Tuesday, July 31, 2012

The 14-percent advantage of eating little and then a lot: Putting it in practice

In my previous post I argued that the human body may react to “eating big” as it would to overfeeding, increasing energy expenditure by a certain amount. That increase seems to lead to a reduction in the caloric value of the meals during overfeeding; a reduction that seems to gravitate around 14 percent of the overfed amount.

And what is the overfed amount? Let us assume that your daily calorie intake to maintain your current body weight is 2,000 calories. However, one day you consume 1,000 calories, and the next 3,000 – adding up to 4,000 calories in 2 days. This amounts to 2,000 calories per day on average, the weight maintenance amount; but the extra 1,000 on the second day is perceived by your body as overfeeding. So 140 calories are “lost”.

The mechanisms by which this could happen are not entirely clear. Some studies contain clues; one example is the 2002 study conducted with mice by Anson and colleagues (), from which the graphs below were taken.

In the graphs above AL refers to ad libitum feeding, LDF to limited daily feeding (40 percent less than AL), IF to intermittent (alternate-day) fasting, and PF to pair-fed mice that were provided daily with a food allotment equal to the average daily intake of mice in the IF group. PF was added as a control condition; in practice, the 2-day food consumption was about the same in AL, IF and PF.

After a 20-week period, intermittent fasting was associated with the lowest blood glucose and insulin concentrations (graphs a and b), and the highest concentrations of insulin growth factor 1 and ketones (graphs c and d). These seem to be fairly positive outcomes. In humans, they would normally be associated with metabolic improvements and body fat loss.

Let us go back to the 14 percent advantage of eating little and then a lot; a pattern of eating that can be implemented though intermittent fasting, as well as other approaches.

So, as we have seen in the previous post (), it seems that if you consume the same number of calories, but you do that while alternating between underfeeding and overfeeding, you actually “absorb” 14 percent fewer calories – with that percentage applied to the extra calorie intake above the amount needed for weight maintenance.

And here is a critical point, which I already hinted at in the previous post (): energy expenditure is not significantly reduced by underfeeding, as long as it is short-term underfeeding – e.g., about 24 h or less. So you don’t “gain back” the calories due to a possible reduction in energy expenditure in the (relatively short) underfeeding period.

What do 140 calories mean in terms of fat loss? Just divide that amount by 9 to get an estimate; about 15 g of fat lost. This is about 1 lb per month, and 12 lbs per year. Does one lose muscle due to this, in addition to body fat? A period of underfeeding of about 24 h or less should not be enough to lead to loss of muscle, as long as one doesn’t do glycogen-depleting exercise during that period ().

Sounds good? It actually gets better. Underfeeding tends to increase the body’s receptivity to both micronutrients and macronutrients. This applies to protein, carbohydrates, vitamins etc. For example, the activity of liver and muscle glycogen synthase is significantly increased by underfeeding (the scientific term is “phosphorylation”), particularly carbohydrate underfeeding, effectively raising the insulin sensitivity of those tissues.

The same happens, in general terms, with a host of other tissues and nutrients; often mediated by enzymes. This means that after a short period of underfeeding your body is primed to absorb micronutrients and macronutrients more effectively, even as it uses up some extra calories – leading to a 14 percent increase in energy expenditure.

There are many ways in which this can be achieved. Intermittent fasting is one of them; with 16-h to 24-h fasts, for example. Intermittent calorie restriction is another; e.g., with a 1/3 and 2/3 calorie consumption pattern across two-day periods. Yet another is intermittent carbohydrate restriction, with other macronutrients kept more or less constant.

If the same amount of food is consumed, there is evidence suggesting that such practices would lead to body weight preservation with improved body composition – same body weight, but reduced fat mass. This is what the study by Anson and colleagues, mentioned earlier, suggested ().

A 2005 study by Heilbronn and colleagues on alternate day fasting by humans suggested a small decrease in body weight (); although the loss was clearly mostly of fat mass. Interestingly, this study with nonobese humans suggested a massive decrease in fasting insulin, much like the mice study by Anson and colleagues.

Having said all of the above, there are several people who gain body fat by alternating between eating little and a lot. Why is that? The most likely reason is that when they eat a lot their caloric intake exceeds the increased energy expenditure.


Moira Morales said...

Very Interesting theory , I think I read about it some years ago, but I don't remember where.

Anonymous said...

Great stuff Ned.

David Isaak said...

Very enjoyable, and I think you are on to something.

Just let me throw in a caveat regarding feeding experiments with mice, though. People are quick to point out that mice aren't humans in terms of nutritional pathways, but they often miss the fact that our lifespans are so different.

For example, if a mouse is forced to fast for a day, it is about 0.1%of the typical mouse life span. A comparable fast for a human, if we assume as 70-year lifespan, would be a 35-day fast.

It's worse than apples and oranges. A lot of our biochemical reactions run at a similar pace, so there are similarities, but what "a day" means is quite different.

So I'm much more taken with your arguments from human overfeeding studies than from mouse lore. (Although it reinforces the argument when they both seem to point in the same direction.)

dearieme said...

I don't know whether you can get internet access to the Beeb, but BBC2 is broadcasting a show this evening (9:00 - 10:00 pm BST) called "Horizon: Eat, Fast, and Live Longer" about 'alternate day fasting'.

Ned Kock said...

That is a good point David. There are a number of other reasons why research on non-human animals cannot be easily extrapolated to humans. One is the size of our brain and its hunger for glucose; our brain consumes about 5 g/h of glucose, with that number going down and glucose being replaced by ketones as liver glycogen becomes depleted.

Ned Kock said...

Thanks dearieme.

Anonymous said...

Mice are VERY SIMILAR to humans with regard to boidy weight regulation. There is a REASON scientists such as Dr. Friedman use them:

ONLY Internet FOLKLORE says other otherweise about mice. Mice make AWESOME model organisms. Dr. Fredman specifically CONFIRMED most of his findings in mice WITH HUMANS.

Scientists even refer to them as "furry humans" because they are MOST similar to use GENETICALLY. THEY ARE EXCELLENT.

THE CLOSEST TO HUMASN out of ALL model organisms.

Reputable site:


The misinformation by uneducated Internet people is SCARY.

Great blog, Ned.

Ned Kock said...

Extrapolation has to be done very carefully Razz. For example, doubling lifespan in mice would be very unlikely to be comparable with doubling lifespan in humans. That’s what David was referring to.

Thanks for commenting.

adante said...

I'm not familiar with the mechanisms here so just wanted to confirm that the 150 calories lost in your example really would be 100% fat? That seems like a very impressive recomp ratio.

Ned Kock said...

Hi Adante. There is no guarantee that it will all be body fat, but alternate under/overfeeding tends to push your body in the direction of using stored energy at intervals that do not require muscle breakdown for glucose generation. In that sense, most if not all of the stored energy used, should come from body fat.

David Isaak said...

Hi, Razz--

Last time I checked, my BSc, MA, and PhD were all still valid. You are welcome to call me, for example, an "idiot." But "uneducated" won't do.

I think you are misunderstanding the thrust of the link you cite.

Yes, mice as models for diseases in other mammals are extremely valuable. But most researchers acr acutely aware of their limitations as well. Google : mouse models limitations : and you will find a voluminous literature on the topic.

I don't know where you get the idea that mice are the most similar animals to humans genetically of all model organisms. (Though I admit that by defining "model organisms" as you like you could probably prove anything.) Mice are more similar to us than fruit flies, certainly, but any primate is more similar to us genetically than any rodent.

You might want to take a look at the following paper from the Proceedings of the National Academy of Sciences on the problems with modeling human morbidity with lab rodents.


But I suppose the National Academy--like me, apparently--are uneducated.

In addition, YOU might want to LIMIT your use of ALL CAPS in your text...it's SCARY.

Anonymous said...

plus, "the closest" is still not "the same". This is one reason why so many drugs look so promising in the animal testing phase, but fail when they try them on humans


Anonymous said...

I do the following:
3 day fast, about 20-22 hour, and heavy high intensity weight trainig, and eating after it.
2 day underfeeding, only prot and fat lunch and full meal dinner. Overeating at weekend.
I dont see muscle loss but do you think fasted trainig can cause?