Reconsidering exercise selection with EMG: poor agreement between ranking hip exercises with gluteal EMG and muscle force

Key Points

1. Surface electromyography (sEMG) is often thought to be a proxy for exercise effectiveness, but this notion has been questioned.
2. This study compared sEMG and estimated muscle force for the gluteus maximus and gluteus medius during eight hip exercises performed by 10-14 female footballers.
3. Correlations between sEMG amplitude and estimated muscle forces ranged from 0.29 to 0.56, suggesting sEMG should not be relied on in isolation for exercise selection.

BACKGROUND AND OBJECTIVE

Surface electromyography (sEMG), a measure of muscle activation, is commonly thought to be a proxy for how effectively exercises target muscles. For example, two systematic reviews identified a total of 71 studies assessing sEMG of the gluteal muscles (1,2). Despite sEMG’s popularity, its validity for predicting muscular adaptations has been questioned. A 2018 review (3) and a follow-up 2022 commentary (4) argued that sEMG does not actually predict strength or hypertrophy gains.

An alternative measure for exercise effectiveness might be muscle force, which can be estimated with neuromusculoskeletal modeling. The purpose of this study was to compare sEMG amplitude and estimated muscle force for the gluteus maximus and gluteus medius during eight hip exercises. The authors hypothesized that the relationship between the two metrics would be poor.

Surface electromyography, a measure of muscle activation, is commonly thought to be a proxy for how effectively exercises target muscles.

Clinicians should not rely solely on surface electromyography for selecting exercises to target the gluteal muscles.

METHODS

Data was analyzed from a previous study (5) of 14 healthy female footballers with a median (interquartile range) age of 24.1 (6.5). Participants performed eight common injury prevention and rehabilitation exercises for the hip: single-leg squats, split squats, single-leg Romanian deadlifts, single-leg hip thrusts, banded side-steps, side-lying leg raises, hip hikes, and side planks (see video 1 for examples). Most exercises were done with both body weight and 12-repetition maximum (12RM) loads.

During the exercises, sEMG data was collected from the gluteal muscles along with three-dimensional motion capture and force plate data. These inputs were used in a neuromusculoskeletal model to estimate gluteal muscle forces. For both sEMG amplitudes and estimated muscle forces, normalized peak and mean values were calculated for each exercise and muscle. The relationships between these metrics were assessed with Spearman’s correlations and linear mixed-effects models.

VIDEO 1 – HIP EXERCISES

RESULTS

Correlations between normalized peak and mean sEMG amplitudes and estimated gluteal muscle forces ranged from 0.29 to 0.56. The only correlation that was statistically significant was between normalized mean sEMG amplitude and estimated muscle forces for the gluteus medius. Normalized sEMG amplitudes explained between 5% and 26% of the variance in estimated muscle forces (see Table 1).

To illustrate these discrepancies, consider 12RM hip hikes (for gluteus maximus) and 12RM side-lying leg raises (for gluteus medius), which both ranked high in sEMG amplitude but low in estimated muscle force. On the other hand, 12RM split squats (for gluteus maximus) and 12RM single-leg Romanian deadlifts (for gluteus medius) both ranked relatively low in sEMG amplitudes but high in estimated muscle force (see Figure 1).

LIMITATIONS

This study was limited by a relatively small sample of participants (10-14), with only two muscles assessed. Nevertheless, the robustness of the analyses was aided by a large number of repetitions across multiple exercises and loading conditions, resulting in ~1000 data points for analyses. Given the focus on gluteal muscles, caution is warranted if generalizing these findings to other muscle groups.

CLINICAL IMPLICATIONS

Clinicians are often under the impression that sEMG is a valid indicator of how effectively an exercise targets a muscle for strength and hypertrophy. As such, they may select exercises based on the ample body of evidence ranking exercises by sEMG for various muscles (e.g. the gluteal muscles [1,2]).

Recent critiques have questioned these premises (3,4), and the data from the present study supports these concerns. The authors deemed the relationships between sEMG amplitudes and estimated muscle forces for the gluteal muscles “weak” (ρ < 0.7). It’s worth noting, however, that a Spearman’s ρ of 0.7 is a stringent threshold. Although ρ < 0.7 was considered weak by this study’s criteria, other interpretation frameworks might classify the observed correlations – especially for the gluteus medius – as “moderate.”

All in all, sEMG amplitudes and estimated muscle forces may be somewhat related, but their relationship is likely weaker than many clinicians assume. These findings suggest clinicians should not rely solely on sEMG for selecting exercises to target the gluteal muscles. When available, clinicians should prioritize exercises supported by longitudinal evidence of strength and hypertrophy gains. In the absence of such data, priority can be given to exercises that allow for relatively heavy loading, especially with a bias towards longer muscle lengths (6).

+STUDY REFERENCE

Collings T, Bourne M, Barrett R, Meinders E, Gonçalves B, Shield A, Diamond L (2025). Reconsidering exercise selection with EMG: Poor agreement between ranking hip exercises with gluteal EMG and muscle force. Medicine & Science in Sports & Exercise. Advance online publication.

SUPPORTING REFERENCE

1. Neto, W.K., Soares, E.G., Vieira, T.L., Aguiar, R., Chola, T.A., Sampaio, V.L., & Gama, E.F. (2020). Gluteus maximus activation during common strength and hypertrophy exercises: A systematic review. Journal of Sports Science and Medicine, 19(1), 195–203.
2. Moore, D., Semciw, A.I., & Pizzari, T. (2020). A systematic review and metaanalysis of common therapeutic exercises that generate highest muscle activity in the gluteus medius and gluteus minimus segments. International Journal of Sports Physical Therapy, 15(6), 856–881.
3. Vigotsky, A. D., Halperin, I., Lehman, G. J., Trajano, G. S., & Vieira, T. M. (2018). Interpreting signal amplitudes in surface electromyography studies in sport and rehabilitation sciences. Frontiers in Physiology, 8, 985.
4. Vigotsky, A. D., Halperin, I., Trajano, G. S., & Vieira, T. M. (2022). Longing for a longitudinal proxy: Acutely measured surface EMG amplitude is not a validated predictor of muscle hypertrophy. Sports Medicine, 52(2), 193–199.
5. Collings, T.J., Bourne, M.N., Barrett, R.S., Meinders, E., Gonçalves, B.A.M., Shield, A.J., & Diamond, L.E. (2023). Gluteal muscle forces during hip-focused injury prevention and rehabilitation exercises. Medicine & Science in Sports & Exercise, 55(4), 650–660.
6. Schoenfeld, B. J., Grgic, J., Ogborn, D., & Krieger, J. W. (2017). Strength and hypertrophy adaptations between low- vs. high-load resistance training: A systematic review and meta-analysis. Journal of Strength and Conditioning Research, 31(12), 3508–3523.