Why would any individual concerned with fitness, performance, appearance, and health voluntarily choose to not eat? Obviously because by not eating they expect to be able to positively impact one or several of those dimensions ( fitness, appearance, health etc. ) without doing overt harm to the others. Recently, however, several studies have been published linking short term starvation with increased autophagy in fast twitch muscle fibers. If this is true, then it is of significant concern because fast twitch muscle fibers are the ones that exhibit the highest capacity for hypertrophy, as well as the fibers recruited when maximal force output is called for. These studies are especially worrisome because they suggest that metabolically speaking, fast twitch muscle fibers seem to be considered highly expendable, even more so than their slow twitch brethren. Let’s take a look at one such study, brought to my attention by reader Rachid, that makes precisely these claims to see what we can find out, because if these findings are accurate, then we would have to seriously reevaluate our rationale for fasting.
After reading this abstract, I would not fault you if you came away with the impression that these researchers determined that rat metabolism sacrifices fast twitch muscle fiber in lieu of slow twitch fiber under fasting conditions. After all, “preferential atrophy of fast-twitch muscle” is pretty difficult to misinterpret, and the word “atrophy” itself is ominous. My dictionary tells me that atrophy means to waste away, typically due to degeneration of cells. It sounds, on the face of it, like bad news all around.
B y now, though, if you’ve read some of my previous posts, you should have come away with the realization that the devil is always in the details, and far more often than should be the case, paper abstracts are downright misleading. So, let’s look at some of those details and see whether we can exorcise some devils. Going straight to the meat of the matter, the results section of the paper, here is what we find:
By my count, there are two bombshells in that paragraph. See if you can spot them. I’ll wait … on the internet, we have infinite time. Click on the bombshells below to reveal them when you are ready.
Once you collect your jaw from the floor, ponder what in the world must have been going on in these rats to make them a full 1/3 lighter than their unfasted companions after a mere three days of food deprivation. But, before you go down that speculation rabbit hole, you might want to decide whether that really matters to you because, frankly, you do not need to know a single thing about metabolism in order to understand that this simply is not the way that things work in human beings.
I will freely admit that the first time I read that number, I had to go to the results table because I fully expected that it was a typographical error that had slipped by the editors. But, no, the computation is correct:
Note that the F-3d 31% weight loss computation is the result of taking the 98g post weight of these rats and comparing it to the control rats post weight of 142g. If you actually compare the pre and post weights for all of the groups, the control group gained 4% body mass (!!), the 1 day fasted group lost 14%, the 2 day fasted group lost 24%, and lastly, the 3 day fasted animals lost 28% of their mass.
Let’s suspend disbelief and assume that the above weight loss figures are accurate and representative of what we would observe in humans under equivalent conditions. This would have several and immediate effects. The principal effect would be the complete and utter collapse of the diet book, infomercial, supplement, and magazine industries. You would never need to see another article entitled “How the celebrity flavour of the week lost 7 lbs. in 7 days” because the simple answer would always be … fast for one day and you will lose 14% of your body mass. No need to suffer through eating only grapefruit for an entire week when you can simply get it over with in a mere 24 hours. This alone makes me wish that things did work this way! But, the fact of the matter remains that we are awash in all of these plagues … diet books, supplement pitches, and strident magazine articles gushing about the latest diet fad. We can surmise from this that fasting probably does not work quite like this for people.
Do human beings ever experience weight loss that is this rapid? Yes, we do, but only as a result of severe pathologies that are fatal unless immediately treated. Think cholera, AIDS, and advanced cancer. If you have ever fasted for 24 hours and wound up losing over 10% of your body mass, please contact me. I would love to hear from you. Oh, and by the way, just to appease the lawyers … I am in no way advocating cholera as a weight loss technique.
Wait, what? Aren’t these the same researchers that lead off the discussion of their results with the following:
All my life I have been told that you cannot have your cake and eat it too. My understanding of this idiom was that certain states of reality are mutually exclusive, and one must choose one or the other, but not both. You either possess your cake, “have it”, or you do not, because you have eaten it. It turns out that this was all a pack of lies because Ogata et al seem to have discovered a new phenomenon whereby at the cellular level, you can both have your protein and also use it as an energy source. I’m sure that it is only a matter of time before this principle is demonstrated at a macroscopic level, and when that happens, I’m going to give my parents an earful! I also plan to retroactively demand to have all the cake that I have eaten since I was three. Revenge will be cloyingly sweet!
On some level, though, you need to take a step back and applaud the judo like mastery of that paragraph. Everything it says is absolutely true … skeletal muscle does contain large amounts of protein, protein can be used as an energy source, and they did observe atrophy in their experiment. The implication is left up to the brain of the reader … that the atrophy was caused by the use of skeletal muscle protein as an energy source leading to “muscle type specificity in protein degradation”.
This implication is completely false and a direct contradiction of their actual results, but they carefully word the paper so as to never make this provably false assertion. Once the seed of the implication has been planted in the mind of the reader, the issue of no change in the muscle protein content after fasting is never touched upon again. Instead, the remainder of the discussion centers solely on the increased rates of authophagy observed in fast twitch fibers relative to slow twitch and the arcane details of autophagy regulation.
When all is said and done, it is left entirely up to us to explain the increased autophagy activity that was observed coupled with the pronounced weight loss. We certainly cannot rely on Ogata et al at all, because they seemingly have a different agenda than furthering our understanding. If you will indulge me, I would like to take stab at it.
Why would fast twitch muscle fibers exhibit different autophagy rates under fasting conditions than slow twitch fibers? Clearly because the different muscle fiber types have different energy requirements and different metabolic machinery to address their specific needs. It is well known that fast twitch muscle fibers are predominantly anaerobic and glycolitic, whereas slow twitch are predominantly aerobic. This difference in respiratory pathways speaks directly to the energy consumption patterns of the muscle under consideration … anaerobic respiration produces modest amounts of energy, but very very quickly, whereas aerobic respiration produces much more energy but at a much slower pace.
Practically speaking, the aerobic vs. anaerobic respiratory pathway dichotomy is going to manifest as differences in oxygen carrier proteins ( myoglobin ), and the mitochondria that use this oxygen in aerobic respiration ( the TCA / Krebs cycle ). Slow twitch oxidative fibers will have much more of both, relative to fast twitch fibers. Moreover, when it comes to rodents, the dominant fast twitch fiber type is type IIb, which is the most anaerobic, ”white” muscle that exists. As a consequence, type IIb muscle fibers are the least dense in mitochondria and myoglobin.
But if these fast twitch muscle cells do not contain much in the way of myoglobin and mitochondria, what do they contain? It’s a safe bet that to a large extent, type IIb fibers fill the additional space in the cell with their preferred energy substrate, glycogen. Keep this in mind, it will be critical shortly.
So, on an otherwise unremarkable day, our fast and slow twitch muscle cells determine via various signaling pathways that their external energy sources are not meeting their energy demands. Within short order, the cells begin to reconfigure themselves to accommodate their new reality. Cellular machinery that was dedicated to nutrient absorption and storage begins to be dismantled in order to be repurposed as machinery for the mobilization and utilization of previously stored energy. The rate at which this will happen will depend on the energy requirements of the particular cell, with those cells that have higher energy needs exhibiting a higher rate of this reconfiguration. I’ll let you come to your own conclusion as to which of the fiber types has the higher energy requirements.
As a concrete example of this process, consider that most muscle cells depend on glycogen synthase to take up freely available glucose during times of surplus and join them together to form long strands of the glycogen storage polymer. In circumstances where there is a lack of free glucose as a substrate, it makes little sense to keep large quantities of glycogen synthase around, especially if the lack of free glucose occurs in conjunction with a lack of free amino acids. Under such circumstances, it is perfectly sensible to break down glycogen synthase via autophagy into its constituent amino acids. These amino acids are then recycled and used to synthesize additional glycogen phosphorylase, the enzyme that cleaves a glucose molecule from the end of a glycogen polymer to make it available for glycolysis and cellular energy generation. Without the critical first step of autophagy, none of this is possible and the cell would quickly die in such a nutrient restricted environment.
From the above, it should be readily apparent that high rates of autophagy may be entirely indicative of a massive cellular reconfiguration, without involving any use of the proceeds of autophagy ( free amino acids ) as an energy substrate. It is this realization that explains why the massive weight loss observed occurred with no loss of cellular protein, even in the face of high rates of autophagy. The cells were apparently relying on other sources of energy, either glycogen or fats. The high rate of autophagy was the result of cellular reorganization to make these sources of energy more readily available.
Slet’s come back to glycogen as an energy substrate. Why all the fuss? Because unlike fat, glycogen must be stored associated with fluid, which means that every gram of stored glycogen comes with an associated 2 to 3 additional grams of water. It is precisely for this reason that the body stores the vast majority of its excess energy as fat. However, as glycogen is consumed, the cell no longer needs to retain the additional water so it will excrete it, and in so doing it will get lighter … much lighter, as it turns out.
Armed with this information, we can make a prediction … those cells with a higher proportion of their energy derived from glycogen ( the predominantly anaerobically respirating ones ) will experience the highest level of weight loss / atrophy during times of starvation as they excrete significant amounts of excess water. We are extremely fortunate that we have the work of Ogata and colleagues to confirm this prediction, at least with respect to rat plantaris and soleus muscle cells.
We can actually go further with the above analytical framework. Since three days of fasting had no effect on the protein content of muscle cells, reversing the so-called atrophy is merely a question of restoring glycogen levels, thereby essentially rehydrating the cells! Sounds suspiciously close to a carb re-feed …
In the end, I actually really like this paper by Ogata et al., but for reasons quite apart from what the original authors most likely intended. This paper unequivocally demonstrates that autophagy is highly conservative when it comes to protein. There was no net protein loss in either fast or slow twitch muscles over a period of three days of fasting, despite the fact that significant loss of mass occurred, and despite massive increases in marker proteins (LC3-II) for increased autophagy. Going forward, this information will be invaluable in assessing other similar scholarly articles pertaining to autophagy. Those studies that do not attempt to measure and quantify the nature of the mass that was lost will be immediately suspect.
Rob, a reader from Ireland, pointed me to a worrying post regarding the potentially negative aspects of intermittent fasting and asked about my opinon of it. Ok, I’m being generous here by throwing in the modifier potentially, because the post’s author definitely holds the viewpoint that fasting will kill you … I’m paraphrasing here and exaggerating a bit for effect, but not much since the actual quote from the posts’s conclusion is:
It has been said that extraordinary claims require extraordinary evidence, so, without further ado, let’s dive in to the “proof” of Kiefer’s pudding.
If you Can’t Dazzle Them With Brilliance, Baffle Them With Bullsh*t
Try as I might, I can’t follow Kiefer’s argument, mostly because I find it muddled and contradictory. Things start off on the wrong foot when he asks the reader to suspend all disbelief and “imagine a purely destructive process that’s critical for building the very thing it destroys.” Call me pedantic, limited, or lacking in imagination, but to me a purely destructive process only destroys, it doesn’t build. If it destroys and builds, then it’s not purely destructive, but rather, complex, subtle, and nuanced.
If you are in a charitable mood, you might be tempted to let Kiefer off the hook and chalk all of this up to enthusiastic overstatement – maybe he didn’t really mean that autophagy is purely catabolic. Except that apparently, he really does mean it as he repeats this mantra throughout his post : “by definition and function, autophagy destroys tissue, making it purely catabolic.” Not to nitpick, but even if we were to concede that autophagy is purely catabolic (it isn’t), autophagy does not destroy tissue. Autophagy is active at the cellular level, while tissues are collections of cells. This is a very important distinction for reasons that will become apparent later, I raise this issue here to point out that Kiefer might be a bit sloppy in either his understanding, or explanations, or both.
This stance on autophagy as purely catabolic is especially surprising when you realize that in his citations, Kiefer refers to a paper by Masiero et al. entitled “Autophagy is required to maintain muscle mass.“ I understand that Kiefer is a busy man as he is trying to build his nutritional conselling business, and he may not have had the time to read the full paper, or maybe even the abstract. But I would hope that he at least had the time to read the title! Trying to characterize something that plays a critical role in maintaining muscle mass as purely catabolic smacks of expending a lot of effort in pounding a square peg into a round hole.
But why is Kiefer doing this? After all, he’s not dumb, far from it, as he tells us himself, “he’s smart as f*ck.” Mostly, he’s trying really hard to create an association in your mind between autophagy and pure catabolism so that in the subsequent discussion, whenever you see the word autophagy, he would ideally like you to picture your body melting into a puddle of catabolic goo. Because if he can sell you this purely catabolic disease, he can then sell you his cure, and it will only cost you $89. The last thing Kiefer wants is for you to think of autophagy as complex, subtle, nuanced, and worst of all, tightly regulated:
More recent research has altered the original view of autophagy as a nonselective process by revealing the existence of molecules involved in selectivity of autophagosome cargos. … Autophagy is tightly regulated to prevent its unbalanced activation, which may cause damage to cells. Studies have identified at least two signaling complexes that participate in the fine-tuning of autophagic activity under conditions of stress and other physiological circumstances in which autophagy is needed.
Selective autophagy includes a pathway for targeting specific hydrolase proteins ( the Cvt pathway ), one for processing protein aggregates (aggrephagy), another for processing ribosomes (ribophagy), yet another one for mitochondria (mitophagy), as well as pexophagy, a pathway for specifically targeting peroxisomes. That is certainly a lot of pretty specific pathways. As Shakespeare might have said, “if autophagy be the food of the nutrient deprived cell, play on, give me excess of it!”
Over time, I’ve learned a thing or two from my wife, especially at mealtimes, when she habitually leaves her favourites to savour slowly at the end, so I’ve saved my favourite selective autophagy pathway for last – macrolipophagy:
If at this point you want to take some time to bask in the glow of the work of Singh et al., by all means, please take all the time you need. I’ll wait.
All of the mentioned pathways make use of macroautophagy, that is, the formation of a bubble-like structure within the cell that encloses the target of autophagy in what’s been termed the phagosome. The phagosome then fuses with another structure containing enzymes that lyse, or breakdown, the phagosomal contents. This is quite important to keep in mind, especially when Kiefer gets a hold of the phagosome and waterboards it into confessing to crimes it did not commit:
Kiefer’s mental model of autophagy seems to be that of a process that is turned off and on much like one operates a light switch. It isn’t. It is an ongoing process, the rate of which rises and falls in accordance with the cellular context and needs, but never actually “turns off.” But let’s ignore that for the time being in order the take a closer look at the paper by Mizushima et al cited as reference 7. We will be looking for a substantive demonstration that macroautophagy is something that treats fast-twitch muscle as exceptional, in the manner described by Kiefer.
In 2003, Mizushima and team published a paper where they revealed that they had figured out a way to make the autophagosome glow in the dark under UV lights. Like all children when presented with a cool new toy, they promptly set about making mice with glow in the dark autophagosomes so they could look at them and count them. Since they knew that a good way to increase the number of phagosomes is to restrict nutrients, they tested their mice in the fed state, after 24 hours of starvation, and lastly, after 48 hours to determine what changes in the number of autophagosomes per unit of surface area could be detected. This is what they found:
The first thing that ought to strike you is that each graph looks different from the others. Different tissues exhibit different responses to food deprivation. This is not unexpected as we know different tissues have different energetic needs as well as different internal structures and therefore different substrates available as autophagy targets.
The second thing that ought to jump out at you is that at time 0, in the fed state, all tissues exhibit varying levels of basal autophagy. Some tissues, like the pancreas, exhibit quite a high level of basal autophagy. In contrast, the heart has some of the lowest basal levels of autophagy recorded by the researchers, although even here they are not zero. So much for “turning off” autophagy.
The third thing that you ought to notice is that in almost all tissues that respond to starvation, including fast-twitch muscle ( EDL M. ) the rate of autophagy after 48 hours is actually lower than after 24 hours. The notable exceptions here are the heart and the soleus muscle, a muscle predominantly consisting of slow-twitch fibers. So, we can toss “macroautophagy stays on for 24 hours or longer only in fast-twitch muscle if we don’t eat” into the trash can too.
Lastly, note that in terms of the Thymus and Lens graphs, none of the observed differences were statistically significant, so, for all intents and purposes we can consider autophagy to be independent of the nutrient availability in these tissues over the timeframe under consideration ( i.e. nutrient deprivation over 48 hours. )
All of these observations put together ought to make you wonder how it is that one can claim that “macroautophagy is transient in nearly all tissues except fast twitch muscle” after having read this paper? It certainly gives me pause. The paper does show that for the most part, cells deprived of external nutrients turn to internal stores, which we knew was the case before we started, so, there is nothing novel here.
In and of itself, however, the fact that we observe the exepected increased autophagic activity, doesn’t tell us very much at all. What we need to know is: what materials are being broken down? Which autophagy pathway is being invoked? Are we talking about damaged organelles, toxic protein aggregates, intracellular parasites, in other words, expendable cellular garbage, or something else, like large supramolecular structures such as whole undamaged organelles of the cell? And if we are talking about whole undamaged organelles of the cell, how many of them are actually LDs, the cellular lipid transporters?
Suffice it to say that up to this point, I am left unconvinced by Kiefer’s argument, to put it mildly, and we’ve hardly scratched the surface. However, I do think that I’ve given you a lot to read and think about, so I’ll stop for now, or else this article won’t ever see the light of day. I’ve already rewritten it four times. It’s about time to publish it.
In closing, I would like to thank Rob for bringing this post to my attention. Ordinarily I would have replied to his email directly to thank him, but it turns out that my contact page was buggy, and did not provide me with his email address. This is now fixed, so I am doubly indebted to him. Cheers, mate!