Skip navigation

Calculating training load using your heart rate monitor: Adjusted TRIMP

Tue, Jun  9, 2009 - By Mike Muha

When The Master Skier magazine printed "Training with TRIMP" by Dan Heil (Mid Season 2008-2009 issue, pp 33,40), I was pumped. Heil introduced a way to calculate Training Load using a method called TRIMP or TRaining IMPulse, and I'd been looking for just such a method.

Training Load is defined as Intensity x Duration. The harder and/or longer you workout, the higher the training load.  Training load can also be the cumulative load if measured over time. Add the training load of each day over a week, and you have the training load for the week.

Training load is different than Training Volume, the number of hours you train. Two athletes can have the same volume of training (e.g., 300 hours per year), but have a very different cumulative training load.

Why track training load?

The first reason to track training load is to make sure training load increases over time. Getting fitter as a cross country skier (or runner, or biker, or whatever) requires progressive increases in training load. The increase stresses the cells, tissues and organs and forces them to adapt to the higher levels of stress. During recovery between sessions, the actual training effect occurs: the body recovers from the breakdown by rebuilding a stronger infrastructure.

A second reason is to ensure you don't over-train. All good athletes include rest as an active aspect of their training. The traditional 4-week training cycle (three week of progressively harder training followed by an easier "recovery" week) is an example of varying training load over time. My own experiences with coach by Torbjorn Karlsen plus books like "Cross Country Skiing: Olympic Handbook of Sports Medicine" argue that you should have training blocks of high load followed by an easier week of lower load to recover. And more recently, Bryan Cook of CXC Skiing in a video for CXC Academy suggests that alternating high training load weeks / low training load weeks is a good model to follow for master skiers.

Unless you have a coach who intuitively can plan your training, how can you tell if you training loads are progressively increasing or if you're training too much?

TRIMP or TRaining IMPluse

Unfortunately, training load has not been that easy to measure until the TRIMP model was introduced by Bannister et al. in an article entitled "A systems model of training for athletic performance" (Australian Journal of Sports Medicine 7:57-61, 1975).  Essentially, Bannister says to multiply your average heart rate during a training session by the length of the training session (in minutes).

The problem with TRIMP is that scores can be similar for very different workouts. I took two training sessions, a 58 minute bike commute at an easy pace and an hour-long running interval session, and computed TRIMP. As you can see in the table, the scores ARE very similar even though the interval session was much harder:

The Original TRIMP Model: Not the best...

Session Minutes Average

April 24, 2009 (Zone 2 distance session) 58 121 7,018
May 23, 2009 (4x4:00+ interval session) 58 129 7,740

Diff     722
Diff %     10%


Adjusted TRIMP

To fix this issue, Foster et al. in "A new approach to monitoring exercise training" (Journal of Strength and Conditioning, 2001, 15(1), 109-115) divided a workout into 5 training zones based on percentages of peak heart rate: 50-60% (zone 1),60-70% (zone 2); 70-80% (zone 3); 80-90% (zone 4); and 90-100% (zone 5). After the exercise session, the heart rate monitor was download and the cumulative time in each zone is calculated. The time in each zone is them multiplied by that value for that zone and the results summated.

I substituted the training zones determined by my last VO2 max test done by CXC's Yuriy Gusev at the NordicSkiRacer/CXC clinic last fall, then computed an Adjusted TRIMP with data from my Polar heart rate monitor. Here's the adjusted computation for the May 23 Interval Session:

Calculating Adjusted TRIMP
(May 23, 2009 Interval Session)


 Time in Zone


1 x 22 = 22
2 x 20 = 40
3 x 6 = 18
4 x 12 = 48
5 x 0 = 0
Sum of Loads = 128
Training Load = 128

The results show a much larger differences between the two workouts in terms of training load:

Adjust TRIMP based on training zones: Much better!

Session Minutes Zoned

April 24, 2009 (Zone 2 distance session) 58 70
May 23, 2009 (4x4:00+ interval session) 58 128

Diff   58
Diff %   82.9%

I've modified my training log to compute the Adjusted TRIMP (which I simply call "Training Load") for every workout for which I have heart rate data. Now I can track the training load of each session and, if you look at the bottom of each training log week, you'll see the weekly training load.

Of course, you can do this yourself with a spreadsheet...

Problems with Adjusted TRIMP

The Adjusted TRIMP does not actually compute the total training load over time for a few reasons. First, strength workouts are not covered. Sure, you could wear a heart rate monitor while pumping iron, but you'll be surprised how low your heart rate averages. Certainly the strength session has a much higher training load than heart rate alone would indicate.

Second, for some workouts, you simply don't have a heart rate monitor available. You're swimming, or the battery is dead, or you left the monitor at home - there will be workouts you do without the monitor.

Third, the activity makes a difference. I think the training load of a two-hour run may be higher than a two-hour rollerski, which is higher than a two-hour bike ride at the same intensity. Running is very hard on the legs - a two hour run takes my legs some time to recover. Rollerskiing uses both the upper body and lower body - that should produce more load than biking which only activates the legs, with the upper body supported by the bike seat. Maybe I'm wrong, but on the face of it, different training activities would seem to be easier or harder at the same intensity.

There is a solution for these problems: Calculating training load based on the perceived exertion of the training session, also call the "RPE Method." Unfortunately, that's another article. If you're in a hurry, read Foster's A New Approach to Monitoring Exercise Training for everything you need to know.

For now, I recognize that Adjusted TRIMP is not a perfect method for determining training load, but it's much better that what I had before, and I plan to use it going forward.


For an excellent and very detailed instruction of the principles of training, see Chapter 3 "Training for Cross Country Skiing" in the IOC Medical Commission Publication "Cross Country Skiing: Olympic Handbook of Sports Medicine," edited by Heikki Rusko. Unfortunately, it's a pricey book. You can read a subsection of the chapter on Google Books for free.

Banister EW, Calvert TW, Savage MV, Bach A. A system model of training for athletic performance. Australian Journal of Sports Medicine. 1975;7:170–176. Unfortunately, I have not been able to find a copy online.

Foster, Carl, Florhaug, Jessica A., Franklin, Jodi, Gottschall, Lori, Hrovatin, Lauri A., Parjker, Suzanne, Doleshal, Pamela, and Doge, Christopher. A New Approach to Monitoring Exercise Training. Journal of Strength and Conditioning Research, 2001, 15(1), 109-115.