March 14, 2012


The interplay of exercise and ketosis – Part II

Read Time 10 minutes

You’ll recall from last week’s post I did a self-experiment to see if I could learn something about the interplay of exercise and ketosis, at least in myself. To understand this discussion, you’ll want to have read Part I of this post.

However, before getting to this, I want to digress and briefly address two unrelated issues:

  1. Some of you (about 67 or 68 as of this writing) have sent me various links to news reports released yesterday reporting on a study out of Harvard’s School of Public Health.  I was planning to eventually write a post about how observational epidemiology is effectively at the heart of the nutritional crises we face – virtually every nutrition-based recommendation (e.g., eat fiber, don’t eat fat, salt is bad for you, red meat is bad for you) we hear is based on this sort of work.  Given this study, and the press it’s getting, I will be writing the post on observational epidemiology next weekHowever, I’m going to ask you all to undertake a little “homework assignment.”  Before next week I would suggest you read this article by Gary Taubes from the New York Times Magazine in 2007 which deals with this exact problem.
  2. I confirmed this week that someone (i.e., me) can actually eat too much of my wife’s ice cream (recipe already posted herepretty please with lard on top no more requests for it).  On two consecutive nights I ate about 4 or 5 bowls of the stuff.  Holy cow did I feel like hell for a few hours.  The amazing part is that I did this on two consecutive nights.  Talk about addictive potential.  Don’t say I didn’t warn you…

OK, back to the purpose of this post: How is ketosis impacting my ability to exercise? Here is the summary of the results from my personal experiment:


Exercise summary data

Let’s take a closer look at what may have been going on in each workout and see what we can learn.



This workout probably produced the most lactate of the three workouts (we don’t know for sure because I only measured immediate pre- and post- levels without measuring in-workout levels).  My glucose level rose by nearly 40% during this workout despite the fact that I did not consume anything.

How does this happen?  Our bodies store glucose in the liver and in muscles in a “storage” form (a long chain of joined glucose molecules) called glycogen.

Whenever our bodies cannot access sufficient cellular oxygen, our metabolism shifts to a less efficient form of energy acquisition called anaerobic catabolism.   During these periods of activity, we cannot oxidize fat or glycogen (i.e., use oxygen to harness the full chemical potential of fat or carbohydrate molecules).  I will be writing in much more detail about these ideas in the next month or so, so don’t worry if these ideas seem a bit foreign right now.  Just know that sometimes our bodies can convert fat or glucose to energy (efficiently), and sometimes we can only convert glucose to energy (inefficiently).

Because of my ketosis, and the metabolic flexibility that accompanies it, I only “require” that my body turn to glucose for energy under the most “stressful” forms of exercise – like I was doing a lot of during this workout.  But keep in mind, my muscles CANNOT export one gram of the glucose they store, so any glucose in my bloodstream is either ingested (which I didn’t do) or coming from my liver, which CAN export glucose.

Furthermore, the stress of a workout like this results in my adrenal glands releasing a set of chemicals called catecholamines, which cause my liver to export even more of its stored glucose via a process called hepatic glucose output (HGO).

[As an aside, one of the major defects in type-2 diabetes is the inability of insulin to suppress HGO.  In other words, even when not under the catecholamine stress that “should” lead to HGO, their livers constantly export glucose, which contributes to elevated blood glucose levels.  The very popular drug, metformin, used often in type-2 diabetes, blocks this process.]

While I did experience a pretty large rise in lactate (almost 3x), my ketones still went up a bit.  This could imply a few things:

  • Elevated lactate levels do not directly inhibit beta-hydroxybutyrate (B-OHB)
  • Mild elevations in glucose do not directly inhibit B-OHB
  • Mild elevations in glucose do not directly inhibit B-OHB, if insulin is being suppressed (as is the case during vigorous exercise)
  • B-OHB was suppressed, but we are only appreciating the net effect, which was a small increase (i.e., because of my MCT oil and activity, B-OHB levels were rising dramatically, but the rise was blunted by some other factor, such as HGO, insulin, and/or lactate)

More questions than answers from this workout, so on to the next workout.



Despite this being a tough ride at several points, on average it was less stressful than the other two workouts and I spent the greater fraction of time in my aerobic to tempo (zone 2 to zone 3) zones.

A ride like this, however, is a great example of the advantages of improved metabolic flexibility that accompanies nutritional ketosis.  My average heart rate during this 6 hour ride was 141.  Prior to becoming ketotic, at a HR of 141 my respiratory quotient (RQ) was about 0.98, which meant I was almost 100% dependent on glycogen (glucose) for energy.  Today, at a HR of 141 (with the same power output), my RQ is about 0.7 to 0.75, which means at the same HR and same power output as prior to ketosis, I now rely on glycogen for only about 10% of my energy needs, and the remaining 90% comes from access to my internal fat stores.

This is an important point.  I will devote future posts to this topic in more detail, but I wanted to use this opportunity to mention it.

So what happened physiologically on this ride?

  • My glucose levels fell, probably because I was slowly accessing glycogen stores for peak efforts (once my HR reaches 162 I become 50% dependent on glycogen) throughout the ride (e.g., peak climbing efforts, hard sections on flats), but my liver was not “called on” to dump out a massive amount of glucose in response to a catecholamine surge (and if it was, at some point during the ride, that amount of glucose had been used up by the time I was finished).
  • B-OHB levels increased by about 2.5x – to 4.4. mM, which is pretty high for me.  My highest recorded B-OHB level was 5.1 mM (also after a long ride).  This confirms what my RQ data indicate — my body almost entirely relies on fat oxidation for energy for activity at this intensity.  In the process, B-OHB is generated in large quantities, both for my brain and also my skeletal muscles (e.g., leg muscles). In reality, cardiac myocytes (heart muscle cells) also “like” B-OHB more than glucose and probably also access it when it is abundant.
  • Lactate levels by the end of the ride were effectively unchanged though. Based on “feel,” I suspect I hit peak lactate levels of 8 to 10 mM on this ride during peak efforts, but I had ample time to clear it.

A few observations:

  • I consumed 67 gm of carbohydrate on this ride (of which 50 gm was Generation UCAN’s super starch), yet this did not appear to negatively impact my ability to generate ketones.  Technically, we can’t be sure this is the case, since I would have needed a “control” to know this (e.g., my metabolic and genetic twin doing and eating everything the same as I did, but without the consumption of super starch and/or without the bike ride).  It’s possible that super starch did slightly inhibit ketosis and that my B-OHB level would have been, say, 5.0 mM instead of 4.4 mM.  Metabolic studies of super starch show that it has a minimal impact on insulin secretion and blood glucose levels, hence the name “super” starch.
  • Whatever impact peak levels of lactate production and hepatic glucose output had during the ride, they seem blunted by the end of the ride (and the ride did finish with a modestly difficult 1.4 mile climb at 6-7% grade, which I rode at a HR of about 150).

Since neither lactate levels nor glucose levels (nor insulin levels by extension) were elevated, I can’t really draw any conclusion about whether one factor, more than any other, suppressed production of B-OHB, so on to the next workout.


High intensity training

This sort of workout spans the creatine-phosphate (CP) system and the anaerobic energy system, and probably involves the aerobic energy system the least. I’ll write a lot about these later, but for now just know the CP system is good for very short bursts of energy (say 10-20 seconds) and recall the previous discussion of aerobic and anaerobic catabolism.  In other words, this is the type of workout where my nutritional state of ketosis offers the least advantage.

  • This workout saw the greatest increase in glucose level, about 70%.  It is important to recall that during this workout I ingested water with a small amount of branched chain amino acids (BCAA’s – valine, leucine, isoleucine) and super starch, about 4 gm and 10 gm, respectively. I do not believe either accounted for the sharp rise in blood glucose and, again, I believe hepatic glucose output in response to a strong catecholamine surge attributed to this increase.
  • Lactate levels also rose, though probably less so than during a peak swim effort.  This suggests more of the effort in this workout was fueled by the CP system (versus the anaerobic system, which probably played a larger role in the swim workout).
  • This was the only workout that saw a fall in B-OHB levels, which now offers some insight into what might be impacting B-OHB production.

Contrasting this workout with the swim workout draws a pleasant contrast: both saw a similar rise in lactate, but one saw twice the rise in blood glucose.  In the former, B-OHB was unchanged (actually rose slightly), while in the latter, B-OHB fell by over a third.

This suggests – but certainly does not prove – that it is not lactate per se that inhibits ketone (B-OHB) production, but rather glucose and/or insulin.  It is possible the BCAA played a role, and if I was thinking straight, I would not have consumed anything during this workout to remove variables. But I have a very hard time believing 3 or 4 gm of BCAA could suppress B-OHB. When you see hoof prints in the sand, you should probably think of horses before you think of zebras.

Conversely, there is some evidence that lactate promotes re-esterification of fatty acids into triglycerides within adipose cells.  What does that mean in English? High levels of lactate take free fatty acids and help promote putting them back into storage form.  This would prevent free fatty acids from making their way to the liver where they could be turned into ketones (e.g., B-OHB). In other words, we may be missing this effect because of my sampling error – I only sampled twice per workout, rather than multiple times throughout the workout.


So what did I learn, overall?

I think it’s safe to say I did not definitively answer any questions, which is not surprising given the number of confounding factors, lack of controls, and sample size of one.  However, I think I did learn a few things.


Lesson 1

The metabolic advantages of nutritional ketosis seemed most apparent during my bike ride, evidenced by my ability to access internal fat stores across a much broader range of physiologic stress than a non-ketotic individual.  (More on this in Lesson 4.)


Lesson 2

The swim and high intensity dry-land workouts suggested that my state of nutritional ketosis did not completely impair my ability to store or export hepatic glucose.  This is a very important point!  Why?  Because, it runs counter to the “conventional wisdom” of low-carb (or ketotic) nutrition with respect to physical performance.  We are “told” that without carbohydrates we can’t synthesize glycogen (i.e., we can’t store glucose).  However, those who promote this idea fail to realize that glycerol (the backbone of triglycerides) is turned into glycogen, along with amino acids, not to mention the 20 to 40 gm of carbohydrates I consume each day (since my brain doesn’t need them).  We know muscles still store glycogen in ketosis, as this has been well studied and documented via muscle biopsies by Phinney, Volek, and others.  But, my little self-experiment actually adds a layer to this.  Because muscle can’t export glucose (muscle lacks the enzyme glucose-1-phosphatase), we know that the increase in my blood glucose was accounted for by HGO – my liver exporting its glycogen.  In other words, ketosis does not appear to completely impair hepatic glycogen formation or export.  Again, we’d need controls to try to assess how much, if any, hepatic glycogen formation and/or export is inhibited.  It’s hard to make the argument that being in ketosis is allowing me to swim and do high intensity training with greater aptitude, and as I’ve commented in the past, I feel I’m about 5-10% “off” where I was prior to ketosis for these specific activities, but at the same time, I could be doing more to optimize around them (e.g., spend less time on my bike which invariably detracts from them, supplement with creatine which may support shorter, more explosive movements), which I am not.


Lesson 3

Consuming “massive” amounts of super starch (50 gm on the ride), did not seem to adversely affect my ketotic state.  My total carbohydrate intake for that day, including what I consumed for the other 18 hours of the day, was probably close to 90 gm (50 gm of super starch plus 40 gm of carbs from the other food I ate).  This suggests one or two possibilities:

  • Because of the molecular structure of super starch (I’ll be discussing this in the future, so please hold questions) and the concomitant metabolic profile that follows from this structure, it may not inhibit ketosis like other carbohydrate, and/or
  • During periods of profound physical stress insulin secretion is being sufficiently inhibited that higher-than-normal amounts of carbohydrate can be tolerated without negatively impacting ketone production.

This is pretty straightforward to test, even in myself.  I just haven’t done so yet.


Lesson 4

While it’s probably the case that my liver has less glycogen (i.e., stored glucose) at any point in time, relative to what would be present if I were eating a high-carb diet, it’s not clear this matters, at least for some types of workouts.  Why? Take the following example:

  • Someone my size can probably store about 100 gm of hepatic (liver) glycogen and about 300 gm of muscle glycogen at “full” capacity. This represents about 1600 calories worth of glucose – the most I can store at any one time.
  • Before I was ketotic, my RQ at 60% max VO2 (about 2,500 mL of O2 per min consumption) was nearly 1.00, so at that level of power output (a pace I can hold for hours from a cardiovascular fitness standpoint) I required 95% of my energy to come from glycogen.  So, how long do my glycogen stores last?  2,500 mL of O2 per minute translates to about 750 calories per hour, so I would be good for about 2 hours and 15 minutes on my glycogen stores.
  • Contrast this with my ketotic state.  Let’s assume my glycogen stores are now only half what they were before.  Muscle biopsy data suggests this is probably an overly conservative estimate, but let us assume this to be the case. Now I only store 50 mg of hepatic glycogen and 150 gm of muscle glycogen, about 800 calories worth of glucose.
  • In ketosis, my RQ at 60% max VO2 is 0.77 (at last check), telling me I am getting only 22% of my energy from glucose and the remaining 78% from fat.  So, how long do my depleted glycogen stores last? Nearly 5 hours.  Why?  Because I barely access glucose at the SAME level of oxygen consumption and the same power output. 

I know what you’re thinking…why is this an advantage? Just consume more glucose as you ride! It’s not that simple, but you’ll have to wait until my upcoming post, “What does exercise have to do with being in the ICU” to find out.

Going back to the black sheep example I open Part I of this post with, we know that at least one person in nutritional ketosis seems to make enough liver and muscle glycogen to support even the most demanding of his energetic needs.


Photo by Troy Oldham on Unsplash

Disclaimer: This blog is for general informational purposes only and does not constitute the practice of medicine, nursing or other professional health care services, including the giving of medical advice, and no doctor/patient relationship is formed. The use of information on this blog or materials linked from this blog is at the user's own risk. The content of this blog is not intended to be a substitute for professional medical advice, diagnosis, or treatment. Users should not disregard, or delay in obtaining, medical advice for any medical condition they may have, and should seek the assistance of their health care professionals for any such conditions.


  1. Doc,

    I can do all the cardio I want without problem, but when I add weightlifting (being mindful to only do it a couple of times per week, not over-doing it, etc.), I ‘hit a wall’ and feel run down. Any idea why this would be? I’ve reached my goal weight, and am very active, but I’d love to add some more muscle but not if lifting weights is going to make me feel like garbage again : (


    • Mark, it definitely took me a while to get my “juice” back in the weight room once eliminating carbs. Over 12 weeks. In reality, I am probably 5-10% weaker now than I was on a “normal” carb diet. For me the tradeoff is worth it, but it might not be for you. That said, I think you may need more time to adapt. Also, you need to make sure you’ve got all the other variables tweaked (e.g., sodium, potassium, magnesium, BCAA, glutamine).

  2. Thank you so much for this and all your other posts! So wonderful to have somebody actually providing the science behind the Carbohydrate hypothesis in a blog form (I’m looking at your Mr. Taubes! 😉 )

    I exercise constantly, and one of the questions I am constantly asked is, “Don’t I need to supplement with some Carbs?” I generally don’t know. I know there are diets offered by Lyle McDonald called the “Targeted Ketogenic Diet” and “Cyclic Ketogenic Diet” or TKD and CKD respectively. However, McDonald has generally dismissed the science offered up by Mr. Taubes which I find run directly counter to the very things McDonald espouses.

    As Mark posted earlier, I sometimes hit a wall a little earlier when lifting heavy weights while trying to stay in ketosis, which I would be able to push through with some carbs previously. However, I still continue to gain strength and muscle mass, because I eat when I am hungry until I am full.

    The one thing i wonder about is whether it is possible to gain weight while eating lo-carb and what is the mechanism? Does it essentially have to be an insulin response via Gluconeogenesis?

    • Try experimenting with creatine for 4-week cycle? You can definitely gain weight weight eating low carb (I’m assuming you mean lean mass), but it is a bit tougher in low insulin environment. Remember, insulin is an anabolic hormone.

  3. It seems like from your experience and from what Steve Phinney’s studies show, ketosis is the optimal state for an endurance athlete. Any idea if this type of diet has been adopted by a significant number of professional triathletes/marathoners/cyclists? Phinney is also big on never breaking ketosis, so no nonsense like carb cycling or carb loading. Do you agree with this?

    • I think the perfect athletic event for ketosis would be an Ironman triathlon or marathon running or ultramarathoning. Anyone who needs to be *just* under threshold for really long periods of time will capture most of the benefits of ketosis. But in my experience (and I believe Steve’s), do you have to give up some sprint capacity, so I would not recommend it even for the guy trying to win the Tour de France, as you still need to sprint when the pack moves.
      Carb cycling does not make sense if your goal is to really optimizing your body around ketosis. Perhaps it has benefits for other reasons.

    • Can’t you work around the sprint capacity issue by doing occasional carb refeeding? That’s what I do. I am very low carb during the week for marathon training, and will eat more carbs for my long runs and occasionally carb up for speed training. Am I still going to lack spring capacity on race day even if I’m carbed up?

      • I could, but it would mean going into and out of ketosis constantly. Also, the carbs would take away from my sub-threshold performance. It’s all a trade-off. No “right” way to do it. I’ll keep fiddling with it over time.

    • First, I’ll admit I’m really new to the notion of MET (Seebohar) and Ketosis based endurance training. I just stumbled on this site, but have been reading and working on following Bob’s work for a couple months now. I haven’t tested myself to the level that Peter has, but I am definitely convinced that this is a good model for me. And I am an Ironman, Ultramarathon, and long distance solo cyclist.

      In reply to Peter’s comment about Ironman, Marathons, and Ultramarathons being the “perfect athletic event for ketosis”… I would take one small exception with Peter’s suggestion that Ketosis is ideal for Marathons… specifically “elite” Marathon runners, here’s why:

      At the current world class marathon times of 2:05:00-ish, I think we are really seeing a breakthrough in the pace and timing of (world class) marathons where they can realistically be completed at essentially near-AT pace ~RQ in the >0.8…>0.9 zone. I haven’t run the math on this, and so Peter may have a lot more data and numbers to disprove my assumption. But my instinct tells me that these world class runners are actually running for a short enough duration of time that they can fuel almost entirely off of the calories available in their muscle and liver glycogen stores. Consider that a 130lbs runner will only require about 2200 calories to run at 5 min/miles for 2 hours.

      Of course I totally agree that for mere mortals who are going to run 3+ hour marathons, or more like 4+ hour marthons, then Ketosis or near-Ketosis is a far more preferable metabolic “mode” to race in.

      Anyway… like I said, I just found this site, but so far, I’m loving everything I’m reading. Great stuff!

  4. Peter,

    Thanks for this excellent post. I used to scour books searching for a reasonably clear explanation of energy transport during various phases of exercise as a long-distance runner. It seemed as if most authors merely parrot the same series of unexamined and rather unexplained literature on the subject. I feel like I have a better feel for energy transport systems after this post, not just from the standpoint of being in a ketotic state (which I’m not…I’m low carb, but haven’t taken the final plunge), but just in general.

    An interesting take-away as a former marathoner is your pre and post ketotic energy transfer as it related to your aerobic cycling data. Think “hitting the wall” in terms of the 2 hr and 15 min store of glycogen pre-ketotic and then the post-ketotic 5 hrs +.

    One thought that comes to mind. Not that I care about race performance anymore (I’m not out for a gold medal or anything close, just my health), but could one selectively use simple carbs for a race event in order to maximize performance without any change to the ketotic state?— or, if not fully ketotic (like me), in any case avoid the ill effects of carb’s in general, such as insulin production. In other words, like the endurance athlete drinking a few cups of coffee, could hitting some carbs on race day be beneficial without ill effects if you merely see them as a sports supplement? I realize that you use UCAN (which I plan on trying myself), but it comes to mind, why bother?–why not use any simple carb if your metabolic system is already suppressing insulin in the midst of an endurance activity.

    Again, great blog and information. By the way, my wife and I made the ice cream and we are addicted.


    • Greg, if my goal were to win a gold medal I would probably have to change my diet. If you think about it (I’ll be doing an entire post on this), carbohydrates are in some ways like a performance enhancing drug — help with certain aspects of performance, but have some chronic harm associated with use.

      UCAN offers many advantages over, say, fructose, glucose, or maltodextrin. I’ll write about this at some point, also.

    • Peter, I would love to see the article comparing carbs to performance enhancing drugs! One thing that has always bugged me is when people say that our bodies must prefer (be designed for) carbs since we use those first for energy. The other option is that those carbs are seen as toxic by the body and are used first just to get rid of them. Fat could still be the preferred energy source.

  5. “When you see hoof prints in the sand, you should probably think of horses before you think of zebras.”

    Unless you’re in Africa.

    (Fantastic entry/entries! I’m SO glad you’re doing — and reporting! — these great N=1 experiments. Thank you for your blog!)

  6. Great post.

    Is there any value then in trying to deplete glycogen as much as possible via weight training or some other means to improve insulin sensitivity? I, for one, was very surprised at how much glycogen the body can store even when in ketosis.

    • Neil, I can’t think of one off hand, except that it would allow you to ingest more carbs later to replace glycogen stores (vs. de novo lipogensis). Of course, this would need to be balanced against the downside of having very low glycogen levels.

  7. Hey Peter, as usual great post…I think actually reading it as I had just read J. Stanton’s post debunking the aforementioned Harvard study (which I highly recommend reading for anyone on the insulin hypothesis side of the debate, or anyone, for that matter) it’s going to take a further retreading to completely absorb, but, eh, something for after my nap.
    Oh, and as far as the topic of losing some strength in the gym and sprint speed through being ketogenic: I absolutely refuse to believe that our fore bearer’s ability to outrun a sabretooth or out wrestle a grizzly would be predicated upon them having recently consumed some berries or something, there must be some piece of the puzzle that we’re missing. Mmmm, berries. Oh, and I find the reaction to the ice creamer recipe amusing…you really should just edit your previous post to include it in the body, I mean, low carb ice cream to an audience of, mostly, die hard low carbers (who’re therefore largely self selected for being carbohydrate addicts), that’s just chumming the waters, lol.

  8. Thanks for the link to the 2007 Taubes article! Although I frequently google his name to see what’s new, I had somehow missed that one.

  9. “Today, at a HR of 141 (with the same power output), my RQ is about 0.7 to 0.75, which means at the same HR and same power output as prior to ketosismes from access to my internal fat stores.” Guessing due to your weight loss your weight to power output actually increased. In addition to vo2 testing, any testing done on your ftp- if your familiar with that.

  10. Peter-
    Thanks for assembling so much interesting information in such a useful way. In your interpretation of your latest experiment, I wonder what role you think that gluconeogenesis plays in the glucose levels you measure. I would imagine that your hepatic glycogen stores could be much lower than you estimate, and instead that your liver is actually manufacturing glucose to meet your body’s needs. Under this scenario, your capacity to exercise would be rate limited by how fast you can break down amino acids and lactate to form glucose, not a finite store of glycogen in the liver. The path to ketosis involves depleting the liver glycogen stores. Is there some experiment you can imagine that would validate that your hepatic glycogen levels get restored while in ketosis?

    • Yes, this is also a possible explanation. My guess, if I had to make one, would be a bit of these, but still primarily HGO for this workout. Given how short this workout is, I think contribution from the Cori Cycle (lactate) and amino acid breakdown would not account for much of the rise I saw. BUT, we can’t know without testing this! This actually would not be difficult to indirectly test (i.e., without a liver biopsy, which would be an unjustifiable risk in my opinion) using labeled glucose. Great point.

  11. Peter, I used to be in the same category as you: lots of activity, endurance sports (marathons, mountainbiking) and rock-climbing. I worked out a lot and stayed lean and fit till about 30. Then with every year I’d get fatter even though I tried to work out even harder and I was fit by all possible standards (rather than fat % 🙂 ). Gary’s book opened my eyes and changed my life. I find your blog now incredibly helpful as well, esp. with your take at combining low-carbs with sports. Thanks for that!

    Now to my question: even though I adopted low-carb / paleo diet some 2 years ago, I’ve been on a strict keto-diet only for some 3 months now. My main source of fat is coconut oil. I supplement it with meat. It works great, I feel great, running or cycling long is not a problem, doing sprints or hill runs is not a problem. Hard climbing is. My main game now is bouldering and it requires greater power output which is, so it seems, glycogen dependant. What would be your recommendation in that regard? Would increased does of L-Glutamin help? I’ve read Lyle McDonald’s book on ketogenic diet and started experimenting with Cycled KD, i.e. I eat some starches (mostly sweet potatos and vegies) on the weekend but then I always struggle a bit with going very-low-carb again in the week. Any suggestions how to make ketogenic diet work for power sports?

  12. Regarding the ice cream addiction potential… there really is something about cream in general that is unique. I could easily pour myself an 8-oz. glass of heavy cream, drink it down, and have another. That’s like 1600 calories. Try doing that with butter, or coconut oil. Ick. Cream is special, and for that reason I need to stay away! 🙂

    Cream plus coffee is even better!

  13. You say:
    >Because muscle can’t export glucose (muscle lacks the enzyme glucose-1-phosphatase), <

    But, 1) then how do they access the glycogen at all, since it is my understanding that g-1-phosphatase is how the polymer is hydrolyzed; and 2) "can't export" should be due rather to some kind of membrane transporter (like Glut maybe or is that only for import…?) shouldn't it?

    • Muscle can IMPORT glucose via GLUT-4, but once it makes glycogen of it, it cannot export it, because it can’t make glucose again. Glycogen is broken down into glucose-1-P, which cannot leave the cell. Glucose-1-phosphatase (present in the liver), take the P off G-1-P so glucose can be exported.

    • Hi. I want to ask to Peter. In you article Ketosis – advantaged or misunderstood state? (Part I), you have wrote the enzyme is glucose-6-phosphatase, no glucose-1-phosphatase. What is right? thanks

  14. CALL ME CRAZY– but I’ve been on a search for your wife’s low carb ice cream recipe and although I see you reference to look for it in previous response, I can’t find it! Can you please make this easier. I love all of your posts– keep it up.

  15. I couldn’t find the ice cream recipe by using your search field. Apparently, the search feature doesn’t peruse the “comments”, only the articles. Is this a behavior you choose, or is it hard-wired into the blog’s software?

    I did eventually find the recipe, so no need to provide a map to it yet once again, but I thought I might bring to your attention that, if you have a choice, there might be something to be said for letting the search feature look through the comments as well as the articles, particularly in a blog like yours where this is so much additional and relevant information posted in the comments.

    • This is a harder problem to solve than I imagined. Every time I find a plug-in that can search comments it has 10 horrible features that render the rest of the search features utterly useless.

      BUT — don’t be afraid to bust out the old CTRL-F trick. I was zipping around the comments using this today.

    • The best way to search comments is use Google and do a site search that includes the term you’re looking for: ice cream recipe

    • Peter: I did try to use Control-F, but it didn’t work because the ice cream recipe was hidden from, and therefore unavailable to, the browser’s “Find” capability by being back a page or two in the three pages of comments.

      Lacie: I tried your idea but it didn’t work. The Google search you suggested did indeed take me to some discussion of those three words, but it wasn’t so easy to find the recipe and I’m not even sure the recipe itself was among the results that Google gave. To eliminate confounding variables, I Googled one of the more unique words from the recipe: espresso

      …and Google couldn’t find the recipe. Try it for yourself. Thanks for the idea. If I could get the idea actually working robustly it wouldn’t be a bad fix to the problem.

  16. Thanks for this series on ketosis and training. There is a lot to ponder and assimilate, and as someone who just discovered you, I’ll be sifting through these posts a few more times I’m sure. On first hit, though, a couple of questions:

    When I first read the change in RQ that you experienced, I thought that the change was in the range for training effect alone and that I wouldn’t account for it from ketosis. Then it occurred to me that you were probably trained previous to starting a ketogenic diet. I quickly looked through your history and also saw the max test, but I’m still not quite certain of the whole timeline thing and whether or not you were trained to the level of this current post previous to the onset of ketosis. I’m trying to ferret out the training effect from the ketosis effect. It seems that you were trained but that first VO2 peak isn’t consistent then. Could you clarify this for me?

    I ask because the metabolic advantage thing is, of course, the thing.
    And also because when you look at the above substrate graphs of the three workouts, and you consider ketones a proxy for ffa’s, then the graphs look pretty regular vs. intensity. So again, I’m trying to ferret out the possible advantage or “improved metabolic flexibility” as you call it, of ketosis over a mixed diet.

    Just a comment on fuel utilization as you mention you’re going to write a post about it later. And please forgive me if I’m misunderstanding what you’re saying and I’m about to be an arse and tell you something you already know, but it caught me when you said, “Whenever our bodies can’t access sufficient cellular oxygen, our metabolism shifts”…. There’s never a time, irrespective of exercise intensity, that we’re unable to access oxidative metabolism. Even at VO2max there’s sufficient oxygen pressure in the mitochondria to effect oxidative phosphorylation. Non-oxidative or RAPID glycolysis (anaerobic) is simply additive. As intensity increases, along with an increased recruitment of FT glycolytic fibers, oxidative routes just can’t keep up with the rate of muscle demand for ATP, so the short glycolytic pathway is increasingly added as a supplement, not as a replacement, to oxidative routes. While cranking out a relatively puny amount of ATP per glucose, it nonetheless does it fast; ergo, the name rapid glycolysis. So this is why a true VO2max achieves an R greater than (the symbols don’t seem to be working in the editor); 1 – you have all the CO2 from oxidation PLUS the CO2 from the buffering of lactic acid from rapid glycolysis. A true VO2max reflects the oxidative pathways cranking at full tilt along with the supplemental glycolytic pathway adding it’s supplement. If this is how you understood it to be, please forgive the lecture, I misunderstood you, but your statement about the “shift” and inability to access oxidative metabolism made me think you were describing the moving out of one pathway into the other, instead of the addition of one pathway onto the other. As an aside to this though, I was also wondering about your decrease in VO2max on your second GXT, and thinking that maybe they didn’t take you all the way to R greter than;1 before calling it a max.

    • Lorraine, great questions. Let me clarify:

      1. Fitness/training were effectively the same between the tests so, yes, the RQ changes were virtually all due to the nutritional shift. Keep in mind, I’m not the first to write or talk about this. Bob Seebohar has written about this, and Steve Phinney has on several occasions going back to 1982. When you remove carbs from the diet (even shy of ketosis), the body increases its ability to access fat stores.

      2. Yes, I understand your last point, and so I’m oversimplifying a bit technically. The practical implication, however, is still that at RQ >= 1, the amount of substrate going through the mitochondria is very small, which implies fatty acid is not much of a substrate at all, hence the “glycogen dependence” idea.

      3. As for the difference in VO2 max (both in absolute VO2 volume and on per kg) — great question??? I definitely felt (and still feel) like I have a bit less glycolytic power, but I was still surprise when the test stopped. I guess it’s possible we should have not further, but it’s not clear I had much left.

  17. Can you elaborate a little on what you mean by, “But keep in mind, my muscles CANNOT export one gram of the glucose they store, so any glucose in my bloodstream is either ingested (which I didn’t do) or coming from my liver, which CAN export glucose.”?

    I don’t really understand the difference between using muscle glycogen vs. liver glycogen. Is the reason that you need energy from HGO because your muscle glycogen stores are depleted? When you say they can’t export glucose, does that mean they can only fuel their own muscles but not redistribute fuel to other muscles/cells that need it?

    • See previous response to this post on March 15. Yes, that is the implication — muscle glycogen can ONLY be used to supply glucose the muscle.

  18. Thanks, Peter. On the RQ again, did you control for weight change like Phinney? I’m only trying to get at the real impact of ketosis vs some other confounder because your results are so dramatic. Phinney’s studies were small, but likewise impressive, so any additional n’s are useful. Haven’t read Seebohar, but will.

    Yeah that glycolyitic peak power thing is a little puzzling, if only because something that improves performance even thru HIT with such effective glycogen sparing, fails when approaching max.

    Thanks again.

    • The weight difference was trivial, plus in only impacts the per kg numbers. Same for training difference. I would guess virtually all of the difference was due to eliminating carbs and reducing insulin levels to unmeasurable amounts (literally). Phinney’s study used a cross-over, so even though it was small, it’s quite powerful.

  19. N=1 with a glucometer yielded some serious swings in glucose for me while pricking myself in the finger intermittently during a HIIT session of 1-minute intervals of 100%+ VO2 effort followed by 3 minutes of recovery in between.

    Would be interesting to get a streaming reading of some of these numbers rather than just pre-and-post.

    I know Tim Ferriss wrote about — actually I can probably find it with a search — here it is:

    DexCom Seven Plus ( The DexCom Seven Plus is the continuous glucose monitor I used and abused. It is an implant that gives you the approximate data of 288 fingertip blood samples per day.

    It would be great if the technology could yield some good devices that could track real-time biomarkers such as insulin, lipids, etc. so you could track your workouts — and everything else — that would probably help elucidate some of this stuff further.

Facebook icon Twitter icon Instagram icon Pinterest icon Google+ icon YouTube icon LinkedIn icon Contact icon