Critiquing Gary Taubes, Part 2: Atkins Redux

Pre­vi­ously: Main­stream Nutri­tion Science on Obesity

Edit: In ret­ro­spect, I think it maybe should have com­bined this post with part 3. Un­for­tu­nately, the prob­lem of what to do with ex­ist­ing com­ments makes that hard to fix now.

Taubes first made a name for him­self as a low-carb ad­vo­cate in 2002 with a New York Times ar­ti­cle ti­tled “What if It’s All Been a Big Fat Lie?” When I first read this ar­ti­cle, I was get­ting ex­tremely sus­pi­cious by the sec­ond para­graph (em­pha­sis added):

If the mem­bers of the Amer­i­can med­i­cal es­tab­lish­ment were to have a col­lec­tive find-your­self-stand­ing-naked-in-Times-Square-type night­mare, this might be it. They spend 30 years ridi­cul­ing Robert Atk­ins, au­thor of the phe­nom­e­nally-best-sel­l­ing ″Dr. Atk­ins’ Diet Revolu­tion″ and ″Dr. Atk­ins’ New Diet Revolu­tion,″ ac­cus­ing the Man­hat­tan doc­tor of quack­ery and fraud, only to dis­cover that the un­re­pen­tant Atk­ins was right all along. Or maybe it’s this: they find that their very own dietary recom­men­da­tions—eat less fat and more car­bo­hy­drates—are the cause of the ram­pag­ing epi­demic of obe­sity in Amer­ica. Or, just pos­si­bly this: they find out both of the above are true.

When Atk­ins first pub­lished his ″Diet Revolu­tion″ in 1972, Amer­i­cans were just com­ing to terms with the propo­si­tion that fat—par­tic­u­larly the sat­u­rated fat of meat and dairy prod­ucts—was the pri­mary nu­tri­tional evil in the Amer­i­can diet. Atk­ins man­aged to sell mil­lions of copies of a book promis­ing that we would lose weight eat­ing steak, eggs and but­ter to our heart’s de­sire, be­cause it was the car­bo­hy­drates, the pasta, rice, bagels and sugar, that caused obe­sity and even heart dis­ease. Fat, he said, was harm­less.

Atk­ins al­lowed his read­ers to eat ″truly lux­u­ri­ous foods with­out limit,″ as he put it, ″lob­ster with but­ter sauce, steak with béar­naise sauce . . . ba­con cheese­burg­ers,″ but al­lowed no starches or re­fined car­bo­hy­drates, which means no sug­ars or any­thing made from flour. Atk­ins banned even fruit juices, and per­mit­ted only a mod­icum of veg­eta­bles, al­though the lat­ter were ne­go­tiable as the diet pro­gressed.

It’s one thing to claim that, all else equal, low-carb diets have ad­van­tages over low-fat diets. It’s an­other thing to claim you can eat un­limited amounts of fatty foods with­out gain­ing weight.

I’d heard of Atk­ins be­fore but didn’t know much about him. I got cu­ri­ous to know more about the man Taubes was cast­ing as the hero who just may have been “right all along,” so I popped over to the Wikipe­dia ar­ti­cle on the diet, which says:

Many peo­ple be­lieve that the Atk­ins Diet pro­motes eat­ing un­limited amounts of fatty meats and cheeses. This was al­lowed and pro­moted in early edi­tions of the book. In the newest re­vi­sion, not writ­ten by the now de­ceased Dr. Atk­ins, this is not pro­moted. The Atk­ins Diet does not im­pose caloric re­stric­tion, or definite limits on pro­teins, with Atk­ins say­ing in his book that this plan is “not a li­cense to gorge,” but rather that eat­ing pro­tein un­til sa­ti­ated is pro­moted. The di­rec­tor of re­search and ed­u­ca­tion for Atk­ins Nutri­tion­als, Col­lette Heimow­itz, has said, “The me­dia and op­po­nents of Atk­ins of­ten sen­sa­tion­al­ise and sim­plify the diet as the all-the-steak-you-can-eat diet. This has never been true”. How­ever, this new ap­proach by Atk­ins Nutri­tion­als is of­ten at odds with the ear­lier writ­ings of Dr. Atk­ins.

The last sen­tence of this para­graph is helpfully marked “cita­tion needed,” leav­ing an un­re­solved con­flict be­tween what­ever Wikipe­dia ed­i­tor wrote the para­graph and what the Atk­ins folks (at least now) claim. I or­dered a used copy of the origi­nal 1972 edi­tion of Atk­ins’ book through Ama­zon, and what I found sup­ports the Wikipe­dia ed­i­tor. The folks cur­rently in charge of Atk­ins Nutri­tion­als are white-wash­ing.

The sen­sa­tional “truly lux­u­ri­ous food with­out limit” quote in Taubes’ ar­ti­cle, for ex­am­ple, can be found on page 15 and comes with no con­text that would make it more rea­son­able. In fact, lest any­one mi­s­un­der­stand it, it’s fol­lowed by a state­ment that “As long as you don’t take in car­bo­hy­drates, you can eat any amount of this ‘fat­ten­ing’ food and it won’t put a sin­gle ounce of fat on you.” (In the book, this is ital­i­cized for em­pha­sis.)

Atk­ins ac­knowl­edged that most of the peo­ple who used his diet ended up eat­ing less over­all, but claimed that some of his pa­tients had lost sig­nifi­cant amounts of weight eat­ing 3,000 calories per day or more. In one case, Atk­ins claimed, a man had lost fifty pounds on a diet of 5,000 calories per day. He at­tempted to ex­plain this by in­vok­ing the fact that ex­tremely low-car­bo­hy­drate diets will cause peo­ple to ex­crete ke­tones (which Atk­ins referred to as “in­com­pletely burned calories”) in their urine. How­ever, as a state­ment on the Atk­ins diet put out by the Amer­i­can Med­i­cal As­so­ci­a­tion ex­plains:

When ke­tone ex­cre­tion in­ci­dent to such diets has ac­tu­ally been mea­sured, it has been found to range be­tween 0.5 and 10 gm/​24 hr. Stud­ies car­ried out on starv­ing nondieabetic per­sons in­di­cate that at most about 20 gm of ke­tones per day may be ex­creted in the urine. And, as Folin and De­nis have show, the to­tal ace­tone ex­cre­tion with the breath is quan­ti­ta­tively in­signifi­cant; at most, 1 gm/​day. Since caloric value of Ke­tones is about 4.5 kcal/​gm, it is clear that, in sub­jects on ke­to­genic diets, ke­tone losses in the urine rarely, if ever, ex­ceed 100 kcal/​day, a quan­tity that could not pos­si­bly ac­count for the dra­matic re­sults claimed for such diets.

As far as I can tell, no­body to­day defends Atk­ins’ origi­nal “ke­tones in the urine” ex­pla­na­tion for how his diet sup­pos­edly works. It’s not en­tirely clear to me what was go­ing on with the pa­tients Atk­ins claimed lost weight on a high-calorie diet, but it wouldn’t be sur­pris­ing if a minor­ity of his pa­tients had sim­ply mis­judged their caloric in­take. In spite of this, Taubes still ap­pears to want to defend Atk­ins’ most ex­treme claims about peo­ple be­ing able to eat un­limited fat with­out gain­ing weight.

This isn’t en­tirely ob­vi­ous when you read his books Good Calories, Bad Calories or Why We Get Fat, which go for a slightly less sen­sa­tional pre­sen­ta­tion than the Times ar­ti­cle. Nev­er­the­less, in the epi­logue to Good Calories, Bad Calories, he claims that “Die­tary fat, whether sat­u­rated or not, is not a cause of obe­sity, heart dis­ease, or any other chronic dis­ease of civ­i­liza­tion.” There’s a sense in which that claim might be some­what plau­si­ble, if he meant that it’s to­tal calories, not fat per se, that’s the main culprit in all those prob­lems. But Taubes also puts a lot of en­ergy (no pun in­tended) into at­tack­ing the main­stream em­pha­sis on calories.

Why We Get Fat, for ex­am­ple, con­tains claims such as:

[A 1965 New York Times ar­ti­cle claimed that] “It is a med­i­cal fact that no dieter can lose weight un­less he cuts down on ex­cess calories, ei­ther by tak­ing in fewer of them, or by burn­ing them up.” We now know that this is not a med­i­cal fact, but the nu­tri­tion­ists didn’t in 1965, and most of them still don’t. (p. 161)

But we know now what hap­pens when we re­strict car­bo­hy­drates, and why this leads to weight loss and par­tic­u­larly fat loss, in­de­pen­dent of the calories we con­sume from dietary fat and pro­tein… If you re­strict only car­bo­hy­drates, you can always eat more pro­tein and fat if you feel the urge, since they have no effect on fat ac­cu­mu­la­tion. (p. 174)

No effect? That’s a strong claim. And as we’ll see in the next two posts, Taubes’ ev­i­dence for this claim ends up con­sist­ing largely on a se­ries of mis­rep­re­sen­ta­tions of main­stream nu­tri­tion sci­ence, which al­low him to pre­sent his views as the only al­ter­na­tive once he’s knocked down his straw men.