SYNTH-86 · Guide

Foot strike

Leave a pain-free runner's landing alone: most distance runners, elite marathoners included, land on the heel; neither habitual pattern is more economical, an imposed forefoot strike costs economy and moves load from the knee to the ankle and Achilles, and the injury link is unsettled — change it only for a named problem, gradually

Blimited evidenceMostly from recreational to world-class athletesWritten before October 2026, being re-checked

Default (what to prescribe first)

1

A pain-free runner keeps the landing they have. The meta-analysis of changing strike concludes that changing it cannot be recommended for an uninjured rearfoot striker (REF-1113, A-); a narrative critique concludes that a switch does not improve economy or reduce injury risk (REF-1115, C+).

2

Heel striking is the norm, not a fault. Pooled across distance-running studies, 79% land on the rearfoot early in a run and 86% later (REF-1107, A-).

In an elite half marathon 74.9% were rearfoot strikers and 1.4% forefoot (REF-1104, B); at the 2017 World Championship marathon rearfoot striking never fell below 54% of men or 67% of women at any distance, and the top four men were rearfoot strikers (REF-1105, B); 93.67% of a recreational marathon field heel struck (REF-1106, B).

3

The fastest runners land further forward. In the 2017 World Championship 10,000 m final, run above 22 km/h, no rearfoot strikes were seen (REF-1108, B); in the elite half marathon midfoot striking was more common in the faster groups (REF-1104). Inference: that is likely a consequence of speed, not a technique to copy at easy pace.

4

If knee load is the problem, try step rate before strike. Raising step rate by 5-10% at the same speed cut energy absorbed at the knee (REF-620, B); a higher step rate lowered joint loading across 10 studies (REF-883, A-).

Decision (situation → what to do)

SituationDo thisSource and how sure
Healthy runner, any landingLeave itREF-1113 A-; REF-1115 C+
"Will a forefoot strike make me faster or more economical?"No. Habitual rearfoot and non-rearfoot strikers did not differ in economy; imposing it on a rearfoot striker worsened economy at 10.8 and 12.6 km/hREF-1113 A- (SMD −1.67 and −1.26, immediate); REF-1109 B (VO2 and %CHO lower with the habitual rearfoot pattern at 3.0 and 3.5 m/s in rearfoot runners; habitual patterns equal)
Knee pain (patellofemoral) in a rearfoot strikerStep rate +5-10% first; a forefoot cue only through a clinician, knowing it moves load to the Achilles; gait retraining added nothing over education in one trialREF-620 B; REF-1111 B (habitual forefoot strikers: patellofemoral force 4.3 vs 5.1 BW, Achilles force 6.3 vs 5.1 BW); imposed forefoot raised ankle load in rearfoot runners (REF-1110, B) and non-rearfoot striking raised the plantar-flexor moment (REF-1113); REF-917 B+; the sequence is practice
Achilles or calf trouble in a midfoot or forefoot strikerDo not push further forward; treat the load (SYNTH-73) (practice)REF-1118 B (midfoot strike distinguished Achilles injuries, OR 2.27; forefoot strike posterior lower-leg injuries, OR 2.59 — among injured runners only)
Switching to minimal shoes or barefoot runningOver months, not weeksREF-1120 B (10 of 19 showed increased bone marrow oedema in at least one bone after a 10-week transition); timing is practice
Runner worried about injuryAvoid any single run more than 10% longer than the longest of the past 30 days (REF-1102); strike is not the shown leverREF-1117 B+ (827 military cadets: no injury difference by foot strike or cadence)

Evidence

Prevalence: pooled 79% → 86% rearfoot, 11% changed with distance and 84% of them toward rearfoot; no study found a performance benefit of rearfoot, five found one for non-rearfoot, one in women but not men, four none — "inconclusive" (REF-1107). World Championship marathon: no difference between the top and bottom half of men; more of the top half of women kept a non-rearfoot strike (REF-1105).
Economy: REF-1113, REF-1109 as above.
Where the load goes: total lower-limb work did not differ between habitual patterns; knee moments and work were higher in rearfoot, ankle in forefoot strikers (REF-1110, B); rearfoot strikers had higher vertical loading rates (REF-1112, A-; REF-1113); non-rearfoot striking lowered patellofemoral stress and raised the plantar-flexor moment (REF-1113).
Injury: rearfoot strikers reported about twice the rate of repetitive stress injuries in one university team, with no difference in traumatic injuries (REF-1114, B-); retrospective evidence of lower repetitive stress injury with non-rearfoot striking is limited and prospective comparisons are lacking (REF-1113); low evidence for any relationship (REF-1116, B+); prospectively, longest ground contact time, not strike, raised overuse injury risk (HR 2.15, REF-1117).
Retraining that reduced injuries: impact-loading feedback cut 12-month injuries from 38% to 16% in novice runners (HR 0.38, REF-1119, B+) — it targeted vertical loading rate, not foot strike.
Barefoot context: barefoot forefoot strikers produce smaller collision forces, but injuries were not measured (REF-714, B); lifelong barefoot populations show no difference in relative injury rates, on limited evidence and with injuries distributed differently (REF-718, B+).

Practical Application

"Heel striking is wrong" → say: most distance runners, the top four men in one World Championship marathon among them, land on the heel; changing a pain-free landing has no shown benefit and an immediate economy cost in lab tests.
"How did Kipchoge land?" — the base has no measurement of him; answer with the race data instead.
For knee pain, offer the step-rate cue and refer persistent pain (SYNTH-73); if a clinician chooses a forefoot cue, add calf and Achilles capacity and build gradually (practice).
Never present forefoot striking as injury-proof: it relocates load, and the injury evidence for it is retrospective.

Limitations

Race data are snapshots at one or a few points; habitual and race patterns can differ.
The injury evidence linking strike to injury is mostly retrospective and low quality; prospective cohorts are military or short.
Economy comparisons are acute; whether a long, gradual switch changes economy is not settled here.
No entry measures elite marathoners' economy or injury by strike pattern.
How this entry was checked and changed
Why this level

(prevalence on an A- meta-analysis and B race videos; economy on an A- meta-analysis and a B lab study; load on A- meta-analyses plus B lab studies; injury on low-quality retrospective evidence, a B+ prospective cohort, a B case-control analysis and one B+ RCT of a different intervention; the advice for a painful runner is practice)