IPC Recruitment & Landing Patterns
When one side of the pelvis is caught in a phase of gait, what does that leg do when it is asked to recruit a muscle, to produce force, to land, and to hold a line? Each participant is their own control: the leg locked in an intrinsic pelvic condition (IPC B, C or D) is compared with the functional leg of the same person, across quadriceps strength, the interpolated twitch technique, peak ground reaction force, and three-dimensional landing kinematics.
Four intrinsic pelvic conditions, read as phases of gait
On passive assessment each side of the pelvis is graded by the phase of gait it can or cannot leave. A person can carry different grades on the two sides; here the graded side is compared with the other side of the same person.
Does the caught leg lose voluntary quadriceps force?
Maximal voluntary contraction (MVC) of the quadriceps on each side. Every IPC B participant is weaker on the impaired side (8–27%). IPC C splits both ways — two weaker, two stronger, one equal — so on this evidence the heel-off lock is not a recruitment story.
Table view
Asymmetry is the difference as a percentage of the stronger limb. Whether that deficit is a recruitment problem or a muscle problem is the next question.
Is the deficit in the drive to the muscle, or in the muscle itself?
The interpolated twitch technique separates the two. Each quadriceps was stimulated electrically to find the current that produces a maximal twitch and the current that evokes a twitch equal to 5% of that limb's MVC. If the impaired muscle responds to stimulation as well as the functional one but produces less voluntarily, the gap is recruitment — inhibition — not contractile capacity.
Table view
Left panel: the voluntary deficit from Question 1. Right panel: force evoked per mA at the 5% target (5%-of-MVC twitch ÷ current), impaired relative to functional. In two of the three IPC B participants (B01, B02) the impaired quadriceps produced less on voluntary effort yet responded to stimulation better than the functional side — the dissociation that inhibition predicts; B03 is weaker on both, which reads as a muscle-side rather than a drive-side deficit. In IPC C one participant (C04) shows the same dissociation; in the other four the two panels move together or the impaired side is the stronger on both, so no group-level recruitment deficit is visible. Six of eight impaired limbs evoked more force per mA than their functional partner; the maximal-twitch current showed no pattern (higher on the impaired side in two, lower in three, equal in three). Evoked force per mA is a proxy, sensitive to electrode placement and skin impedance, and C03's 5%-twitch currents were recorded as 7.5 and 8.5 and are read here as 75 and 85 mA.
Which leg takes the peak force in a countermovement jump — and does a rigid pole change it?
Peak vertical ground reaction force per limb, jumping freely and then with a rigid pole across the shoulders. Asymmetry is largest in IPC B, moves with the pole in IPC C, and is absent in IPC D.
Table view
Mean peak vertical GRF per limb (N) across the trials retained in the original analysis; asymmetry = (functional − impaired) ÷ larger. Without the pole the heel-off (long) leg takes the greater force in three of six IPC C participants; with the pole four of six shift load to the functional leg. The heel-strike pair land within 3% of symmetric either way.
Where does the caught leg finish the landing?
Hip adduction angle at the end of the landing. In every IPC C participant the impaired limb finishes adducted (+8° to +17°) while the functional limb finishes neutral or abducted. Sagittal ranges are near mirror-symmetric within each person, so it is the frontal plane that separates the two legs.
Table view
B02 is the one IPC B participant whose functional limb finishes more adducted than the impaired one — also the only participant who lands harder on the impaired leg without the pole, and a bilateral supinator on screening.
Which condition lands stiff and upright?
Hip-flexion range during the landing, averaged across both limbs, for all 25 kinematic records. A, B, C and E sit around 55–60°; two of the three IPC D records land with barely 30°.
Table view
Range = end − start of hip flexion during the landing. IPC A here is the original analysis block, which includes B03 and C05 before their reclassification.
Three locks, three different signatures
Impaired side weaker by 8–27%.
B01, B02 weaker on effort but not on stimulation; B03 weaker on both.
~26% asymmetry; one loads the functional leg, two the impaired one.
Impaired limb more adducted in two; one reversed.
Normal ranges; LLD in one of three (8 mm).
2 weaker, 2 stronger, 1 equal on the impaired side.
Only C04 shows the effort–stimulation dissociation; the rest move together.
Long leg loads more in 3 of 6 without the pole; 4 of 6 shift to the functional leg with it.
Impaired limb lands adducted (+8 to +17°); functional neutral or abducted.
6–11 mm discrepancy; sagittal ranges symmetric.
—
—
Within 3% with or without the pole.
Kinematic records not linked to individuals.
~30° hip flexion against ~55–60° elsewhere.
How to read this — honestly
Small groups. Three, five and two participants per condition. Everything on this page is direction and size, not statistical significance; no hypothesis tests are reported because none would be meaningful at this scale. It is pattern-spotting from a master's dissertation dataset, shared so the negatives stay visible alongside the positives.
Within-person comparison. The strength of the design is that each participant's impaired leg is compared with their own functional leg, measured in the same session, so age, training and body mass cancel. The weakness is that "impaired" and "functional" were assigned by passive clinical assessment, and where the record was ambiguous the assessment sheet, not the strength data, decided the side.
What the twitch technique did and did not measure. The interpolated twitch protocol established maximal twitch current, MVC and the 5%-of-MVC twitch for each limb, and three superimposed-twitch sets were completed. The superimposed increments themselves were not retained in the archived record, so voluntary activation (%) is not reported; Question 2 uses the evoked force per milliamp at the 5% target as a proxy for how the muscle responds to stimulation. The IPC B dissociation — weaker on effort, not on stimulation — is what inhibition predicts, but a proxy is not the measurement, and three participants are not a sample.
Two reclassifications. Two participants (B03, C05) were graded IPC A in the original tables and are shown under B and C following a later review of the screening record. Their landing kinematics remain in the original "A" block in Question 5.
Trials retained. For the IPC C and original-A blocks the archived peak-GRF means use the second and third trials only, following the original analysis; the other blocks use all three.
What would make this conclusive
Keep the superimposed twitch. Recording the increment on every set converts the IPC B finding from "weaker" into "under-activated" or "not" — the single measurement that would settle the inhibition question.
Ten per condition, both sexes. Enough to put confidence intervals on the frontal-plane finding and to test whether the IPC C hip-adduction landing survives once leg-length discrepancy is controlled.
Pelvis kinematics in the same trial. The conditions are defined at the pelvis; measuring innominate motion during the landing would show whether the lock is still present under load or is a resting-state finding.
Re-test after release. If the pelvic lock is the constraint, restoring it should move the impaired-leg values toward the functional leg's within the same person — the cleanest test of the whole framework.
Methods and references
Source. Data were collected for the author's master's dissertation at the University of Chichester: screening (functional leg-length discrepancy, Downing's test, ligament stress tests, foot alignment, intrinsic pelvic condition), interpolated twitch technique on the quadriceps, ten countermovement vertical jumps and five with a bar across the shoulders on a force platform, and three-dimensional motion capture (torso, pelvis, thigh, shank and foot marker set). One participant was excluded before analysis as a deliberate control case.
- Abbott, J. (year). Dissertation title. MSc dissertation, University of Chichester. Add ethics approval reference if you wish to cite it.
- Merton, P. A. (1954). Voluntary strength and fatigue. Journal of Physiology, 123(3), 553–564.
- Allen, G. M., Gandevia, S. C., & McKenzie, D. K. (1995). Reliability of measurements of muscle strength and voluntary activation using twitch interpolation. Muscle & Nerve, 18(6), 593–600.
- Shield, A., & Zhou, S. (2004). Assessing voluntary muscle activation with the twitch interpolation technique. Sports Medicine, 34(4), 253–267.
- Hewett, T. E., Myer, G. D., Ford, K. R., et al. (2005). Biomechanical measures of neuromuscular control and valgus loading of the knee predict anterior cruciate ligament injury risk in female athletes: a prospective study. American Journal of Sports Medicine, 33(4), 492–501.
- Abbott, J. (2026). LLD Gait Risk Explorer and Failed Load Transfer Explorer [Living Research]. drjoabbott.com/living-research/ — the retrospective cohorts in which the same intrinsic pelvic condition grades are applied.
Participant codes (B01–B03, C01–C06, D01–D02) are study codes; the key is held with the source data and is not published. Hover any mark for its values; each figure has a table view.