Carbless in Seattle

Hi, I’m Adrienne and I’m going to tell you about my unwitting participation in a Quantified Self experiment. So we’re going to start with a little bit of a background around energy store and metabolism.

So the body uses glycogen to fuel muscles and brain activity. physical activity depletes these glycogen stores, and after about two to three hours of continuous moderate to intensive exercise your glycogen storage start to deplete. You might experience something called a bonk, which is an extreme fatigue, but eating carbohydrates raise those glycogen stores.

I recently experience this fatigue while I was riding my bike in downtown Seattle. Now this was not intense activity. I’d been riding for about 10 minutes and I was going up a slight hill. So it turned out that this was not a single bout of exercise but depletion over time.

A few days before I had started changing my diet, eating all vegetables, lean chicken, seafood, and nut on the guidance of my doctor. Not only had I excluded all of the things lie cakes and cookies but grains such as bread and pasta. Now, three days into this I was feeling terrible. This raised some questions. Not only was I exhausted going up the hill but curious. Am I actually slower in my commute, or does it just feel different. I was also curios how long this was going to last. Really overall I wanted to know how does this low-carb eating impact my athletic performance.

So what did I do?

In October I had fully committed to riding my bike to work at least three times per week, and at that point I was tracking those commutes using Runkeeper. It was December I stopped eating all carbs for a short period of time and I was tracking my non-adherence to this using the Lift habit tracker. And then in April I put all this analysis together to share with you.

Now going into this I had a hypothesis that when eating a low-carb diet my time to commute to work will be longer.

So the first thing that i did was pull out all the data and then plot it. Now this is just the bike commute through Monday to Fridays between October and April. Now a quick look at this shows some pretty outliers. Once I’ve cleaned up the data and something very nice to work with and now I shared how terrible I felt the first few days and I was tracking non-adherence.

So the blues lines here show the days that I consumed carbohydrates. Now this might mean I had pizza for dinner. The green line shows the day that I started this experiment and the grey shaded area is the baseline before I change what I was eating.

So first I started with a sanity check. I felt like every time I ate carbs, the next time I rode my bike to work it felt a lot faster. So here’s a quick sanity check I did, where I calculated the time between the commute before I ate carbs and the commute after I ate carbs. So it’s really interesting to see that all of these times are faster, and not only that they’re substantially faster; six minutes on a 33 minute ride is a big difference.

However there are a few days between, and event five where I only had an improvement of about 11 seconds and that reflects the methodology that I used to analyze it. It doesn’t show the 25 mile bike ride I did in the meantime.

So going back to this commute data I was encouraged that I saw this improvement after a bout of eating carbs, but that is still not convincing. It hit me as I was sitting at the drawbridge on my way to work that I have little control over the entire duration of the commute. Not only do I hit red lights and construction, this is rainy season is Seattle.

So I reviewed the entire commute and extracted out a section that was free of construction or red lights. It’s called the Dexter climb, and it has somewhat of a reputation in Seattle. A climb of about 40m over a kilometer so not too bad but it’s very distinct.

I extracted all my data from Runkeeper and plot out my duration of the segment, and this is exciting. This is encouraging because now once I’ve taken my data and separated it out ino phases. The first again the grey phase is the baseline. The first green is the first phase of extreme low carb intake. The red is a period of more moderate carb intake; you’ll notice that’s where all the blue lines are. And then again I went to a very low carb intake and these patterns of dots look different.

Infact when I plot a trend line, these trend lines are very different and you’ll see in the very first period where I had very low carb intake my ride was almost 30 seconds longer for the climb.

So what’s also encouraging is in that second very low carb intake phase I improved a lot.

So what did I learn?

My carbohydrate really did influence my athletic performance, even just measuring it took me to ride my bike to work. But I also learned that it’s reversible. Once I increased my carb intake my performance was infact improved again.

It also doesn’t last forever. That second time I went into the very low carb phase I adapted very quickly. It also confirms that the fatigue was real and science works.

I will be participating in a breakout this afternoon and would love to see you guys.