Countering Anti-Energy Balance, CI=CO Arguments ~ Part I The 3500 Cal = 1 Pound Fat

Calories In - Calories Out = Change in Energy Stored

I continue to be amazed by the number of people who argue against this basic physical law.  This series will discuss the various versions by which folks seek to argue against this.  

This installment will consider the argument that 1 pound of fat = 3500 calories is incorrect.

= 3500 Cals??

Many will point to the fact that someone can overeat, say, 5000 cals/day for a period of time and not gain the exact amount of weight predicted.  Going the other way, folks can eat a 500 cal/day deficit for a period of time and not lose a full pound in a week or even lose more than a pound in that week.  These observations are often put forth as "proof" that a pound of fat doesn't equal 3500 calories and/or that CICO (calories in calories out) theory is fatally flawed.

There are any number of explanations, however, why these observations don't disprove CICO.    Before I list them, let me state a few things here.  For starters, I don't know of anyone "in the business" who would claim that 1 pound = 3500 cal is an exact value.  It's not, and it is intended to be a guideline to estimate expected losses or gains.  Secondly, not all body fat is the same fatty acid.  The actual caloric content of a lipid depends on the chain length and saturation.  The 9 cal/g factor for fats is an average approximation.  But that said, it would hold that if we expend 3500 cals more than we take in, and for the sake of simplicity presume we're completely glycogen depleted so that all of the deficit is made up for by burning stored lipid, you will lose approximately 1 pound of stored lipid.  And if this is the other way around and consume 3500 excess calories in a glycogen replete state so that all of the surplus is stored as lipid, we'll gain approximately 1 pound of stored lipid.  But already one can see how this likely won't translate to the bathroom scale.  Because:

• Glycogen stores are fluid and for every gram of glycogen there are three grams of associated water.  This explains short term gains and losses.  But in the scheme of body weight regulation, Hall & Chow even leave glycogen out of their model.

• We never lose or gain JUST fat.  Lean body mass changes as well.  LBM contains significantly fewer calories per pound (can't find a number on that).

• We will expend some calories to absorb and metabolize calories.  So 3500 calories in doesn't equal 3500 calories available to do work or be stored.  I don't think anyone disputes this, thermogenesis and all that stuff.   But that is accounted for in the CO side of the equation.

• If our caloric excess is primarily carbohydrate, energy must be expended to convert that to fat.  Again, this is accounted for on the CO side of the equation if one were to do an energy balance (IOW 3500 cal excess is not 3500 if we expend, say, 500 of those calories to convert carb to fat).  This was shown in this study, where carbohydrate overfeeding led to storage of only 75-85% of energy consumed vs. 90-95% for fat overfeeding.

• Are we talking a pound of "fat" or a pound of lipid?  As I've been discussing here and here fat tissue is 14% water and only 75% lipid (the rest is the cells, connective tissue, etc.).  We expand and shrink fat cells, but we never get rid of them (well, unless we get liposuction and such).  But presuming the water correlates with lipid (that is a bit of a leap of faith), we lose or gain about 1.2 lbs "fat" for every 1 lb of lipid.  Another way to look at this is that a pound of fat actually contains only 2625 calories.  This doesn't violate energy balance in any way.

• Similarly, lean mass is almost 80% water, so if we presume LBM is mostly fluxing carb + protein, the energy content of 1 pound of LBM "solids" could be estimated to be 1556 cal, these are associated with 4 pounds of water, so a pound of lean tissue contains just over 300 cals (roughly 1/10th that of fat tissue).  

• So to summarize the two prior bullet points, tissue weight gains and losses are disproportionate for whether energy is stored in lean vs. adipose tissues, and water weight gains and losses are both considerable (14% for fat, 79% for lean).  Put together, it's no wonder the bathroom scale doesn't reflect the 3500 cal per pound figure.

• Lastly, we all know that CI and CO are not independent terms.  Our bodies can and do rev up metabolism a bit to counter overfeeding, and they do dial it down to conserve energy when in deficit, especially prolonged deficit.  

Bottom line:  A pound of lipid is approximately equal to 3500 calories.  A pound on the bathroom scale is not a pound of lipid.  A 3500 calorie deficit over the long run will deplete about one pound of lipid stores, and a 3500 calorie surplus in energy in v. energy out, will add about one pound to your lipid stores.