Monday, January 10, 2011

Book review: Gary Taubes' Why We Get Fat


It's rare that I've had the occasion to read a book whose premises I agree with (that we eat way too many carbs, that they in turn impact on our weights, and that weight-wise exercise isn't much to write home about), but whose arguments make me cringe. Gary Taubes' Why We Get Fat met that billing.

Let me start out by stating that I'm quite low-carb friendly and that I readily agree that science has proven that saturated fat has been wrongly demonized by the medical establishment for decades, including somewhat by me when I co-wrote my book in 2006/7 (a guy's allowed to learn, and it was in this spirit that I approached reading Taubes' book). Furthermore, I also agree that carbohydrates, more specifically the refined highly processed ones, contribute dramatically to both obesity and chronic disease and their reduction may well have a role to play in most folks' weight management efforts, and that a myopic view of dietary fat as causal to chronic disease and obesity has likely in and of itself, by means of a consequent dietary shift to carbohydrates, contributed dramatically to the rise in the societal prevalence of chronic disease and obesity.

All that said, I found Why We Get Fat to be an extremely difficult read. Not because the writing wasn't engaging. On the contrary, Taubes is an excellent writer. I found the book difficult to read because for reasons I can't understand, Taubes seems to have decided to abandon journalistic and scientific integrity in place of observational data, straw men and logical fallacy.

Taubes' manifesto is straight forward. Carbohydrates make us fat and they do so independently of the first law of thermodynamics. Forget about calories, you can eat as many or as few of those as you'd like, ultimately weight is purely about carbohydrates.

Why We Get Fat's observational data comes hard and heavy right out of the gates. Taubes posits that because there have been examples throughout history of impoverished peoples with high rates of obesity that the concept of a toxic environment (cheap calories and minimal exercise) being causal to our modern day weight woes must be false. Does Taubes really think that obesity has a singular cause? That there's only one pathway to weight gain? That because he can find obese impoverished people the environment's not involved? Apparently he does.

Straw men and logical fallacy? Try this one on for size. He tackles an observation made by Williams and Wood, researchers who'd studied exercise and weight. They found that even marathon runners tend to gain weight over time and suggested that in order to avoid that gain they'd have to run further each and every year. Taubes extrapolates to suggest that any middle aged runner wanting to stay lean will have to run half marathons five days a week to resist weight gain. Sounds ridiculous, right? He then concludes on that basis the calories-in/calories-out hypothesis couldn't possibly be true. But couldn't calories in and out still matter in these runners? Don't our lives change as we get older - less time, more disposable income, more responsibilities - all things that may cause increased reliance on both convenience and celebratory calories. And doesn't metabolism naturally slow as we age due to age-related sarcopenia - wouldn't that also lead a person to steadily gain weight even if all other things (intake and exercise based output) remained constant? Or couldn't the runners consistently be consuming more calories than they burn as a consequence of either exercise-induced hunger or overcompensated reward based intake, and that coupled with natural aging's effect on metabolism causes ever increasing weight gain in the absence of changing calories in or calories out?

Next Taubes' tackles the classic 20 calorie a day mismatch leading to 2 pounds of weight gain a year (roughly the average gain per year of North American adults) and then presents it literally which of course appears ludicrous and ridiculous. Are 3 extra chips a day causing you to gain weight? Of course not. Therefore according to Taubes' next straw man, calories in and calories out can't be causal for weight. Forget about the fact that calories can come in clumps. 3,500 calories in extra Christmas time treats doesn't sound to me to be too off the mark and those alone would take care of 175 of Taubes' 20 calorie days. Tack on a couple of birthday celebrations, a vacation and a visit or two to a Chinese buffet and we're more than done.

Next comes cows. Taubes shows us pictures of two cows. One an Aberdeen Angus - fatty and delicious looking. The other a Jersey milk cow - lean and chewy looking. He then goes on to state that it's not possible that breeders have simply manipulated genes to make the Aberdeen cow hungry and the Jersey cow active. Instead Taubes argues that genetic breeding has impacted on these cows partitioning of fat (one to flesh and one to milk production). While I don't disagree with him, his conclusion that their genes don't determine how many calories the animals consume, but rather what their bodies do with those calories doesn't in any way support or refute the notion of calories in or calories out. A quick Google image search had me staring at some pretty hefty Jersey cows and some Anguses with very visible rib cages, and I'm fairly confident that were female Angus cows given hormones and milked daily, they'd also have big full udders. Could it be that cows do in fact behave both according to the laws of thermodynamics and also according to their genetic makeups? That's where my money'd be in Vegas.

Taubes' own cognitive dissonance appears mid book and it led me to be briefly more hopeful. Taubes was going on about the first law of thermodynamics (TFLOT) (that energy can neither be created nor destroyed - the cornerstone of the calories in/out hypothesis of weight). He actually seemed to agree with TFLOT but then tried to suggest that talking about "overeating" is not the same thing as talking about "energy" and that the important question to ask is, "WHY",

"Why do we take in more energy than we expend? Why do we overeat? Why do we get fatter?"
I agree, those are tremendously important questions and moreover I agree that macronutrient distributions (one's daily dietary spread of carbs, fats and proteins) for many folks likely play a big role in those whys.

Sadly Taubes takes a strange road to walk in following up on those whys. Taubes rages against the calories-in/calorie-out hypothesis stating that one would be, "hard-pressed to find (a concept) more damaging". He states the calorie concept has done, "incalculable harm", that it has fueled the obesity as a function of sloth and lack of willpower that lays blame squarely on each obese person's ample shoulders.

I wholly disagree with him. It's not the concept that's done so much harm, it's the misuse of the concept, its oversimplification and its grossly unfair, individualized, blame-based application. The danger and the harm lies solely with the failure of society and medicine to ask why are people consuming so many more calories - a failure Taubes so rightly pointed out just a few pages prior.

Amazingly, despite his embrace of the concept that different bodies do different things with the calories presented to them (as evidenced by his discussion of the fat partitioning of cows), next he comes at the reader with this painful series of questions,
"It's the environment we live in that makes us fat, we're being told, not just our weakness of will. Then why don't lean people get fat in this toxic environment? Is the answer only willpower?
Could Taubes possibly be suggesting that genetic variation only impacts upon fat distribution and that it's silent in terms of dietary choices, inclination towards activity, non-exercise induced thermogenesis, the thermic effect of food, resting energy expenditures, etc? Does Taubes truly consider us all to represent a single genetic and metabolic lineage that should see us all uniformly responding the same way to a specific environment?

Taubes next talks of differences in fat distributions seen between men and women to support his assertion that the amount of fat is "exquisitely" regulated. To me all it suggests is that distribution is regulated. Then Taubes launches into a discussion of wild animals and points to the fact that hippos and whales don't get diabetes and "never get obese" as further proof for his theories. He states that no matter how abundant their food supply, wild animals will maintain a stable weight and never become obese. Really? Do we track the weights of wild animals? Is there an animal NHANES database on squirrels, frogs and crows, or is Taubes, a man who purports to be a "firm believer in science" really just relying on his own observational assessments of animals and at the same time extrapolating wild animals to be useful to this discussion as small little human models appropriate for comparisons? Even putting aside a lack of actual data to support his thesis, even were it to be true that animals never become obese, could it not also possibly be the case that wild animals, most of who don't have very good health care plans, don't have the luxury of living long enough or healthfully enough to maintain gains in weight and that disease and early death help keep their weights down?

And what of human "wild animals". Up until a hundred years or so ago, obesity was a rarity for free-living humans as well, and I don't think Taubes will dispute the fact that genes tend not to change dramatically in just 3 generations. Clearly something else is at play - but for Taubes it's not excess consumption of calories that make us fat. Instead he believes carbs make us fat independently of thermodynamics, and in turn our fat itself drives us to overeat.

To support his point Taubes continues to lean on anthropomorphizations and frighteningly non-scientific comparisons with other species. He talks of elephants and blue whales eating huge amounts because they're huge. He talks of mice that are bred to be obese who when starved to death, still have more adipose tissue at autopsy, but to me all that proves is the notion that the distribution of calories is regulated by the body, not that calories don't count. Then he suggests that marathon runners run not because they choose to, or that they want to embrace what they see as a healthy lifestyle, but rather because their muscle tissue is regulated to take up more calories and they're literally driven unconsciously to burn off those calories by, "a very powerful impulse to be physically active". Who cares that there's no evidence to suggest that's true - that sound awesome. The reason I'm not a marathon runner is that my actual body isn't forcing me to lace up my shoes. And here I thought it was because I don't have much free time. But wait a second, I'm pretty damn lean and my diet's in the neighbourhood of 45-50% carbs, how can that be if carbs make people fat independently of energy consumption? But forget about me, I'm a case study of one. Forget too about the thousands of folks in the National Weight Control Registry who manage to control their calories and maintain their weights on low-fat (and hence high carb) diets. Forget too about the multiple studies that have looked at isocaloric restriction and haven't found significant differences between different dietary macronutrient distributions. No, instead I think it'd be fair given Taubes' penchant for observational data for you to do a quick observational exercise of your own. Think for a moment about all your leanest friends and relations. Are they marathon runners, obsessive exercisers or hard-core low-carb dieters? Perhaps some are, but I'm willing to wager, the vast majority aren't. What's up with them? How do they (and I) fit into Taubes' hypothesis?

And then there are simple non-truths. At one point Taubes states,
"You don't get fat because your metabolism slows, your metabolism slows because you're getting fat."
But as anyone who measures resting energy expenditures knows, metabolism actually rises as one gains weight due to the very simple fact that the more of a person there is, the more calories that person burns, and that significant weight loss almost invariably results in significant decreases in resting energy expenditures. Perhaps Taubes is referring to the individual who stays the same weight but loses muscle in place of fat - but of course as a seasoned writer, he knows that's not how readers will interpret his statement.

So what's the cause of everything according to Taubes? Oh yeah, carbs. In fact he states that the reason any diet works isn't because of caloric restriction, but rather it's due to carb restriction, with the corollary also being true - weight gain's not a consequence of caloric intake, but rather carb intake. Tell that to Twinkie Diet guy Mark Haub who lost 34% of his body weight eating 1,800 controlled, processed carby, junk food calories a day purchased from convenience stores.

To hammer his point home Taubes ignores the fact that correlation doesn't prove causality and talks of how it was during the 60s and 70s that doctors stopped believing in the low-carb route to weight management and how this coincided with the obesity epidemic as a means to prove his thesis. Of course nothing else has changed since the 60s or 70s to promote weight gain has it? I mean the world's exactly the same as back then, isn't it? Same number of restaurants, same dollars spent outside the home on food, same time spent preparing home cooked meals, same number of food commercials, same fast paced electronically tethered existence, same portion sizes, same number of available foods in the grocery store, same cost per calorie of food, same everything, right? No? You think that some or all of those things and dozens more might impact on weight? Me too.

Next he launches into antique medical textbooks and their takes on low-carb diets as proof of his manifesto. Does the fact that antique medical texts through until the mid 19th century recommend phlebotomy to treat asthma, cancer and pneumonia mean that we should start practicing it again?

Taubes next argument centres around the suggestion that carbs were not evolutionarily a part of our natural diet and that consequently we did not adapt to high carbohydrate diets. While I don't disagree with the suggestion, unlike Taubes I can't make the stretch that the foods we became adapted to eating during the millennia that our life expectancies were in our 30s and 40s necessarily have a bearing on our health and long lifespans today. He points to indigenous peoples of the world and comments on how when they kill an animal they eat, "virtually all" of its fat as proof that's a wise way to live and that fat's healthy for us, and while there's no doubt that the data on fat suggest it's not particularly scary, the fact that hunger-gatherers try to make the most out of every meal they catch doesn't impress me much to prove anything other than hunting's tough. Further to his discussion of evolutionary dietary adaptations he notes of our modern refined sugars and carbohydrates,
"That a diet would be healthier without them seems manifestly obvious."
And here I thought scientists were supposed to rely on evidence, not what seems obvious to draw conclusions. Sadly what seems, "manifestly obvious" isn't always a good or true idea as was evidenced by our reliance on the logic of the obvious in recommending decades of hormone replacement therapy to postmenopausal women.

Next up is his discussion of the Maasai tribesmen who he reports don't have cancer, heart disease or obesity and traditionally were very-low carb eaters. Of course they were also nomads who wandered and hunted all day long with life expectancies of 42 years for men and 45 for women - hardly a lifestyle or lifespan we should aspire towards and certainly potential explanations in and of themselves for the Maasai's lack of heart disease and cancer - both diseases tightly associated with age and at least partially preventable by means of exercise.

So how does Taubes' explain the impact of carbs on weight? Amazingly he states,
"We know the laws of physics have nothing to do with it."
Ultimately he embraces the notion that carbs make you fat regardless of the calories in/out hypothesis, rather than discuss such possibilities as carbs making you fat by having a lesser impact on satiety. Mid book I had hoped this was where he was going, but sadly, for reasons that are backed up by observation, inference, logical fallacy and straw men, apparently he's decided that living creatures are magical beings that live independently of the laws of physics and thermodynamics.

Taubes doesn't just rely on non-scientific argument, he also appears to be comfortable in ascribing his beliefs to other people and to omitting facts when its convenient. In discussing the World Cancer Research Fund's Food, Nutrition, Physical Activity and Cancer Report that concluded obesity and cancer were tightly linked he calmly states,
"If the expert authors of the report had paid attention to the science of fat accumulation ... they would have concluded the obvious: that the same carbohydrates that make us fat are the ones that ultimately cause these cancers"
Yeah, I'm sure none of these folks know anything about the physiology of fat. On omission - there's no doubt that Taubes knows that between 5-10lbs of weight are lost on a low-carb diet due to the mobilization of the water stored with glycogen, yet when discussing the A TO Z Weight Loss study, he doesn't bother to mention that the 9lbs lost in the first 3 months of the study almost certainly included a significant percentage from the water lost due to the participants' adherence to ketogenic carb limitations and that some of their regain as they added carbs back in was also likely just water weight.

Finally Taubes talks of the individuals who have cut their carbs down to nothing and still can't lose weight. He quotes Wolfgang Lutz, an Austrian low-carb practitioner from the 50s as stating that those patients must have "reached a point of no return". Makes sense to Taubes because he doesn't believe that the laws of physics apply to people and therefore it wouldn't be consequential to him (or presumably Lutz) if perhaps those folks were simply consuming too many fat and protein based calories.

Why We Get Fat is certainly a book that will appeal to the masses as it pseudo-scientifically preaches that carbs are a magic food and that if you eat almost none of them - the diet he recommends includes 20 grams (less than an ounce) a day - you'll magically lose weight. Perhaps more appealingly, Taubes and Why We Get Fat also preach that you can eat as much fat and protein and you want and you'll never gain. Of course that's pretty much identical to the original Atkins' diet, and virtually all of the diets that Taubes himself references. You think that maybe, were it that easy the world would already be skinny? That low-carb would have continued its huge surge from the early 2000s (or the mid 1800s)? Why didn't it? Not because it doesn't work as for many folks it does - by means of folks on low-carb diets naturally eating, wait for it, smaller numbers of daily calories because they're not as hungry. No, ultimately I think that low-carb diets didn't continue to surge because most folks don't want to adhere to them as by their very definition they meet the classic definition of a "diet" - blind restriction and deprivation - things most folks don't want to live with for a lifetime.

At the end of the day Why We Get Fat's likely fate is to serve as the book for the next century's Taubes to point at, just as Taubes pointed at Banting's, in discussing how science has overlooked the horrors of carbs. Nothing new to see here, it's just another magic diet book.

The question that bothered me most throughout the book wasn't about carbs or thermodynamics, but rather why has Taubes chosen to argue his points like he's a Grade 9 student writing a high school science project rather than a well respected, scientific journalist? Some have suggested it's simply to sell books and have acerbically changed his name to Gary Taube$. I don't know, I think he was probably already doing pretty well for himself before this book was published. Perhaps the best explanation for why Taubes seems to have abandoned thoughtful journalism in place of this mess is his comment near the end of the book,
"One problem here is that when people, experts or not, decide to review the evidence on an issue dear to their hearts (me included), they tend to see what they want to see. This is human nature, but it doesn't lead to trustworthy conclusions."
Well said, and readers of this book, are best advised not to forget it.