Estimated Training Load: What the Numbers Mean

What "training load" means

Training load is a way of putting a number on how much work a session or week contained. Apps and watches show it under names such as training load, strain, training stress or effort score. The idea is simple: add up how long and how hard you exercised. The formulas, however, differ by brand and are often not published in full, so the same week of activity can produce different numbers on different platforms. This page is information only and does not set any training targets.

Two kinds of load

Devices measure the inputs (time, movement, heart rate) and then calculate a score from them. The final number is a proprietary or published formula output, not a direct measurement of fatigue, fitness or injury risk.

The best-studied low-tech method: session RPE

The session RPE method multiplies a rating of how hard the whole session felt by its duration. A 2001 study compared it with a heart-rate-based method during cycling and basketball and concluded it was a valid way to quantify training across many types of exercise. [1] A later review of 36 studies found good validity, reliability and internal consistency in people of different ages and levels, mostly athletes, and said some experts suggest combining it with heart rate. [2] Label: established for athletes as a way to describe how hard training was; it does not show that a given load is right for a person. See perceived exertion.

Acute:chronic workload ratio (ACWR)

Some products compare recent load ("acute," often a week) with longer-term load ("chronic," often a month). The ratio is popular in team sports and has been used to flag injury risk.

• A systematic review of 27 studies found high variability in how ratios were calculated and in the reference groups used, and concluded that issues with the method must be addressed before it is used with confidence to reduce injury risk. [3]
• A methods critique by Impellizzeri and colleagues argued that no study has properly estimated causal effects and concluded that there is no evidence supporting use of ACWR in load-management systems or training recommendations to reduce injury risk. [4]

Label for ACWR as an injury predictor: insufficient evidence and, as a basis for training advice, no good evidence of benefit in the sources we read.

Consumer scores and wearables

We did not read manufacturers' method papers and we found no independent validation study of any specific proprietary "strain" or "training stress" score in our search of the research database. Proper evidence not found for these named scores. What we can say is that consumer wearables vary in accuracy: a living umbrella review of 24 systematic reviews found that only about 11% of released devices had been validated for at least one outcome, and that physical-activity-intensity measurements had mean absolute errors of 29% to 80% depending on intensity. [5] Inputs that are off produce a load number that is off. See how to read a wearable validation study.

Accuracy is not benefit

A load number may track what you did reasonably well and still not prove that using it makes you healthier, fitter or less injured. We found no trial in our search showing that following a consumer load score improves those outcomes.