REF-1164 · Schaefer & Bittmann, 2017 · BMC Sports Science, Medicine and Rehabilitation · Paper · PubMed · Free full text
Is holding a weight harder than pushing a wall?
At 80% MVC in 10 adults, a held elbow position stayed still 19 s vs 41 s when pushing; total time at that force differed less (50 vs 59 s). One small arm study.
Schaefer and Bittmann asked whether "holding" an isometric position against an external force (holding isometric muscle action, HIMA, yielding) differs from "pushing" against a stable resistance (pushing isometric muscle action, PIMA, overcoming).
Ten healthy sports students performed both with the elbow extensors at 80% of maximal voluntary contraction (MVC), using a pneumatic cylinder that could either push against the arm or be pushed by it.
In endurance trials held until the position could not be kept, the longest phase without any yielding of the rod lasted 19.1 s in holding vs 41.4 s in pushing (p = .005), and the whole isometric plateau 50.0 s vs 58.8 s (p = .029). The longest stable phase made up 31.6% vs 59.4% of the plateau. Mechanomyography differed only in a few measures.
The authors conclude there are probably two forms of isometric action and that muscles exhaust earlier when holding, perhaps through metabolic fatigue or more complex neural control.
Protocol
Participants: 10 healthy adults (5 women: 24.4 y, 59.9 kg, MVC 18.25 Nm; 5 men: 24 y, 71.8 kg, MVC 31.166 Nm); all right-handed; university sports students plus one high-school student.
Set-up: seated, elbow at 90 degrees, dominant triceps; interface at the wrist connected to a pneumatic cylinder set to 80% of the MVC pressure with the rod half extended.
Pushing (PIMA, overcoming): push the rod to 80% MVC and keep it in position (the rod yields if overcome, extends if the subject gives way).
Holding (HIMA, yielding): the rod pushes toward elbow flexion; the subject holds 90 degrees against it. Pushing the rod back in counted as a failed trial.
Trials: 2 MVCs; 3 x 15-s trials of each action at 80% MVC (randomised start, blocked, 60 s rest); then endurance trials held as long as possible, alternating actions after a randomised start, 120 s rest (two per action).
Stop rule: rod deviation of more than 3 cm (about 1.27 degrees) from the middle position.
Measures: pressure, force, acceleration, mechanomyography of triceps and external oblique, mechanotendography of the triceps tendon; paired t-tests or Wilcoxon, no correction for multiple tests.
Key Results
| Endurance measure at 80% MVC (n = 10) | Pushing (PIMA) | Holding (HIMA) | p |
|---|---|---|---|
| Total isometric phase | 58.8 ± 15.5 s | 50.0 ± 10.3 s | .029 |
| Mean of single isometric phases | 24.5 ± 25.1 s | 7.7 ± 3.3 s | .049 |
| Longest single isometric phase | 41.4 ± 24.9 s | 19.1 ± 7.9 s | .005 |
| Longest phase as share of total | 59.4% | 31.6% | .000 |
| Other measure | Result |
|---|---|
| Force in 15-s trials | holding 99.31% of pushing (p = .412) |
| Force in endurance trials | holding 96.35% of pushing (p = .013) despite equal pressure |
| MMG triceps amplitude, last 10% of trial (n = 8) | higher in pushing (p = .012) |
| MTG triceps tendon power, 10-29 Hz / 8-15 Hz | higher in holding (p = .048 / p = .037) |
| MMG / MTG mean frequency | no difference |
Mechanisms
Measured: holding broke into shorter still phases (mean 7.7 vs 24.5 s). Authors' observation (subjective, not quantified): the breaks were sudden short yields of the rod, after which the position was regained at a slightly different elbow angle.
Authors' interpretation: two candidate causes — earlier metabolic fatigue (compressed vessels, slower removal of metabolites when the muscle oscillates less) and more complex neural control, holding resembling eccentric control and pushing resembling concentric. Neither lactate, oxygenation nor neural control was measured; the authors call the metabolic idea speculative.
Measured, against the hypothesis being clean: during endurance trials the force while holding was about 4% lower than while pushing, which should have made holding easier, so the authors argue it strengthens their conclusion.
Practical Application
Is holding a weight harder than pushing a wall? At 80% MVC in 10 adults, a held elbow position stayed still 19 s vs 41 s when pushing; total time at that force differed less (50 vs 59 s). One small arm study.
Expect shorter usable holds when the load is trying to move the joint than when pushing against something fixed.
When prescribing yielding holds (a wall sit, a held split squat with load), set shorter holds or lower %MVC than for pushing against a fixed bar. The magnitude is from one small arm study; that it applies to leg holds is the base's inference.
Holding and pushing are not interchangeable in fatigue and may not be in training effect either (REF-1163).
Limitations
Ten people, one session, one muscle (elbow extensors), one intensity (80% MVC); earlier work found differences mainly at low intensities, so generalisation across intensities is uncertain.
The headline 19 vs 41 s is the longest phase without any rod movement, not total time to task failure; the total plateau differed by about 15% (50.0 vs 58.8 s).
Custom pneumatic device; friction caused stick-slip, and force (not pressure) was lower in holding during endurance trials. Mechanotendography is not an established method (the authors say so).
MVC may be underestimated (no back or foot fixation). Two participants' pushing trials were stopped before exhaustion, so the end-of-trial MMG analysis used n = 8.
No correction for multiple comparisons, by the authors' choice.
Funding: none; no competing interests.
How this entry was checked and changed
Why this level
within-subject comparison, n=10; one muscle group, no correction for multiple tests (deliberately), outcomes from a new device and new metrics, no training outcome; read from the full text
Full text (checked 2026-10-16)
Ten healthy subjects (m = 5, f = 5); female MVC 18.25 Nm (±1.96), male MVC 31.166 Nm (±9.78) (full text).
Pneumatic system adjusted to 80% of the MVC; abort criterion a deviation of the push rod of more than 3 cm, representing 1.27° (full text).
Longest single isometric phase PIMA 41.384 vs HIMA 19.082 s; total isometric phase 58.848 vs 49.996 s; p .029, .049, .005, .000 (full text).
Longest single isometric phase approximately 41.4 s (59.4% of total) in PIMA vs 19.1 s (31.6%) in HIMA (full text).
Force during HIMA endurance trials 96.35% of PIMA (t(9) = 3,087; p = .013); 15-s trials 99.31% (p = .412) (full text).
MMG triceps amplitude at exhaustion higher during PIMA (n = 8; W = −2.521, p = .012) (full text).
MTG power 10 to 29 Hz higher in HIMA (t(9) = −2.284, p = .048); 8 to 15 Hz p = .037 (full text).
"No funding was used for this study or the production of this article"; no competing interests (full text).