17 Comments

Nice, it's good to see this laid out! I'll need to start using the term 'glycogen stores' rather than 'water weight' (especially since that's more useful as a term for fluid retention).

Expand full comment
author

Agreed, I think 'water weight' is an unfortunate choice. I think that people in bodybuilding circles use "glycogen stores" instead.

Expand full comment

so what you're really saying (and what authoritarian regimes want to hide from us) is that we're all, on average, equipped to sustain a 2 month hunger strike?

More seriously, could this be the reason why "all diets work" short term but very few hold long term? The implications of an initial 2000 calorie deficit leading to a 2.25kg weight loss is that running even a small deficit for a short period of time will lead to short-term, impressive gains on the scale, but will be reverted just as fast.

Also, do you think this is affected by what you eat? I assume if you eat less sugar, the body has a harder time making glycogen. Actually, it's perhaps the whole point of the keto diet: keeping your glycogen forever low so that your body can't use that juicy buffer and is forced to tap into fat stores?

Expand full comment
author
May 4, 2023·edited May 4, 2023Author

> so what you're really saying (and what authoritarian regimes want to hide from us) is that we're all, on average, equipped to sustain a 2 month hunger strike?

At last, someone who can read between the lines and see the true message I'm trying to convey! :)

Regarding "all diets work", I tend to think this is not the explanation. Or at least not the full explanation. When I tried the potato diet (https://dynomight.net/potato-diet/) I did lose weight very quickly at the beginning, but I kept losing weight after that and after a few weeks I lost much more than I think could be explained by glycogen.

Instead, my guess is that "all diets work" because they all restrict food choice, and it's human nature to eat less when you have access to less variety or less palatable food. And also (for some of the diets) perhaps because they reduce caloric density. I'd guess these gains don't stick because once you return to your usual unlimited amounts of ultra-palatable food you naturally eat until you are at some biological set point—probably very close to your weight before you started dieting.

Regarding your last point, in principle eating sugar vs. protein shouldn't matter as long as you are consuming/expending the same amount of calories per day. But it does seem plausible keeping your glycogen permanently low might have some significant effect on appetite, etc. I don't know enough about this to say anything useful, though.

Expand full comment

It's clearly affected by what you eat. I currently eat a zero carb, zero fiber, super low protein (<30g/day) diet and my water weight is way down. Just eating normal amounts of protein will make me gain several pounds of water weight, not to mention any carbs.

And yea this is probably the reason for the "all diets work" illusion, which is clearly false in the long run.

Expand full comment

I might be wrong about this, but don't the glycogen stores in your muscles only get depleted if they're used to power those specific muscles? So if you don't use your quads during your fast, the glycogen in your quads won't get used up. Thus, your weight loss pattern while fasting will depend on both your amount of exercise and which muscles you exercise.

Expand full comment
author

This paper suggests you are at least partially right! https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6019055/

> Glycogen particles are distributed within the muscle cell to support the local energy needs of the cell during exercise (see Figure 2).

But I'm not sure how to square this with the fact that pretty much everyone seems to lose a lot of weight when they start a deficit. And the paper isn't very conclusive about how quickly we should expect these stores to be depleted in a caloric deficit. Maybe the "normal" uses of the muscles will still be enough to deplete these stores fairly quickly?

Definitely the reality is a bit nuanced. It's not really true that you "first" burn all your glycogen stores and "then" burn fat. It's more that you burn a mix where the percentage of glycogen burned decreases over time.

Expand full comment

I was gonna make this last comment, that things are a bit more complicated, and you dont deplete all your glycogen before losing any fat. I'd recommend to add this caveat somewhere in the article.

Nice read anyway

Expand full comment
author

Good call, done. (I actually had something like this in my original draft, but deleted it in response to my beta reader's usual "the words there are too many words it burns" reaction.)

Expand full comment

I think that glycogen is not the only thing that is temporary and related to water storage. I feel and look a lot squishier and I gain 2-3cm around my waist when I gain "water weight." Pretty sure the fat actually takes up some of that water somehow.

Expand full comment

Good writeup! I talk about the more practical applications of this in dieting here: https://exfatloss.substack.com/p/water-weight-temporary-fluctuations

Expand full comment

This was great.

I recently learned that 1 lb of weight is not 3500 calories as many people believe, but is variable depending on a lot of factors. I am guessing a lot of what you have here plays into that. Very interesting stuff.

Expand full comment

This is very nicely laid out, thank you! I hadn't known about the glycogen explicitly before, and the notion fills a lot of gaps for me. Particularly why I can drop 10 lbs in a week of hard labor, or 6 lbs from a day of food poisoning. That always struck me as odd.

Expand full comment

What about muscle mass? Is that included in the “body fat” category?

Expand full comment

Re: “muscle mass,” if you were to do DXA scans before and after losing, e.g., 7lb of “water weight” (let’s just pretend you lose zero fat), the DXA would show that you have the same fat mass and 7lb less “lean mass” (i.e., everything that isn’t fat or bone). So your body fat percentage would actually rise. Now, typically when we lose weight, we are losing some combination of fat and lean mass, but it is worth keeping in mind that what we really want to do is lose fat, not lean mass. So, from a health perspective, I don’t see the point in trying to lose water weight simply because it will make the weight on the scale go down.

Expand full comment

A couple years ago I did a low-carb diet and did experience this quick drop in water weight. I’m not sure how much of it is due to glycogen depletion as you describe here, versus the loss of water that happens when you enter ketosis with a loss of sodium, and the attendant flushing out of water from the kidneys to maintain sodium balance in the blood.

Expand full comment

Nitpick: refering to kcal as "calories" means favoring a term with two interpretations varying by three orders of magnitude over a shorter, unambiguous, SI-compatible term. In the EU, the food energy content is given both in kJ/100g and kcal/100g, and people generally can interpret them. Following your SI (imperial) unit scheme for weight, stating the energy as "~8000kJ (~2000 Calories)" or "2000kcal (~2000 Calories)" might be preferable.

From my understanding, glycogen and body fat could be likened to a capacitor and a (secondary) battery. Storing glucose (electrical energy) in glycogen (a capacitor) is highly efficient, but takes a lot of weight because the storage density is low. So instead your body (an electrical car) stores its long term energy reserves in fat tissue (a battery), which has a much lower energy conversion efficiency, but a much higher energy density.

Which leaves the question what the efficiencies of both methods of energy storage are. For glycogen, I would assume that the efficiency is probably high, polymerizing/depolymerizing glucose molecules without changing their inner structure does not sound costly. 70-95%, perhaps?

For fat, it likely depends on the starting point. Turning food fats into body fats does not feel to be terrible hard. Turning carbohydrates into fats seems much more difficult. I would assume that this is overall less effective. I would guess perhaps 40-80% of the energy of glucose can be transformed into fat (with a part of the waste energy possibly being captured in NADH or ATP or whatever weird non-SI currencies cells use for energy).

Expand full comment