There Are No Nonresponders to Resistance-Type Exercise Training in Older Men and Women

110 at 12 weeks; 85 at 24 weeks

Men and women over 65, pooled from prior supervised resistance-training trials in the Netherlands. Not children; the age gap to this database's 4-18 range is total.

Retrospective pooled analysis of pre-post change in earlier training studies. NO non-exercise control arm and no randomization, which puts it in grade D by this database's hierarchy (pre-post without control). The decisive weakness is the responder CRITERION: a participant counts as a responder if they improved on AT LEAST ONE of four outcomes (lean mass, fibre size, 1-RM strength, chair-rise time). With four noisy measures and no control arm, "improved on at least one" is close to arithmetically guaranteed — so "there are no nonresponders" is substantially a property of the classification rule, not a discovery about physiology. The paper is included because it is the most-cited statement of the no-nonresponders position, not because the inference is sound.

Every participant showed a positive adaptive response on at least one training outcome; the authors conclude there are no nonresponders to resistance training in older adults. Group means all moved: lean body mass +0.9 +/- 0.1 kg at 12 wk and +1.1 +/- 0.2 kg at 24 wk; leg-press 1-RM +33 +/- 2 kg at 12 wk and +50 +/- 3 kg at 24 wk; chair-rise time -1.3 s at 12 wk and -2.3 s at 24 wk. But the RANGES are the real content and they are enormous and two-tailed: lean mass -3.3 to +5.4 kg, leg press -36 to +87 kg, type I fibre size -4458 to +3386 um2, chair-rise +21.6 to -12.5 s. Substantial numbers of participants got WORSE on individual outcomes. Responsiveness rose with intervention duration.

Neutral. Without a control arm the negative tail cannot be separated into true non-response, measurement error, and normal ageing — which is precisely why "no nonresponders" and "large true response heterogeneity" are both unsupported by this design.