Running Cold - Does it Burn More Calories?

From Personal Science Wiki
Jump to navigation Jump to search
Project Infobox Question-icon.png
Self researcher(s) Nick Alexander
Related tools Runkeeper, phone
Related topics Diet and weight loss, Sports and fitness, Activity tracking, Food tracking, Weather

Builds on project(s)
Has inspired Projects (0)
Show and Tell Talk Infobox
Featured image Running-cold-does-it-burn-more-calories.jpg
Date 2013/10/11
Event name 2013 QS Global Conference
Slides Running-cold-does-it-burn-more-calories.pdf
UI icon information.png This content was automatically imported. See here how to improve it if any information is missing or out outdated.

Running Cold - Does it Burn More Calories? is a Show & Tell talk by Nick Alexander that has been imported from the Quantified Self Show & Tell library.The talk was given on 2013/10/11 and is about Diet and weight loss, Sports and fitness, Activity tracking, Food tracking, and Weather.

Description[edit | edit source]

A description of this project as introduced by Quantified Self follows:

Nick Alexander wanted to find out if running in colder weather would make him burn more calories. So he wanted to try and keep everything the same as much as he could, including his route, the duration it took him to run, and the clothes he wore throughout the year. What he found out was that the more fit one becomes, he/she burns fewer calories doing the same activity which could also be a compounding. In this talk, Nick explains his experimental setup and what he found after tracking over 30 runs and crunching the numbers.

Video and transcript[edit | edit source]

A transcript of this talk is below:

Nick Alexander - Running Cold - Does it Burn More Calories

So hi I’m Nick Alexander and I wanted to find out if running in colder weather would make me burn more calories. So I wanted to try and keep everything the same as much as I could, including my route, the duration it took me to run, the clothes I wore throughout the year. So I live just outside of Chicago where winter is more than just a frame of mind, and there was snow on the ground. It was 35 degrees and I’m wearing short sleeves and this very quickly seems like a terrible idea. It’s been about a year and a half since I last run seriously or exercise seriously and it takes me 11 minutes, and I’m going as fast as I can because it is so cold. I get back and as soon as I look at the data though it kind of had a motivating effect for me. So I wear a body mediafit armband, which tracks my rate of calorie burn, so that was kind of the main variable I want to track. I also took a look at the environmental temperature as well as my duration for the run because I thought that could be a compounding factor as far as calorie burn was concerned. So I did track a couple of other things like community and wind speed and that was mostly again compounding factors to make sure I kind of knew with certainty what was having an effect. So why on earth would I want to do this? There’s a guy out there named Ray Farnese, he’s a former NASA scientist and did some experiments and found you burn a lot more calories when you exercise in the cold, so I kind of wanted to test that claim to see if it would work for me. And then also the other kind of motivation and this is probably more why I wanted to do this was I need an excuse to exercise. I needed to get more active and doing an experiment would kind of motivate me in a way that the exercise itself would not. So this is kind of very low barrier of entry. It takes me 15 minutes to do any individual trial, so it’s not a big time commitment. And you know I had the calorie tracker and I also wanted to know how to do experiments. I wanted to become more active as far as QS experiments and I thought I could learn some things that I could use later for additional experiments in the future. So one of the early issues that I faced was just time issues. So I guess you know 5:30 PM on my IPhone, the same as 5:30 PM on my body media armband, so how do I reconcile that. So I wrote to body media to ask what would they typically do, and they gave me some really good pointers. One was don’t use your first or last data points. So if you want to use seven minutes of data, record nine and throw out the first and last to kind of avoid start and end effects. And then also start and end trials with two minutes of no steps. So that does then show you very clearly the boundaries of the test in your data, even if it’s shifted a little bit over based on where the time stamps are. So those are kind of two things that really kind of helped me to more effectively track my data. So as far as things I want to keep constant was the route and therefor distance and elevation, my clothing, as much as possible my run speed. As far as variables, track a whole bunch of stuff but the main thing was calorie burn, environmental temperature, and to a certain extent run duration. So here you can see this is the run duration in blue for each of my trials. Orange is the trend line and you can see I stabilized after nine and a half minutes after I got more fit, and in red here we have the outdoor temperature so it was about 35 degrees. Went up to 91 and started to come back down, so I got kind of both sides of the trend a little bit on the backside, so that was good and kind of helped me to tease out some of the correlations. So in terms of talking about tools, when I first started I used the stopwatch app on my IPhone to, literally just track my time it took me to get to each of the intersections along my route to see if I was on track. That was before I found out there was way better ways of doing that with running apps, so I did that latter. Bought a media armband for calories, and then for weather I would when I literally got back I was just Google search what’s the temperature right now. But I found Google was kind of drawing it from three different sources and I didn’t necessarily know which one it was giving me at any given time. So I switched over to Wunderground, which had all sorts of historical weather information and it was also from KARR airport which is near where I live. So as far as running app, I tried MapMyRun; really didn’t like the interface. Tried Run Keeper, which I loved the interface but it didn’t work with the heartrate monitor that I started using later on. Eventually wound up with DigiFit, which gives me really good granular data downloads, and allowed me to separately track my cooldown period as a kind of a separate factor so I could try and analyze later and worked with my heartrate monitor. So here’s some actual data from my run. So these are all my runs superimposed on each other. On the X we have how many minutes we are into the run, and on the Y the number of calorie burned in a minute. Red is hotter days, blue is colder days. So you can see a little bit how it lines up. And there’s kind of two major sections. One is the cooldown period which I haven’t really analyzed a whole lot yet. The main thing I was interested was my seven minute calorie burn, which started at the end of my first minute up to minute seven. So that’s where I kind of did a lot of my analysis in that kind of section. So my seven minute calorie burn here’s in green and you can kind of see how it fluctuates. And the other line is the temperature, so you can expect to see there are a little bit of what looked like negative correlations in the data which was kind of encouraging. Although the other thing I should point out is that the range as far as the green calorie burn is only about eight calories on this scale, so the effect doesn’t seem to be very much even if it is correlated. So I did pop this into a statistical tool called Stat Wing which helped me to find some correlations. So I found that temperature is negatively correlated with my seven minute calorie burn. While at the same time run duration has no significant relationship with seven minute calorie burn. The same thing was true for peak calories which are the kind of measure of the highest minute of calorie burn on an individual run. So it does appear that tentatively my hypothesis looks like it’s going to be correct, you know and I was able to distinguish it from run duration. You know all that said the effect size is pretty small and it maybe five calories difference per mile run which is not very much. So more importantly what I’ve kind of learned was something about my motivations which is there is a period which is a little bit warmer and I got cold again, and I realized I don’t want to run today it’s so cold. But then I realized there’s not that many cold days left in the year. I need that data point; cold days are premium. I have to run today. So it kind of motivated me in a way that running never will so it was kind of one of the interesting things that I learned. So another one was to kind of refine my experimental process; I kind of threw out the first six or seven data points as I was kind of still figuring out the best ways to track things. So one of the improvements I’d like to do is you know I found out the more fit you become you burn fewer calories doing the same activity which could also be a compounding factor. Although now that the temperature has started to go down a little bit that could decouple, so I’m going to continue to collect data and make sure that the correlation stays there. The other thing was I’d love to have longer runs. When you have nine minutes, you know one minute of difference is 10%. You know so it would be nice to have longer runs. At the beginning it was cold and I just wanted to get through it.

Yeah, it’s an ongoing experiment so this is how to reach me if you have any questions or comments in the future.

About the presenter[edit | edit source]

Nick Alexander gave this talk.