Menu

HomeHow it worksContact UsContact Us

A Held Number Is Not a Held Rate: Testing Whether Reduced Training Really Maintains Adaptation

O'Leary R. · Preprints.org · 2026

This study examined whether traditional comparisons of muscle strength and size before and after reduced training actually demonstrate that adaptations are being maintained, rather than just slowly declining. The researchers found that most published studies don't take enough measurements or use the right statistical approach to truly distinguish between maintained adaptations and gradual losses, and they propose a more rigorous three-step testing procedure requiring multiple equally-spaced measurements per person.
Takeaway: When evaluating whether a reduced training dose maintains your muscle gains, look for studies with multiple measurement points spread over time, not just before-and-after snapshots.
Abstract (source)

Studies of reduced training often conclude that a lower dose maintains muscle strength and size because values after reduced training do not differ from values after resistance training. A maintained level does not show how fast the adaptation is being lost. This study proposed a three-step procedure for testing whether reduced training maintains strength and hypertrophy adaptations, applied to published group means from two resistance training trials and 8,000 simulated datasets. First, a shape check: under exponential detraining, each later equal-interval loss must be smaller. Second, if the detraining arm passes, a constraint linking its loss rate to that under reduced training. Third, a calibrated estimate of the maintenance dose, checked on a held-out training frequency.

Significance: was set at p ≤ 0.05. The detraining arm of older adults violated the shape check in the reported means, and the constraint then depended on an unmeasured floor. With training every other week, the calibration implied a 1.63% loss of quadriceps cross-sectional area against 5.90% reported (a

Result: known before analysis). Leg press strength fell 5 kg, then 10 kg, over successive six-week periods, whereas exponential detraining requires shrinking losses; published spreads cannot establish this for the population. Simulated false rejection of a correct model was 3% to 8%. A maintained level of strength or muscle size does not establish that the adaptation is being retained. Testing maintenance needs measurements at the start of reduced training and at two or more equally spaced later times, with each participant's changes reported.

Read the original →