Yvonne:
Thanks for the good question. CTL (Chronic Training Load) is a proxy for fitness developed by Training Peaks. CTL is a nonlinear, weighted, running average of the daily TSS (Training Stress Score) for the past 42 days (6 weeks).
TSS can be determined in a number of ways. In our mountain sports we are limited to using the heart rate calculation of TSS (hrTSS) as opposed to running TSS (rTSS) which is pace based and only works for runners in mostly flat ground or power TSS developed for cyclists using power meters. hrTSS is by far the least accurate of these 3 measures of training load because heart rate is not a good predictor of training load. HR response to exercise intensity is non linear and very individual. But we’re stuck with it as the best tool for our sports.
Some things to keep in mind about CTL then:
1) It is an imperfect metric.
2) It is the best one we have ever found for monitoring training.
3) It relies on the hrTSS
4) hrTSS relies on an accurate measure of your Anaerobic Threshold (not aerobic threshold)
5) You must accurately define your HR zones in the TP zone calculator
6) Failure to do either step 4 or 5 and the hrTSS will become even less accurate.
7) When hrTSS is not accurate CTL will no longer be meaningful
So, Yvonne? Have you converted all your aerobic workouts to hrTSS? Are you assigning adequate TSS to the strength workouts? HR is useless for calculating training stress from strength training so you have use other methods which I have discussed at length in he forum. Look under Training Peaks. Have you (re) established your anaerobic threshold? If not you should do that. The test is explained in detail here; https://uphillathlete.com/diy-anaerobic-test/
Then have you set all your zones correctly?
Do all these things and your CTL should be more reflective of your fitness. Even if it is not bear in mind the point #1 I made above. CTL is an imperfect metric. If you are familiar with statistics think if it like correlation (r*2). I’d give CTL an r*2=.8-.9 or about 80-90% of the observed effect can be attributed to the cause (a rough definition of “correlation”).
Scott