REF-1481 · Kassiano et al., 2026 · Medicine & Science in Sports & Exercise · Paper · PubMed

Does a fixed gym load stop working?

In untrained women's arms over 8 weeks, adding load at 12 reps grew the triceps about twice as much as holding load and reps (0.79 vs 0.41 cm); holding still grew. Triceps only; abstract only.

Main caveat: Abstract only. Step size, effort, adherence, blinding of the sonographer, how a woman's two arms were allocated, and the statistical model are unknown.
moderate evidenceMeasured on non-athletes and recreational people

Kassiano and colleagues tested progressive overload directly. Fifty-five untrained young women (n = 55) took part, and each participant's arm was randomly assigned to one of three conditions: progressive overload (PO, n = 37), non-progressive overload (N-PO, n = 32) or a time-matched non-exercise control (n = 41).

The training conditions did unilateral elbow extension for three sets of 8-12 reps, 3 days a week for 8 weeks. PO "increased the load whenever it reached the upper limit of the repetition range", while N-PO "remained training with the same load and number of repetitions throughout the 8 wk".

Triceps thickness by ultrasound at 60% and 70% of upper-arm length rose more in PO than in N-PO, and more in both than in control: summed sites PO 0.79 cm (+22.9%), N-PO 0.41 cm (+11.6%), control 0.05 cm (+1.8%) (P < 0.001).

The authors conclude growth is more pronounced with progression and that training without it still suffices over this period in this group, and call the longer term an open question. Our reading: this is the only one of these ten entries that holds load and reps flat, the question the guide faces in a stalled log.

It is the base's real-world example of a "top of the range, then add" rule, but the abstract does not say how much load was added.

Protocol

Design: participant arms randomly assigned to one of three conditions. Within-subject or not: the arm was the unit of allocation, so a participant contributes two arms, and 37 + 32 + 41 = 110 arms = 2 × 55 (our arithmetic). The abstract does not state how a woman's two arms were paired, whether they could share a condition, or how the analysis handled this.
Who: n = 55 untrained young women.
Training: unilateral elbow extension, 3 sets of 8-12 reps, 3 d·wk⁻¹, 8 weeks.
Increment rule, exactly as given: PO raised the load whenever it reached the upper limit of the rep range (that is 12). N-PO kept the same load and number of reps for all 8 weeks. The size of the load step and any effort target are not in the abstract.
Outcomes: triceps brachii thickness by ultrasonography at 60% (TB60) and 70% (TB70) of upper-arm length and their sum (ΣTB). No strength outcome is reported in the abstract.
Funding and interests: not stated in the abstract record.

Key Results

SitePON-POControl
TB600.40 cm, +21.4%0.22 cm, +11.3%0.02 cm, +1.3%
TB700.38 cm, +25.0%0.19 cm, +12.0%0.03 cm, +2.5%
Sum of sites0.79 cm, +22.9%0.41 cm, +11.6%0.05 cm, +1.8%
Order at every site: PO > N-PO > control, P < 0.001 as given; no pairwise interval is in the abstract.
Ratio: PO's summed gain is about 1.9 times N-PO's (our arithmetic).

Mechanisms

Measured: muscle thickness; no effort, volume load, or neural data are in the abstract.
Authors' interpretation: progression makes growth more pronounced; holding the stimulus still suffices in young untrained women over 8 weeks; whether it stays effective for longer is open.
Inference: holding load and reps while strength grows means sets end farther from failure over time, so the two arms may differ in effort as well as load. REF-1483's preprint cites the same concern, that reps in reserve grow when load is not adjusted. The abstract doesn't report effort, so progression and proximity to failure can't be separated here.

Practical Application

Does a fixed gym load stop working? In untrained women's arms over 8 weeks, adding load at 12 reps grew the triceps about twice as much as holding load and reps (0.79 vs 0.41 cm); holding still grew. Triceps only; abstract only.
A rule that has been tried: set a rep range, add load when the top of the range is reached. The trial supports moving over not moving. It does not test this rule against any other rule, and the step size isn't in the abstract.
A flat lift still does something, at least early: the N-PO arm grew well beyond control. So holding on a busy block isn't wasted work, in this group over 8 weeks; do not extend to trained people or months.
Keep it in proportion: the outcome is muscle size in a small arm muscle. Not from this paper, and a suggestion for the new guide rather than a finding: for an endurance athlete, first ask whether a flat lift needs to progress at all this block.
The base already uses this paper: PHYS-39 cites it by DOI; this entry gives it a refId, and its wording ("about twice as much") holds.

Limitations

Abstract only. Step size, effort, adherence, blinding of the sonographer, how a woman's two arms were allocated, and the statistical model are unknown. If every woman had two different conditions, the counts imply only 14 women carried PO and N-PO together (our arithmetic, on an assumption the abstract does not state), so the direct within-person comparison could be smaller than 55 suggests.
No interval. Percentages and P values only; no confidence interval for the PO minus N-PO difference.
One outcome. Muscle thickness of one small muscle, at two sites; no strength or function in the abstract.
Who and how long. Untrained young women, one single-joint exercise, 8 weeks. The authors flag the long term as open.
One research group. Replication elsewhere is not in the abstract.
How this entry was checked and changed
Why this level

the rubric would allow A- for a randomised trial of ≥40; held at B+ because only the abstract was read, the allocation of a woman's two arms and the analysis are not stated, and the outcome is the thickness of one small muscle, measured by ultrasound, not performance