Does an isometric hold before a jump help?
Over 25 studies it raised jump height slightly vs no primer (ES 0.35), like a heavy dynamic set; no clear effect on strength or power (few studies). Low certainty; untested for endurance.
He et al. pooled 25 studies (782 participants) in which an isometric contraction was used as a conditioning activity before strength, power or jump tests, with a three-level model to handle several time points per study.
Against control conditions, isometric conditioning raised vertical jump height (ES 0.349 after removing four influential effects; 0.406 before, 95% prediction interval crossing zero). Within the isometric groups jump height rose by a trivial ES 0.180. Against heavy dynamic conditioning (9 effects) there was no difference (ES -0.061).
Power output and muscular strength showed no significant effect.
Subgroups with larger effects were 8-12 s of total contraction per session, 6-10 "volume units" (1 unit = 1 s of maximal contraction), 4-8 min rest, multiple repetitions, knee angles of 90-140 degrees (shallower squats) and athletes with at least 2 years of strength training; only joint angle and volume reached significance as moderators, and the spline peaks were about 8.2 s of contraction, 7.4 units and 7.7 min of rest.
Both holding and pushing isometrics worked. Certainty was low.
Included studies
Search: PubMed, Web of Science, EBSCO and Scopus to 30 October 2024, English, peer-reviewed; healthy participants; acute isometric conditioning (holding or pushing) with a pre-test and a post-test within 60 min; outcomes strength (MVC, peak force, torque, 1RM), power, or jump height (CMJ, squat jump, drop jump). Pseudo-isometrics excluded.
Studies: 25 (22 from databases, 3 from other sources); 782 participants, mean age 23.5 ± 5.1 y (17-52); 19 strength-trained samples, 4 untrained, 2 mixed; 24 lower-limb, 3 upper-limb (bench press).
Kind of isometric: pushing (overcoming) in 18 studies, holding (yielding) in 6; isometric squat, mid-thigh pull, leg press, knee extension, plantar flexion, hip hinge; intensity 20-100% MVC (18 maximal, 8 submaximal); contraction duration 5-60 s per session (sustained duration x reps x sets); volume 5-60 units; knee angles 60-140 degrees.
Comparators: control (no conditioning), dynamic isotonic conditioning (for example heavy squats), or pre-post only.
Quality: modified PEDro 14-20 (mean 16.8); randomisation reported in all but detailed in 8; no study blinded testers; 22/25 had one or no familiarisation session; 12 reported an a priori sample size.
Key Results
| Comparison | ES (95% CI) | p | Heterogeneity / notes |
|---|---|---|---|
| Jump height, isometric vs control (full model) | 0.406 (0.152 to 0.660) | 0.002 | I² 60.65%; 95% PI -0.560 to 1.372 |
| Jump height, isometric vs control (4 influential effects removed) | 0.349 (0.128 to 0.569) | 0.003 | 95% PI -0.365 to 1.062 |
| Jump height, isometric vs dynamic | -0.061 (-0.352 to 0.230) | 0.642 | 9 effects; I² 0% |
| Jump height, pre vs post within isometric (revised) | 0.180 (0.078 to 0.282) | 0.0008 | I² 7.67% |
| Power output, within isometric | -0.130 (-0.777 to 0.517) | 0.684 | no between-group analysis possible |
| Strength, isometric vs control | 0.542 (-0.217 to 1.301) | 0.141 | |
| Strength, within isometric | 0.233 (-0.386 to 0.852) | 0.454 | |
| Moderator: knee angle 90-140 vs ≤ 90 degrees (vs control) | difference 0.500 (0.015 to 0.984) | F = 4.325, p = 0.043 | the only angle contrast that held |
| Spline peaks (within isometric, jump) | 8.2 s contraction (ES 0.30); 7.4 units (ES 0.28); 7.7 min rest (ES 0.23) | rest curve illustrative, sparse beyond 3-10 min | |
| GRADE certainty | low for all comparisons | Egger significant at level 2 for the within-group jump analysis |
Mechanisms
Authors' interpretation: classic potentiation (myosin light-chain phosphorylation) may contribute at short rests, with muscle temperature and fluid shifts later; none of these was measured in the included studies, and the authors call the mechanisms provisional.
Authors' interpretation: potentiation raises rate of force development and velocity more than maximal force, which could explain why jump height moved and maximal strength did not.
Authors' interpretation: isometric contractions cause more central and less peripheral fatigue than dynamic ones; central fatigue clears faster, so the effect may appear earlier than after heavy dynamic sets.
Authors' interpretation: deeper-knee holds (longer muscle length) cause more occlusion and fatigue, which may blunt the effect; speculative.
Inference: "8-12 s" is total contraction per session, so it means one or two short maximal holds, not one long hold.
Practical Application
Does an isometric hold before a jump help? Over 25 studies it raised jump height slightly vs no primer (ES 0.35), like a heavy dynamic set; no clear effect on strength or power (few studies). Low certainty; untested for endurance.
Best-looking signals: about 8-12 s of total contraction per session, 4-8 min of rest and shallow knee angles (90-140 degrees).
How to try it (base's practice built on the moderator signals): after a general warm-up, two to three maximal isometric pushes of about 3-5 s against an immovable bar at a shallow squat angle, then 4-8 min before the effort. Test it in training first; individual responses vary and the prediction interval includes no effect or harm.
It may suit stronger, strength-trained athletes more (subgroup effects tended to be larger with ≥ 2 years of lifting, but the difference between experience levels was not significant; male-dominated sample).
Limitations
Low GRADE certainty throughout; testers never blinded; familiarisation often minimal; randomisation details mostly missing.
Registration on PROSPERO came after searching and screening (before extraction), so it is a transparency measure rather than a fully prospective protocol (authors' own statement).
Strength, power and isometric-vs-dynamic comparisons rest on few studies and were sensitive to influential effects; the main jump estimate was retained after removing four influential effects.
Many moderator analyses with conceptually linked moderators and no multivariable model; submaximal intensities were almost all holding contractions prescribed from dynamic 1RM, so intensity and isometric type are confounded.
Male-dominated (601 of 782); the female subgroup effect was not significant and may reflect lower relative strength.
Funding: "The authors have nothing to report"; no conflicts of interest declared.
How this entry was checked and changed
Why this level
meta-analysis of small, mostly crossover acute studies; GRADE certainty low for every comparison, risk of bias high or unclear in several domains, moderators exploratory; read from the full text
Full text (checked 2026-10-16)
The 25 studies enrolled 782 participants (601 men, 181 women; mean age 23.5 ± 5.1 y, range 17–52); PIMA was used in 18 studies and HIMA in 6 (full text).
Intensities ranged 20%–100% MVC; contraction duration spanned 5–60 s; reported knee angle range 60°–140° (full text).
Isometric vs control, jump height: ES = 0.406, 95% CI [0.152, 0.660], 95% PI [−0.560, 1.372], I2 60.65% (full text).
Revised model: ES = 0.349, p = 0.003, 95% CI [0.128, 0.569], 95% PI [−0.365, 1.062] (full text).
Isometric vs dynamic: ES = −0.061, p = 0.642, 95% CI [−0.352, 0.230] (full text).
Within isometric, revised: ES = 0.180, p = 0.0008, 95% CI [0.078, 0.282] (full text).
Power output within isometric ES = −0.130, p = 0.684; strength vs control ES = 0.542, p = 0.141; strength within isometric ES = 0.233, p = 0.454 (full text).
Joint angle moderator F = 4.325, p = 0.043, difference = 0.500, 95% CI [0.015, 0.984] (full text).
Spline peaks: rest around 7.7 min (ES = 0.23); contraction duration around 8.2 s per session (ES = 0.30); volume around 7.4 units per session (ES = 0.28) (full text).
Modified PEDro scores ranged from 14 to 20, with a mean score of 16.8 (full text).
Funding: "The authors have nothing to report"; no conflicts of interest (full text).