Maintenance Phase Reality Check: "The Keto Diet"
I’ve been slow with posts over the past few months, because things have been busy - and also because I felt like I got my point across that MP is absolutely unreliable at best and dangerous at worst. But then that All Fired Up episode got me pretty fired up (haha) and I decided I wanted to revisit some of the other MP episodes. In the past, I’ve gone through and done really intense fact checking, but 1) that’s a huge undertaking and 2) some of the things I pointed out were considered “nitpicking” or irrelevant (which I sometimes agree with, but I felt like I needed to include everything for the sake of comprehensiveness). With this episode, I am going to focus up front on the things that I think are most problematic and add some nit-picky “fun” fact check pieces at the bottom.
Overall, I think this episode is a textbook example of the cherry-picking that MP engages in and highlights the lack of actual scientific content in their research. Exhibit A is the articles they put in their “Episode Notes” for this episode; not a single one is a primary source, and several of them directly contradict things that Aubrey states in the episode. I’ll note that the episode on calories has the same problem.
At this point, I should say that I do not believe the keto diet has any magical weight loss properties. People who lose body fat on the keto diet do so through a caloric deficit, same as any diet. Some people find dietary fat keeps them satiated for longer durations of time than carbs or protein, so in that sense, perhaps it works well for keeping people feeling fuller in a caloric deficit. But as of this moment, I don’t think there are any definitive data to suggest that for the average person (i.e., no diabetes, no epilepsy) the keto diet has any specific health benefits vs a “healthy” (yes, loaded term) diet that doesn’t restrict carbohydrates. I also haven’t seen any head to head trials of the keto diet vs low carb diets for diabetes/blood sugar control, so while we know keto has benefits for diabetic patients, there’s no way to say if those benefits are greater than those of a low carb diet (which has also been shown to have benefits). The preliminary research in some cancer patients is promising, but is far from definitive. That being said, the biological rationale for further examination is sound.
There is a lot to criticize about the way our culture approaches diets, about the messaging around the keto diet, about the way we blame individuals for their body size, and about the way we default to physically and psychologically harmful levels of restricting food types and amounts when attempting to change our bodies. These critiques need to be made in a way that is thoughtful, compassionate, and grounded in scientific evidence. MP and other similar voices in the activist domain have compassion in spades, but their strategy of starting with a desired conclusion and then massaging facts to support it are concerningly Roganesque.
Alright, let’s tackle the “laundry list” of risks that Aubrey claims the keto diet is associated with, in order of appearance in the episode .
1. Losing “more muscle mass than on other diets”
This is false. I have seen no data comparing keto to other diets with respect to this. And the source in the show notes directly disputes the fact that there is a substantial loss of muscle mass on the keto diet at all. It cites a study of 20 “obese” individuals who followed a very low calorie keto diet for 4 months and found that the weight loss was almost entirely fat mass; muscle mass was not impacted, and muscle strength did not decrease.
Some other evidence:
A review of studies on the keto diet that found that some studies reported a decrease in fat-free mass while others didn’t.
A review of 44 studies
A clinical trial of 53 “obese” patients randomized to either a very low calorie keto diet or a regular low calorie diet. There was no significant change in lean body mass for either group over the 12 month follow-up period while fat mass decreased substantially (and weight loss was significantly greater in the keto diet group compared to the regular low calorie diet group).
2. “Really significant liver, kidney and gallbladder problems”
Specifically, Aubrey says, with respect to the liver, “if you have any liver issues, like fatty liver disease, like cirrhosis, whatever, adding so much fat to metabolize could make those worse. And if you don't have them, it can increase your risk for future liver issues.” So non-alcoholic fatty liver disease (NAFLD) is not as simple as too much fat consumption. Beyond lipid metabolism, it’s also linked to insulin resistance and impaired glucose metabolism. Because of this, the keto diet may actually be beneficial for individuals with liver disease, but the studies are small and are typically very low calorie keto diets (ref 1, ref 2, ref 3)
With respect to kidneys, Aubrey says, “your kidneys metabolize protein. So, when folks don't get the protein balance just right, the keto diet can cause major kidney issues, it can also cause kidney stones.” This is a strange statement, because although protein is important for a variety of reasons, the kidney doesn’t need a “balance” of protein. A protein-restricted diet is actually recommended for chronic kidney disease (if you’re not on dialysis). Regarding kidney stones, there is also virtually no research here and none of the studies are comparative (i.e., they do not look at the incidence of kidney stones in individuals not on the keto diet). I did find one poor quality meta-analysis that reported a cumulative incidence of kidney stones of 7.9% among adults. Almost all of the included studies were in populations of patients with epilepsy, so it’s not generalizable to the general population. And interestingly, there is actually some evidence that the keto diet may be beneficial for folks with chronic kidney disease.
3. Constipation and diarrhea
I’m not going to dispute this, but I will say that this is a consequence of basically any big dietary change. Aubrey’s main point, that the keto diet is low in fiber, is not likely to be a culprit here, given that the average American doesn’t consume enough fiber, anyway. While yes, it’s definitely a problem not to eat enough fiber, the keto diet is not going to be the only reason someone isn’t eating enough fiber. In fact, if you just Google “keto diet fiber”, you’ll see a whole bunch of ideas about how to get enough fiber on the keto diet.
4. Reduced cognitive function
Again, evidence is scarce. I found a study that found no effect of the keto diet on cognition, but I also found some that found enhanced cognition while on the keto diet. A systematic literature review found that 80% of the 27 studies in humans reported a benefit associated with the keto diet and none of the other 20% reported a detrimental effect.
5. Increased risk of diabetes and heart disease
I have to laugh at this one, because the source that they put in their show notes to support this is a press release about a study done in mice. This gets me, because they are constantly going on about how studies in mice don’t necessarily apply to humans (true, but it’s a step toward it - that’s why it’s called “translational research”). Anyway, there is no conclusive evidence from studies in humans that the keto diet increases the risk of diabetes nor heart disease. In fact, there is actually research that suggests the keto diet may be beneficial for patients with diabetes (this study came out after MP put out this episode). This tracks, given that Type 2 diabetes, like NAFLD discussed above, is a consequence of insulin resistance, which is the impaired biologic response of target tissues to insulin stimulation.
6. Mortality
Aubrey cites two papers here that are both incorrectly described. I’m going to start with the second one: Aubrey cites a study in the Lancet, saying it “found that people who followed diets that were low in carbs and high in animal proteins, typical of the keto diet, had a higher risk of early death compared to those who consumed carbs in moderation.” The paper that Aubrey is referring to here utilized methods that Aubrey and Michael are quick to dismiss when the study is about obesity and health, such as self-reported diet (note that I do not think this necessarily invalidates the findings, but there are clearly many other methodological issues in this study that make the findings questionable). As an aside, it’s worth noting that Walter Willet is a co-author on this study, because Aubrey and Michael have roundly criticized his work on obesity and mortality in another episode.
Anyway, this study included over 15,000 adults aged 45-64 in the United States when the Atherosclerosis Risk in Communities (ARIC) study was recruiting from 1987-1989. So already we can see that this is perhaps not generalizable to folks who are 45-64 now. Additionally, food frequency questionnaires (FFQ) were collected only twice, once at enrollment and once ~6 years later. Follow-up continued until 2013. This means there was a substantial duration of time after the second FFQ in which any dietary change would not be captured by the study. Carbohydrate intake was divided into quantiles. This means that the “low carbohydrate” group was not necessarily consuming what we would classify as a “low carb” diet today. They were simply consuming a lower quantity than the rest of the study population. The group characterized by the lowest carbohydrate intake consumed a median of 37% of their caloric intake from carbohydrates, compared to the 5-10% on the keto diet and <26% for “low carb” diets. As Feinman et al. aptly point out, “First, in terms of relevance and accuracy, Seidelmann and colleagues did not investigate a low-carbohydrate diet as a specific intervention, contrary to the general understanding and expectation of patients and health-care providers. Instead, carbohydrate consumption was evaluated using data from the Atherosclerosis Risk in Communities (ARIC) study, which was not designed for this purpose.” Many other academics and journalists also weighed in, which helps highlight how problematic this study was. I highly recommend reading the short responses by Nina Teicholz, Dr. Jocelyn Tan-Shalaby, and Drs. Amelia Finaret and Marc Bellemare, which summarize the issues with this paper very succinctly.
The first study Aubrey cites suffers from many of the same issues as the one above. Aubrey says, “This is a quote from Health magazine. ‘A 25,000-person study presented at the European Society of Cardiology Congress suggested that people on the lowest carb diets had the highest risk of dying from cancer, cardiovascular conditions, and all other causes.’” I think it goes without saying that “Health Magazine” is not a rigorous scientific publication, but it’s also disappointing that Aubrey didn’t even look at the primary source (Mazidi et al.). This study, similar to the one above, did not look at low carb diets. The group characterized by the lowest carbohydrate intake consumed a median of 39% of their caloric intake from carbohydrates (again, compared to the 5-10% on the keto diet). The data source for the primary analysis (not the meta-analysis) was NHANES - the self-reported diet data that Michael and Aubrey have dismissed in the past, and data on diet was collected at baseline only. In fact, since then, two additional papers (Angelotti et al. and Shan et al.) (using NHANES data and more rigorous methods) have refuted these findings. As Angelotti et al. explained, “This discrepancy between prior studies is likely due to differences in assessment of dietary intake and exclusion criteria. Mazidi et al. used a single day of dietary recall to evaluate acute intake rather than usual intake, and was thus subject to within-person random error and overestimates the range of dietary intakes in the population (mean carbohydrate intake of extreme quartiles was 39% and 64%). Rather, it is recommended that longitudinal studies employ multiple dietary recalls collected from most participants to reduce measurement bias, and analyze these data using usual intake methodologies such as the NCI method, as did Shan et al. (mean carbohydrate intake of extreme quintiles was 46 and 58%). To further control for bias, Shan et al. excluded participants with a history of heart disease or cancer, as well as those that died during the first year of follow-up, whereas Mazidi et al. did not. The methodology and results of the present study (mean carbohydrate intake of 41% in the restricted carbohydrate group and 50% in the recommended carbohydrate group) are consistent with Shan et al…” The point is, neither of these studies was looking at low carb diets, let alone the keto diet. Additionally, they are not high quality and both would be dismissed by Aubrey and Michael if they were about obesity and mortality.
Other issues in the episode:
Yes, absolutely. It's not enough to have some ketones in your blood. You have to have a critical mass to get to ketosis. That's why there's this extraordinary restriction of carbohydrates. You're having less than one slice of bread a day, and you're just having a ton of fat, as much as you can.
This may seem nitpicky, but I think it’s important to get definitions right and this is not entirely true. It is kind of like saying you have to have a lot of money to be rich - being rich means having a lot of money. It’s not that you have to have a “critical mass” to get to ketosis, it’s that “ketosis” is defined by ketone body concentrations in the blood above a certain threshold (most commonly 0.5 mmol/L serum beta-hydroxybutyrate [BOHB] for “nutritional ketosis”).
Yeah, so you pee onto these strips to tell you how many ketones are in your urine. What I've read from the research is, that isn't actually a measure of anything because it doesn't really matter if they're present in your urine, it matters if they're present in your blood. So, the only way to really test for that is with a blood test, and most people don't get blood tests.
This is a great example of how a lack of scientific expertise can lead to black and white statements instead of acknowledgment of nuance. I’m not sure what research Aubrey was reading (again, would be great if there were some citations here!), but her conclusion is incorrect and overly simplistic. Here’s some more detailed information: urine sticks and blood tests measure different things - urine sticks measure acetoacetate, a surrogate marker for BOHB in the blood. The primary issues with using urinalysis to assess ketosis are:
1) Urinary dilution – hyperglycemia can cause something called osmotic diuresis, in which the body produces large volumes of urine. This may result in underassessment of urinary ketones. Note that this is primarily an issue for individuals with diabetes, not those attempting to reach nutritional ketosis.
2) More relevant for nutritional ketosis is that the kidneys adapt to ketosis over time, reabsorbing the acetoacetate rather than excreting it. This, of course, means that the body could be in ketosis without measurable ketones in the urine.
3) Ketones in urine are not necessarily an indicator of current ketosis, since the urine could have been in the bladder for a number of hours. So a positive urine sample could mean your body was in ketosis 2 hours ago, but isn’t anymore.
The point is, saying that urine isn’t a measure of anything is missing some important nuance.
The first time anyone tested and studied fasting as a treatment for epilepsy was in 1911. It was in France. Epileptic people were put on a very low-calorie vegetarian diet, which included phases of both fasting and of purging. We're in full eating disorder territory here.
First, I’m calling this out because I’ve seen folks in the MP subreddit call things eating disorders that are not (i.e., the low FODMAP diet). It’s not an eating disorder if it is a prescribed therapy. This is not an eating disorder. Second, this was considered a “detox” diet (literally, the title of the paper is “The fight against epilepsy through detoxification and nutritional rehabilitation”) and is not related to the keto diet at all except that I guess it is nutritional therapy for epilepsy. A lot of sources erroneously cite this as a paper on fasting, but it’s really a wacko “detox” diet. Anyway, here’s what the regimen included: four days of daily administration of sodium sulphate (a laxative) and unlimited beverage consumption without food followed by 4 days of a vegetarian diet “with a reduction of half the food intake” (translated from the original using DeepL). So, basically intermittent fasting with reduced caloric intake and laxatives on top. Sounds terrible. Also, to make things worse, this was done on patients from an asylum (yikes, research ethics have changed a lot!).
Of the 20 patients who took part in that study, two of them showed reduced symptoms. About half of them ended up falling off of the diet altogether, because it was so incredibly restrictive and hard to stay on.
Not quite. It was supposed to be 21 patients but only 6 patients ended up doing the diet for more than one cycle and only 2 of those actually followed the diet. There was also apparently no quantitative result published – the authors just said that the seizures were less frequent and less severe. So anyway, this is a terrible paper and was not based on any evidence whatsoever.
Two-time Academy Award winning actress, Meryl Streep, returns to television.
Not a fact check, but just to provide some context, Meryl had previously narrated a film by Jim Abrahams called, An Introduction to the Ketogenic Diet. It’s not really as random as it seems – Meryl and Jim were friends (their kids went to school together). The movie itself is based on a true story from a letter Meryl and Jim got in response to the informational video. It’s a bit wild for Michael and Aubrey to suggest that this is over-the-top and oppose the implication that “doctors don’t actually care what’s best for their patients” when they frequently express that same sentiment when it comes to fat patients. I also think we can all agree that portraying doctors pushing against a new treatment =/= doctors not caring about their patients.
Yeah, totally. That's fine if you want to eat less bread or less pasta or whatever. Eat however you want, eat whatever you want. The number one gateway into disordered eating and full-on eating disorders is diets. That's the other thing to know, is I think people have this idea that like a diet is just a benign thing that you do for a while and that there aren't really consequences to it. As we will see, there are really significant consequences to diets broadly, but to this one in particular. Fucking tread lightly. Tread lightly, everybody.
A recurring theme on MP is the use of the word “diet” as a catch-all for any dietary change intended to promote weight loss. This is unfortunately just perpetuation of diet culture itself. There is no evidence that “diets” are the “number one gateway” into eating disorders. In fact, there are several randomized trials that found that caloric restriction did not increase disordered eating behaviors. For example, a study from 2004 randomized 123 women to either 1) a very low meal replacement diet for 12 weeks with a slow transition to a reduced calorie diet of conventional foods, 2) just a reduced calorie diet of conventional foods, 3) no diet and encouragement NOT to limit caloric intake. A few of the women in the group that was randomized to meal replacements reported a binge episode at Week 20 (compared to zero in the other two groups) but this difference did not remain at the end of the study (Weeks 40 and 65), when diets were normalized again. The difference in binge eating at Week 20 is not surprising, given that the prescribed diet was incredibly restrictive and unhealthy, but the point is that these women didn’t develop eating disorders and, more importantly, the group that just had some caloric restriction (what Aubrey and Michael might deem a “diet”) didn’t have any binge episodes. In other words, caloric restriction is not a path to an eating disorder for every human. This is not to say that it isn’t a slippery slope for some people! But it is possible to have a healthy relationship with food and still reduce your intake if you want to lose weight. As far as risk factors for eating disorders, we don’t know enough to say that there is a “number one gateway”. Eating disorders are heterogeneous and risk factors vary by type of disorder (i.e., bulimia, anorexia nervosa, binge eating disorder). We do know that genetics play a strong role (ref 1, ref 2). I recommend also checking out these two recent publications that discuss the interplay of the current food landscape and eating disorders:
The Mayo Clinic runs the first trial of the ketogenic diet in 1921. Their findings at that time was that a strict ketogenic diet reduced seizures in 95% of epileptic kids, which is incredible. It's still on the books as a treatment, and it's still something that folks use.
The Mayo Clinic did not run a trial. Dr. Russell Wilder published a paper in 1921 that reported on a series of patients with epilepsy treated by Dr. Geyelin at Presbyterian Hospital in New York who had successfully responded to a fasting diet. Wilder then detailed the ketogenic diet and proposed that it might be therapeutic, as well. In a second paper, Wilder reported on three patients who followed the keto diet (see below). This was what we call a “case series” - not a clinical trial. I am not sure where Aubrey pulled the 95% number from.
Essentially, what happens from here is that this leads to this major spike in research around the ketogenic diet as treatment for epilepsy. So, there's a study that comes out in 1998, once again from Johns Hopkins University. They publish an update in 2001. This was a larger study. There were 150 participants. Basically, their findings were a more modest version of what had been discovered in the 20s, that the diet could reduce seizures in epileptic kids, but that the biggest stumbling block was just being able to stick to a very restrictive and often very costly diets. After a year of this study, almost half of their subjects had dropped off of the diet. So, they couldn't really measure its effects because folks couldn't stick with it.
The “update” was not an update but a separate study and it was published in 1998. Both studies found a remarkable decrease in seizures among patients who adhered to the diet. The results from both studies were very similar - the larger study did not have more “modest” findings. In the study of 51 patients published in JAMA Neurology, 22% of patients still on the diet at one year had a >90% reduction in seizures. In the study of 150 patients in Pediatrics, 27% of the patients still on the diet at one year had a >90% reduction in seizures.
Currently, the ketogenic diet is approved for use by the NHS in the UK. It's covered by many US insurance plans. It's pretty mainstream, although it's usually these days considered last resort for kids with drug-resistant epilepsy, that you get to drugs before you get to a dietary sort of fix. They know that it works, but they don't know why it works, and they don't know why it only works for some kids, like why it isn't universal applicable.
Not entirely sure what Aubrey means that it is “covered by many US insurance plans” but it’s important to clarify that the keto diet itself is not covered by insurance. As in, your food is not covered. However, as is common with medical nutrition therapy, there are ketogenic “medical foods” products that are covered. Coverage varies by supplier and insurance plan.
Mike: I don't know. That's always the scariest thing, is people, anybody under 18, but especially littler kids. Argh.
Aubrey: You don't like experimenting on children?
Conspiracy land. No one is experimenting on children for weight loss. Also, children need treatment, too. We literally have to do experiments including children in order to know how things work in children.
Aubrey: Basically, there is a study published in Science in 2013. There's a press release attached to the release of this study. The press release says that the keto diet can “slow the aging process and may one day allow scientists to better treat or prevent age-related disease, including heart disease Alzheimer’s and many forms of cancer.”
Mike: Wait, this was in Science?
Aubrey: Yep.
Mike: This sounds like some Dr. Oz citation needed shit.
It appears that this was ripped directly from the Men’s Health website, because the actual press release does not say this. It is much more careful with language. It reads:
Scientists at the Gladstone Institutes have identified a novel mechanism by which a type of low-carb, low-calorie diet—called a “ketogenic diet”—could delay the effects of aging. This fundamental discovery reveals how such a diet could slow the aging process and may one day allow scientists to better treat or prevent age-related diseases, including heart disease, Alzheimer’s disease and many forms of cancer.
It very clearly does NOT say that the keto diet can do anything. This is another example of MP misrepresenting the way science communication has occurred and conflating the media representation of a study with the way something is being discussed in the scientific community.
Yeah, it was an era of science telling you what you want to hear and you being like, “That sounds great. Moving on.”
Not science, but bro science influencers who did not actually have any connection to actual science. This is a very important distinction.
Aubrey: Here's a quote from Men's Health magazine about the episode with Dr. D'Agostino. It says, “Ferriss told the story of a friend with testicular cancer, who would fast for three days, entering into ketosis before chemotherapy.”
Aubrey: [laughs] Just wait for it, buddy. “D'Agostino noted that anyone with cancer needs medical supervision of their diet, but also said, if you put your physiology into a state of fasting ketosis, that puts tremendous metabolic stress on cancer cells that are highly dependent for survival and growth on high levels of glucose and insulin. By subtracting them from those growth needs, they can die and you could potentially purge yourself of some precancerous cells.”
Mike: Ah, I hate this shit. There's always this perfunctory, like, “No, I'm not saying it's going to cure cancer,” followed by “basically, it'll cure cancer.” And then, if anybody calls you on it, you're like, “Well, I never said it was going to cure cancer.” But, of course, that's the impression you're leaving people with.
Instead of using Men’s Health as a source, I looked at the actual transcript for this episode and this is a misrepresentation of what was said. I am absolutely NOT a proponent of Ferriss or D’Agostino, but spreading misinformation like this on a podcast that is supposed to debunk things is problematic. Ferriss did not say that his friend entered ketosis - he simply said that his friend fasted for 3 days. Then he asked, “...are there implications for fasting or nutritional ketosis or exogenous ketones for helping to prevent or mitigate the onset of those type [sic] of neurodegenerative diseases?” In response, D’Agostino did not say what Aubrey quotes him as saying or really anything close to it. What he did say is in a completely different section of the transcript and is referring to people without cancer. He said, “And I always − and I discussed this with my colleague, Thomas Seyfried, he thinks that if you don’t have cancer and you do a therapeutic fast once or twice a year or maybe three times a year if it’s a shorter fast that it would purge any precancerous cells that may be living in your body.” He continues, “So, if you put your physiology into a state of fasting ketosis, that puts tremendous metabolic stress on cancer cells that are highly dependent for survival and growth on high levels of glucose and insulin. By subtracting them of those growth needs, they can trigger apoptosis autophagy, and you could potentially purge yourself of some precancerous cells.” Again, this is NOT about curing cancer. Also important to note that D’Agostino is not an oncologist or a cancer research expert. That being said, there are a lot of very smart people who strongly believe that fasting has a role in therapy for cancer. That is why they are researching it! Just like we do with drugs. Additionally, it’s worth noting that while I think D’Agostino is in influencer mode here, what he’s saying isn’t totally absurd. There is some evidence that supports what he is saying here - it’s just almost entirely in animal or in vitro models:
Fasting reshapes tissue-specific niches to improve NK cell-mediated anti-tumor immunity: Immunity
Intermittent and Periodic Fasting, Hormones, and Cancer Prevention - PMC (nih.gov)
Safety and feasibility of fasting in combination with platinum-based chemotherapy - PMC (nih.gov)
Dr. D'Agostino now says, “Cancer is, of course, much more complicated than just like, “Starve it of sugar and you'll be fine.”
This is ripped almost verbatim from the Men’s Health article but without the context. Actual quote from Men’s Health:
When asked about that statement, D’Agostino concedes, “This episode’s title is unfortunate,’ but he points out that his research does suggest keto can help slow the progression of some cancers, though it speeds up others. ‘It’s much more complicated than ‘starve your cancer of sugar’,” he says.
And all the while, you can pretend that all you're doing is asking questions. It's this built-in defense mechanism where it's like, “We're not necessarily saying everybody should go on the keto diet. We're just dedicating hours every month to saying, well, will it improve sports performance? Will it cure cancer?” And so, the amount of attention that you're placing on his one very specific thing is in many ways more important than what you're actually saying about it.
This is how research works. If you believe something has a chance to be beneficial, you spend a lot of time thinking about it. This is no different than people who are obsessed with mindfulness/meditation as a cure-all.
“Within a year of the Rogan podcast, Orian Research estimated keto as a $5 billion industry, and because people on keto often lack nutrients like vitamin C, magnesium and fiber, there's been a supplement gold rush for brands behind products that make staying on the diet easier.”
What’s wild to me is that this company is no longer functioning, as far as I can tell. Their website is broken. This report is not findable anywhere and all references to it online lead back to this Men’s Health article. So I’m a bit skeptical.
Mike: Yeah, it's also just totally ascientific. There's not really data on not wearing shoes.
This is totally irrelevant to the keto diet, but I do just want to point out that there absolutely is research on not wearing shoes.
Your point about MP has definitely been proven; however, I beg you never stop ripping it to shreds
I once very much enjoyed Michael Hobbes. The title of his series "You're Wrong About" demonstrates why. He made me feel enlightened, not about things _I'd_ been wrong about - ho, ho, no - but about things _everyone else_ was wrong about!
The wide-ranging topics of that series invited Gell-Mann Amnesia. Now he's working on a prolonged series about one topic, and it's been startling to discover experts in the field taking his work apart in detail.