replace them with carbohydrates back then. Now they say

01:43:02.360 monounsaturated or unsaturated fats. And you're now eating a carbohydrate rich diet. And in theory,

01:43:09.620 you're going to minimize your risk of heart disease. The flip side was that carbohydrates

01:43:14.460 are fattening. That's the simplest way to describe what's now called the carbohydrate insulin model.

01:43:20.260 To those of us who gain weight easily, it's the carbohydrates that do it. And we can't eat carbs.

01:43:25.180 But if you don't eat carbs, what are you going to replace those with? And your choices are protein

01:43:29.700 and fat. And protein usually comes with fat attached in real foods. And so you're going to increase your

01:43:35.380 fat consumption one way or the other. And now you're eating a diet that in theory is supposed to kill

01:43:39.620 you. And that was the one of the many arguments against Atkins was it's going to trigger heart disease

01:43:44.920 and you're going to drop dead. When I first tried it as an experiment, I was kept waiting to drop

01:43:50.080 dead. And I'm still waiting. I mean, I still have a piece of bacon and pat of butter and I wait for

01:43:55.220 my heart to blow up. So one of the issues I had to deal with in the book is how did we come to believe

01:44:00.540 that dietary fat's the problem? So the first third of the book is the evolution of the, it's the soft

01:44:07.560 science of dietary fat at book length. How did this theory evolve? What was the real data? What were the

01:44:13.360 problems with the data? Why is it likely wrong? And then the second part of the book is as doing the

01:44:18.980 research, I realized that there had always been a competing hypothesis, which is that the chronic

01:44:23.760 diseases that associate with modern diets and lifestyles are driven by the carbohydrate content

01:44:28.440 of the diet. So this started as a British hypothesis, a British nutritionist, and then it should have

01:44:34.580 embraced in the 1960s, the research by people like Jerry Riven and others on insulin resistance. And all of

01:44:41.360 that targeted the quality and the type of carbohydrates we're eating, the glycemic index and the fructose

01:44:47.940 kind. And I didn't know anything about that when I started this research. And so the first third of

01:44:53.040 good calories, bad calories is the deconstruction of the fat hypothesis. The second third is the

01:44:59.340 replacement with a carbohydrate centric hypothesis. And then the third third of the second half of the

01:45:05.780 book actually is this question of obesity, because it's clear that whatever causes obesity is so closely

01:45:10.800 associated our disease and diabetes, type 2 diabetes, that whatever makes us fat also causes these

01:45:16.720 chronic diseases that associate with us. And the question is, what makes us fat? And now I, doing the

01:45:23.500 research, I also came upon a whole other stream of research that had kind of been ignored. First, by

01:45:30.400 pre-World War II European researchers saying obesity is not an energy balance problem. It's not caused by

01:45:37.120 people eating too much. It's a hormonal regulatory disorder. And then post-World War II, by people

01:45:42.740 studying primarily the science of hunger, saying, in effect, hunger is caused by fuel availability.

01:45:49.360 If your liver thinks there's enough fuel available, then you're not going to be hungry. And when your

01:45:53.360 liver starts thinking fuel is being crimped, then it'll release inhibitions on food-seeking behavior.

01:45:59.040 And that work done by people in this field of physiological psychology all implicated the

01:46:05.360 insulin as the hormone determining fuel availability, the primary hormone. And then there was also,

01:46:11.600 and this I think is what gets me most. Until 1930, virtually no one studied fat tissue. In all the

01:46:18.320 sort of burgeoning fields of physiology and endocrinology, nobody really cared. The assumption was it was inert,

01:46:24.000 it was for padding and cushioning. And if you starve people, somehow the fat tissue would shrink. They

01:46:29.840 had no way to study fat tissue or fat accumulation. And then beginning in 1933 with this work of this

01:46:35.520 German émigré Schoenheimer at Columbia, you start having people studying fat tissue. They become aware

01:46:42.480 that mobilization, deposition goes into fat, goes on constantly. Like even when you're in between meals,

01:46:50.000 you're starving, your body is still depositing calories as fat and mobilizing calories as fat.

01:46:56.320 And through the 1940s and 50s, they begin to work out the details as more and more tools become

01:47:01.920 available to study this problem. And by the mid 1960s, you have very well established science of

01:47:09.040 what we could call intermediary metabolism, which is what your body does with the foods,

01:47:14.080 the proteins, fats and carbohydrates after you eat them. And then what your body continues to do to

01:47:20.160 make fuel available as necessary to your body. And all of this is left out of the science of obesity.

01:47:26.880 As soon as the obesity researchers start saying, we've demonstrated that obesity is caused by energy

01:47:32.880 balance, imbalance, the impolitical way of saying that is people eat too much. That's a fact of life.

01:47:39.360 We can see it in our animal models, because when we create an obese animal model, it tends to be

01:47:44.880 what they call hyperphagic, which is very hungry. So we have an association between hunger and obesity,

01:47:50.320 and we assume causality. And they just kind of ignore the entire science of fat metabolism.

01:47:57.280 And even to this day, when I try to make that point, it's excruciating. Nobody wants to hear it.

01:48:04.400 Nobody wants to deal with it. And, you know, if I had to prove it to an editor, I could. Ideally,

01:48:13.040 I would have the fact checker fly out to my office at Oakland and say, let's go through these textbooks

01:48:17.680 from 1965 onward. And I'll show you the complete and utter absence of any discussion of the science of

01:48:25.600 fat metabolism, fat storage from the obesity discussion. And from the 1960s onward,

01:48:31.360 the obesity researchers are dominant. Well, obesity research in the 1960s is dominated by

01:48:36.480 psychiatrists and psychologists trying to get fat people to eat less. People like Mickey Stunkard,

01:48:41.360 who was a wonderful man, but his expertise was psychology. And this idea that obesity is a hormonal

01:48:47.920 dysregulation of fat storage, fatty acid oxidation and storage, a fuel partitioning disorder gets left out

01:48:54.960 of the field. Who do you most credit for that changing, that new course in the 1930s thereabout?

01:49:02.560 Where did this idea that is the dominant point of view with respect to obesity being caused by caloric

01:49:09.360 imbalance rather than being the result? Because there has to be luck sometimes, right? Like sometimes

01:49:13.920 there's a dominant personality wins out or something like that. You think of a field like diabetes,

01:49:19.280 where Elliot Jocelyn opens the first diabetes clinic in the U.S. dedicated. Diabetes clinic in

01:49:25.280 the U.S. around 1900. And by 1916, he's seen a thousand patients, which is about 950 more diabetic

01:49:32.160 patients than any other doctor in America. So he writes the first textbook in 1916 and he does another

01:49:38.480 edition in 1917 and 1923 and 28 and 33. And we're now on, I think, edition 14 of Jocelyn's diabetes

01:49:46.400 mellitus. Jocelyn's became the determinant of truth, the arbitrator of truth in the diabetes

01:49:53.840 world. One researcher, one physician, a wonderful physician, not a great scientist. Because of his

01:49:59.840 position in the field, this doesn't happen in other, I don't know if it happened, it didn't happen in

01:50:04.080 physics because there are too many bright people arguing about the data and not enough people outside

01:50:09.360 care. But in medicine, you get fields in which single individuals, like the famous breast cancer

01:50:15.040 surgeon from Hopkins. Halstead. Yeah. It's maybe a kind of unique phenomena in science.

01:50:22.160 In obesity research, there was no such thing as obesity research pre-1930. If when you read the

01:50:28.480 literature and you go back and look at the literature, there's half a dozen physicians

01:50:32.320 around the world who are writing articles about obesity and often with some statement of causality.

01:50:37.920 And there's kind of two competing hypotheses. One is fat people just eat too much. And you know that's

01:50:43.200 true because you have an association between gluttony and obesity, right? They watch Shakespeare. We know

01:50:48.400 Falstaff was obese. He had this zest for living. He ate people out of house and home. So therefore,

01:50:54.640 you assume causality from an association. That's classic pathological science. And the other is that

01:51:00.640 people who suffer from obesity have some kind of hormonal, some kind of dysregulation where they're

01:51:06.080 going to get fat regardless of how much they eat. So 1930, Lewis Newberg comes along. He's a physician

01:51:13.040 at the University of Michigan. He does what he considers the first experimental test of the

01:51:20.320 hormonal regulatory disorder hypothesis that you get fat no matter how much you eat or how little.

01:51:26.000 He runs the tests on six or seven people. He discusses it at a medical meeting in Boston. So this

01:51:33.040 guy is now, we'll call him, if not the Newton of the field, the Eddington of modern, you know,

01:51:40.320 this is the guy's name that everyone, if obesity research was a functioning science, every obesity

01:51:45.360 researcher would be insulted if they, you thought they didn't know who Lewis Newberg was. He publishes

01:51:50.720 a series of papers in 1930 based on 3031, based on the one experiment with like seven patients where he

01:51:57.360 basically starves lean people and patients with obesity and says they lose weight at roughly the

01:52:03.760 same rate when starved, therefore the obese people, and I'm not making this up, got fat because they

01:52:11.440 ate too much. He saw this as refuting the conventional hormonal regulatory hypothesis. Nobody's ever done an

01:52:19.680 experiment in obesity before. So Newberg's papers begin to be taken as gospel. And even though there

01:52:27.920 are European researchers, particularly Julius Bauer, who's one of the pioneers of the science of

01:52:32.800 endocrinology at the University of Vienna in Austria, which the Germans and Austrians at this period of

01:52:38.400 time are the sort of apex of the height of medical science, he's writing that this is naive and can't

01:52:44.320 possibly be right based on a series of what I think are very logical and undeniable observations.

01:52:50.480 Even back in 1920, as soon as Newberg publishes his article, Bauer coauthors an article, the first

01:52:56.160 thing he ever writes in English explaining why Newberg's argument is inane, this idea that people

01:53:01.680 get fat because they take in more calories than they expend. 1938, the first animal model of obesity

01:53:07.600 is pioneered at the laboratory of a fellow named Ranson, the leading neuroanatomist of the era.

01:53:14.320 And by his graduate student, Albert Hetherington. And they take what's called a stereotaxic instrument

01:53:20.240 where you could use it to direct the needle into the brain of an animal. And it's been used on dogs,

01:53:26.240 Hetherington pioneers, so he could use it on rats. And he can reproducibly create obesity if they

01:53:32.240 lesion the ventromedial hypothalamus of the rat. Hetherington has a, he's a postdoc. There's a fellow

01:53:37.840 postdoc at the lab named John Brobeck, who 1939 gets accepted at Yale for medical school.

01:53:44.160 He's got his doctorate. He goes back to New Haven. He needs to get a job to help finance medical

01:53:50.160 school. He gets a job in a laboratory and he realizes it's a stereotaxic instrument and he

01:53:55.360 can do the same experiments that Hetherington did. And Brobeck starts doing the VMH lesion experiments

01:54:01.120 at Yale. And Brobeck comes, creates what he believes he discovers, what he calls hyperphagic obesity.

01:54:07.680 So you lesion the VMH of the rat, the rat gets obese and it gets crazy hungry. And so you now have

01:54:14.880 the association between hunger and obesity. I've read that the rats got so hungry that even as you're

01:54:22.160 taking the anesthesia off them, they're basically trying to eat as they're coming out of anesthesia.

01:54:27.200 I interviewed recently a researcher who did these experiments in the seventies. He compared it to

01:54:32.000 gasping for breath. So like if you are drowning and you suddenly come up above the water and you're

01:54:39.440 the animals are eating like that. And it's a fascinating observation because you can explain

01:54:44.640 it pretty easily. And in fact, if you don't allow the anesthesia to wear off, if you allow the animal

01:54:50.240 to eat while it's still under the influence of the anesthesia, it might choke to death. And they often

01:54:54.560 did. And so the researchers had to learn to give the animal a couple hours despite its

01:54:59.280 evident starvation. Brobeck has read Newberg. Okay. So Brobeck interprets this experiments as

01:55:08.640 I created hyperphagia, extreme hunger in the animals. That's the discovery. And the animal got

01:55:14.000 obese. The fact that the animal got obese because it was hyperphagic is just boring. Animals get fat

01:55:19.600 because they eat too much. I know that because that's what Newberg says. Back in Chicago at Ransom's,

01:55:25.440 let's pause for one sec because that's a very important point. I think this gets to something

01:55:30.480 you've already talked about twice, but I want to make sure the reader doesn't miss the subtlety

01:55:34.480 of this or the listener rather. The observation is undeniable. If you lesion the part of the brain

01:55:42.080 in question, the animal eats in an uncontrolled manner. However, if your incoming hypothesis is that

01:55:50.320 overeating leads to obesity, you will interpret it as lesioning that part of the brain leads to

01:55:56.960 overeating. If your hypothesis is that fat accumulation, dysregulated fat accumulation leads

01:56:04.080 to obesity. You would interpret that finding as lesioning that part of the brain leads to dysregulated

01:56:10.560 fat accumulation, which then causes overeating. That's a very subtle difference. I try to be

01:56:18.080 fair to the people of the era to see how easy it could be to make that mistake.

01:56:23.120 Well, and now let's put it in the context of, remember the cold fusion, what's your control?

01:56:28.000 So if your hypothesis is that eating too much causes obesity in these animals when they're

01:56:33.840 lesion, the proper control is control for the eating too much. So what I was going to say back

01:56:40.960 in Chicago, Hetherington and Ransom had animals that got obese independent of the hyperphagia. So some of

01:56:47.200 their animals got fat anyway. Actually, some of Brobeck's animals also got fat despite not eating

01:56:54.240 any more than lean animals. Did he ever pair feed his animals with non-lesioned animals?

01:56:59.600 Brobeck did. He did an experiment where he pair fed them with non-lesioned animals. And if I remember

01:57:06.080 correctly, it was a dozen animals, nine of them did not get fatter, but three of them did despite

01:57:12.240 being pair fed. So Brobeck, because he had a preexisting hypothesis, assumed that there must

01:57:17.760 have been something wrong with these three animals, or he did the lesion incorrectly, or maybe they were

01:57:22.400 eating at different times. So he left them out of the analysis. He mentions it in their paper,

01:57:27.600 but they're the counter argument. They're the evidence for this counter hypothesis, which something

01:57:33.040 about the lesion dysregulates fat accumulation. What do you think explains the other nine?

01:57:37.360 You can imagine that if you dysregulate the fat, so let's say we take an animal that normally would

01:57:42.800 grow naturally and we create a dysregulation in their fat tissue such that it's going to store

01:57:48.800 excess fat. It might not be able to do that if you restrict its calories. It's that simple. We all know

01:57:55.920 from personal experience, we can lose weight by starving ourselves. So we can somehow starve our

01:58:00.720 predisposition to get fatter by restricting the amount of food we eat. The counter argument to

01:58:06.400 the hyperphagia causes obesity. Evidence is obesity being caused in the absence of hyperphagia. So what

01:58:14.320 Ransom and Hetherington argued, Ransom was the leading neuroanatomist of the day. He had just come

01:58:20.080 from studying diabetes insipidus. And in diabetes insipidus, you can also cause it by lesioning a

01:58:25.280 different part of the hypothalamus. You have animals that are extremely thirsty. I forget the

01:58:30.160 technical term. Yeah. So like any diet, they have endless thirst and they're peeing constantly.

01:58:35.200 So the conventional thinking might be the reason they're peeing so much is because they're so

01:58:39.360 thirsty. And the counter hypothesis is that the thirst is a response. The lesion causes them to

01:58:48.320 pee constantly. They're losing body water. That makes them thirsty. That's become unequivocally the

01:58:53.680 case. We now know that diabetes insipidus is the result of the loss of the hormone called

01:58:59.280 antidiuretic hormone or DDAVP. The hyperdrinking is in response to replenishing the water that they're

01:59:05.680 losing because they can't concentrate urine. Exactly. So Ransom thinks, well, if they're

01:59:10.960 losing, if you're losing water and drinking in response, maybe in these cases, the animal is losing

01:59:16.720 calories, energy into its fat tissue. And they're eating in response. The hyperphagia is a response

01:59:23.040 to the loss of energy. And some of his animals, the one that got obese are also sedentary. So he said,

01:59:29.120 maybe the ones that got obese without hyperphagia were sedentary. He said, maybe the reason they're

01:59:34.560 sedentary is because again, they're losing calories into the fat tissue. The fat tissue is creating a sort

01:59:41.520 of vacuum. And the animal is responding either by eating more or exercising less. So changes in energy

01:59:47.920 balance are a response to the fat tissue being driven to accumulate calories or fat calories. 0.61

01:59:56.320 They published this paper. Three months later, this is 1942, Ransom has a heart attack and dies.

02:00:02.720 He was eating too much fat.

02:00:04.240 Clearly. Hetherington as postdoc, this is the middle of World War II, joins the Air Force

02:00:09.920 and leaves the research behind. And the only voice left in the field is John Brobeck. And so Brobeck

02:00:16.480 dismisses their hypothesis, not really even understanding it. It's clear from the literature.

02:00:22.080 He still thinks they're trying to blame obesity on sedentary behavior because he's trapped in this

02:00:27.760 energy balance thinking. He doesn't realize that they're trying to blame both the sedentary behavior

02:00:33.040 and the hyperphagia on the drive to accumulate fat, which is the fundamental different paradigm.

02:00:39.520 So Brobeck continues to write about hyperphagia causing obesity. By the 1950s,

02:00:46.160 between Brobeck's work and Newberg's work, the conventional wisdom is all fat people get fat

02:00:52.480 because they eat too much. And there are textbooks that have a statement to that effect in virtually

02:00:57.840 those identical words. By the 1960s, obesity, as I said, is being the leading figures in the field are

02:01:04.560 psychologists and psychiatrists who are trying to explain why fat people eat too much and how to

02:01:10.480 stop them from eating. And there's a whole world of physiologists and biochemists studying fat metabolism,

02:01:16.960 learning about all the ways it's regulated, learning how its deposition and mobilization

02:01:21.920 of fat go on independent of the nutritional state of the organism. That's a direct quote from one of the

02:01:27.520 seminal papers. And that what you should be studying is not how much people eat and exercise, but why

02:01:32.880 they accumulate so much fat. What's the best explanation for why a lesion to the VMH, which is

02:01:39.600 a very central, very, very specific, narrow part of the brain, why would that have such a broad

02:01:47.440 peripheral consequence of fat hoarding within all of the adipose tissue throughout the body? What is

02:01:54.640 the mechanistic or best mechanistic explanation for that in that model? Well, in my model, first

02:02:01.600 observable effect from the VMH lesion is hyperinsulinemia. So you lesion the brain and the

02:02:07.280 animal hyper secretes insulin in response to even thinking about food. Now, if you think about it,

02:02:12.880 an animal that's hyper secreting insulin isn't going to be able to. So the insulin is signaling

02:02:17.920 its fat tissue to take up fat and to store it for food. It's in storage for fuel. It's inhibiting

02:02:24.320 the process of lipolysis. It's also inhibiting the oxidation of fatty acids and the muscle tissue

02:02:32.080 through the malonyl-CoA pathway. So think about what happens now to these rats or people with tumors.

02:02:37.920 You have the surgery. You hypersecrete insulin. You wake up from the surgery without carbohydrate

02:02:45.040 supplies and unable to oxidize dietary fat. So that would explain the hunger. It's as though you've

02:02:53.760 created a starvation state in the animal instantaneously. You know, normally it's going

02:02:58.720 to take a while for the animal to get to the point where it has to start cannibalizing its own

02:03:04.560 muscle for fuel if you're starving it. But now you've done it instantaneously with the lesion.

02:03:10.400 And that would explain the gasping for food that these animals manifest in the hyperphagia.

02:03:16.640 What's the relative increase in insulin that those animals experienced?

02:03:20.800 You know, I don't have those numbers.

02:03:23.200 That's an important thing, right? I mean, if that's the conduit through which a central

02:03:26.960 lesion has such a broad peripheral implication, I would expect an enormous increase in insulin to

02:03:34.240 produce that. A level that wouldn't be otherwise physiologically described. I guess I'm struggling

02:03:39.280 with how that alone could do it. Well, also, if you think about it,

02:03:42.800 until 1960, they had no way to measure insulin in the bloodstream.

02:03:46.560 Right, but we could do it today, is what I'm saying.

02:03:48.320 Yeah, I just don't know the number. I mean, it was clearly enormous. And it was observed. And post-1960,

02:03:54.320 you get yellow and Burson's radium amino assay. So now you could actually measure insulin levels

02:03:59.040 from blood samples accurately and with sensitivity. And pretty quickly afterwards, that's when they

02:04:05.200 realized that the BMH lesion animals were hyperinsulinemic. How did the OB-OB mice also give us

02:04:11.040 a clue into both ways to interpret the same observation? Yeah, so the OB-OB mouse is a mouse

02:04:16.480 that's discovered at the Jackson Laboratories in Maine. It's a mutant that manifests obesity,

02:04:22.560 dramatic obesity. And it turns out there's two different, depending on what strain you breed

02:04:27.120 these animals into. The DB-DB mouse is diabetic, obese and diabetic. The OB-OB mouse is just obese.

02:04:35.760 The work is done at the Jackson Laboratory. What's interesting is the assumption is always right that

02:04:41.840 these animals get fat, that they're missing some kind of satiety hormone. The theories through the

02:04:48.160 1950s, based on Newberg's work and this idea that obese people get fat, not because they

02:04:53.520 partition fuel preferentially into fat tissue, but because they take in more energy than they expend.

02:05:00.240 The dominant, the only hypothesis of obesity in the 1950s, there's a lipostat hypothesis from a guy

02:05:07.040 named Kennedy and a glucostat hypothesis from Jean Maier. There's a thermostat hypothesis with body heat.

02:05:13.680 They're all hypotheses trying to explain eating too much in the animal. And none of these pan out.

02:05:19.440 Not only do they not pan out, you've got these animals that get obese. If you think they're caused

02:05:24.400 by eating too much, again, what's the correct control? Just pair feed the animal. Don't let the

02:05:29.360 animal eat more than a lean animal. So this experiment is done numerous times with leptin, the OB-OB mice.

02:05:38.080 If you take a lean animal, measure how much it's eating, give half of that to the OB-OB mouse. So 0.95

02:05:45.680 you're literally semi-starving it. It gets obese anyway. One of the fellows who published these

02:05:50.800 results, Douglas Coleman, but Jeff Friedman gets credit for doing the work that identified leptin

02:05:56.160 as a hormone that's missing in the OB-OB animal or dysfunctional. And he did it based on Coleman's

02:06:02.000 parabiosis experiments at Jackson Lab with the OB-OB and DB-DB animals. So Jeff Friedman comes along,

02:06:09.200 the young, ambitious researcher. He wants to look for the obesity hormone, the satiety hormone that's

02:06:16.480 missing in the OB-OB mice. He comes with this assumption that what's missing is a signal from

02:06:22.080 the fat tissue to the brain telling the brain not to eat too much. When he discovers leptin,

02:06:28.160 he interprets that signal as being the satiety hormone. It's been identified as the satiety

02:06:36.560 hormone ever since. When people write books with titles like the hungry brain, they're assuming that

02:06:42.560 the absence of leptin makes the brain hungry. The person gets hungry, eats too much, and that's why

02:06:48.800 they get fat. And yet, you always have this observation in the field that the animal gets fat

02:06:54.400 even when it's half-starved. The point is, I'm doing this book on diabetes now. I was curious,

02:06:59.520 what is it about the DB-DB animal? Why would the absence of a leptin receptor cause diabetes?

02:07:06.480 And I never really thought about it. So you look into the literature. We actually asked Rudy

02:07:10.800 Libell and Rudy said, well, basically it depends what background you breed the animal onto. And then I

02:07:16.480 went back and I found where Douglas Coleman discussed it. And so the issue is both animals are obese.

02:07:22.240 Both animals are hyperinsulinemic from weaning onward. So in the DB-DB animals, the background

02:07:29.840 strain can't sustain the hyperinsulinemia. So its pancreas fails and you manifest frank diabetes.

02:07:35.520 The OBOB strain can continue to keep pumping out the necessary insulin. But now you've got leptin

02:07:40.800 basically triggering hyperinsulinemia. And it's always been known these animals are hyperinsulinemic.

02:07:46.080 And now you're back to the Bauer-Ranson hypothesis. I didn't say, by the way,

02:07:50.880 that when Ranson and Hetherington interpreted the data the way they did, they cited Julius Bauer.

02:07:57.360 So Brobeck cites Newberg. He's got a preconceived opinion. And Ranson had his own preconceived opinion

02:08:03.120 based on the diabetes insipidus work. And Julius Bauer's paper supported that. But we grew up with

02:08:08.960 one hypothesis, which is the Brobeck eating too much hypothesis. And it influenced how the leptin work

02:08:15.840 was interpreted. It's influenced virtually every experiment afterwards. And then some body like

02:08:21.680 me comes along and says, well, wait a minute, all this evidence supports the alternative hypothesis.

02:08:27.040 And you have not been considering that for 60, 70 years. How can you even trust anything you've done

02:08:33.440 when you weren't aware that there was a competing hypothesis? It's an awkward place to be in.

02:08:38.160 Now, your work has brought a lot of attention to this alternative hypothesis over the past decade,

02:08:42.560 and there have been a number of people who have tried to test it. How would you reconcile

02:08:48.080 the findings that have not demonstrated what would be predicted by the carbohydrate insulin model?

02:08:54.400 This then gets into the kind of questions we talked about earlier about judging the value of the

02:08:59.440 experiments, how rigorously they're done, and the biases and preconceptions of the researchers.

02:09:05.680 So, and again, I've been having these arguments on Twitter this week, and I have to stay the hell

02:09:09.840 off Twitter. And you know this, we funded these people at NUSI. We funded two groups of researchers,

02:09:14.720 one who was inherently basically had a hypothesis that dietary fat and fat balance was a driver of

02:09:22.320 obesity. And we had one that had a carbohydrate insulin model like we did. And the researchers who

02:09:28.480 believe the conventional wisdom interpreted their results as supporting the conventional wisdom and

02:09:34.240 refuting the carbohydrate insulin model. And the researchers who believe the carbohydrate insulin

02:09:39.760 model interpreted their results as supporting that model.

02:09:42.960 So let's pause there for a second, because I simply don't see how this field can make any progress.

02:09:49.040 And I'm trying to understand how this field can make progress in a way that physics does.

02:09:53.440 When, as you point out, everybody in this field is biased, everybody has a point of view,

02:09:59.200 and everyone seems to do experiments that simply confirm their point of view.

02:10:04.080 There's one other thing that makes this more difficult, which is the inherent

02:10:07.040 messiness of biology. As complicated as physics is, I think biology is more complicated. We don't

02:10:12.160 have a standard model of biology the way we have a standard model of physics. So there are more

02:10:16.880 unknown unknowns in biology. And then secondly, the experiments are far more difficult to control.

02:10:22.320 I think there's more noise in the biologic experiments. And then you couple that with

02:10:27.440 everything we just said. I just don't understand what a path forward looks like towards a reconciliation.

02:10:33.120 I don't know if reconciliation is a word you're looking for. You want to know which hypothesis is

02:10:37.120 right. Yeah. I mean, a scientific reconciliation. We're recapitulating the discussions we had,

02:10:41.680 whatever, eight, 10 years ago now, when Lucy was starting. One of my favorite stories is back around

02:10:47.840 2009. After good calories, bad calories came out, I was invited to lecture at the Pennington

02:10:52.880 Biomedical Research Center, which is the largest obesity research center in the country. And I gave

02:10:57.680 my lecture, Why We Get Fat, Adiposity 101. And I suggested that the energy balance hypothesis

02:11:04.640 thinking was, I like to say, not even wrong, stealing from Wolfgang Pauli. And then why all the reasons why

02:11:10.720 it should be replaced with this hormonal regulatory disorder, focusing on insulin. And after my talk,

02:11:16.320 one of the faculty raised his hand very politely, a gentleman who was probably then as old as I am

02:11:22.240 now, maybe 65. And he said, excuse me, Mr. Taubes, would it be correct to assume that you think we

02:11:29.360 are all idiots? Because the argument I was making is that they embrace the wrong paradigm. And it was 1.00

02:11:36.240 a tough question, because partially the answer, I want to say, well, that's one way to look at it. I

02:11:40.640 can't say that. So what I said to them is, I think the problem here is that when you entered the field,

02:11:45.760 there was a paradigm, a way of thinking about obesity that seemed so intuitively obvious,

02:11:50.240 this idea that it's an energy balance disorder, that you never questioned it. And certainly your

02:11:55.120 mentors didn't question it. And so you were assumed this had been well-tested and well-proven

02:12:01.280 and unambiguous and that it deserved to be dogma. And it hadn't been, and it didn't. And that's been a

02:12:06.960 problem ever since. Now we come along 50 years later, 40 years later, and we have to get people to

02:12:13.360 a huge proportion of the community to entertain the possibility that their fundamental belief

02:12:18.240 system is wrong. So this isn't a subtle shift in thinking. This is, you're operating under the

02:12:23.680 wrong paradigm. You think the earth is flat and it's round, or you think the sun rotates around

02:12:28.640 the earth, but it's the other way around. 99.9999, thrown as many nines as you want,

02:12:34.480 percent of the time people say that they're quacks.

02:12:36.800 So why are you not?

02:12:38.080 Uh, that's a good question. I often argue this with my old friend, our old colleague,

02:12:44.080 Mark Friedman. And Mark says, well, we're not because we're right. And I say, well,

02:12:47.840 every quack thinks they're right, Mark. That's not evidence that we're not quacks. 0.98

02:12:52.480 That's the same tautology that you argue against.

02:12:55.840 Yeah. Well, the other argument is an even better one. I'm not a doctor. So technically,

02:12:59.680 I can't be a quack. The best I could hope for is whack job.

02:13:02.960 I would offer a more compelling reason if you turn out to not be a quack, which is the 0.98

02:13:08.480 application of the current model is failing. That's probably the most compelling reason to

02:13:14.080 continue to question it, I would say. So in other words, if we believed that the earth was the center

02:13:20.640 of the universe and every time we tried to launch a rocket into space, it blew up because we failed

02:13:27.280 to understand gravity and orbits, I would hope we would then say, God, what if the earth is actually

02:13:34.400 moving even though it doesn't feel like it? That's where all these discussions start is failure to

02:13:39.440 prevent obesity, failure to treat obesity. The obesity and diabetes epidemics are out of control.

02:13:44.720 Let us question these fundamental hypotheses. How do we test them? I'm thinking at the moment,

02:13:51.440 having done that, and move to the next stage where we now have competing researchers arguing. One says

02:13:57.840 I'm right. The other says I'm right. I side with the one who thinks like I do. How do we get people to

02:14:03.520 care? How do we move this? It's still this issue. How do you get to move it outside the diet issue,

02:14:08.560 as we discussed over the years? The conventional wisdom comes with a lot of explanations for why

02:14:13.920 people don't lose weight on various dietary interventions, and some of those are true. You know,

02:14:19.280 people don't comply with the interventions. They're like, dude, nobody sticks with the diet. So

02:14:24.160 maybe everybody's different. And so maybe we have to invest huge sums into individualized nutritional

02:14:31.680 therapy. Precision nutrition. Yeah. That's the way the NIH is going. So the idea is, well,

02:14:37.920 these people are arguing it's carbohydrates. Yeah, we don't believe that. So we're not going to spend

02:14:42.640 any money researching that. There are people believing it's fat. There are people who think it's meat.

02:14:47.720 In an ideal world, and we discussed this for eight years, right, the federal government would say,

02:14:52.520 we, COVID is going to pass, and obesity and diabetes are going to go back to being the long

02:14:57.480 pole, as using NASA terminology. They're costing a billion dollars a year in direct medical costs. 0.54

02:15:04.120 We have to understand how we failed, because we have to stop blaming it on industry and blaming it on the

02:15:10.120 individuals. And so let us put together teams of researchers, red teams, blue teams,

02:15:17.320 multiple colors, and hash this out and try and figure out how we failed to control these epidemics.

02:15:22.760 And until we do that, we're not going to reject any reasonable hypothesis.

02:15:26.520 But how would you even populate said teams? I mean, to your point, there really aren't that many people

02:15:31.720 people that would still stand by this idea that you propose.

02:15:37.400 I mean, if we're going to get practical here, Peter, we're not going to make any progress whatsoever.

02:15:42.120 Dreams, it's like populating a jury in a trial. You pick people, unbiased scientists who are good

02:15:47.960 at what they do, have demonstrated that they're good at what they do. You can't pick obesity

02:15:52.120 researchers because they have biases. You certainly can't pick nutritionists because they have biases.

02:15:58.040 I question whether any of them are ever taught to think critically enough about scientific progress.

02:16:03.240 I mean, think about the kind of effort that went into COVID. The amount of money that was

02:16:07.400 spent on research to prevent this disease from killing, what, one-tenth of the number of people

02:16:14.600 who die from chronic diseases related to obesity and diabetes every year. I mean, I'm not,

02:16:19.960 I'm glad they did it. But you make that kind of effort, you can solve this problem. Step one is

02:16:25.560 saying, look, we failed. The conventional wisdom fails. It's clear it fails because we have obesity

02:16:31.960 and diabetes epidemics that haven't been stemmed in any way. We have to question our assumptions.

02:16:37.000 Who actually does that job? There's a lot of very intelligent, critical thinkers out there.

02:16:41.240 The question is, how do you get them to care? Now, some of them would argue that,

02:16:45.560 hey, we know the answer and the palatability of food, the availability of food,

02:16:50.600 the affordability of food are driving this equation. And the convenience of modern life

02:16:56.360 is making inactivity even a greater and greater issue. And so you might argue, well, look, we

02:17:02.360 haven't put all the steps in place to address those issues. The answer is going to be one through policy.

02:17:08.600 We have to make foods that don't taste as good. You know what I mean?

02:17:11.960 Yeah, yeah.

02:17:12.440 I'll give you a funny story. My son came home the other day and he said,

02:17:16.600 mommy, one of the kids at school today had something for school and I really want it too.

02:17:21.080 And she was like, okay, what is it? He's like, I don't know. It came in a blue bag and it was a

02:17:25.160 triangle. And of course it was Cool Ranch Doritos. So what does Jill go and do? She goes and buys Cool 1.00

02:17:30.920 Ranch Doritos for Reese for lunch, for a little snack. And you know, there's a kid who really likes to eat

02:17:36.200 good food. And so every day he gets a little Ziploc with like five Cool Ranch Doritos. The problem is,

02:17:42.120 I can't eat just five Cool Ranch Doritos. So the other day I'm inhaling half a bag of Cool Ranch

02:17:48.520 Doritos. And all I'm thinking is this is almost as impressive as the Apollo 11 program from an

02:17:54.360 engineering perspective, like the way they made this thing, the crunch, the taste, the lingering

02:18:01.320 flavor. It's unbelievable. Like there's no denying how good this stuff is. So why did I just sit there

02:18:07.080 and eat a whole bag of that thing? You know, that's a very good question, but here's the counter

02:18:11.960 argument. I gave a talk a couple months ago in Tahoe and it was just as the ability to give

02:18:18.440 talks in person was winding down. And the fellow invited me to give the talk is from Texas. Afterwards,

02:18:24.200 sent me a gift box of Wagyu steaks from a butcher in New York. And these are the Wagyus that cost,

02:18:32.200 it's about $170 a pound. One of them is called Wagyu sashimi. You're just looking at it,

02:18:38.200 you know, it's by weight as much fat as protein and by calories and farm, it's going to be 80%

02:18:45.160 fat. And I actually made one of these for lunch and I looked up how to do it. So you get the skillet

02:18:51.880 and you use the fat on the edge to grease the skillet and hot. And then it's a minute on each

02:18:58.120 side and it was eight ounces and I couldn't finish it. It was so filling. My son, Harry said,

02:19:05.160 he got a taste. He said, this is beef butter, dad. I still finished it. So it was delicious. However,

02:19:12.760 even if I could afford it, I still have some in the freezer. I didn't, it wasn't like a Dorito issue.

02:19:18.760 So yeah, it's not quite as accessible, but the question would then be, did I get fatter? Did I

02:19:24.120 somehow dysregulate my fat tissue by eating that the way you might be doing eating the Dorito? Is

02:19:30.440 it the palability that drives your particular hyperphagia or is it the peripheral response to

02:19:37.160 the macronutrient composition that then drives your hyperphagia when it comes to Doritos? I mean,

02:19:43.880 I just think if you look at the response time, it has to be more central than peripheral. It's

02:19:49.000 barely exiting my esophagus while I'm inhaling these things. But still the response is going to

02:19:55.560 be a response to an expected peripheral change. So, you know, your body is evolution and almost

02:20:02.520 homeostasis are very good at what they do. They know what's about to happen. So they're going to

02:20:06.600 prepare your body for that to happen. Think about your experience when you were 205 pounds and swimming,

02:20:15.400 working out three hours a day and eating as healthy as any human being could eat.

02:20:19.720 The question is, again, why then? And what's changed between then and now? And that wasn't

02:20:26.440 a Dorito-induced effect. You have this idea that some people are just going to get fat

02:20:30.680 are pre-programmed. The quote is from George Bernard Shaw. One of the leads said,

02:20:36.280 we're just going to get fat no matter how much we eat. So if that's the case, what triggers that?

02:20:40.520 The Doritos, I would say trigger it, whereas the Wagyu, which is far more calories, does not.

02:20:45.480 How do we get people to care? How do you get people right or wrong? There's a lot of people.

02:20:49.880 Clearly, the conventional wisdom is what you just argued. We know what causes it. It's too

02:20:54.920 much food available. It's very easy to blame industry. You know, Michael Moss has written

02:20:59.640 two terrific books that get a lot of publicity because he's blaming industry. He's not blaming the

02:21:04.600 scientists. He's not blaming the administrators of the government. He's blaming the industry.

02:21:09.640 It's very easy. When I was blaming the industry, people love what I do. As soon as I shift to

02:21:13.960 blaming the scientific community for doing unacceptable science, then I have trouble

02:21:18.920 getting the message heard. I don't know what the answer is. You know, it's something I think about

02:21:22.760 every day. And when people are programmed to assume you're wrong, how do you get them to accept the

02:21:29.080 possibility that you're not and that somebody has to study this? And I don't know what the answer is.

02:21:35.560 I do know it's the challenge. I said this before during the NUSI years. It's like you decide you're

02:21:41.560 going to tilt at windmills for life. You've got to get used to the fact that the windmills are going

02:21:46.600 to kick you in the ass when you ride by. I don't know. I think people do care, Gary. I mean, 0.99

02:21:51.160 I guess it depends how you define care. People certainly care about obesity. Are you really

02:21:55.640 asking the question, not how do we get people to care, but how do we get people to question?

02:21:59.080 Yeah. And by people, I mean people in a position of authority who have the ability to put the

02:22:05.960 necessary funds and effort towards addressing this question. And I think one of the problems is that

02:22:11.640 the way we fund research in this country, for instance, is a kind of, you know, you get together

02:22:16.960 study groups of mostly like-minded individuals and they give out small... These are all the issues

02:22:22.880 we discussed at NUSI. They give out small R01 grants of $500,000 a year for five years to what

02:22:30.320 Kuhn would have called normal science. So we have our conventional wisdom, our dominant paradigm. We're

02:22:36.480 just going to continue to fund whatever research people have to do to create bricks to fit in that

02:22:43.040 paradigm. There is no method by which paradigms are questioned and research programs can shift.

02:22:49.520 If there's an assumption that it could happen naturally, then science is self-correcting.

02:22:54.000 But it's the way we fund research in this country doesn't actually allow that to happen.

02:22:58.560 If there's money to be made, then capitalism will kick in and people will take advantage of

02:23:04.320 the opportunities that other people might be missing. And that's why you have operations like

02:23:08.960 Virta Health, which is doing very well, advocating using nutritional ketosis to treat type 2 diabetes.

02:23:15.840 And you have other people who have started operations like diet doctor too. There is no mechanism.

02:23:21.440 So when I say people, I mean like, how do I get Francis Collins to care? And the journalist

02:23:26.320 living in Oakland talking on a podcast, you know, it's just, it's so many levels from influence.

02:23:33.120 And even if you did get him to care, could you get the NIH to do anything about it?

02:23:37.840 We're having this conversation the day after there was an article in science by two researchers who

02:23:43.840 we know very well, arguing that this carbohydrate insulin model of obesity is just wrong. It failed.

02:23:50.880 It's interesting. We tested it. It failed. That's what we were talking about when we discussed,

02:23:55.680 why not? How do I reconcile my beliefs with that? My response was, okay, you leave out this

02:24:02.240 people who tested it and that supported the model. You're looking at only a small proportion of the

02:24:09.040 evidence here. But even if you're right, we have a problem. You've avoided the elephant in the living

02:24:14.480 room, which are the obesity and diabetes epidemics. What do we do about it?

02:24:18.160 The paper that came out yesterday by John Smeakman and Kevin Hall is a relatively short paper,

02:24:23.360 but an interesting summation of what would be viewed as a possible refutation of this

02:24:29.440 carbohydrate insulin model. I think of it less as a public health paper. I don't see them trying to

02:24:34.000 address obesity and I don't see them offering or proposing to offer a solution to obesity. But I think

02:24:39.920 they're really trying to understand a mechanistic thing. And I don't know John, but I do know Kevin.

02:24:44.480 And my view of Kevin is this is a guy who came in as an outsider. So he has at least the benefit of

02:24:49.840 not having some of the limitations that others might have who have come up through the field

02:24:55.040 without the historical context. And well, maybe Kevin doesn't have that historical context just

02:24:59.440 as no one would. At least he came in, I think, as a physicist looking at this from first principles,

02:25:04.640 which you could argue could be an advantage to him, right? Yeah. In fact, if you'll recall,

02:25:09.120 that's how I was introduced to Kevin by Mitch Lazar at Penn because he thought we saw the problem in

02:25:15.360 similar ways, which we did back then. And then I introduced him to you, smart, young researcher

02:25:22.320 without biases at the time. Through the course of our interaction with the Energy Balance Consortium,

02:25:28.000 I think our belief system, it gets complicated. If you'll recall, his initial claim to fame as a

02:25:35.520 researcher, what got him in his position at NIH, as I understand it, is the model that he has of

02:25:41.120 metabolism and how that influences body weight. And at one point I had asked him to, you know,

02:25:47.040 he had suggested we could use his model to test my beliefs and his model rejected them.

02:25:54.480 And I argued, well, you've rejected them because we're working in an area that your model has no

02:25:58.960 data. So why don't we do the experiment to generate that data? And before Nussi hired him,

02:26:05.680 I was arguing to Kevin that he could do this experiment with an NM1 or 2 at NIH with the

02:26:11.600 metabolic chamber. As soon as they started interpreting their experiment, one of the

02:26:18.000 fundamental problems with pathological science is you establish what you believe based on premature

02:26:23.840 data. So you don't, rather than entertaining multiple hypotheses as possible explaining what

02:26:29.040 you see, you kind of collapse down onto one and then you go public with that prematurely. And now

02:26:34.880 you're trying to support that hypothesis rather than test that hypothesis. So once Kevin started to

02:26:44.320 interpret the evidence from the Energy Balance Consortium as a refutation of the carbohydrate-insulin model,

02:26:53.040 I think it's basically, and supportive of his computer model, I think it basically began to create

02:27:01.040 his bias and lock it in. And then the more conflict he and I had, which you witnessed, the more that

02:27:08.800 locked in his bias as it might have locked in mine. The other possibility is that he's just right and

02:27:13.680 I'm wrong, which if we could do more experiments, we might find out. One of the critiques of that paper

02:27:19.920 was on different types of methodologies that are used to measure energy expenditure,

02:27:24.720 which is a very important thing to study and understand when trying to test these hypotheses.

02:27:30.560 You've already alluded to one method, which is indirect calorimetry, where people sit inside of a

02:27:36.400 medical grade, hermatically sealed chamber that can, at very minute levels, measure inhaled carbon

02:27:43.120 dioxide and calculate the consumption of oxygen. And the ratio of those things allows us to calculate

02:27:48.560 very precisely the amount of energy consumed. An alternative method to doing that is something

02:27:53.600 called doubly labeled water, where a person drinks water that has two different types. One is heavily

02:27:59.840 labeled oxygen, the other deuterium. And by collecting urine over a period of days following

02:28:06.080 that ingestion, you can also estimate energy expenditure, but do so in a free living environment.

02:28:10.640 At the outset, you alluded to two experiments that were done in parallel.

02:28:14.240 The experiment that was done by the group that had the hypothesis that energy balance determines

02:28:19.200 obesity used the more precise of these calorimetry. The group that studied this, who had the

02:28:24.960 preconceived idea that the carbohydrate insulin model was correct, did a longer outpatient study

02:28:30.080 and used this other method. One of the critiques of the paper is that the method of doubly labeled water

02:28:36.480 is not valid in people who are carbohydrate restricted. And that explains the difference in these findings,

02:28:42.800 because otherwise they're very difficult to reconcile, right?

02:28:45.120 Right. Even the first Kevin Hall's energy balance paper, remember they saw a signal,

02:28:50.080 so they're measuring energy two different ways. I think of science differently than Kevin does.

02:28:55.120 You measure energy by the chamber, indirect calorimetry, and by DLW, energy expenditure. Remember,

02:29:00.080 the reason we're measuring energy expenditure is because the researchers didn't believe in the

02:29:05.200 course of four weeks you could see an effect on fat mass. So measuring fat mass directly by

02:29:10.560 DEXA wasn't a possibility. And I'm still curious, because there was also evidence. Remember back

02:29:15.760 then, DEXA was going to be confounded by water loss? And so can you ever use DEXA reliably for the

02:29:22.960 kind of measurements that are made over, say, two weeks when you're transitioning somebody from a low

02:29:27.920 fat to a very low carbohydrate diet? I don't think so.

02:29:31.600 I don't think so either. And yet that's been done in the last two papers out of NIH.

02:29:36.720 So anyway, the point is you've got two ways to measure energy. They give you

02:29:40.240 two different results. One result is more consistent with the energy balance prediction.

02:29:46.160 The other result is inconsistent with the energy balance and consistent with the carbohydrate

02:29:50.960 insulin model. One way to interpret this data is to say we measured energy expenditure two different

02:29:57.200 ways. We got two different results. Therefore, our results were not very robust. We tend to trust the

02:30:04.080 chamber. But if the chamber were right, we would have expected to see the same thing with the DLW and

02:30:10.080 we didn't. Unless the DLW is wrong and the chamber is the gold standard.

02:30:14.880 Which is a possibility. But the first thinking is we get divergent results here. We have other issues

02:30:22.880 with the experiment, non-randomized. So let's design a different experiment.

02:30:27.840 You know, Gary, it's clear that there's probably still enough smoke to question whether there's

02:30:33.200 some fire with respect to these unanswered questions. Do you think this is an answerable

02:30:37.440 question? Well, I hope so. Clearly, I keep doing what I'm doing because I'm hoping I can motivate

02:30:42.320 people to think more deeply about this and to resolve this issue. I don't think progress will

02:30:48.480 ever be made. Meaningful progress will be made on obesity without understanding the fundamental

02:30:54.080 cause and without elucidating how treatments are working. Although I have to say the GLP-1 agonists

02:30:59.920 are fascinating. So there's a public health problem. There's a scientific issue. Things have changed.

02:31:05.520 So in the 20 years that we've done this, when I started this, conventional wisdom was that low-carb,

02:31:10.960 high-fat diets like Atkins, what we called keto in pre-2010 or so, that these diets were

02:31:18.160 deadly, that they would cause heart disease, that they would ultimately make you fatter.

02:31:22.080 Today, for instance, the American Diabetes Association recommends these diets for type 2

02:31:27.600 diabetes, which means one-tenth of the public. They're everywhere. They're viral. The world is

02:31:32.240 saturated with books on keto. And even this article in Science, we've been discussing,

02:31:37.440 acknowledged that they can be beneficial for weight control. And nobody's talking anymore

02:31:42.000 about them causing heart disease. In fact, they're probably the most studied diets in history.

02:31:46.560 And if you go to clinicaltrials.gov, you'll find over 200 trials in the works looking at

02:31:52.080 ketogenic diets for everything from epilepsy, diabetes, cancer, cancer, and Alzheimer's.

02:31:59.760 Yeah, far beyond that. Traumatic brain injury. I mean, you've interviewed a lot of these people.

02:32:04.560 So in one sense, we've made an enormous amount of progress in liberating the

02:32:11.040 intervention that comes out of this way of thinking, such that it can now be used for

02:32:16.720 people. And anyone who struggles with the weight or blood sugar control can now know that they can

02:32:25.120 try these diets. And it's not hard to find a physician to help them, guide them through it. And

02:32:30.720 from the personal perspective, we know that they won't kill you, which is what we assumed going in 20

02:32:36.240 years ago. And so it would be nice if the research community understood obesity the way I think they

02:32:43.040 should. So it's a way of saying it would be nice if I was right about all this. But even if I'm dead

02:32:48.240 wrong, the ability to use this dietary intervention to improve your health has now become widespread,

02:32:56.640 mostly accepted. And clearly, there are teams of researchers all over the world who find the

02:33:02.720 clinical efficacy of these diets fascinating. So the world is changing. It might not be changing

02:33:08.640 as quickly as I would like, and it may not change as much as I would like. But like I said, even the

02:33:13.520 science article the other day, there are statements in there that would have been considered unacceptable

02:33:18.560 20 years ago, including the acceptance that insulin plays a major role in fat accumulation. And so obesity,

02:33:25.760 it's just not the role they were arguing. It's just not the role that I think it is. And David Ludwig

02:33:30.320 thinks it is. Well, Gary, we've been going at this for a while. This was kind of a tour de force of

02:33:35.280 science and a history of science in many ways. I'm sure folks found your journey interesting,

02:33:40.000 especially given that many people will know you from your more recent work. And I think

02:33:43.120 many people will not know you from the foundation of work that led to your curiosity here. I've always

02:33:47.360 found that to be the most interesting part of your personal story, by the way, is the way that you

02:33:51.280 stumbled into health sciences almost accidentally. I've always found it interesting. And obviously,

02:33:57.440 I think a lot of good has come of it. So Gary, thanks so much for making the time,

02:34:00.640 especially in this case, a lot of time, probably more than your typical podcast.

02:34:03.840 Thank you for having me. It's been a pleasure. And it's great to have an opportunity to

02:34:08.320 not just catch up, but talk at you for three and a half hours.

02:34:11.520 Thank you for listening to this week's episode of The Drive. If you're interested in diving deeper

02:34:18.400 into any topics we discuss, we've created a membership program that allows us to bring you more

02:34:23.120 in-depth exclusive content without relying on paid ads. It's our goal to ensure members get back

02:34:28.720 much more than the price of the subscription. Now to that end, membership benefits include a bunch

02:34:33.680 of things. One, totally kick-ass comprehensive podcast show notes that detail every topic, paper,

02:34:39.520 person, thing we discuss on each episode. The word on the street is nobody's show notes rival these.

02:34:45.600 Monthly AMA episodes or Ask Me Anything episodes, hearing these episodes completely.

02:34:50.000 Access to our private podcast feed that allows you to hear everything without having to listen

02:34:55.680 to spiels like this. The Qualies, which are a super short podcast that we release every Tuesday

02:35:01.280 through Friday, highlighting the best questions, topics, and tactics discussed on previous episodes

02:35:06.000 of The Drive. This is a great way to catch up on previous episodes without having to go back and

02:35:10.960 necessarily listen to everyone. Steep discounts on products that I believe in, but for which I'm not

02:35:16.720 getting paid to endorse, and a whole bunch of other benefits that we continue to trickle in

02:35:21.440 as time goes on. If you want to learn more and access these member-only benefits, you can head

02:35:25.520 over to peteratiamd.com forward slash subscribe. You can find me on Twitter, Instagram, and Facebook,

02:35:32.880 all with the ID peteratiamd. You can also leave us a review on Apple Podcasts or whatever podcast player you

02:35:40.000 listen on. This podcast is for general informational purposes only and does not constitute the practice

02:35:45.520 of medicine, nursing, or other professional healthcare services, including the giving of

02:35:50.480 medical advice. No doctor-patient relationship is formed. The use of this information and the

02:35:56.080 materials linked to this podcast is at the user's own risk. The content on this podcast is not intended

02:36:02.400 to be a substitute for professional medical advice, diagnosis, or treatment. Users should

02:36:08.160 not disregard or delay in obtaining medical advice from any medical condition they have, and they should

02:36:14.560 seek the assistance of their healthcare professionals for any such conditions. Finally, I take conflicts of

02:36:20.720 interest very seriously. For all of my disclosures and the companies I invest in or advise, please visit

02:36:27.040 peteratiamd.com forward slash about where I keep an up-to-date and active list of such companies.

02:36:57.040 Thank you.

The Peter Attia Drive - June 28, 2021

#167 - Gary Taubes： Bad science and challenging the conventional wisdom of obesity

Episode Stats

Summary

Transcript