Bad Science Receives Another Spanking

Around the time the USDA’s new dietary guidelines were released, a group of researchers published a rebuttal in the journal Nutrition, which I covered in a post titled The Dietary Guidelines Committee Receives The Spanking It Deserves.

One of the authors of that rebuttal, Dr. Richard Feinman, has meted out another well-deserved spanking on his blog.  This time the unruly children produced a piece of bad science they claim demonstrates that high-protein / low carbohydrate diets aren’t any better at managing type 2 diabetes than high-carbohydrate diets.  (That would certainly come as a shock the many readers who’ve emailed or commented to tell me how much their glucose control has improved since switching to a low-carb diet.)

I’ll just give you one highlight from Dr. Feinman’s post:  the “low carbohydrate” diet was, according to the self-reported intake of the subjects, 41.8% carbohydrate.  That might qualify as a Zone diet, but it’s certainly not a low-carbohydrate diet.  And here’s the real kicker:  I read the study, which shows that the “high carbohydrate” group consumed 48.2% of their calories as carbohydrates.

So a measly 6.4% reduction in carbohydrates didn’t make a big difference?  Well, I am shocked.  I guess we can all stop looking at carbohydrate restriction as a possible treatment for type 2 diabetes now.  Better fill those prescriptions for Metformin and be done with it.

There’s more to the spanking than that, however, so I’d urge to you read Dr. Feinman’s full post.  It’s always fun to watch a real scientist discipline the children.

Share

21 thoughts on “Bad Science Receives Another Spanking

  1. David

    “So a measly 7.1% reduction in carbohydrates didn’t make a big difference?”

    Actually there was less of a difference between the high protein and high carb group, a -6.4% difference between carb intake of the 2 groups (41.8% vs 48.2%). Regardless, not very meaningful changes when the high protein group went from 44.3% carb at baseline to 41.8% at 1 year, and the high carb group went from 45.5% to 48.2% over the same period.

    One small positive, even though overall fat intake was low, at least it was only ~20% PUFA.

    Argh! That’s what I get for writing quickly and not proofing my numbers. Thanks for the correction.

    Reply
  2. dlm

    As a diagnosed (12 HbAIC) type II diabetic (now at 6 HbAIC):: test my glucose with or without carbs — with carbs I am considered diabetic, without carbs I am considered without diabetes. No drugs involved, just sugar and/or starch.

    Think how many people could live without the drugs if only the ADA just admitted they’ve got it wrong.

    Reply
  3. Underground

    Put your helmet on, this is going to hurt.

    I just have to wonder if it’s incompetence, duplicity, or sheer blind acceptance that drives these people?

    I wish I knew. Something about a study like this doesn’t pass the smell test.

    Reply
  4. cbw

    Why, oh why, do so many who are (or could be) so intelligent prefer to stay so stupid?

    I mean, really, why even bother with doing a study with a 7% difference in your groups? What a waste.

    If only the spanking could have been a way more public flogging. That would have been the most fun ever! Sigh.

    But still, thank you Dr. Feinman, and Tom, for showing it to us — it’s awesome for any amount of wee-wee to get filtered out of the pool!

    The upshot of the study is that diet doesn’t help. That leaves us with drugs. That’s why I’m suspicious; it’s as if someone intended to prove diet doesn’t help.

    Reply
  5. Lori

    How much disclosure of financial interests are researchers required to give? Stock holdings in pharmaceutical companies? Membership in a family trust with such holdings? Honorariums? Family member on the board of directors? Loans that are “forgiven” (more common than you might think)? Unless their blood sugar is crashing too fast for them to think straight, researchers *have* to know they’re trying to sell us a bunch of horse pucky.

    From the study:

    J. E. Shaw has received grants, honoraria and speakers’ fees from: GlaxoSmithKline, Lilly Pharmaceuticals, Bristol
    Myers Squibb, Astra Zeneca, Pfizer, Merck Sharp and Dolme, and Novo Nordisk.

    Reply
  6. gollum

    Speculative theories ahead:

    Last time I looked at it, diabetes diagnosis was made on (roughly) the following criteria/methods:

    1. Textbook symptoms of insane fluid intake/excretion, ravenous appetite, hyperphagia but weight loss, pruritus etc
    2. Sugar in urine, Blood sugar

    (There is an advanced test for 2., the oral glucose tolerance test, but if you fail that it basically means your system is shot already, and it is usually only done when diabetes is really suspected)

    The textbook symptoms in 1, especially the weight loss, are more like DM1, or DM2 when your pancreas finally dies, and doctors are generally not caring enough unless you present as a textbook case.
    Criteria 2 just mean your carb metabolism is damaged so much that it is not funny any more.

    With some people, (especially phaenotype behemoth?), the pancreas works just so well that it all ends up in fatty tissue and liver, for years and years. (Hypothesizing. But I think they call it “metabolic syndrome”)

    This is like, you have cancer, and they let you run five miles, and if you fail that, you obviously must have adult-onset immobilitis. If you still manage to slep 5 miles, nothing is wrong with you and here is a referal to a real good shrink.

    Note also that you can probably ace the criterion 2 if your metabolism still works somewhat and you go low-carb, i.e. your liver is filled with like 7 % of its carb load capacity, then you can buffer lots of glucose away. Though I faintly recall reading something about the glucose/insulin tolerance parameter technically low on zero carb diets.

    Reply
  7. Milton

    Geez. A day after I say that I don’t feel that researchers are doing this deliberately, we get this report. Feinman makes a very compelling argument for the charge that they dismissed or discarded anything that went against the point that they wanted to make. Even for a well-meaning person, that’s not science.

    And LOL to anyone calling a diet with 41% of calories coming from carbs “low carb.” I get around 35% of my calories from carbs in any given day, and I’d never consider my diet low-carb. It might be low-carb compared to how I used to eat, but I don’t think we should measure our portions by using overdose levels as a baseline.

    There’s something fishy about this one in particular … as in data they apparently gathered but chose not to share, except as group totals.

    Reply
  8. Firebird

    “I just have to wonder if it’s incompetence, duplicity, or sheer blind acceptance that drives these people?”

    Money.

    Reply
  9. eddie watts

    the problem with the glucose tolerance test, in which they make you drink a glass of OJ i seem to recall? (surely a pure glucose drink would work better, but whatever)
    is that it is likely to show you as diabetic if you eat low carb consistently anyway. after all my daily carb intake is 20-50g on non-training days and maybe 150g on a training day. so a sudden intake of sugar that probably exceeds many of my meals in one go is likely to cause a negative response anyway.

    good write up though 😀

    Reply
  10. Nowhereman

    “The upshot of the study is that diet doesn’t help. That leaves us with drugs. That’s why I’m suspicious; it’s as if someone intended to prove diet doesn’t help.”

    Bingo, Tom, everyone. Follow the money trail here and see where it leads you. Who funded this study? I’d be really interested in finding out.

    The lead researcher has received funds from several drug companies.

    Reply
  11. Peggy Cihocki

    “J. E. Shaw has received grants, honoraria and speakers’ fees from: GlaxoSmithKline, Lilly Pharmaceuticals, Bristol
    Myers Squibb, Astra Zeneca, Pfizer, Merck Sharp and Dolme, and Novo Nordisk.”

    There you go.

    Thanks, Tom and Dr. Feinman (I read his blog on it, too.) If only the mainstream media health writers were as discerning!

    We can dream, can’t we?

    Reply
  12. 0mer

    The more I read about nutrition “science” the more I realize it’s about advancing careers and agendas than advancing patients’ health.

    Bingo.

    Reply
  13. namine

    For a moment there I thought you meant the great quantum physicist Richard Feynman. He saw the state of science from the after life, he woke up from the grave and wrote a paper to stop it.

    Here’s a creative photo of him:
    http://i.imgur.com/GpbYj.jpg

    Not the same Feynman — this one is Feinman — but he would agree with the sentiments about science.

    Reply
  14. Laurie

    “the problem with the glucose tolerance test, in which they make you drink a glass of OJ i seem to recall? (surely a pure glucose drink would work better, but whatever)
    is that it is likely to show you as diabetic if you eat low carb consistently anyway. after all my daily carb intake is 20-50g on non-training days and maybe 150g on a training day. so a sudden intake of sugar that probably exceeds many of my meals in one go is likely to cause a negative response anyway.”

    I had a GTT about 10 years ago (pre-low carb.) I was given glucose to drink, not OJ. You are right in that if you eat low carb on a regular basis, the results of a GTT will not be valid. People who eat low carb are told (or are supposed to be told) to eat high carb (at least 150 grams per day, I believe) for at least 3 days prior to the test for this reason.

    Reply
  15. Debbie

    Hmm. well lets not totally knock metformin. I’m no a big fan of drugs, but I do think metformin is one of the “good guys” as far as drugs go, and at about $8/month very affordable for most! I follow a low carb diet, but even a strict low carb diet doesn’t get my blood sugar down into normal range. So I take metformin too. I have to see where it goes. After a couple years of low carb I was diagnosed with hypothyroidism. Apparently a very low carb diet can cause this. The great Broda Barnes himself said patients should stay out of ketosis and take in at least 50g of carbs daily to prevent thyroid slowdown – so I’m trying to increase my carbs now to see if it helps, though I’m also now taking Armour thyroid medicine, one of the other “good guy” drugs, and less than $20/month for a prescription. 🙂

    Metformin and other drugs serve a purpose. Some people will need them, no matter what diet they adopt. The problem is that we have doctors telling people to eat lots of carbs and simply beat down the resulting high blood sugar with drugs. The goal should be to get as many people off the drugs as we can.

    Reply
  16. Goober

    Low carb does NOT cause hypothryoidism. In fact, it is recommended that one go on a low carb diet as T3/T4 control how your body metabolizes carbohydrates (i.e., poorly when you don’t have enough of those hormones). In fact, there may be a link to the imflammatory nature of insulin to the efficiency of the thyroid. Meaning that chronic inflammation, caused by high levels of insulin is causing the thyroid to malfunction.

    Since I have started a low carb diet, my thyroid numbers have improved. I’m taking a lower dose of Synthroid. The great Broda Barnes sounds like a typical physcian that makes assumptions based on bad science.

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *