What we really learn from Ebbeling et.al.

I think Carson Chow's observation that the TEE's measured in the supposedly weight stable phase of this diet, that varied from an average of 140 to >500 cal/day more than the average intake pretty much negates drawing any sort of meaningful conclusions from the  Ebbeling et.al. study.  It really is a waste of time and mental energy to discuss a highly suspect 300 cal/day difference in TEE measured by state of the art methods, but highly susceptible to error as it was not measured directly.

But what can we learn from this study?  Well, as with the Biggest Loser study we learn that:

• Obese people, on average, eat more than the average American eats even today, and even more than the average American ate in the 70's -- if baseline TEE is anywhere near accurate they were consuming around 3200 cal/day.  
• When you take people eating 3200 cal/day and restrict calories to 60% of weight maintaining levels (avg around 1900 cal/day) they lose weight and fat.  Period.  They lose weight because they are in calorie deficit.  Period.
• Calorie counting didn't fail for weight loss.  (Somebody alert Gatewing!)
• When you lose weight, your metabolic rate tends to go down
• After losing an average of 30 lbs, intake to stabilize a roughly 200 lb weight averaged around 2600 cals/day (SD almost 700), still quite high.  
• Even when you pay people and provide and prepare all of their food for free for almost 8 months, some have difficulty with complying.
• There is a phenomenal amount of individual variability in EE, both resting and total.   Measured across two weeks, one person had a TEE  almost 800 cal/day above their baseline while another had a TEE almost 1200 cal/day below baseline.  That's huge.  I imagine that the individual whose REE declined by almost 500 cal/day will be having a tougher time maintaining his/her weight loss than those lucky few who saw increases. 
• All subjects lost weight consuming over 200g carb/day
• Subjects maintained weight eating almost 400g carb per day average on the LF diet.  No insidious weight gain ensued.  

Now for the bottom line take home message.   Unless the researchers are lying to us by omission., within the macronutrient ranges investigated in this study:

The quantity of calories mattered,

The quality of those calories did not matter

Think about this one long and hard Gatewing and Gary Taubes and anyone still suffering from CDS (calorie denial syndrome), and perhaps go watch and rewatch the BBC program on YouTube I linked to HERE -- and pay especial attention to the metabolic chamber study with the twins.  

Because in the Ebbeling et.al. study, the weight stable caloric intake of the participants was first estimated using weight loss phase intake and rate of weight loss.  Their intake was ramped up accordingly and then the subjects stepped on precise and accurate simple direct measuring device ... a scale ...  every day  ... for around 4 weeks, and intake was altered in response to gains or losses until their weight remained stable.  The researchers allowed for varying timeframes to stabilize weight.  Although they don't expand on the criteria they used to determine this (e.g. X days to within +/- Y kg change), but before going into the test phase these subjects were in caloric balance and weight stable by simple, direct measures.   If they were not, there really wasn't any point in doing this study, so I tend to think the researchers took care in this phase.  

Then they kept calories constant for a further 12 weeks, varying the macronutrient ratio.  Heck, they even varied protein in a way that should have favored the LC diet by thermogenic quotients.  But they kept calories constant and at the end of each 4 weeks on each diet, the average weight of the group stayed as near as dead-nosed constant at around 200 lbs ...  plus, we're not talking about people who don't have more fat to liberate here either.   Their fasting leptin and cortisol levels bounced around, they stayed weight stable.  Their insulin sensitivity bounced around (levels are not reported but no doubt those bounced around too), and they stayed weight stable.   Their inflammation markers bounced around, and they stayed weight stable.   The authors make mention that calories were held constant regardless of weight fluctuations in the test phase.  

So this is why I say either this study has demonstrated for once and for all that there is NO metabolic advantage to a LC diet, and weight loss/gain is ALL about the calories ... or the researchers are dishonest and withheld information that the subjects weights bounced around enough to influence the outcomes.  I don't think this is likely given that the average weight of the group stayed so consistent, and if they did this is unacceptable deception on the parts of the researchers.  I'm going to presume we're dealing with honest, well-meaning scientists here.  Eades, Fat Head and now Peter D, and all the rest certainly didn't question them as they tweeted and blogged exuberantly over some supposed  300 cal/day metabolic advantage.

They determined weight maintenance calories the old fashioned way -- 21 little "n=1" experiments if you will -- measured and adjusted intake until the number on the scale stayed constant within some acceptable variation.   Simple.  Nothing fancy.  No smoke and mirrors or statistical gymnastics.  And then, they varied the macros and controlled for calories.  And the average weight of these 21 people stayed virtually the same for the next three months.   Simple.  Nothing fancy.  No smoke and mirrors or statistical gymnastics.

CICO has been upheld.  The EL of ELMM worked as it has every time it's "practiced".  Amen and goodnight :D