Associations between changes in leg extensor muscle power and physical function after supervised exercise

Key points

1. 01Larger gains in leg extensor power predicted greater improvements in sit-to-stand, stair climbing, walking speed, and self-reported daily function.
2. 02Associations were statistically significant for all five physical function outcomes after adjusting for age, sex, pain, adherence, and baseline scores.
3. 03The association did not significantly differ between the two exercise types (progressive resistance training vs. neuromuscular exercise), suggesting the finding is exercise-type independent.
4. 04Both exercise programs produced small-to-moderate increases in leg extensor power (mean ~0.22 W/kg for PRT and ~0.22 W/kg overall), with no statistically significant difference between groups.
5. 05Leg extensor power explained only a small-to-moderate share of variance in physical function outcomes (R2 0.09 to 0.38), meaning many other factors also contribute.

How it was conducted

Design

Secondary analysis of a 12-week multicenter cluster-randomized controlled trial (The Hip Booster Trial)

Participants

147 adults with clinically diagnosed hip OA (82 allocated to PRT, 78 to NEMEX; 147 had complete follow-up data for crude analyses)

Groups

High-velocity progressive resistance training (PRT) vs. neuromuscular exercise (NEMEX), both supervised for 12 weeks

Primary exposure

Change in leg extensor muscle power (LEP) of the affected limb, normalized to body weight (watt/kg), measured with the Nottingham Leg Extensor Power Rig

Outcomes

30-s chair stand test (30s-CST), 9-step timed stair climb test (9step-TSCT), 40-m fast-paced walk test (40m-FPWT), and HOOS ADL and Sport/Recreation subscales

Analysis

Simple and multivariate linear regression (crude and adjusted for age, sex, baseline LEP, baseline function, baseline pain, adherence, and group allocation); interaction term tested for exercise-type moderation

What they found

• Adjusted beta for LEP and 30s-CST: 2.34 [95% CI 1.33; 3.35] repetitions per W/kg change, R2 = 0.13, p < 0.05
• Adjusted beta for LEP and 9step-TSCT: -1.47 [95% CI -2.09; -0.85] seconds per W/kg, R2 = 0.38, p < 0.05
• Adjusted beta for LEP and 40m-FPWT: -2.20 [95% CI -3.30; -1.11] seconds per W/kg, R2 = 0.09, p < 0.05
• Adjusted beta for LEP and HOOS ADL function: 8.63 [95% CI 3.16; 14.10] points per W/kg, R2 = 0.23, p < 0.05
• Adjusted beta for LEP and HOOS Sport/Recreation: 10.57 [95% CI 2.32; 18.82] points per W/kg, R2 = 0.21, p < 0.05
• Mean change in LEP (affected limb) for total cohort: 0.22 [95% CI 0.15; 0.28] W/kg; PRT 0.25 [0.16; 0.34] W/kg vs. NEMEX 0.18 [0.09; 0.28] W/kg; difference 0.06 [-0.07; 0.19] W/kg
• Mean change in 30s-CST for total cohort: 1.5 [95% CI 1.1; 1.9] repetitions; no significant difference between PRT and NEMEX (p = 0.21)
• No significant interaction between LEP and exercise group for any outcome (interaction p-values: 0.14 for 30s-CST, 0.67 for 9step-TSCT, 0.33 for 40m-FPWT, 0.97 for HOOS ADL, 0.98 for HOOS Sport/Recreation)
• To achieve a minimal important difference in 30s-CST, one patient would need to improve LEP by approximately 0.27 W/kg more than another patient; for HOOS ADL (MID 8.6 points) the difference needed is 1.0 W/kg; for HOOS Sport/Recreation (MID 8.0 points) the difference needed is 0.76 W/kg

Limitations

• Secondary analysis of a randomized trial; causality cannot be established from associations alone.
• Regression models are approximations and describe differences only between otherwise identical patients, limiting individual-level predictions.
• Some patients were excluded from adjusted analyses due to missing adherence data, potentially introducing selection bias.
• Statistical power was reduced when testing the interaction between LEP change and exercise type, leading to wide confidence intervals and uncertain group-specific estimates.

Why it matters

For patients

If you have hip osteoarthritis, improving your leg muscle power through either resistance or neuromuscular exercise is likely to help you walk faster, climb stairs more easily, and perform daily activities better.

For clinicians

Both high-velocity progressive resistance training and neuromuscular exercise produce similar gains in leg extensor power and physical function; targeting muscle power is worthwhile but should be combined with attention to psychological and other physiological factors rather than treated as the sole focus.

For readers

This secondary analysis provides longitudinal evidence that leg extensor power is a plausible mechanism for exercise-induced physical function gains in hip OA, complementing prior cross-sectional data, though the modest R2 values indicate it is one of several contributors.