Monthly Archives: February 2019

The endurance athlete’s guide to fueling and weight loss part 5: Hunger etc..

Please read the previous posts if you haven’t seen them before…

We are moving closer to the post where I hope to give useful advice on the different tactics you might use to improve fueling or lose some weight. Or perhaps both.

This was going to be a short post as I was having trouble coming up with a good way to talk about hunger, but I came across some new (to me) information that I hope will be informative.

Hunger

Hunger lies at the intersection of energy balance, brain function, psychology, and group dynamics. Oh, and evolutionary biology. That makes it complex and hard to understand, and therefore not very amenable to easy explanations or simplification. It’s much more complex than – for example – how blood glucose is controlled. I’m therefore going to have to simplify quite a lot, and there will be a number of areas where it’s just not clear (to me) what is going on. That said, let’s get started.

The evolutionary purpose of hunger is to drive us to eat in ways that maintain our energy stores – and in particular, our fat stores – at a certain level. Or perhaps within a certain range. What factors could set the low and high limits of that range? (note 1).

Let’s start at the lower end – what drives the lower end of the fat storage range?

  • If we have too little stored energy, we won’t be able to survive when food becomes scarce.
  • If we are female, we need extra stored energy to be able to build and feed a child.
  • And at the upper range?

  • If we have too much stored energy, we may not be able to move around effectively, which could compromise our survival.
  • There may also be impact based upon climate, but I’m going to ignore that for this discussion.

    Another way to describe the range is “enough fat, but not too much fat”. Where “enough” and “too much” have definitions that are a bit squishy. But you get the idea…

    The regulation of hunger – and therefore the regulation of energy intake – is driven by two hormones, leptin and ghrelin. To oversimplify things:

    Leptin is produced by fat cells, and serves to reduce hunger. Generally speaking, the larger the fat storage, the higher the leptin levels will be. You can think of leptin levels like the gas gauge on a car; if your fat tank is empty, leptin reads low, if you fat tank is full, leptin reads high.

    Ghrelin is produced by the stomach, and serves to enhance hunger. It’s a shorter term signal.

    Based on a simple understanding of how the hormones work, we would expect that Ghrelin levels would be proportional to how long it was since we ate; the longer we went without food, the more hungry we would be.

    Here’s a graph of Ghrelin values over a typical day:

    (note 2)

    The solid line is the average while the dotted lines show the upper and lower range.

    That’s not what I expected. Ghrelin levels are lowest right when we get up, which is when we have gone the longest without eating and would therefore expect them to be highest. It turns out that ghrelin levels have a few somewhat interesting features:

  • They are adaptive based upon when we usually eat and our circadian rhythms.
  • They increase when we start to eat. This is familiar to most of us; not being really hungry but finding out that as we smell dinner or start eating, we are suddenly hungry.
  • There’s another strange feature of ghrelin related to fasting. Here’s a study (note 3) that looked at ghrelin levels over an 84 hour fast:

    image

    That’s just weird. The mean ghrelin levels decrease from day do day, which means you are actually quite a bit less hungry on day three of fast than you are in day one. While weird, this is pretty well established by research.

    Also note that there seems to be a greater reduction for women, for reasons that are not well understood.

    Why our bodies behave this way is not really known. The best theory I know is one from an evolutionary standpoint; while it is good to be hungry if food is available, it can quickly become counter-productive if food is scarce.

    Moving onto leptin, what do leptin levels look like? (note 4):

    image

    That is what we generally expect – at least there seems to be a decent linear relationship between BMI and leptin levels. It’s a bit messy, probably because BMI does not correlate perfectly to fat mass, and likely because of individual variations as well. The differences between male and female leptin levels are asserted to be caused by a) women having more fat mass at a given BMI and b) women having more of the kind of fat cells that produce more leptin than men, though I don’t think the question is truly settled.

    Disfunction

    Leptin is supposed to inhibit significant weight gain; if you gain excess fat, your leptin levels rise, your hunger drops, and you lose fat mass until your leptin levels drop. And that seems to work for some people, especially those who are young, choose their parents well, and male. But it’s pretty obvious it does not work well in a lot of cases.

    Is there anything known about the disfunction? Well, a bit…

    Leptin and ghrelin levels based on types of food

    I found a very nice experiment (note 5) that looks into leptin and ghrelin response based on different kinds of food. Take a group of people, have them fast overnight, and then give them one of three drinks:

  • 500 calories with 80%/10%/10% from carbs, fat, and protein
  • 500 calories with 10%/80%/10% from carbs, fat, and protein
  • 500 calories with 10%/10%/80% from carbs, fat, and protein
  • In other words, a carb-heavy (glucose-heavy), fat-heavy, and protein-heavy drink (note 6). This is done in what is called a “crossover” study, which means that each subject had all three drinks on different days.

    You sample their blood before they have the drink, and then every 30 minutes afterwards, and measure a bunch of different things. Based on how ghrelin works, I would expect that eating would suppress the ghrelin levels and then over time, they would rebound to their previous levels.

    What happens?

    image


    All three produce a significant suppression of ghrelin production, but carbohydrate produces the biggest reduction. Interestingly, however, after about two hours the carbohydrate ghrelin level goes shooting up and after 3 hours it is higher than the fat or protein curves and soon after becomes higher than the initial ghrelin level.

    Or, to put this another way, five hours after eating 500 calories of mostly glucose we would be *more hungry* than we were at the start.

    The authors write, “Our finding of a rebound of total and especially acyl-ghrelin above baseline after high-carbohydrate meals could provide some physiological basis for claims made by low-carbohydrate diet advocates that ingesting carbohydrates prompts an early hunger rebound”.

    Indeed.

    They did measure subjective appetite which showed no effect, though unfortunately they had technical difficulties with the appetite reporting system and that data was therefore not published.

    I’d also like to note that there are many experiments that measure satiety (the inverse of hunger) and show that carbohydrates lead to more satiety than fats or protein. And they do, if you only measure them for 2-3 hours.

    The experiment also measured blood glucose over time:

    image

    If I eyeball the two charts, it looks like ghrelin production starts to go up steeply about the time blood glucose drops quickly at 140 minutes, and is highest when the blood glucose is the lowest. The pattern here matches what is known as either “reactive hypoglycemia” or “postprandial hypoglycemia” – basically the blood glucose drops below initial levels a few hours after eating.

    It’s important to note that the drinks were dominated by a specific macro and drinks with mixed macros may show unexpected results. Though 500 calories is not really that much and these were consumed in a fasted state, and as we know that is the time when the body is best able to handle a lot of glucose.

    You can find the leptin chart in the paper if you’d like to see it; there were small drops over time but nothing very striking.

    Fructose vs Glucose

    Is there a difference between fructose and glucose? Let’s look at another experiment (note 7).

    In this experiment, we take 12 normal-weight women and feed them three meals containing 55%, 30%, and 15% of carbohydrate/fat/protein and take blood samples every 30-60 minutes.

    Of the 55% of the calories that come from carbs, 30% either comes from a fructose-sweetened or glucose-sweetened beverage.  Take a normal meal pattern and make the carbs either fructose heavy or glucose heavy. And sample their blood periodically:

    image

    Wow. Lunch and dinner show large spikes in ghrelin for both drinks, but the late-night spike of the fructose drink is much higher. Also notice the difference in levels at 8am the next day; the level of ghrelin in those who had glucose is quite low, but it’s pretty high for those who had fructose.

    Fructose gives us bigger positive ghrelin peak than glucose.

    They also measured blood glucose levels:

    image

    Based on what we know about glucose and fructose metabolism, that is what we would expect; because the fructose is processed in the liver, there is much less glucose. We would expect that the liver processes the fructose to triglycerides. Is there data to support that?

    image

    Yep. Vastly higher triglyceride levels show the fructose being converted to fat and released into the bloodstream, and those levels persist through the night.

    They also measured leptin levels:

    image

    Leptin levels rose much less for the high-fructose meal, which means there was less inhibition of appetite.

    Overall, fructose led to a greater increase in ghrelin (higher appetite) and a lesser increase in leptin (less appetite suppression).

    So, carbohydrates in general aren’t great, and fructose is worse.

    The best theory I’ve seen around why fructose behaves so differently is that significant fructose supplies were rare in historical times, and therefore it was advantageous for humans to eat as much as possible when they found them. That sounds reasonable, though like many studies in evolutionary biology they are hard to support.

    Leptin resistance

    The mystery of why leptin isn’t behaving as we would expect – why people still eat a lot even with high leptin levels – has been labeled “leptin resistance”, by analogy with insulin resistance; the idea is that for some reason the brain is not sensitive to the levels of leptin.

    Whether there is actual resistance or whether there are other factors that are overpowering the leptin signal is not clear.

    There are a number of theories round what is actually happening. Among them are:

  • There is a defect in transporting leptin from the bloodstream into brain cells across the blood/brain barrier.
  • The cells within the brain become less sensitive to leptin.
  • Dietary fructose leading to elevated triglycerides reducing leptin transport across the blood/brain barrier.
  • Here’s two papers if you want more information (note 8) (note 9).

    Dopamine

    I’m including this section for completeness. There is some research on how dopamine is affected by sugar ingestion, and while I think there something going on there that partially explains why sugar is addictive – at least for some people – I’m not confident enough in my understanding and the quality of the research to have much to offer.

    I do offer a few papers:

    Hunger Summary

    The main points from the preceding section on hunger:

    • Carbs – and especially fructose – seem to interfere with the hunger control system.

    • Hunger is not directly related to how long it’s been since you ate; it has a daily rhythm and decreases when fasting.

    Energy balance and weight loss

    If you have read official guidelines and advice about weight loss, you can generally boil them down to one bit of advice:

    Eat less and move more

    This is often summed up as the “Calories in / Calories out” (CICO) model; eat less means fewer calories in, move more means more calories out, and the result will be weight loss.

    It is also typical to see appeals toward the Laws of Thermodynamics. The more rabid adherents to the model treat it as if they have discovered one of the deep secrets of the universe.

    If you have read the earlier posts and have learned anything about biochemistry, I’m sincerely hoping that you suspect that the reality might just be a *tiny* bit more complicated than the simple world of “eat less and move more”. Especially since that advice fails for a large number of people, at least for the long term.

    There is truth to the CICO model in one sense, if you are losing weight you are burning more calories than you are taking in. And vice versa. But that’s the result, not the driver; the driver is the biochemistry at work.

    The key to understand how things really work – and why CICO isn’t very useful in many cases – is related to how the body responds to a reduction in food intake. The body essentially has three options to balance things out:

    1. It can burn stored fat.
    2. It can tear down muscles (ie “lean mass”) and burn that.
    3. It can reduce its energy use.

      We are hoping that it would do #1 – after all, the whole point of the fat storage system is to provide an energy reserve when food is scarce. But remembering the earlier posts, there are a couple of things that can get in the way of that. First off, we need to be good at burning fat in general. And second, we need to be in a hormonal state where fat burning is possible.

      If either of those aren’t true – or are true only to a limited extent – then we are stuck with tearing down muscle and reducing energy use. And, in fact, that is what we see in a lot of diet studies; people will lose lean mass – sometimes a significant amount – and people report being cold, tired, and hungry all the time.

      And they don’t really lose all that much weight.

      If we can get rid of the conditions that are blocking fat metabolism – and, since we are athletes, up the amount of fat we burn when exercising – then the body should naturally start burning more stored fat, and we should lose weight. Or, to put it another way, we are going to focus on the fat burning side of the house.

      This post is already quite long so I’ve tried to limit the detail in this section; if you want more I highly recommend Peter Attia’s post on fat flux.

      Summary

      There is a lot more that I could write about WRT hunger and energy balance, but I think this is enough for now.

      Our strategy is going to make dietary modifications to reduce hunger and improve our ability to use fat to generate energy while exercising, with a goal to lose weight/improve our ability to perform in long events.

      Which takes us to the tactics portion of the series. What do I think you should actually *do* to implement these strategies, so that you can – with any luck – see the benefits that I’m talking about.

      That will be post #6. When I get that done, I’m planning on doing at least one post to cover any questions.

      Notes:

      1. I’ve tried to base this section on what I’ve learned about evolutionary pressure, fat stores, and hunger.
      2. From Jason Fung’s excellent discussion on hunger and fasting here.
      3. Fasting unmasks a strong inverse association between ghrelin and cortisol in serum: studies in obese and normal-weight subjects
      4. Mechanisms behind gender differences in circulating leptin levels. This result is widely replicated in other studies.
      5. Acyl and Total Ghrelin Are Suppressed Strongly by Ingested Proteins, Weakly by Lipids, and Biphasically by Carbohydrates
      6. The beverages were mostly composed of a glucose beverage, whey protein/nonfat milk, and heavy whipping cream for the carb/protein/fat drinks.
      7. Dietary Fructose Reduces Circulating Insulin and Leptin, Attenuates Postprandial Suppression of Ghrelin, and Increases Triglycerides in Women
      8. Leptin resistance: a prediposing factor for diet-induced obesity
      9. Mechanisms of Leptin Action and Leptin Resistance
      10. MarksDailyApple, by Mark Sisson’s. Mark is an advocate for a way of eating named “Primal”.

      The endurance athlete’s guide to fueling and weight loss part 4: Better fat burning in actual athletes

      Please read the introduction and earlier posts if you haven’t…

      In the last post, I finished by asking what factor or factor might cause the difference between these two athletes:

      image_thumb2

      image_thumb5

      The differences might be genetic, differences in diet, or differences in training. Or maybe something else. What do you think is at play?

      We don’t know all the factors, but it turns out that this athlete does not have a sweet tooth, so he eats closer to the recommended cyclist diet; lots of carbs, but not a ton of sugar. At least in his base diet; I don’t know what he eats before/during/after training. And he’s better at burning fat.

      Hmm. It’s almost as if there might be a dietary effect here, that the availability of carbohydrate in the diet and on the bike might have an effect on fat burning ability. How could we test that?

      Well, let’s put two cyclists – how about these two cyclists? – in a situation where there is much less carbohydrate available, let them train for 10 weeks, remeasure their VO2Max, and see what changes. Since we want to maximize the effect, we’ll put them on a very low carb “keto” diet, which is something like 30 grams of carbs per day.

      What happens?

      Well, athlete #1 is not happy on the keto diet, and that’s really no surprise; he’s not good at producing energy from fat *at all*. If you take away his carbs he will feel horrible on the bike. He ends up making the following adjustments from has previous diet:

    1. He minimizes all sources of sugar
    2. He limits bread, rice, pasta, and potatoes to once or twice a week.
    3. He reduces the food he eats on the bike to an occasional banana plus water
    4. Not really a low-carb diet, but certainly a *lower than before carb* diet.

      Ten weeks of training go by, he retakes the test, and we generate a new graph (previous results are dotted):

      image

      This athlete is now a significantly better fat burner; rather than hitting only 25% of calories from fat, he’s averaging 40% or so across most of his range. His beta oxidation system is better.

      Athlete 2 started with the same very-low-carb diet, and he didn’t stick to it strictly either, but he did mostly get rid of grains and fruit from his diet, and he ended up lower-carb than athlete 1.

      10 weeks go by, and we get this:

      image 

      Yowsa! He nearly doubled the amount of energy he got from fat in parts of his range, and averages over 70% for most of the range. There is also a significant shift to the right, and he produces a higher maximum relative power. That’s great, the low carb diet made him much more powerful!

      Not so fast. Remember that this is *relative power*, and we know that he lost some weight as part of his 10 weeks training, so most of the improvement is likely from the reduction in weight. Though the training may also have helped.

      What did we see across these two athletes?

      First – and most importantly – we saw that a dietary switch made significant changes in the fat metabolism ability for both athletes. This really isn’t very surprising knowing what we know about the underlying biochemistry, but it does show very clearly that beta oxidation capability can be trained.

      Second, we saw a correlation between the amount of carbohydrate in the diet and the overall ability to metabolize fat.

      I do want to add a few caveats. The first is that two cyclists is a very low sample size, and the second is that we can’t tell the difference between changes due to base diet and changes due to food before/during/after. And there may be a genetic component at play here.

      Returning to my RAMROD example from the last post, let’s add in two more lines, corresponding to getting 50% of calories from carbs and supplementing 200 cal/hour (yellow) and getting 25% from carbs and supplementing 50 cal/hour (blue). These lines are closer to the “after” graphs of the two cyclists.

      image

      If I can get up to 50% fat utilization and supplement at 200 calories per hour, I should be able to go 12 hours without running out of glycogen. And if I can get up to 75% from fat, I can supplement at only 50 calories per hour and still easily make it to 12 hours.  I can maybe get to 12 hours without eating anything at all…

      It may be better than that. Remembering back to the first post where we talked about triglycerides, and how those are composed of a glycerol backbone with three fatty acids attached to it. If we are burning lots of fatty acids in beta oxidation, that means the fat cells are breaking apart triglycerides, and that means we have a lot of extra glycerol around.

      Since the body doesn’t want to waste energy, it will try to use the glycerol for something useful, and it can turn it into glucose through gluconeogenesis. That means if we are burning a lot of fat, we will get some glucose out of it to support the glycolysis side. And no, I don’t know how much “some” turns out to be, this is hard to study.

      Some more data

      Is there more data out there that would be interesting?

      Jeff Volek and Stephen Phinney have done a number of studies looking at low-carb/keto diets and athletes (note 1), and here’s one that I think is relevant to this topic:

      Metabolic characteristics of keto-adapted ultra-endurance runners

      One of the problems with dietary studies is that the studies are expensive and time is limited, so it’s very hard to do studies where you change an athlete’s diet and check to see how it affects her for the next 12 months. That is why many of the studies around keto diets and athletes are extremely short; less than 3 weeks. Knowing what we know about how long it takes to achieve training gains in general, it’s pretty clear that 3 weeks is on the short side to see adaptations, so I don’t think most of those studies are very good.

      In this study, Volek and Phinney instead looked at two groups of elite ultra-endurance runners, 10 who followed a high-carb diet, and 10 who followed a low-carb diet. Since each group was on their habitual diet, it’s a pretty good bet they were adapted to it pretty well. I will note at the outset that this is an observational study and therefore there is the possibility that the runners who chose low-carb are genetically better at burning fat than the high-carb ones.

      From the hypothesis about trainability, what difference would we expect to see in fat burning rates between the two groups? Here’s the first graph:

      image

      The graph shows the peak fat oxidation during a 3-hour run for all of the athletes. Every low carb athlete is significantly better at burning fat than even the best fat-burning high carb athletes. If we look at the averages (circles to right), the average in the low carb group was 2.3 times higher than the high-carb group.

      I’d also like to note that the average for the low carb group was 1.54 grams/minute. That would be 92.4 grams/hour, or a whopping 92.4 * 9 = 832 calories / hour from fat burning alone.

      Where did that peak fat oxidation occur? Here’s another graph:

      image

      Here we are looking at the relative intensity of that fat peak for each runner; the HC group peak was at 54.9% of VO2max, and the LC group peak was at 70.3%. Not only are the low carb athletes burning more fat, they are hitting their peak at a higher intensity.

      It is pretty clear that the low carb athletes are burning vastly more fat during a long run than the high carb athletes. And this study was with elite athletes doing ultra-distance events, which means even the high-carb athletes are likely to be decent fat burners.

      There are some other interesting graphs in the paper that I recommend looking at (note 2), and we may come back to it later.

      Summary

      We found out that there seems to be a strong training effect and we can expect to improve the rate at which we burn fat through training. We also learned that if we can do that, it can make fueling more straightforward on long rides; we may need to still supplement, but likely not as much.

      In the next post, we’re going to take a little excursion into talking about weight loss and hunger.

      Part 5: Hunger etc.

      Notes:

      1. Phinney and Volek have written a couple of books that cover this same subject area: “The Art and Science of Low Carbohydrate Living” and “The Art and Science of Low Carbohydrate Performance”.
      2. Section 3.2 talks about submaximal substrate utilization, and features two charts that show the average oxidation rates (in grams/minute) for both fats and carbohydrates. The low carb group was very steady across the whole 180 minutes of the run; both fat and carbohydrate utilization is nearly a flat line. The high carb group saw a drop in carbohydrate and increase in fat over time; they also saw a slight reduction from 13.2 cal/minute at the start of the run to 12.0 cal/minute at the end.


      The endurance athlete’s guide to fueling and weight loss part 2: Energy Systems

      Please read the introduction and earlier posts before reading this one.

      In our last episode, we learned about how fat and carbohydrate get into our systems and the fundamental asymmetry between those two systems.

      In this episode, we will look at how fat and glucose are used for energy. My focus is going to be on muscles since these posts are for athletes; there are similar concepts that apply to other tissues but they are beyond the scope of this discussion (see note 1)

      I’m going to have to dive into some biochemistry to create a framework for further discussion; I have tried to simplify it as much as possible, and the post is relatively short.

      This is a summary of the overall processes at work for aerobic energy production; there are others that are at work for non-aerobic production.

      Each of the boxes is a complex series of chemical steps (note 2)

      image

      We’ll start at the bottom and work our way up.

      Citric acid cycle

      Our goal is to take fat and carbohydrates and create ATP, which is what powers our muscles. At the center of the diagram is a compound named acetyl coenzyme A, abbreviated as “Acetyl CoA”. You can think of Acetyl CoA as the common energy compound produced from either glucose or fatty acids. That’s not absolutely correct, but it’s correct enough.

      The Acetyl CoA feeds into the citric acid cycle (sometimes known as “Kreb’s Cycle” after one of the discoverers). If you want to see the details, here’s a link:

      Image result for don't press this button image

      You pressed it, didn’t you?

      Biochemistry is just ridiculously complex.

      All of this takes place inside of each of most of our cells. You probably studied cells when you were in biology class, and the diagrams looked something like this:

      Image result for simple cell nucleus mitochondria

      Specifically, the citric acid cycle happens within the mitochondria, which you can see are those oval-shaped structures within the cell. I generally think of cells as being these little spheres or flat discs, and some are, but muscle cells are a bit different:

      Image result for muscle cell nucleus mitochondria

      Not at all like a globule or a flat disc; I especially like the nucleus stuck on the side as an afterthought.

      Muscle cells need a lot of energy so they have tons of Mitochondria. Thousands of them.

      And how are we going to get all the Acetyl CoA that we need to drive those muscles?

      Glycolysis and Beta Oxidation

      Glycolysis is the the process that converts Glucose to Acetyl CoA. Beta oxidation is the process that convert fatty acids to Acetyl CoA. Both of these processes happen in the mitochondria.

      There is a very important point to be made about glycolysis and beta oxidation. Even though they both occur in the mitochondria and even though they both feed into the citric acid cycle, they use different chemical pathways and therefore use different machinery.

      The citric acid cycle, glycolysis, and beta oxidation can all be trained. Not only can the body build more mitochondria, it can improve the rate at which each bit of machinery in the mitochondria works. That is one of the adaptations that makes us better at aerobic exercise.

      But the body only makes improvement to the parts of the machinery that are being stressed, so if it’s glycolysis that is being stressed, improvements will be made in the glucose metabolism, and similarly if beta oxidation is stressed, improvements will show up in fat metabolism.

      This point is going to be fundamental to our next discussion, so I’ll state it again in a different way – you can be an athlete with a very powerful glycolysis pathway and a crappy beta oxidation pathway. And vice versa.

      Another interesting outcome is that to achieve higher performance, you need to improve either glycolysis or beta oxidation *and* the citric acid cycle. You can do a bunch of work to improve your beta oxidation and you’ll burn more fat and fewer carbs, but you won’t see higher performance if the citric acid cycle is unchanged (note 3).

      Summary

      This was a short post and I’m not sure I really need a summary, but the short one is that both fat burning and glucose burning have separate chemical reactions (glycolysis and beta oxidation) that feed into a shared set of chemical reactions (the citric acid cycle) that ultimately give the energy to drive the muscles.

      That’s all for this post. The next post will take this post and apply it in real-world situations.

      Part 3: Carbohydrate and fat use in actual athletes

      Notes

      1. That would take us into ketone bodies and their usage in different tissues.
      2. Like, ridiculously complex. In glycolysis, to get from glucose to pyruvate – which is fed into the citric acid cycle – takes a series of 10 different chemical transformations. Beta oxidation goes through 5 steps but repeats them for every two carbon atoms on the fatty acid, and it’s more complex for unsaturated fatty acids. The citric acid cycle has 9 steps.
      3. This depends a bit on what you mean by “performance” and it ties into fueling, which will come up very soon in a future post.