What does a fast farmers carry or sled pull look like?
In strongmen, better performers took longer, quicker strides at top speed. Cross-sectional, short maximal efforts, no training test; no sled push studied.
Main caveat: Cross-sectional only. Higher vs lower performers in small groups; no group sizes or loads are printed in the review for these comparisons.
Hindle and colleagues searched five databases to 25 March 2019 for biomechanical studies of strongman exercises; 877 results gave 11 eligible articles, all cross-sectional. Higher performers in the farmer's walk and heavy sled pull had longer, quicker strides and shorter ground contact, mainly at maximum velocity.
Greater total system force (body mass plus 1RM squat) went with performance (farmer's walk r = 0.64, vehicle pull r = 0.68). Compared with unloaded walking, the farmer's walk had a shorter stride, a higher stride rate, more forward trunk tilt and much greater vertical force. Compared with a back squat, the sled pull produced more horizontal and less vertical force.
The authors identify grip strength as a likely limit in the farmer's walk, as their own suggestion. Our reading: the first usable numbers in the base on a carry and a sled pull, and they answer "what does a fast one look like" in strength athletes.
They do not answer how to train a carry or sled for HYROX: a faster person has longer, quicker strides almost by definition, the efforts are short and maximal, and the differences at the acceleration phase do not all point the same way.
Protocol
Design: systematic review (PRISMA; PROSPERO registration "not deemed to be relevant"); n = 11 articles from 877 results; search of AusportMed, CINAHL, Embase, Medline, SPORTDiscus up to 25 March 2019, plus reference lists and forward citations.
Eligibility: peer-reviewed; anthropometric, kinematic, kinetic, muscular or spatiotemporal measures of athletes performing strongman exercises; excluded injured athletes, sled loads below body mass and sprint-focused sled studies.
Exercises: atlas stone lift, farmer's walk, heavy sled/vehicle pull, log lift, keg walk, suitcase carry, tire flip, yoke walk.
Quality: a 16-item checklist developed by the authors: 9 articles low risk of bias, 2 satisfactory (our count of Table 1). Eight of the 11 have the review's co-authors Keogh or Winwood among their authors (our count; Renals [37] includes Keogh); screening was by BH, ST and EO.
Funding and interests: Australian Government Research Training Program Scholarship to the first author; no competing interests.
Correction: 10.1186/s40798-020-0239-3 says the symbol ☼ in and below Tables 2, 3, 4 and 5 did not display properly and reissues the tables; no stated change to findings.
Key Results
Higher (HP) vs lower performers (LP), as tabulated (mean ± SD; effect size). † marks a significant HP-LP difference in the paper:
| Task, phase | Measure | HP | LP | ES |
|---|---|---|---|---|
| Farmer's walk, HP max velocity vs LP sub-max velocity by distance; velocity row each at own max (Keogh [34]) | Ground contact (s) | 0.29 ± 0.02† | 0.34 ± 0.03 | −1.96 |
| Stride rate (Hz) | 2.01 ± 0.13† | 1.83 ± 0.04 | 1.88 | |
| Stride length (m) | 1.83 ± 0.04† | 1.40 ± 0.17 | 3.48 | |
| Velocity (m/s) | 3.66 ± 0.17† | 2.83 ± 0.36 | 2.95 | |
| Farmer's walk, acceleration | Ground contact (s) | 0.39 ± 0.04† | 0.32 ± 0.03 | 1.98 |
| Stride length (m) | 1.38 ± 0.16 | 1.33 ± 0.09 | 0.39 | |
| Velocity (m/s) | 2.61 ± 0.38 | 2.19 ± 0.27 | 1.27 | |
| Heavy sled pull, max velocity (Keogh [32]) | Ground contact (s) | 0.33 ± 0.04† | 0.76 ± 0.37 | −1.63 |
| Stride rate | 1.63 ± 0.12† | 1.10 ± 0.42 | 1.72 | |
| Stride length (m) | 1.29 ± 0.26† | 0.80 ± 0.16 | 2.27 | |
| Velocity (m/s) | 2.08 ± 0.08† | 0.99 ± 0.50 | 3.04 | |
| Heavy sled pull, acceleration | Ground contact (s) | 0.42 ± 0.19 | 0.57 ± 0.23 | −0.71 |
| Stride length (m) | 0.85 ± 0.25† | 0.65 ± 0.04 | 1.12 | |
| Velocity (m/s) | 1.22 ± 0.20† | 0.79 ± 0.32 | 1.61 |
Correlations (Winwood [8], novices by its title): total system force with farmer's walk r = 0.64, vehicle pull r = 0.68; flexed arm girth r = 0.46 and r = 0.74; muscle mass with farmer's walk r = 0.49.
Farmer's walk vs unloaded walk (Winwood [7]): mean vertical force 2540 ± 376 vs 1030 ± 247 N; stride 1.04 ± 0.12 vs 1.43 ± 0.11 m; rate 1.42 ± 0.17 vs 0.88 ± 0.06 Hz; trunk at foot strike 78.0 ± 3.00 vs 90.0 ± 2.00°.
Heavy sled pull vs back squat (Winwood [38]): mean anterior force 555 ± 107 vs 43.0 ± 22.0 N; mean vertical force 1330 ± 364 vs 2580 ± 648 N.
Spine and trunk in carries (McGill [35]; "three athletes" by the review's account of that study; no SD printed): muscular compressive load 7900 N for the farmer's walk, 7800 N for the yoke walk, and 5800 and 6700 N for suitcase carries.
Mechanisms
Measured: spatiotemporal, joint-angle and force differences between higher and lower performers, between phases and between exercises.
Authors' interpretation: better carrying means limiting the loss of stride length while keeping or raising rate; grip strength may limit the farmer's walk, as chalk is the usual only aid; a more horizontal trunk in the sled pull may be how athletes direct force horizontally, which they say needs quantitative confirmation.
Inference: velocity is stride length times stride rate, so "faster athletes had longer or quicker strides" is partly arithmetic. Who is stronger, and so can lengthen strides under load, cannot be separated from technique in these designs.
Inference: these are short, maximal efforts. Whether the same pattern holds for an athlete walking 200 m after several kilometres and stations is untested.
Practical Application
What does a fast farmers carry or sled pull look like? In strongmen, better performers took longer, quicker strides at top speed. Cross-sectional, short maximal efforts, no training test; no sled push studied.
Use it to describe, not to prescribe. It supports saying that better performers move with longer, quicker strides, and that total strength plus body mass goes with short heavy events. It does not support coaching a longer stride to go faster at a HYROX station; no study tried that.
Heavier people and strength: the "total system force" correlations include body mass in the measure and come from strength athletes, so they do not transfer to an athlete who must also run 8 km.
Grip: the authors' suggestion that grip limits the carry, and that carries are used to train it, is a discussion point; the review holds no grip-endurance data.
Limitations
Cross-sectional only. Higher vs lower performers in small groups; no group sizes or loads are printed in the review for these comparisons.
The headline hides phases. The abstract's "reduced ground contact time" holds at maximum velocity; in the farmer's walk acceleration phase higher performers had longer ground contact (0.39 vs 0.32 s, ES 1.98), and several acceleration differences are not significant.
Unlike phases compared. The farmer's walk table marks some HP-LP comparisons as between phases "based on distance" (HP maximum velocity vs LP sub-maximal velocity). The table's unit for sled pull stride rate is printed as seconds; the abstract and farmer's walk table treat it as a rate.
Author-built checklist and authors assessing a literature partly their own, though screening was by the first author and two others, not Keogh or Winwood; the authors name the checklist as a limitation. Not registered.
Discussion goes beyond the data. Qualitative comparison with other exercises, and suggestions for rugby, rowing, military and firefighting use, are not results.
Not covered: sled push (HYROX's station), pulling a sled in by rope, carries over long distance, fatigued athletes, women as a group, and any training adaptation.
Correction. Checked by identifier: only a display fault of one symbol; every number in the reissued tables also appears in the article text we saved (our check).
How this entry was checked and changed
Why this level
systematic in its search, but of small cross-sectional studies, with a checklist built by the authors, narrative synthesis, and discussion that extends by qualitative comparison
Full text (checked 2026-10-30)
The search of the five databases on 25 March 2019 returned 877 results, of which eleven articles were adherent to the inclusion criteria (full text)
Farmer's walk HP vs LP: ground contact 0.29 ± 0.02 vs 0.34 ± 0.03 s; stride rate 2.01 ± 0.13 vs 1.83 ± 0.04 Hz; stride length 1.83 ± 0.04 vs 1.40 ± 0.17 m (full text)
Heavy sled pull HP vs LP: ground contact 0.33 ± 0.04 vs 0.76 ± 0.37 s; stride rate 1.63 ± 0.12 vs 1.10 ± 0.42; stride length 1.29 ± 0.26 vs 0.80 ± 0.16 m (full text)
Total system force (body mass plus 1RM squat): farmer's walk r = 0.64; vehicle pull r = 0.68 (full text)
HP athletes reached maximum velocity in the final 17–20 m section of the walk, LP in the middle 8.5–11.5 m section (full text)
Sled loads less than the body mass of the athlete were excluded (full text)
Funding: an Australian Government Research Training Program Scholarship to the primary author; no competing interests (full text)