Ketogenic Experiment

From, Running tips
Jump to: navigation, search

I used the Ketogenic Diet for 7 months, running an average of 85 miles/week, peaking at 200 miles/week, including over 50 marathon length runs. I was surprised how little impact the Ketogenic Diet had on my running, and I could maintain my training regime without difficulty. However I was not able to race successfully on the Ketogenic Diet and I found the Ketogenic Diet difficult to comply with. This write up should not be considered as scientific in any way; it is simply my anecdotal experience.

1 Background

As an ultrarunner, I've experimented with Low Carbohydrate Diets (LCD) before with little success. I found that running on LCD let me feeling like I was permanently Glycogen depleted. However, after reading The Art and Science of Low Carbohydrate Performance and Ketogenic Diets: Treatments for Epilepsy and Other Disorders I decided to experiment with a true Ketogenic Diet which is quite different from most general Low Carb Diets.

2 My Ketogenic Diet

  • I was on the Ketogenic Diet from March to November of 2013 (~7 months).
  • For some of the time I was on the Ketogenic Diet I analyzed my food in detail. During this time my average diet was 3,740 Calories/day, 357g fat, 53g carbs, 27g fiber (26g Net Carbohydrates), and 88g protein. This is a Ketogenic Ratio of 3.13:1, and the calories ware 87.6% fat, 2.8% carbohydrate, 9.6% protein.
  • I found the ketogenic diet remarkably hard work and tiresome. The food choices were grim, and it was not really possible to eat anything even vaguely like a normal diet. Even bacon was too low in fat to be eaten freely.
  • I found my body fat dropped to its lowest level ever. However, I'm not sure how much of that is the Ketogenic Diet directly and how much is because I was analyzing everything I was eating, so I was far more aware of any excess calorie intake. Another possibility is that my appetite was dramatically lower on the Ketogenic Diet, especially at the beginning. I could run a marathon distance run and not be hungry for hours afterward.
  • I experimented with Medium Chain Triglycerides, but I found that MCT cause nausea and GI problems, which in turn hindered my running performance.
  • I seemed to have more migraines on the Ketogenic Diet, though my migraines are not predictable enough to be sure.
  • I had more problems with my skin condition on the Ketogenic Diet, though with hindsight I suspect this was due to an insufficient Vitamin C intake. It's hard to get the macronutrients on a KD, so supplements are required.
  • There are reports that Vespa Gel can help performance when on a Ketogenic Diet. The science behind Vespa is extremely poor, with a few animal studies indicating improved performance when taking the equivalent of 100-200 packets of Vespa. However, I performed a simple double blind test on Vespa and found no discernable impact. While I believe that Vespa only acts as a placebo, it's worth remembering that a good placebo can improve performance dramatically.

3 Ketone Levels

One common question I get about my ketogenic experiment is if my ketone levels were actually elevated.

  • I used a blood ketone meter from May to the end of my experiment to ensure my Ketone levels were sufficiently high and to evaluate the factors that impacted my levels. I measured my Ketone levels first thing in the morning and also during some of my long runs.
  • During the time I was measuring my Ketones the average level was 2.0 and ranged from 0.2 to 6.8. This is well above the Ketone Levels that most authorities suggest as adequate for being ketogenic.
  • I also checked my blood glucose levels, and my average morning level was 78.
  • I used urine test strips regularly as well, but found their value limited.

The charts below show the relationship between the values, with the regression line in blue and the 95% confidence intervals shaded. As you can see, for my readings there is better correlation between blood glucose and blood BOHB than there is to urine AcAc. I tried adjusting for hydration using the Urine Specific Gravity values from the test strips, but this made little difference to the correlation. Multiple regression of blood glucose and urine AcAc to blood BOHB did improve the model to r2=0.73.

My blood BOHB against my urine AcAc levels. r2 is 0.43, p < 0.001.
My blood BOHB against my blood glucose levels. r2 is 0.68, p < 0.001

4 Ketoadaptation

The advice from books like |The Art and Science of Low Carbohydrate Performance is that Ketoadaptation takes at least two weeks, and the classic study by Phinney used four weeks to ensure complete adaptation. My experience fitted in with this guidance and for the first couple of weeks running was much harder. My Heart Rate (HR) and Breathing (Respiration Rate, RR) was much higher than I would expect for the given pace. It also seemed like my RR went through cycles; at some points the RR would be what I would expect for the elevated HR, and then increase so that my RR was higher than expected, before lowering again. This would repeat every few minutes. After Ketoadaptation my running felt similar to how I feel on a high carbohydrate diet. However, an analysis reveals that my HR was higher than it would be on a high carbohydrate diet (see "Running Economy" below.)

5 Ketogenic Running

  • I ran an average of 85 miles/week, generally running 4 days/week with an average distance of 20 miles. I ran the marathon distance or longer 58 times.
  • I found I had no problems running marathon distance training runs at an easy pace without any fueling on the Ketogenic Diet, but I can also do that on a high carbohydrate diets.
  • Taking small amounts of carbohydrate during a long run could be done without impacting my Ketone Levels, but too much would cause the levels to drop precipitously. To my disappointment, the extra carbohydrate did not seem to reduce my fatigue or improve performance. I also tried taking extra fat, or mixtures of fat/carbs/protein, which seemed to help a little.
  • While I could run the marathon distance without fueling, running longer distances required some source of calories. I discovered I could still 'crash' on the Ketogenic Diet, though it is a softer, slower crash than a typical blood glucose/Glycogen depletion crash (AKA "hitting the wall"). My suspicion is that I ran out of available fat to use a fuel, even though obviously I had plenty of body fat left. One issue might have been the depletion of intramuscular triglycerides. On a high carbohydrate diet, intramuscular triglycerides provide about the same amount of fat calories as plasma fatty acids (I've not seen any data on their contribution while on a KD.) However, the other issue might be the rate of lipolysis, which is the breakdown of body fat to make it available for the metabolism. It's possible that I was burning fat faster than I was freeing it up.
  • I found that I could put on upper body muscle using weight training at least as easily as I could on a high carbohydrate diet. However, my leg muscle mass was reduced on the Ketogenic Diet.
  • I was able to handle far higher training volumes without it impacting my Mood State, especially when my mileage hit 200 miles/week. (Mood State is a key indicator for Overtraining Syndrome.)
  • I had two runs on the Ketogenic Diet where I experienced a strong feeling of euphoria. The first time the euphoria occurred from mile 16 to 26 of a normal training run. The second time was a shorter 10 mile run with sore legs, where I noted a strange combination of fatigue, soreness and euphoria. I never worked out what caused this euphoria or found a way of reproducing it. Note that while I had a sense of euphoria and strength, my actual pace was not much better than normal. I had one other run where I noted "running 28 miles should not be this easy", and would have run much further had time allowed.

6 Ketogenic Ultramarathons

I took part in three races while on the Ketogenic Diet; the Midnight Boogie, Laurel Valley and Hinson Lake 24.

  • The Midnight Boogie is a 50 mile race in the heat of the North Carolina summer. I started running without calories and felt fine for the first 26 miles, at which point things "went squirrely". I had a few potato chips (crisps in the UK) at mile 26, then I took a peanut butter GU at mile 29, which helped enormously. Over the rest of the race, I had a few more handfuls of chips and two Pocket Fuel gels, the last one about mile 46, finishing the race in 8.54 (8th place). That's not a terrible result, but it was slower than I'd have expected (I've won the race in 7:04 previously.) I estimate that from mile 26 to 50 I consumed about 800 calories, 46g fat, 14g protein and 64g carbs. To put that in perspective, my basal metabolic rate for 9 hours is probably around 800 calories, and my total calorie burn for the race would be around 5,000 calories. After the race my Blood Ketones were 3.1 and my Blood Glucose was 105. The high ketones indicate that I was still ketogenic, but the elevated blood glucose is a little odd. I drank my DIY Electrolyte Drink throughout the race.
  • For Laurel Valley (~35 miles) I took things easy, and sadly I made no useful notes in my training log. I recall that I was not attempting to be competitive, and I hung back to look after another runner who had been stung by a bee.
  • Hinson Lake 24 Hour was a disastrous race. By mile 24 my blood pressure was low, my breathing was strained, my blood ketones were 0.8, and blood glucose 108. At mile 26 I started to suffer from Nausea, but I continued to push on until mile 42 when my blood pressure was too low to stand. After some medically trained runners took care of me for some time, I recovered enough to walk another lap (~1.5 miles), then run a lap, before retiring from the race. You can read more at 2013 Hinson Lake 24 Hour.

7 Running Economy

The graph below is of my Relative Running Economy over the three year period that includes the time I was on the Ketogenic Diet. As you can see my Running Economy was lower on the Ketogenic Diet than with more carbohydrates. This is what I would expect, has fat requires more oxygen to produce a given amount of energy compared with carbohydrate.

Relative Running Economy over a three year period.

8 Conclusion

Here are my personal conclusions based on my interpretation of the scientific evidence available and my own personal experience.

  • I believe that there are some huge benefits as well as equally large risks associated with the ketogenic diet. Therefore, I find myself neither an advocate for the ketogenic diet nor a naysayer. Instead, my recommendation is for each person to weigh up the pros and cons based on their particular situation, and make an informed decision.
  • I recommend medical supervision for the ketogenic diet. It's possible to act as your own doctor for the ketogenic diet as I did, but this increases the risks and requires that you do far more research. You need to be comfortable with handling your own medical tests and supplementation.
  • In talking to several other runners, I suspect that many of them are not actually on a Ketogenic Diet, just a low carb diet. I know of several runners who have done much better on a low carb diet than their prior diet, but that does not mean that low carb is optimal, just that there are worse diets.
  • It takes more oxygen to burn fat than carbohydrate. This simple bit of chemistry is critical for understanding the Ketogenic Diet. If you want to run at a given pace, it will be harder if you're burning fat than if you're burning carbohydrates.
  • I believe the human body is amazingly well designed, and it carefully optimizes the use of carbohydrates and fats, switching gradually from fat to carbohydrate as exercise intensity increases. If there were a simple way of burning fat as efficiently as carbohydrate, humans would make use of this all the time. Instead, it seems that Ketones are a way of surviving famines, something that is quite wonderful.
  • After finishing my Ketogenic Diet, I've found my attitude towards fat is different. I eat more saturated fat, including whole milk rather than skim milk, butter instead of margarine. I've always kept my intake of Omega 3 high, which helps with my skin condition. However, now I avoid Omega 6 oils as much as I can.

9 See Also