Thursday, June 15, 2006

Rest Day Ponderings

Wednesdays are my usual rest days, a day off the bike. In some ways I 'resent' having to take a day off the bike, but I've come to realize that quite often I need to take a day off the bike. This week especially when I've been questioning how my legs have felt and whether I should ease up on my training by taking an interval day out of my week. My legs actually have felt pretty good today, so maybe I'll keep all my intervals in this week as I have a recovery week next week. Then my first century.

So while I was resting, at work no less :) , I did some pondering. Having come into cycling originally as a way to just get some exercise and lose weight, I often obsess about how many calories I'm consuming and burning. I've been using a piece of software called BeNutriFit to track this information and this year I've also started using an Excel spreadsheet so I can fine tune my diet plan.

In the spreadsheet I take the information from BeNutriFit and pair it up with my daily and 7-day moving average weight. I find using the 7-day moving average is better than actual daily weight as there can be some pretty wide swings (several pounds) because of hydration levels. For example, I almost always see a half pound or more increase in weight after a rest day due mainly to drinking a lot of water but not having the increased sweat from my workouts. So the average smooths things out and generally keeps me from freaking out when a day or two shows an uptick in weight.

This allows me to fine tune my calorie consumption based on what my weight is actually doing. So if I see that during a week I've eaten X calories and my weight has gone down one pound, I can estimate that I could have eaten an additional 500 calories per day (3500 calories per pound) if I wanted my weight to stay steady. This is not a perfect system as there is likely some error in my estimates of calories consumed and burned, but assuming the errors in the system are consistent, I can get pretty close to knowing how many calories I'm actually burning compared to my programs estimates.

So I have data for my calories balance going back nearly a year now (first entry in BeNutriFit is July 9th 2005), with my spreadsheet going back to the first of this year. I wondered if I could start to make any generalizations about my calories such as my basic calorie requirement and how much for each hour of training I do. Since I want to get values for both base daily calorie requirements and the requirments for training, I came up with the following formula for a weeks calories:
7x + Ty = C + (Wc)
Where x is the base daily calorie requirement, T is training hours, y is calories per hour of training, C is estimated calories consumed and Wc is the adjustment for change in weight. So if in a given week I trained for 10 hours, consumed 20,000 calories and my weight stayed the same, then we'd have the following formula:
7x + 10y = 20,000
This by itself doesn't really help much as the single formula doesn't allow me to isolate either x or y, but if I have multiple equations this can be done. So that is what I did. I compared overlapping pairs of weekly information and averaged out the results. What I came up with is fairly surprising in some ways.

Base calorie requirements: 2100 calories
Calories per hour of training: 1000 calories

The surprising part of this is the 1000 calories per training hour. BeNutriFit shows that cycling at a speed of 16-19 mph as burning 920 calories per hour, with slower speeds burning less than this. So my estimate seems a bit high, but as it also includes the extra activity level associated with getting the bike ready and so on it isn't that far off. If I plug these values in for my total training time for this year, I should get something pretty close to my estimated calorie burned total and if so, that would be a sort of verification. So lets see what I get:

Training Weeks: 23
Total training hours: 252.73

Using the values and formulas above we get:
((23*7)*2100) + (252.73*1000) = forecast calories burned = 338,100 + 252,730 = 590,830 forecast calories burned.

According to my data, my calories burned for this period, adjusted for the 12.2 lbs I lost during the period, was 592,312 or a variance of 1482 calories over 23 weeks, or 64 per week or 9 calories per day. I'd call that fairly accurate. I'll have to see how it holds up as I move foreward, and maybe I'll look back at my older data and see how it holds up there.

No comments: