2023 — Acceleration Transmission During Cricket Fast Bowling

Research theme

The direct follow-up to Lamb et al. (2022). Same dataset, same 11 bowlers — but this time the question is not just how big the shock is at each level but what kind of shock it is. Impact acceleration has two physically distinct components: a slow part produced by the limbs actually rotating and absorbing (2–16 Hz), and a fast elastic wave that travels through bone and tissue (16–58 Hz). The elastic wave is the part usually blamed for bone microdamage. The study asks whether the elastic wave survives the trip up to the lumbar spine. Method: 11 male fast bowlers (19 ± 2 y, 80.5 ± 9.8 kg, 1.86 ± 0.06 m), nine Vicon Blue Trident IMUs at 1600 Hz on both distal tibiae, both distal thighs, both greater trochanters, and L5, L1 and C7. Six maximum-velocity deliveries with a full run-up; ground contacts from synchronised high-speed video; best three trials (force plate strike at FFC + fastest by speed gun) averaged. Power spectral densities of the tibia signals were inspected to identify the two frequency bands, then PSD integrals in the elastic-wave band computed for every sensor. Two-way (position × contact phase) repeated-measures ANOVAs, Bonferroni-corrected to p < 0.017. Direct measurement, not simulation.

What they measured

• The size of the impact shock at each body level (peak resultant acceleration) at the shank, thigh, hip, lower back (L5), mid back (L1) and neck (C7).
• When that peak arrived relative to ground contact (time to peak).
• How much of the signal is high-frequency “ringing” through the tissues (the PSD integral in the 16–58 Hz elastic wave band) versus the slower 2–16 Hz band produced by joint rotation.
• All of the above at back foot contact (BFC), front foot contact (FFC) and the follow-through (FT) landing.

Findings

1. Position and phase both matter, strongly. Peak resultant acceleration: position F = 113, p < 0.001; phase F = 28.1, p < 0.001. Accelerations fell significantly with each step up the body as far as L5.
2. Back foot contact is the mildest impact. BFC accelerations were significantly less than both FFC and FT. (This is the clean version of a sentence garbled in the 2022 paper — see that file’s caveats.)
3. The body attenuates harder when it is hit harder. A significant position × phase interaction (F = 10.1, p < 0.001) showed greater attenuation between positions following FFC and FT than following BFC — the attenuation system scales with the size of the input.
4. Timing: position had a significant effect on when the peak arrived (F = 26.4, p < 0.001) but phase did not (F = 4.45, p = 0.025, above the corrected 0.017 threshold), nor the interaction (F = 1.15, p = 0.331). General trend: greater delays at more superior positions.
5. The frequency bands: 2–16 Hz = joint rotations, 16–58 Hz = elastic wave. All main and interaction effects for elastic-wave content were significant (8.55 ≤ F ≤ 54.2, p < 0.001), following the same distal-to-proximal pattern as peak acceleration.
6. Most of the shank shock never reaches the low back. Attenuation from distal tibia to L5: 91 ± 2% (BFC), 86 ± 5% (FFC), 88 ± 4% (FT).
7. The headline claim, and it is a striking one: because so little of the tibial acceleration — and specifically so little of the elastic-wave content — reaches L5, the authors suggest “the post-impact elastic wave may be unlikely to contribute to lumbar stress fracture risk.”
8. Important self-imposed limit: the authors state directly that surface accelerations do not represent internal bone loads, because they disregard muscle forces (citing Matijevich et al., 2019). The muscle contribution to lumbar loading is not measured here.

What a coach should look for on video

This paper supports no direct coaching cue on video. It is a signal-processing study on instrumented bowlers; it relates no accelerometer measure to any technique variable, to ball speed, or to injury outcome. Inventing a cue from it would misrepresent it.

What it does change is how a coach or S&C should interpret wearable data:

• If you are using a tibia-mounted accelerometer for workload monitoring, know what it is telling you. The tibia signal is the largest and most reliable — but 86–91% of it does not reach the lower back, and the attenuation fraction varies bowler to bowler (SD ±2% at BFC but ±5% at FFC). Tibial “bone load” or “impact load” scores are a proxy for the impact the leg absorbed, not for lumbar loading.
• Do not build a lumbar stress fracture risk model on impact shock alone. This paper’s own conclusion is that the high-frequency component most associated with bone microdamage is substantially gone before L5. Whatever drives lumbar stress fracture in fast bowlers, this study argues it is probably not the post-impact elastic wave.
• Back foot contact is the low-load impact; front foot contact and the follow-through are the high-load ones. If you count impacts, count those two.

Caveats and limits

• n = 11, all male, mean age 19, all healthy and injury-free — a screening sample, not a risk-stratified one.
• Two-column conference abstract. No results table; peak values and time-to-peak appear only in Figure 1, which is a chart. Absolute g values for this paper are therefore not quoted here — for absolute numbers use the companion Lamb et al. (2022, ISBS), same dataset, which prints them.
• Surface, not bone. The authors say so explicitly. Muscle forces, which are a large part of spinal loading, are not captured at all.
• The “elastic wave is unlikely to contribute to lumbar stress fracture” claim is a suggestion, not a test. No injury outcomes were measured, no bowler was followed prospectively, and no alternative mechanism was tested. Read it as a hypothesis-narrowing statement.
• Frequency bands were identified by visual inspection of tibia power spectra, not by a standardised or validated criterion. Different band edges would change the elastic-wave integrals.
• Six maximal deliveries in a lab. Nothing about fatigue, spell length, or a season’s accumulation.
• The attenuation-to-L5 figures shift slightly between the two 2023-era Lamb outputs. This paper gives 91/86/88% (BFC/FFC/FT) to L5 from the distal tibia; the 2022 ISBS paper gives 88/80/82% to L5 from the ankle. Different reference sensor descriptions of the same site; direction and magnitude agree.

Relationship to other Felton work

• Same dataset and same PhD as Lamb et al. (2022, 40th ISBS) — see Lamb 2022 — surface measured accelerations. That paper gives the magnitudes; this one adds the frequency decomposition and the sharper statistics.
• Companion to Lamb et al. (2023, 19th ISCSB) — see Lamb 2023 — lower limb joint compliance and acceleration transmission — which takes the same phenomenon into a simulation model and tests what happens when the lower limb is made stiffer.
• Continues the criticism of the Felton simulation line: results “highlight the need to consider elastic wave attenuation within inverse and forward dynamics investigations of cricket fast bowling”, i.e. existing rigid-jointed models mis-state internal loads.

CONTRADICTION (within Felton’s own group, flagged in the 2022 paper and inherited here): the finding that acceleration keeps attenuating beyond L5, up to C7, contradicts McErlain-Naylor, King & Allen (2021), who reported no attenuation beyond L5 in drop-jump landings. McErlain-Naylor is a co-author on both. The proposed resolution is that fast bowling delivers a much larger input at the ankle so more residual acceleration is still available to attenuate above L5.

TENSION: this paper’s conclusion that the post-impact elastic wave is “unlikely to contribute to lumbar stress fracture risk” sits awkwardly against the framing that opens every paper in this sub-line — that high ground reaction forces are associated with greater stress fracture risk (Ranson et al., 2008) and that impact shock is the mechanism worth measuring. If the shock does not reach the lumbar spine in a damaging form, the case for accelerometer-based lumbar workload monitoring weakens considerably. The papers do not resolve this.

TENSION (performance vs injury): the attenuation described here is produced by compliance — joints rotating, tissue deforming, the impact being stretched out in time. Felton, Yeadon & King (2020) and Worthington, King & Ranson (2013) find that a straighter, more braced front leg produces more ball speed. Those two are in direct opposition at the front knee. See the folder README and the ISCSB compliance file, which quantifies it.