Anyone who has read Martha's story and put her narrative together with the folks in Phinney's 1980 study will have immediately wondered: How many of Phinney's subjects were lactating? Even just a little bit?
I think we can say, pretty categorically, that none of them were lactating. Gluconeogenesis from lipid is very likely to have been occurring but obviously (now) this can only drop the RQ when the glucose produced is not being oxidised. Clearly my initial idea expressed in the Phinney post is wrong.
Martha is easy, her child took the sugars hence the spectacularly low RQ. Trying to explain why a protein supplemented fast should drop the RQ below 0.69 needs a little more thought.
This is what Phinney thought might be happening under moderate exercise:
"The low RQ value of 0.66 observed during the final exercise test was surprising, as long chain fatty acid oxidation occurs at an RQ of 0.69. (The only common fuel oxidized at a lower ratio is ethanol at 0.67). The answer to this disparity may lie, at least in part, in the rise in serum ketone concentration observed during exercise. As the hepatic production of ketones from long-chain fatty acids occurs at an RQ of zero, a net retention of ketones in body fluids will result in a reduction in observed RQ due to non steady-state conditions. By calculating the increase in the whole body ketone pool associated with exercise, one can account for approximately half of the decrement of CO2 production that would be necessary to explain the decrease in RQ below 0.69. Other factors that could contribute to this low RQ include losses of ketones in the urine and loss of acetone in the breath after decarboxylation of acetoacetate in the blood, as well as CO2 utilized in urea genesis".
However, the non steady state accumulation of ketones does not apply to the at-rest readings from the Eskimo in Heinbecker's study.
I'd like to have a guess at the more "steady state" condition.
Full oxidation of a "typical" protein such as albumin produces a value of around 0.8 for the RQ. So I've invented a single mythical amino acid which gets close to the average RQ of protein. It looks like this:
NH2 - CH - COOH
Two of these amino acids oxidise using nine molecules of oxygen to give one molecule of urea and seven molecules of CO2, giving an RQ of 0.78. If this was replaced with a dietary equivalent the RQ would stay around 0.8 and the RQ of 0.69 from saturated fat would be increased somewhat. If the oxidised amino acid was not replaced the RQ change would be exactly the same but muscle wastage would occur.
What if, as a ketosis induced protein sparing effect, certain non-essential amino acids, were synthesised from urea plus carbon from fats plus a little oxygen. I'm not suggesting for a moment that this is exactly what happens, but the equation must balance whatever pathways might be used.
I've spent quite some time with scraps of paper working out how much oxygen has to be added to a couple of -CH2-CH2- moieties from saturated fat, along with a urea molecule, to reassemble the above pair of amino acids. "Mythical" protein turnover...
It takes 3O2 and liberates one CO2.
Combining this with the 9 O2 and 7 CO2 from oxidation, the whole repalcement of this amino acid would use:
12 O2 and generate 8 CO2 giving and an RQ of 0.67.
So the replacement of one "typical" amino acid using part of the acyl-chain of a saturated fatty acid generates an RQ of 0.67.
That's getting us somewhere below 0.69, what then matters is how general this effect might be which obviously depends on protein turnover, protein intake, protein quality and anything else anyone can think of. The value is pushed further down by the loss of oxygen rich ketone molecules through the breath and urine.
I'm very aware that minor errors in logic or arithmetic might alter the above calculations.
What an RQ well below 0.69 speaks very clearly against is gross muscle catabolism (which pushes the RQ upwards towards 0.80). Clearly, muscle loss does occur but I can see no reason why muscle loss should be an essential pre requisite for fat oxidation during fasting. The ability to minimise muscle loss under fasting (or ketogenic eating) might just provide some advantage on an evolutionary basis.
Marking out amino acid oxidation (ie loss of protein) as an essential pre requisite to fatty acid oxidation (in the absence of carbohydrate) suggests a rather odd view of reality. If it were correct it should show as elevated RQ's above 0.69 in proportion to the amount of amino acid oxidation which might be going on.
Which is not the case.
Edit for raphi: the arithmetic: