Coaches often choose exercises based on which ones show the biggest EMG signal.
But adaptation is driven by mechanical tension (muscle force), not activation alone.
This study tested whether ranking glute exercises by EMG matches rankings by estimated muscle force.
If I pick glute exercises by highest EMG, am I actually picking the ones that load the glutes the most?
What Did the Researchers Do?
Exercises Performed
The researchers had 14 healthy female footballers (18–30 yrs) perform 8 exercises.
- Single-leg squat, split squat, single-leg RDL, single-leg hip thrust, banded side-step, hip hike, side-lying leg raise, side plank.
- Each done wit just bodyweight and ~12RM where applicable.
Data Collection
- 3D motion capture, ground reaction forces, and surface EMG from 12 lower-limb muscles.
- EMG normalized to MVICs per guidelines.
Muscular Modeling
- EMG-assisted neuromusculoskeletal modeling (OpenSim/CEINMS) estimated gluteus maximus and gluteus medius muscle forces.
- The model accounted for force–length–velocity and passive elements.
Analyses
- Rank exercises by peak normalized EMG vs peak normalized muscle force; compare with Spearman correlations.
- Mixed-effects models tested how much EMG explains muscle force, with/without accounting for exercise type and participant as random effects.
What were the results?
Poor Agreement
- Exercises ranked by EMG did not match exercises ranked by estimated muscle force.
- EMG both over- and underestimates where an exercise truly sits if judged by force.
- Across all exercises and loads, peak EMG explained only ~5% of GMax and ~19% of GMed peak force variance.
Context Matters
- When the models accounted for exercise type and participant, EMG explained ~80–85% of peak force.
- Within the same athlete and biomechanically similar tasks, EMG becomes far more meaningful.
What Does This Mean?
- Don’t rank exercises by EMG alone ➔ High EMG isn’t a reliable proxy for high muscle force across different exercise types.
- Force–length–velocity rules the day ➔ Deep hip flexion, longer muscle lengths, and contraction dynamics can increase force (and passive tension) without higher EMG; This is why split squats or single-leg RDLs may “load” glutes more than their EMG suggests.
- Use EMG within-athlete, within-task ➔ EMG can help you compare two similar exercises on the same person with consistent electrode placement; It’s far weaker for cross-exercise or cross-athlete comparisons.
Limitations
- Muscle forces are model estimates (not direct in vivo force).
- However, model-predicted joint moments closely matched inverse dynamics, and parameters were bounded to physiologic ranges.
Coach’s Takeaway
- Pick by purpose, not just peaks ➔ If the goal is hypertrophy/strength, prioritize exercises that place the glutes under high tension at long lengths (e.g., deep split squats, single-leg RDLs) rather than those that merely “light up EMG.”
- Compare like with like ➔ EMG helps fine-tune choices within a category (e.g., which split-squat variant) for the same athlete.
- Program the mechanics ➔ Chase positions and movements that bias favorable force–length–velocity conditions that match your goals.
I hope this helps,
Ramsey
Reference: Collings TJ, Bourne MN, Barrett RS, Meinders E, Gonçalves B, Shield AJ, Diamond LE. (2025). Reconsidering Exercise Selection with EMG: Poor Agreement between Ranking Hip Exercises with Gluteal EMG and Muscle Force. Medicine & Science in Sports & Exercise, 57(9):1829–1837.
