NuSI ... or Just Do It!

Imagine if you will, that a decade or more ago someone told you that if you really wanted to build your journalism career tearing down scientists, you needed to look at diet and nutrition research.  You picked up a copy of Atkins and lost a couple of pounds.  You began cultivating your schtick that simple obesity (the kind of the epidemic) was not about overeating or sedentary behavior, it was something else.  And you received three-quarters of a million dollars to research and write a book.  The book was a best seller and garnered you a cult following in certain circles and you were off to the races giving talks around the globe.  

There were doctors and others who bought into your hypothesis, spreading the word as fact and growing the movement.  But a following from one segment of the population remained elusive:  scientists.   Rather than address the reasons why (serious holes in your hypothesis), you double down and trash scientists more openly and frequently.  You also publish a second book with a dumbed down, even less scientific, hypothesis.  By all indications, this book falls far short of expectations ... and still the scientists are not coming around.  Worse still, many who had believed your hypothesis was plausible, at least worthy of investigation, were abandoning it.

Then a young used-to-be doctor who dropped out of the field after residency reads your book and hooks up with Volek & Phinney and you and declares war on insulin!  You get together and an idea is born.  Fund your own real science to show the world!  Your fears of ending up that old man muttering to yourself because nobody listened to you, perhaps playing the Fat Louisa card game some fan made for you, are abating.  

And then, it happens.  A major funding source materializes.  What to do, what to do.  Do you:
1.  Get right to work and fund a study to test your hypothesis ... because this is really important and a decade is already too long for failed conventional wisdom-based advice making people fatter and sicker.   Or ...
2.  Assemble boards of directors and advisors, talk up scientists, set up a website with nicely produced media, open offices, hire researchers and other support staff and announce the launch a grand initiative:  NuSI.  
Haven't we been hearing for long enough how bad the obesity research is?  About all of those poorly   designed studies, etc.etc.  Well, if you have the money, start out funding that one study and more shall follow.  I mean Gary Taubes is nothing if not certain and convinced what the results would be.  So why not cut to the chase ... Just Do It!!  Heck you can get scientists to do a lot of the work for you for free (grant proposals!), but there's not much work required to tweak existing study protocols to fix them.  I mean they've already done a "massive" review of the literature to tease out the problems with every existing study ... so just fix those problems!  

They might have to first go back and rescue the reputation of free living studies, because the sort of lengthier studies they're hinting at are not practical in metabolic wards.  The sort of "controlled living" study like that recent Ebbeling one seems like a workable compromise (just do some analysis of those diaries and report, in some fashion, the degree of compliance). NuSI would also have to back-pedal on this notion that the macro changes in Grey & Kipnis were insufficient to demonstrate an effect of insulin.   Because as I pointed out previously, you can't blame the obesity epidemic on an increase in carbohydrate content in the diet of only 7 percentage points, and then claim that reducing carbs to 15-20 percentage points below pre-epidemic intake is insufficient to demonstrate the effect.   Of course acknowledging this would be to eliminate the need for NuSI, but let's do the grand study anyway!

Mildly overweight 20-somethings might well be the best population with which to start -- show a tendency to put on weight under the right conditions, but haven't.  They'd also have a little "extra" fat to lose, if underfed w/o becoming underweight (which is one problem with drawing much of anything from the Minnesota Experiment).  Monitor (closely, ward) their habitual intake for an entire month to determine its average quantity and composition.  (Weigh daily to ensure weight stability).   Next design a shake & muffin style diet of equivalent caloric/macro composition and over a period of 3-4 weeks or so, begin with their current cals/comp and transition them to the test diet cals/comp.  Then switch them to a real-food test diet for a long time, 6 mos to 1 year preferably.  Using a non-descript diet to transition should help mask somewhat any placebo effects.  This study would be too long to do in cross-over fashion, so it's probably going to have to be quite large in terms of the number of participants.  But remember we're talking about the grand experiment to settle a century old question!!  

Measure everything, as directly as possible.  Weight, body composition, hormones, glucose, lipids, RMR, TDEE, RQ, inflammation markers, heck, get me a smear of gut biota too!

So here would be my proposal for one way to do this with seven study diets.  First, absolute protein consumption will be held constant throughout, on an individualized basis, at habitual diet levels.   The IC diets would maintain habitual absolute carbohydrate intake.  The LC diets will reduce absolute carb intake to some fraction of habitual levels.  The isocaloric HC would increase carbs from habitual by the same proportion they are decreased in the LC diet, and hypercaloric HC diet would add 20% surplus calories entirely as carbs to habitual carb levels.  The remaining calories in all cases will be accounted for by adjusting fat intake accordingly.  

Here's how this would look like for an average subject habitually eating a 2700 calorie "standard" 15% protein, 35% fat, 50% carb diet where LC = 2/3rds habitual levels.  (shaded block are set values)
  1. Isocaloric HC
  2. Isocaloric LC 
  3. 80% Hypocaloric IC 
  4. 80% Hypocaloric LC
  5. 120% Hypercaloric IC
  6. 120% Hypercaloric LC
  7. 120% Hypercaloric HC
Wanting to cover all the bases, you'd have an eighth Control Group that is transitioned back to eating the habitual diet.  This would control for any effects of all that monitoring, logging, etc.   The types of foods must be comparable ... at the very least, the CG should be fed the same calories/composition of the types of foods used to comprise each "real food" study diet.  In this regard, the inclusion of a CG is most important of all ... even if this study is done in pieces.

Now I expect there to be loud cries of "it's not a real LC diet".  But, as I've pointed out repeatedly now, using a very low carb diet -- by percentage or content -- is inappropriate for testing any hypothesis as to what has caused our obesity epidemic.  Indeed even the macro percent swings produced by the reduction to 2/3rds habitual intake would result in several times the percent differential seen between the American diet from the 1970's to today.  If a shift in macronutrient composition of roughly 7% in favor of carbs v. fat is responsible for the obesity epidemic, then the differentials in carbs in this study should be sufficient to have some measurable effect.  

We don't eat percents, we eat amounts.  So to see if carbohydrates impact fat accumulation independent of calories, we need to hold those constant by amount.  If it's the carbs/insulin and not the calories, TWICHOO predicts:
  • The Control Group will remain weight stable (as also predicted by CW)
  • Study Group 1 will trigger their horizontal growth disorder (HGD) and gain weight/fat while Study Group 2 will lose weight/fat.  At a very minimum they will recompose if not gaining/losing net total body weight.
  • Study Groups 3 would not lose weight since this group is consuming the same carbs as they were habitually consuming that will keep their HGD accumulated fat trapped in their fat tissue.  If weight is lost it will be predominately lean mass.
  • Study Groups 2, 4 & 6 should all lose some comparable amount of fat mass.  They would be the same deficit carbohydrate level at different caloric levels.  If it were carbs and not calories, they should all release their HGD accumulated fat equally.  Any other mass changes would have to be lean mass.
  • Study Group 5 would not gain weight, fat or otherwise.  This is the most direct test of TWICHOO dreams, that dietary fat has no impact on bodyweight/fat because it does not elicit an insulin response, that it is not overeating but {{cringe}} overstoring.  This group is consuming the same amount of carbohydrates they always did.  This is why that Control Group is important.
  • Study Group 6 should lose some fat, or at the very least recompose to have more lean mass.  If Group 5 is the most direct test, Group 6 is the best challenge to TWICHOO dreams.  Do you suppose anyone reading this believes Group 6 will lose fat mass because they are only eating a relatively low level of carbohydrates?  What?  Not low enough?  Well 225g/day is well below the average SAD consumption we are told is responsible for our collective girth ...
  • Study Group 7 should succumb to accelerated HGD vs. baseline, but because of the way the carb content was determined, they aren't consuming that many more carbs than the isocaloric group (473 vs. 450), so their gains should be comparable to Study Group 1's.  Remember, it's not overeating that causes the obesity epidemic, so the fact they are consuming 540 cal/day more than they were habitually consuming should have additional impact.
Does anyone believe this is what would happen here?  I don't even think Gary Taubes believes this would happen or he's managed to completely delude himself through rote and repetition of the "carbs drives insulin drives fat accumulation" meme he's been droning on and on with for the past five or so years.  So, now let me ask you this:  Do you think Gary Taubes really wants this study done?  If the results came out as TWICHOO predicts, it would settle things for once and for all.  Busts of Gary Taubes would don the halls of institutions studying obesity for centuries to come.  The USDA would immediately issue new guidelines based on The Taubeskins Diet ... Panera Bread and IHOP would be put out of business immediately, meatzas would replace pizzas on school lunch menus, and deep fried bacon wrapped dogs would replace hot dogs in a bun.  

Most importantly, Gary Taubes will be taken seriously.  But sadly, there would be no further use for NuSI because for once and for all the question has been answered.   Veil as they may with big talk about bringing together the best minds to design the best studies and fund the best science, the purpose of NuSI is to promote TWICHOO.  To cloak the dying hypothesis with legitimacy so that the franchise may live on.  Only it's even better because they get to feel good about all that philanthropy in the air.

Call me a cynic, but I'm not a big fan of non-profit advocacy groups and foundations that solicit donations in general.  A truly philanthropic act is to give of goods, services and/or funds from one's possessions.  It is not even philanthropy to direct a foundation founded with the fortunes on one's parents, grandparents or whatever.  To me a HeirFoundation head who pays him/herself an inflated salary, writes off their mansion/estate and other assorted living and travel expenses to their foundation, and presides over lavish benefits, self-flagellating dedications and fundraisers is not philanthropic.  The same goes for all of these celebrities and such that advocate for various causes, usually involving soliciting donations from you, the little person.  

So all non-profit means is that the corporation makes no profits, but the individuals working for such organizations often make far more than their for-profit private industry counterparts.  Many are nothing more than capital-sucking middle men.  The various environmental advocacy groups come to mind as we get quite a lot of them soliciting around my area.  Many of these organizations sit on large sums of money that is used, in part, to raise more money.  Someone's got to do the fundraising, someone's got to organize and supervise, and you need payroll personnel, etc.etc.  Most of the time they are raising money to lobby the government to allocate money to do something -- e.g. clean up a dump, invest in the development of alternative energy, fund studies, subsidize products directly or with tax breaks, etc.    It's always bothered me the enormous waste of capital these groups end up being all too often.  What if the organization so eager to see the Toxic Dump cleaned up just raised money to clean up the dang dump and be done with it?  What if instead of trying to lobby for taxpayer funding of alternative energy sources, they just give a chunk of $$ to a university or start up company actually developing some technology?   

Which brings me back to NuSI, because this seems to be what it will be: a capital-sucking middle man.   Or did I miss the part where they plan to build and staff an independent metabolic research facility?  Gary Taubes has known for a very long time the type of studies he now claims to want to fund.  He's said as much in his very first Jimmy Moore podcast and been bemoaning it in front of audiences at medical schools for half a decade now.  Now it might well be prohibitive to build a facility first, but why not seek out a qualified existing research group/facility and just contract them to conduct a study?   It really need not be some grand study (does anyone really expect the first study out of NuSI to be?) but just get to work.  You don't need this fluff and puff show what with advisors and maps with arrows and how you are going to be bringing scientists together (the cream of the crop!) to finally answer the real problems.  It's all big show, and a considerable amount of money is going to be wasted to compensate the members of the organization rather than realizing the main (purported) goal of enabling riskier, longer and larger studies to be undertaken.   

Because in the end, if we know nothing about Gary Taubes, it's that he thinks he knows science, and he sure as heck knows scientists.  From Journalist Gary Taubes Raises Bucks to Disprove His Diet Theory we learn that 
He [Taubes] called me a couple weeks ago to tell me a remarkable story. John Arnold, a wealthy financier, recently contacted Taubes after hearing him talk on the radio about the need for better research on diet and obesity. Arnold ... wanted to help provide funds for such research.  That conversation led to the creation of the Nutrition Science Initiative, or NuSI, ... The Arnold Foundation has put up money for NuSI’s first two years of operation.
So a billionaire has the bucks to fund whatever he wants.  He contacts Taubes and instead of Gary phoning up his creme de la creme science buddies, let's set up another organization.  Private bureaucracy is what it is.  Isn't part of the benefit of private funding to side-step the tedious process of public funding?   There's no need to do a public announcement, call for proposals, ensuring every equal opportunity "t" has been crossed and incentive "i" had been dotted.  Sounds like John Arnold trusts Taubes to pick the project ... or not?  According to the article:
NuSI collaborators—from Duke, the University of Massachusetts and other schools–are now planning rigorous tests of weight-reduction regimens. Most dietary studies, Taubes points out, rely on what subjects say about their eating and exercising rather than measuring what they’re actually doing. Some studies have observed subjects under tightly controlled conditions, with food and energy expenditures precisely monitored, but for obvious reasons such studies generally involve few subjects and don’t last long. Taubes hopes NuSI will carry out highly controlled experiment—perhaps involving special observation chambers that monitor energy expenditures—involving statistically significant numbers of subjects and lasting at least three or four months.
What was it that I was saying about "a bar too high"??  Are the studies already in the planning stage?  Exactly how much planning does it take to repeat any study plagued by whatever-shortcoming and merely fix the shortcoming to do it right??  But if Taubes is for real here, he's not only talking metabolic ward study, he's talking metabolic chambers?  Really??  Really, really???

I do believe that if NuSI does do the studies Gary envisions -- even an inferior mostly free-living study such as I described -- he will indeed be funding the studies that disprove TWICHOO for once and for all.  Therefore I fully support NuSI at this point in time.  Just put your money where the science is.   Drop the pretense of any objectivity, design *the* study to get the ball rolling, recruit a research group to do it, use Arnold's money to fund the friggin thing and GET that ball rolling.   These are critical days, after all.  



an3drew said…
410 Days : 13 Hours
27 Minutes : 45 Seconds
cremes said…
Any chance you can put in a proposal for a grant to do a study? If not, why not?
James Krieger said…
NUSI is starting to sound like those "intelligent design" organizations that try to provide an "alternative hypothesis" to evolutionary theory.
Anonymous said…
Why do that when you can just head up an organization? Why do anything that shines the light directly on you instead of other people's (researchers') shortcomings?

On the other hand, it's just as easy to put yourself in a TV show (paging Dr. Oz) and involve the viewing audience in an experiment not unlike watching a David Copperfield magic trick: sequestered in a studio 'chamber' with TV, books and magazines, and no need to exercise beyond walking in the room (wear a pedometer), chow down on 7,000 calories a day for 14 days - fat and protein and the absolute minimum of carbs (oh, go for it - make it as close to zero as you can get) and weigh in every few days. Call and text friends and family. Watch DVDs. Wait for the weight to drop off of your body. Bring plenty of changes of smaller-size clothes!

That will catch everyone's attention and turn the scientific world on its ear.

I've been waiting for Taubes to do this. It introduces a new kind of reality show. Forget the Biggest Loser. Experiments will be conduced live (a website supplements the usual TV production, so you can check in live on what's happening). Network TV should be interested. Not only Taubes on 7,000 calories a day. Cameras will follow the Kardashians and record everything they eat and what they weigh.
MM said…
I really like your study design. However, I would very much like to see a VLC arm also. Even if 225g/day is lower than the SAD, it is still well above Sisson's "insidious weight-gain" level. I'd like to see the VLC arm to be really low, like 40g/day. So at 2700 calories and 15% protein, you'd have 101g/day protein, 40g/day carbs and 237g/day fat. (One stick of butter is 120g. It sounds totally unpalatable to me. I don't even want to think about the hypercaloric VLC diet.) I suppose if it failed to work as TWICHOO predicts they'd blame it on the "too high" protein level, right?
Sanjeev said…
1 have they stated anywhere
1a what effect sizes will be worth pursuing into the future
1a what effect sizes will make them abandon the BS completely

2 which pieces of Taubes' extremist corpus will each study be designed to prove/disprove

If his feet are not held to the fire and he's required to state all of this UP FRONT

2a I suspect Taubes will trumpet any positive result of any effect size will be as a resounding proof of all of his positions - on insulin, infinite calories, carbohydrates being satan incarnate and all nutrition scientists being morons.
Lerner said…
Ahaaaaa, so that's what the whole NutSI thing is about - Taubes et al have wrangled themselves a steady income from all this smoke and mirrors.

So, don't forget that Section VII of IRS Form 990 for non-profits makes them eventually reveal what compensation they're all getting. You know, for saving the world and all. I did see on that site that Taubes isn't at all trying to prove that he's right, he just wants to help people. Let's see what % of yearly outlays goes to experiments, and what % goes to paying the cast of NutSI characters. Also, be leery of any public claims about 'salary' because these types often makes a lot in so-called 'other compensation'.
CarbSane said…
Ahaaaaa, so that's what the whole NutSI thing is about - Taubes et al have wrangled themselves a steady income from all this smoke and mirrors.

Thank you for writing my Readers' Summary :D

CarbSane said…
You're probably right ... they'll never be happy until the VLC thing is tested. So lets add two more study groups -- VLC to average <50g/day (total) carbs, protein at habitual levels and the rest fat. Hypercaloric will be more fat. Jimmy would be in heaven ;)
CarbSane said…
The Sci American article leads one to believe there's been some work done already ... as did Tim Ferriss elude to bringing scientists from across the country together. If they don't clear the bar they set, WILL anyone hold their feet to the fire? It will really only be the Arnold foundation they have to answer to ...

Gabriella Kadar said…
I'm not volunteering. My quality of life means too much to me.
CarbSane said…
Ya know, there's nothing keeping Taubes or anyone from sponsoring a "Team Taubes" on TBL.
Sanjeev said…
If I had the money one of the first things I'd like to do is a pilot study using cell phone cameras and accelerometers - have the subjects take pictures of everything they eat and wear a fitbit type thing everywhere & @ almost all times

Or this could be done as a side-study of a bigger study - a separate control group (or 2)

If these tools could be validated as useful adjuncts to free living studies (could even be useful in pseudo-ward studies) it would save future costs tremendously.
Well done Evelyn, today you hit the nail on the head, drove it home and tucked it into bed but maybe we should hold off calling this new entity NutSi as catchy as that is, it's getting a little too close to the nomen of Colby Vorland's thoughtful nutritional blog nutsci.
CarbSane said…
Yeah, according to Colby someone tried to buy his Nutsci domain name. I'm betting it was the NuSI folks. Feel bad for Colby if he ever does something brandwise/commercial.
CarbSane said…
Wouldn't a dental nutrient sensor be cool?
Sanjeev said…
is something like that on the horizon?

I've been reading of very minimally-invasive implantable silicon chips loaded with dozens to hundreds of sensors for a long time but they always seem 5 to 15 years away.

I was thinking, if the current granting bodies won't go for making studies cheaper for large numbers of participants and better-monitored using current off-the-shelf, cheap technology, it's what an innovative new agency would go for to establish itself in the field.

Combine with tineye and/or google images type technology to match food against what the study participant just ate and you have the makings of uber-capable diet monitoring for scientific studies and for the general public.
Unknown said…
I emailed NuSci and asked them what the compensation for Dr Attia and Gary Taubes will be. The automated response says I should get a response within a week. What do you think they will say?
Lerner said…
Charles, you broke me up with that :)

My prediction: a digital recording of crickets

No one can expect NutSee scientists to take time out from saving the world to answer. Hey, when Gary gets his Nobel Prize, I think they should play the theme from 'Man of La Mancha'.

But wait, it turned out that Don Quixote thought he was saving the world, while in actuality he was just a megalomaniac having only a tenuous contact with reality. Delusions of grandeur and all that. That's nothing like Gary.

P.S. Sites like have some data on some non-profits, like maybe your local pbs station.

Other ways: How to Find Non Profit Salaries

CarbSane said…
It will be a year or so before the 2012 financial statements become public as is required by law for all 501c3's. Wouldn't looking up this sort of stuff be stalking though? LOL
Lerner said…
"It will be a year or so before the 2012 financial statements become public..."

Yep, that's why I'd said "eventually".

"Wouldn't looking up this sort of stuff be stalking though?"

That's why I was saying it so you didn't have to :)

Josh said…
To borrow a Matt Stone phrase - This phallic infographic seems to suggest that NuSI has a pretty firm boner about grains and sugar being responsible for the obesity epidemic - NuSI - Changes in US Food Availability
Simon Carter said…
Come on, Evelyn, we know we can rely on you to hold their their feet to the fire! :)
Unknown said…

Lack of evidence for high fructose corn syrup as the cause of the
obesity epidemic

High Fructose Corn Syrup: Tasty Toxin or Slandered Sweetener?

The Sweetener Wars – HFCS Strikes Back
RRX said…
Unknown said…
Imo this isn't about drawing a salary it's about maintaining or increasing your stature, if you aren't an actual scientist you at least want to be on the board of something sciency.
CarbSane said…
In Taubes' case you're probably right -- though I'm sure he won't be serving for free. In Attia's case, given how he's a workout freak (2-3 hrs/day) and has a family (wife and kid), this appears to be a career move. A way to do something with the MD, perhaps.
Sanjeev said…
Heck no. Finance & banking is where it's at.

excepting places like Google & Apple It's been a LONG time since science & engineering have been "cool".

Outsiders sometimes think quants on Wall St. have it made.

I've heard the opposite from fellow engineering grads (analysis, systems modelling & numerical methods Masters & PhD level folks) who went into finance R&D / proprietary desks.

Eric Jansen's "geeks versus jocks" article summarized it pretty well.
Lerner said…
Alrighty, let's post predictions on what Taubes will rake in. If he is more than just a figurehead, if he does work on it most weeks but not anywhere near full time, then I'll guess he'll take $150K off the top. Anybody else want to make a guess? Guesses will stay here in the comments like a time capsule, for when the day arrives that the info becomes known.

I'll also guess that the study design, much vaunted as "settling things once and for all", will really be designed as much as possible so as not to fail. However that will be in practical terms, I don't know - but people looking at it will say that it seems oddly designed.