The protective effect of preseason running workload against in-season hamstring strain injuries in elite soccer players

Key points

1. 01Players who ran more sprint distance in preseason had significantly fewer in-season hamstring strain injuries (OR 0.896 per 100 m, P = 0.01).
2. 02A higher number of preseason sprint efforts was also protective (OR 0.761 per 10 reps, P = 0.004).
3. 03Practical thresholds emerged: roughly 840 m of sprint distance and about 34 sprint efforts in preseason flagged lower-risk athletes.
4. 04Overall internal workload and total distance were not significantly associated with injury risk; the protective signal was specific to high-speed sprinting.
5. 05This was an observational cohort with only 17 injuries, so it suggests a pattern rather than proving cause and effect.

How it was conducted

Design

Retrospective cohort study, Level 3 evidence, STROBE reporting

Participants

67 athlete-seasons from 42 elite Japanese soccer players, 2021 to 2023 (age 26.6 plus or minus 4.5 y)

Measurement

GPS workload (Catapult Vector S7, 10 Hz) over the final 4 preseason weeks, plus internal workload from sRPE

Workload variables

Internal workload, total distance, high-speed distance (above 18 km/h), sprint distance and count (above 24 km/h), acceleration/deceleration count

Analysis

Multivariable logistic regression with adjustment (model 1: age, prior HSI, ACL, calf injuries; model 2: propensity score) and ROC analysis

Primary outcome

In-season hamstring strain injury

What they found

• 17 hamstring strain injuries occurred across 67 athlete-seasons (incidence 7.8 per 1000 exposure hours).
• Sprint distance was protective: OR 0.896 per 100 m (95% CI 0.822 to 0.976, P = 0.01), holding after adjustment (model 1 OR 0.892, P = 0.02; model 2 OR 0.899, P = 0.02).
• Sprint count was protective: OR 0.761 per 10 reps (95% CI 0.633 to 0.916, P = 0.004), holding after adjustment (model 1 OR 0.755, P = 0.004; model 2 OR 0.763, P = 0.005).
• High-speed distance trended protective but was not significant (OR 0.865, 95% CI 0.742 to 1.007, P = 0.06); internal workload (P = 0.28) and total distance (P = 0.10) were not significant.
• Sprint distance discriminated injury with AUC 0.721 (95% CI 0.572 to 0.870, P = 0.007); cutoff 842 m gave sensitivity 0.59 and specificity 0.88.
• Sprint count gave AUC 0.739 (95% CI 0.591 to 0.887, P = 0.003); cutoff 33.5 reps gave sensitivity 0.59 and specificity 0.94.
• Injuries most often involved the biceps femoris (10 of 17), with mean time lost 19.2 plus or minus 13.7 days.

Limitations

• Observational cohort design shows association, not proven causation, and unmeasured factors could explain the link.
• Small sample with only 17 injuries limits statistical power and the precision of the cutoff thresholds (wide confidence intervals).
• Preseason loading varied between athletes and some training sessions were unmonitored.
• Most injuries were mild to moderate, so findings may not extend to severe hamstring strains.

Why it matters

For patients

For a competitive soccer player, gradually building up sprinting during preseason may lower the chance of a hamstring injury once the season starts.

For clinicians

GPS sprint metrics in preseason, with rough thresholds near 840 m sprint distance and 34 sprint efforts, can help flag players at higher hamstring injury risk for targeted conditioning.

For readers

A single observational study suggests preseason sprint exposure is protective against hamstring injury, but it cannot prove cause and needs confirmation in larger or randomized work.