REF-1167 · Lum et al., 2020 · Sports · Paper · PubMed · Free full text

Does a mid-thigh pull show strength in movement?

In 47 correlational studies, IMTP force tracked 1RM squat (r 0.71-0.97) and related to jumps and sprints, explaining 12-94% of variance. A proxy, not a 1RM; no endurance data.

B+moderate evidenceMeasured on recreational to elite athletes

Lum, Haff and Barbosa reviewed studies relating force-time measures from multi-joint isometric tests — isometric bench press (IBP), isometric squat (ISqT) and isometric mid-thigh pull (IMTP) — to dynamic performance. Of 105 records, 47 studies qualified: 4 on upper-body, 38 on lower-body and 8 on sport-specific movements.

Lower-body peak force correlated r 0.688-0.864 (ISqT) and 0.705-0.970 (IMTP) with 1RM squat. Shared variance with lower-body dynamic performance ranged widely (ISqT r² 0.085-0.746; IMTP r² 0.120-0.941), and for sport movements r² 0.118-0.700.

Rate of force development and impulse often related to sprint acceleration and jumps, but results varied with joint angles, test standardisation and athlete type. In track cycling, IMTP peak force correlated with Wingate peak power (r 0.74) and split times (r -0.49 to -0.55) in one study.

The authors conclude isometric tests give insight into force capability but list endurance running and swimming as untested, and did not pool data.

Included studies

Search: PubMed, SPORTDiscus and Google Scholar (PICO-based Boolean search) plus reference lists; dates of the search not stated in the full text read.
Inclusion: original peer-reviewed English studies of multi-joint isometric tests vs performance of dynamic movements, in athletes or recreationally active people; single-joint tests excluded.
Studies: 47; quality 13-14 of 27 on Downs and Black for all (correlational designs cannot meet intervention items).
Tests: maximal overcoming (pushing or pulling) isometric efforts against a fixed bar on force plates or dynamometers; outcomes peak force, force at fixed times, rate of force development and impulse.
Dynamic outcomes: 1RM squat, power clean, snatch, bench press; jumps (CMJ, squat jump, drop jump); sprints (5-30 m); change of direction; kayak, shot put, boxing punch force, track cycling, golf club speed.

Key Results

RelationshipRange reportedNotes
IBP force-time vs upper-body dynamic performancer² 0.221-0.6084 studies; IBP peak force vs 1RM bench r 0.570-0.78
ISqT force-time vs lower-body dynamic performancer² 0.085-0.746ISqT peak force vs 1RM squat r 0.688-0.864
IMTP force-time vs lower-body dynamic performancer² 0.120-0.941IMTP peak force vs 1RM squat r 0.705-0.970
Isometric tests vs sport-specific movementsr² 0.118-0.7008 studies
IMTP vs CMJ heightr 0.346-0.820similar for squat jump; varies with test angles
IMTP impulse at 300 ms vs 5-20 m sprint timer -0.74 to -0.78collegiate soccer and rugby players
IMTP peak force vs Wingate peak power / track cycling split timesr 0.740 / r -0.490 to -0.550one study (Stone 2004)
ISqT peak force predicting 1RM squat8.5% difference from actual 1RMone study (Blazevich); no other study has developed a prediction equation with a low standard error

Mechanisms

Authors' interpretation: isometric force-time measures index the force-producing capacity that underlies dynamic movements; the physiological reason they correlate despite different motor-unit recruitment remains unclear.
Authors' interpretation: correlations are strongest when the isometric test angle matches the angle where force peaks in the dynamic task, and with heavier dynamic loads (for example IBP vs 100% vs 60% 1RM bench: r 0.81 vs 0.69).
Inference: correlations across athletes do not show that raising isometric force in one athlete will raise dynamic performance; no training data are reviewed.

Practical Application

Does a mid-thigh pull show strength in movement? In 47 correlational studies, IMTP force tracked 1RM squat (r 0.71-0.97) and related to jumps and sprints, explaining 12-94% of variance. A proxy, not a 1RM; no endurance data.
Correlations across athletes do not show that raising isometric force will raise dynamic or race performance.
Standardise the test position and familiarise. Different knee and hip angles changed both force and its correlation with dynamic tasks; three familiarisation sessions were needed to stabilise the isometric squat in one study the authors cite.
For an endurance athlete, use it to see whether a strength block built force ([REF-1166] for its reliability), not to predict race performance. The only endurance-adjacent data were track-cycling sprint splits; that IMTP gains matter for long events is not shown here.
An isometric test cannot set training loads; keep a dynamic estimate (RM or velocity) for prescription, as the authors note.

Limitations

Correlational, cross-sectional studies only; no pooling, no heterogeneity analysis, no risk-of-bias beyond a generic checklist (all studies scored 13-14/27).
Wide ranges reflect different test angles, filtering and RFD methods, familiarisation and populations; the review does not resolve which matters most.
Few women (the one female sprint sample found no relationship) and no endurance running or swimming outcomes.
Search dates not reported in the text read; Google Scholar as a database limits reproducibility.
Funding: no external funding; no conflicts of interest declared.
How this entry was checked and changed
Why this level

systematic method, but the evidence is correlational, unpooled and drawn mostly from small samples; read from the full text

Full text (checked 2026-10-16)
A total of 47 studies were selected: 4 upper limb, 38 lower limb and 8 on sports specific movements (full text).
IBP r2 = 0.221 to 0.608; ISqT r2 = 0.085 to 0.746; IMTP r2 = 0.120 to 0.941; sporting movements r2 = 0.118 to 0.700 (full text).
1RM back squat vs peak force from ISqT (r = 0.688 to 0.864) or IMTP (r = 0.705 to 0.970) (full text).
Downs and Black Quality Assessment Checklist score ranging from 13 to 14 (full text).
IMTP peak force vs Wingate peak power (r = 0.740) and track cycling split times (r = −0.490 to −0.550) (full text).
Impulse at 300 ms vs 5 and 20-m sprint times (r = −0.740 to −0.780) (full text).
Blazevich: 8.5% difference to actual 1RM load when ISqT peak force was used to predict squat 1RM (full text).
Drake et al.: a minimum of three familiarization sessions required for the ISqT (full text).
"This research received no external funding"; no conflict of interest (full text).