Chronic effects of static and dynamic stretching on hamstrings eccentric strength and functional performance: a randomized controlled trial

Key points

1. 01Static stretching (3 sets x 30 s, 3x/week, 10 sessions) reduced eccentric hamstring torque by 15.4% - a large effect size (d = 1.03)
2. 02Static stretching reduced eccentric torque by ~7.6-7.8% compared with control and dynamic stretching groups (moderate effect sizes of 0.50 and 0.51)
3. 03Triple hop distance decreased by 3.7% after static stretching (p = 0.009, d = 0.29) but was unchanged in the other groups
4. 04Dynamic stretching did not significantly change eccentric torque, hop distance, or 20-m sprint time
5. 05Sprint time (modified 20-m) was unchanged across all three groups

How it was conducted

Design

Randomized controlled trial with three parallel groups; blinded assessor; intention-to-treat analysis

Participants

45 active healthy men, aged 18-28, with limited hamstring flexibility (>=15 degrees active knee extension deficit), no injury in prior 6 months

Groups

No stretching (control, n=15), static stretching (n=15), dynamic stretching (n=15)

Intervention

10 sessions over ~3-4 weeks (3x/week); 3 sets of 30-second holds (static) or 3 sets of 30 repetitions (dynamic), targeting hamstrings

Primary outcome

Isokinetic knee flexor eccentric peak torque at 60 degrees/s, normalized by body mass (Nm/kg x 100)

Secondary outcomes

Triple hop for distance (non-dominant limb) and modified 20-m sprint time

What they found

• Eccentric peak torque: significant group-time interaction (F2,42 = 7.17, p = 0.002); static stretching within-group decrease of 238.63 Nm/kg (95% CI: -50.90 to -26.35), -15.4 +/- 10.4%, d = 1.03 (95% CI: 0.27 to 1.79), p <= 0.0001
• Eccentric torque between-group (static vs. control): -17.48 Nm/kg (95% CI: -55.72 to -20.77), d = 0.50; static vs. dynamic: -18.08 Nm/kg (95% CI: -56.32 to -0.17), d = 0.51
• No stretching group eccentric torque within-group change: -9.28 Nm/kg (95% CI: -22.53 to 3.97), d = 0.22; dynamic stretching: -11.70 Nm/kg (95% CI: -23.98 to 0.56), d = 0.31 (neither significant)
• Triple hop: significant group-time interaction (F2,42 = 5.12, p = 0.01); static stretching within-group decrease of -0.20 m (95% CI: -0.35 to -0.05), -3.7 +/- 4.1%, d = 0.29, p = 0.009
• Triple hop: no stretching within-group change +0.01 m (d = 0.01); dynamic stretching +0.13 m (d = 0.25); neither significant
• Modified 20-m sprint: no significant group-time interaction (F2,42 = 0.93, p = 0.40); effect sizes: no stretching d = 0.03, static d = 0.06, dynamic d = 0.45
• Test reliability: ICC = 0.96 for eccentric peak torque, 0.99 for triple hop, 0.98 for 20-m sprint

Limitations

• All-male, non-competitive active sample limits generalizability to female athletes or competitive sport populations
• Stretching was performed in isolation without warm-up; negative effects may be reduced when combined with a full warm-up routine
• Structural or neural mechanisms explaining torque loss were not directly measured (no muscle-tendon imaging or EMG)
• Relatively short intervention (10 sessions); longer-term effects remain unknown

Why it matters

For patients

Patients doing isolated static stretching for flexibility should be aware it may reduce the hamstring eccentric strength that protects against hamstring tears.

For clinicians

Clinicians should avoid prescribing isolated static stretching as the sole training modality when eccentric hamstring strength or hop performance is a treatment goal; dynamic stretching is a safer alternative.

For readers

This RCT provides controlled evidence that chronic static, but not dynamic, stretching impairs an important injury-protective quality of the hamstrings.