Hamstring muscle architecture and microstructure changes following Nordic hamstring exercise training

Key points

1. 01Hypertrophy was non-uniform across the four hamstrings, largest in semitendinosus (26%) and short head of biceps femoris (22%), and smallest in semimembranosus (6%, not significant).
2. 02Fiber tract length increased significantly across all hamstrings, with the biggest gain in semitendinosus (18%).
3. 03Diffusion metrics (axial, mean, and radial diffusivity) all rose about 4 to 5%, suggesting increased fiber cross-section; fiber tract angle and fractional anisotropy did not change.
4. 04After 3 weeks of detraining, fiber tract length returned to baseline across all muscles, but volume and fiber cross-section gains were mostly retained.
5. 05The knee-dominant nature of the exercise may preferentially recruit semitendinosus and short head of biceps femoris, which have greater moment arms at the knee.

How it was conducted

Design

Single-group longitudinal intervention with MRI at pre-training, post-training, and after detraining

Participants

11 recreationally active adults (5 males, 6 females) with no lower-limb injury or regular Nordic hamstring exercise in the past 12 months

Intervention

9 weeks supervised Nordic hamstring exercise, 3 sessions/week, 4 to 5 sets of 6 to 8 reps (72 to 120 reps/week), followed by 3 weeks detraining

Imaging

3 Tesla MRI using 6-point Dixon for volume and diffusion tensor imaging (b = 0 and 400 s/mm2, 15 directions)

Outcomes

Muscle volume, fiber tract length, fiber tract angle, axial/mean/radial diffusivity (AD, MD, RD), and fractional anisotropy (FA) for BFsh, BFlh, ST, and SM

Analysis

Linear mixed effects models with training status and muscle as fixed effects and subjects random, Tukey post hoc, p < 0.05

What they found

• Volume increased significantly in BFsh (p = 0.025), BFlh (p = 0.007), and ST (p < 0.001, np2 = 0.40); SM was not significant (p = 0.229).
• Percent hypertrophy: BFsh 22%, BFlh 9%, ST 26%, SM 6%.
• Fiber tract length increased significantly in all hamstrings (p = 0.001, np2 = 0.13): BFsh 11%, BFlh 7%, ST 18%, SM 10%.
• Diffusivity rose significantly: AD +5% (p < 0.001, np2 = 0.28), MD +4% (p < 0.001, np2 = 0.30), RD +5% (p < 0.001, np2 = 0.21).
• Fiber tract angle (p = 0.625, np2 = 0.02) and fractional anisotropy (p = 0.322, np2 = 0.03) were unchanged.
• After detraining, ST volume decreased significantly (p < 0.001, about 8%) but remained above pre-training (p < 0.001); BFsh (p = 0.715), BFlh (p = 0.173), and SM (p = 0.945) did not change versus post-training.
• Fiber tract length decreased significantly across all hamstrings after detraining (p = 0.004) and returned to baseline.
• Diffusivity did not change with detraining versus post-training (AD p = 0.496, MD p = 0.231, RD p = 0.327) and remained above pre-training (AD p < 0.001, MD p = 0.001, RD p = 0.001).

Limitations

• Very small sample of 11 recreationally active adults, limiting precision and generalizability.
• Single-group design with no control or comparison exercise, so changes cannot be attributed to the exercise versus other factors with certainty.
• Detraining was only 3 weeks, so longer-term retention or loss of adaptations is unknown.
• DTI-derived fiber tract lengths were shorter in magnitude than 3D ultrasound fascicle lengths in the same cohort, indicating measurement-method differences.

Why it matters

For patients

If you do Nordic hamstring exercise to protect against hamstring injury, the muscle-building benefits fade within a few weeks of stopping, so consistency matters.

For clinicians

Nordic hamstring exercise drives non-uniform hypertrophy favoring semitendinosus and short head of biceps femoris, and fiber-length gains regress quickly, supporting continuous rather than block-only prescription.

For readers

This is a small mechanistic MRI study showing how a single exercise reshapes individual hamstring muscles and how fast some of those changes reverse when training stops.