Facebook Pixel
Brookbush Institute Logo

Friday, December 29, 2023

Stop Calling Deadlifts a Back Exercise

Stop Calling Deadlifts a Back Exercise
Brent Brookbush

Brent Brookbush


Stop Calling Deadlifts a "Back Exercise"

by Dr. Brent Brookbush DPT, PT, MS, CPT, HMS, IMT


The primary joint action of a deadlift is hip extension, not lumbar extension. Your erector spinae, multifidus, latissimus dorsi, and trapezius muscles do not contribute to hip extension. These muscles are active, but they are primarily performing isometric contractions, and isometric contractions are not ideal for increasing muscle size (hypertrophy) or dynamic strength. Saying deadlifts are "good for thickening your back" is as ridiculous as saying "bicep curls are great for rounding your shoulders."

This topic is relevant because it relates to where a deadlift should be programmed in a strength training routine. Is it a lower-body exercise or an upper-body exercise, and if you are doing a split routine, on which day should it be programmed? The answer should be obvious, but for some strange reason, calling a deadlift a "lower-body exercise" is controversial. If a heavy lower-body day is done the day before or after an upper-body day, and your upper-body day includes deadlifts, a frequency and recovery issue is created. For more info, we recently published a new course with a comprehensive systematic research review on Training Frequency and Recovery . It is likely that the glute complex and lumbar erectors (erector spinae and multifidus ) will not have optimal time to recover, the volume of exercise will exceed the body's ability to adapt, over-training these muscles may impede progress, and the risk of injury will have increased without the potential for additional benefits. In short, this mislabeling could result in more work for less gain and a higher risk of injury.

Deadlifts are a great exercise for increasing lower body strength and hypertrophy, especially the hip extensors. They are a poor choice for "back muscle" hypertrophy.

Is circuit training (vertical loading) as effective as conventional trianing (horizontal loading) for hypertrophy, strength, and power? Deadlifts are a great lower body exercise.
Caption: Is circuit training (vertical loading) as effective as conventional trianing (horizontal loading) for hypertrophy, strength, and power? Deadlifts are a great lower body exercise.

Quick Summary

  • The primary joint action during a deadlift should be hip extension, not lumbar extension.
  • The gluteus maximus is the prime mover of hip extension; not the lumbar erectors or latissimus dorsi .
  • Erector spinae EMG activity is similar for most compound lower body exercises (squats, deadlifts, lunges, etc.) when the weight is adjusted to a given repetition maximum (e.g. 1-RM).
  • The latissimus dorsi can only contribute a relatively small amount of force to lumbar extension, and they do not contribute to hip extension.
  • During a deadlift, the lumbar erectors, scapular stabilizers, and latissimus dorsi are maintaining an isometric contraction. Isometric contractions are likely far less effective for hypertrophy, dynamic strength, and power development when compared to dynamic contractions including an eccentric phase.
  • Practical Application: Ideal training frequency is 2-3 sessions/muscle group/week, with 2-3 days rest between sessions for similar muscle groups. Performing heavy deadlifts and/or other high-intensity lower-body exercises on consecutive days is likely to inhibit progress toward strength and hypertrophy.

Brookbush Institute Position Statement: Training Frequency

  • Training frequency and recovery between sessions should be based on the ability to achieve the same or better performance in subsequent sessions. Generally, more days rest should be given between initial sessions of a new routine and then increased to 2-3x/muscle group/week. For experienced exercisers, research suggests that it may not be possible to perform optimal volume and intensity with a frequency of 1 or fewer sessions/muscle group/week; 2 and 3 sessions/week result in similar improvements, and 4 sessions/muscle groups/week may result in less improvement (due to targeting similar muscle groups on consecutive days). In summary, the most efficient and effective recommendation may be a frequency of 2 sessions/muscle group/week, with 2-3 days rest between sessions. However, during the initial sessions of a new routine, 3 sessions/muscle group/2 weeks with 3-7 days rest may be recommended.

Sample Program:

Experienced Exerciser:

  • Goal: Hypertrophy
  • Split: Upper and Lower
  • Routine Construction: Strength/Stability Super-sets and Circuit Training

Acute Variables:

  • Load: Moderate (75-90% of 1-RM); Lighter (60-75% of 1-RM)
  • Reps/set: 6-12 reps; 6-12 reps
  • Sets/exercise (circuits): 2-3 sets/muscle  group
  • Rest between exercises: Circuit training, 1 min rest between exercises
  • Training Time: 25 – 45 minutes (excluding warm-up).


  • Week 1 and 2 (3x/routine/2 weeks):
    • Upper Body: Monday, Friday, and the following Wednesday
    • Lower Body: Wednesday, the following Monday, and Friday
  • Weeks 3 and 4-6: (2x/routine/week)
    • Upper Body: Monday, Thursday
    • Lower Body: Tuesday, Friday

Upper Body Routine:

Lower Body Routine:

Deadlifts as an Erector Spinae Exercise:

The idea that deadlifts would be an ideal exercise for erector spinae (or multifidus ) hypertrophy is flawed for several reasons. First, there is a logical incongruence. Trunk muscles are not generally classified with upper body muscle groups. That is, the erector spinae and multifidus are not commonly categorized with back muscles, just as the rectus abdominis is not commonly categorized with chest muscles. Consider this, if deadlifts are a back and erector spinae exercise, are sled pushes a chest and rectus abdominis exercise? And, would we then program deadlifts on back day, sled push on chest day, and potentially both on an "upper body day? Although all of these ideas may have merit unto themselves, the lack of agreement between these ideas does not. There are several reasons why trunk muscles should be categorized as separate from extremity muscles, including a potential need for more specific techniques, a focus on stability and isometric strength/endurance (and less focus on hypertrophy), and the separate categorization reduces the complexity of more advanced programming models.

Further, research does not support deadlifts as an exceptional erector spinae exercise. Research, actually demonstrates that erector spinae EMG activity is similar for most compound lower body exercises (squats, deadlifts, lunges, etc.) when the weight is adjusted to a given repetition maximum (e.g. 1-RM). If all compound lower body exercises result in similar erector spinae EMG activity, then why specifically choose deadlifts to add to back day? See the studies below:

  • Andersen et al. compared 13 healthy men (age: 21.9 ± 1.6 years) with strength training experience (4.5 ± 1.9 years) performing barbell hip thrusts, barbell deadlifts, and hex bar deadlifts, during a 1-RM, with identical foot placement and distance between the hands. Note, the participants were instructed to lift the barbell while maintaining a neutral back, to avoid the straight-leg deadlift technique, and extend their knees and hips in one movement. Outcome measures included EMG activity of the gluteus maximus biceps femoris , and erector spinae . The findings demonstrated that hip thrusts resulted in the highest EMG activity for the gluteus maximus , deadlifts resulted in the highest EMG activity for the biceps femoris , and all exercises resulted in similar EMG activity for the erector spinae .
    • Andersen V, Fimland MS, Mo DA, Iversen VM, Vederhus T, Hellebø LRR, et al. Electromyographic comparison of barbell deadlift, hex bar deadlift, and hip thrust exercises: a cross-over study. J Strength Cond Res. 2018; 32(3): 587–93. pmid:28151780
  • McAllister et al. compared 12 healthy men (age: 27.1 ± 7.7 years) with strength training experience (age: 8.6 ± 5.5 years) while performing prone leg curls, good mornings, glute-ham raises, and Romanian deadlifts (RDL), with 1-RM loads, and EMG activity assessed for the erector spinae, gluteus medius, semitendinosus, biceps femoris (BF), and medial gastrocnemius. The EMG activity of the erector spinae was similar during prone leg curls, good mornings, and Romanian deadlifts, but was significantly higher during glute-ham raises.
    • McAllister MJ, Hammond KG, Schilling BK, Ferreria LC, Reed JP, Weiss LW. Muscle activation during various hamstring exercises. J Strength Cond Res. 2014; 28(6): 1573–80. pmid:24149748

Deadlifts as a Latismus Dorsi Exercise

Some individuals have suggested that deadlifts are a back exercise because contracting the latissimus dorsi is important for performance. Again, this logic is mostly refuted by the research. The latissimus dorsi can only contribute a relatively small amount of force to lumbar extension and cannot contribute to hip extension. Further, latissimus dorsi engagement is unnecessary to hold the bar close to the shins. Shifting the pelvis back and bending the knees (e.g. "sitting back") is a far more efficient method for aligning the bar, the tibia, and the line of pull, and "sitting back" has the added benefit of reducing the moment arm of hip extension (which may aid in lifting more weight). Note, the research could be interpreted to imply that engaging the latissimus dorsi can enhance performance, but the contribution would be small, would be specific to preventing lumbar flexion (or increasing sacroiliac joint stiffness), and it is unlikely that this type of engagement would result in an increase in latissimus dorsi strength or hypertrophy. The studies below demonstrate the latissimus dorsi is only capable of a small contribution to extension.

  • Bogduk et al. compared the fascicular anatomy of the latissimus dorsi following the dissection of five adult cadavers. The size, attachments, and orientation of each latissimus dorsi fascicle were determined. Maximum force was estimated by applying a force coefficient based on the summation of the physiological cross-sectional area of the fascicles, in conjunction with the moment arms of each fascicle for joint actions at the shoulder joint, lumbar spine, and sacroiliac joints. The findings demonstrated that the estimated maximum total force exerted by the latissimus dorsi is 162 - 529 Newtons (N) on the shoulder, approximately 30 N on the sacroiliac joint, and approximately 6.3 N on the lumbar spine. This study suggests that the force latissimus dorsi's contribution to lumbar extension would be nearly inconsequential.
    • Bogduk, N., Johnson, G., & Spalding, D. (1998). The morphology and biomechanics of latissimus dorsi. Clinical Biomechanics , 13 (6), 377-385
  • Granata et. al. compared 10 resistance-trained males (age: 21.9 ± 1.6 years) while performing standing lumbar extension from a trunk-flexed position of 45°, at a tempo of 0°, 30°, 60°, and 90°/sec, with loads of 0kg, 18kg, and 36kg, while analyzing EMG data gathered from the erector spinae, latissimus dorsi, internal obliques, external obliques, and rectus abdominis. The EMG data were used in conjunction with a biomechanical model to aid in determining the relative contribution of each muscle to generating force. The contribution of flexor muscles (rectus abdominis and obliques) increased with velocity but decreased with load lifted. During extension, the erector spinae produced the gross majority of force (80% or more), the latissimus dorsi contributed significantly less (20% or less), and the internal obliques may have contributed slightly (less than 5%).
    • Granata, K. P., & Marras, W. S. (1995). Coactivity on Dynamic Spinal Loads. Spine , 20 (8), 913-919

Woman and man performing deadlift for lower extremity strength and hypertrophy
Caption: Woman and man performing deadlift for lower extremity strength and hypertrophy

Isometric Contractions Are Not Great for Hypertrophy or Strength

Although we tend to promote exercises that focus on isometric contractions for core muscles (e.g. planks), the reasons are related to function. The ability to maintain a stable core aids in protecting our mobile lumbar spine while transferring force between the ground, lower extremity, and upper extremity. For example, in sports like basketball and football, when we push a defender out of the way, we want to be able to add the strength of our legs to the "pushing motion", which requires that our lumbar spine and pelvis do not bend, give-way, or fail to maintain stability. The same scenario occurs during a deadlift as we use our large lower extremity muscles, especially our hip extensors (gluteus maximus ), while maintaining a stable trunk, to move a load held in our hands.

Exercises that include isometric contractions of the trunk muscles are not generally expected to result in large amounts of trunk muscle hypertrophy. This is related to the complex topic of stimulating hypertrophy, which is a bit beyond the scope of this article, but we will attempt to introduce the major points in this paragraph. Research has demonstrated that eccentric contractions result in more hypertrophy than concentric contractions (when volumes of exercise are similar). This is hypothesized to be related to the amount of muscle damage caused by eccentric contractions, and the strong signal this damage sends to stimulate the repair and growth of contractile proteins (resulting in hypertrophy). Concentric contractions result in far less damage than eccentric contractions, and isometric contractions result in similar or less damage than concentric contractions. At the very least, dynamic contractions involving all 3 contraction types (as is being performed by the gluteus maximus during a deadlift) will result in larger improvements in hypertrophy than on a single contraction type, and this is especially true of isometric contractions (e.g. the erector spinae during a deadlift). Note, isometric contractions may result in some hypertrophy, just far less than eccentric contractions or dynamic contractions.

Isometric contractions may also be less effective for improving strength, depending on the training goal. Isometric contractions are great for isometric strength in the range of motion (ROM) trained, but tend to have fairly poor transfer to dynamic strength or strength in other ROMs. Concentric and eccentric contractions also result in more strength for the type of contraction and range of motion trained, but there is generally more transfer to other contraction types and some transfer to other ranges.

Unfortunately, there seems to be no research available that directly compares the amount of hypertrophy following a multi-week or multi-month training program while performing only concentric, isometric, or eccentric contractions. So, admittedly, the paragraph above is the conclusion that can be made from dozens of related, but ultimately indirect inferences. The publications below demonstrate the trend toward eccentric contractions resulting in more hypertrophy than concentric contractions.

  • Quote from a Meta-analysis by Roig et al. comparing concentric and eccentric training (2009) -"Eccentric training performed at high intensities was shown to be more effective in promoting increases in muscle mass measured as muscle girth. In addition, eccentric training also showed a trend towards increased muscle cross-sectional area measured with magnetic resonance imaging or computerised tomography."
    • Roig M, O'Brien K, Kirk G, Murray R, McKinnon P, Shadgan B, Reid WD. The effects of eccentric versus concentric resistance training on muscle strength and mass in healthy adults: A systematic review with meta-analysis. Br J Sports Med 43: 556–568, 2009.
  • Bamman et al. compared 10 novice male and female exercisers (age: 24.4 ± 0.7 years) 0.5 hours, 1 hour, 2 hours, 4 hours, 8 hours, 1 day, and 2 days post-exercise, following two sessions of concentric-only or eccentric-only squats, with 6-10 days of rest between sessions. Each session included 8 sets/exercise, 8 reps/set, 85% 1-RM for concentric squats, and 110% 1-RM for eccentric squats. The findings demonstrated that only the eccentric squat protocol resulted in a significant decrease in knee extension MVIC, elevated serum concentrations of creatine kinase, and DOMS, with the largest changes exhibited during reassessment on day 2 (16).
    • Bamman, M. M., Shipp, J. R., Jiang, J., Gower, B. A., Hunter, G. R., Goodman, A., ... & Urban, R. J. (2001). Mechanical load increases muscle IGF-I and androgen receptor mRNA concentrations in humans. American Journal of Physiology-Endocrinology and Metabolism, 280(3), E383-E390.
  • An RCT by Farthing et al. compared 24 untrained volunteers (age 18-36 years) performing curls for 8 weeks. The protocol included 12 sessions (4 weeks) of 3x/week, 2-6 sets/session, and 8RM/set, followed by another 12 sessions (4 weeks) of either concentric-only contractions, eccentric-only contractions, a slow tempo of 30°/sec, or a fast tempo of 180°/sec. (4 testing conditions) for 6 sets/session with 6-8 RM/set. The findings demonstrated that CSA increased more following eccentric contractions when compared to concentric contractions, and faster contractions when compared to slower contractions. Note, this study used isokinetic equipment and the findings should not be interpreted as "larger increases in CSA can be achieved from fast eccentrics, or allowing free weights to fall"; however, this study does suggest more CSA may be anticipated from focusing on the eccentric deceleration of loads at faster tempos (25).
    • Farthing, J. P., & Chilibeck, P. D. (2003). The effects of eccentric and concentric training at different velocities on muscle hypertrophy. European journal of applied physiology, 89(6), 578-586.

The Problem is Frequency and Recovery

This is the most important issue that is created by mislabeling deadlifts as a "back exercise." Research implies similar muscle groups should not be targeted with resistance training on consecutive days, and/or that adopting a split routine is necessary to benefit from higher frequency resistance training (3x/week or more). As mentioned above, if a heavy lower-body day is done the day before or after an upper-body day, and your upper-body day includes deadlifts, a frequency and recovery issue is created. It is likely that the glute complex and lumbar erectors (erector spinae and multifidus ) will not have optimal time to recover, the volume of exercise will exceed the body's ability to adapt, over-training these muscles will impede progress, and the risk of injury will have increased without the potential for additional benefits. In short, this mislabeling of deadlifts as a "back exercise" could result in more work for less gain and a higher risk of injury. 

The summary statement in the Brookbush Institute's position stand from Acute Variables: Training Frequency and Recovery Between Sessions , highlights why busting this myth is important for optimizing training and performance programming.

  • "The most efficient and effective training frequency recommendation is likely 2 sessions/muscle group/week, with 2-3 days rest between sessions."

The studies mentioned below demonstrate better results from 3x/week (no consecutive training days) than 4x/week, and that more days of rest results in less lactate accumulation and better-set performance in subsequent training sessions.

  • RCT by Firoozi et al. compared 33 experienced male exercisers (age: 21.9 ± 1.3 years) randomly assigned to a control group (no exercise), a 3x/week group, or a 4x/week group. Participants completed a split routine (exercises included in the program: squats, knee extensions, barbell bench press, barbell incline bench press, dumbbell bench press, barbell arm curls, dumbbell arm curls, arm curls with EZ bar, military press, behind the head military press, dumbbell lateral  raises, dumbbell tricep extensions, lying close grip triceps press, incline triceps extensions, deadlifts, leg curls, leg press, dumbbell lunges, calf raises, lat pulldowns behind the head, barbell rear delt rows, rear deltoid rows, hyperextensions, and abdominal crunches) for 8 weeks, 3 sets/exercise, 8 reps/set, 65-80% of 1-RM loads (periodized), moderate (1-1.5-min) rest between sets, and moderate (2-3 min) rest between exercises. Volume was equated by the total number of exercises performed per session, with each group performing the same exercises. The 3x/week group performed more exercises per routine - Day 1: legs, chest, and biceps; Day 2: shoulders, triceps, and legs; Day 3: legs, back, and core. The 4x/week group performed fewer exercises per routine - Day 1: legs and arms; Day 2: legs and chest; Day 3: legs and shoulders, Day 4: back, legs, and core. Results demonstrated neither group exhibited a significant change in bench press or leg press endurance (reps to failure with 60% 1-RM); however, both groups exhibited significant improvements in 1-RM strength. The 3x/week group exhibited significantly larger increases in bench press and leg press 1-RM strength. Additionally, the 3x/week group exhibited significantly larger increases in thigh circumference, and the 4x/week group exhibited significantly larger increases in chest circumference. The results of this study are puzzling due to the way the exercises were split throughout the routines; for example, the legs and arms are included every day of each program, but specific  exercises for the chest, back, and shoulder are only included 1 day each week. This study may suggest that fewer high-quality sets (and exercises) are more beneficial for upper body training (upper body program for 4x/week group), and that training similar muscle groups on consecutive days is deleterious to hypertrophy (lower body program for 4x/week group).
    • Firoozi, H., Arazi, H. and Asadi, A. (2020) Effects of resistance training program on muscular performance adaptations: comparing three vs. four times per week. Biomedical Human Kinetics, 12, 149-156, doi: 10.2478/bhk-2020-0019
  • Miranda et al. compared 16 experienced male exercisers (age: 26.1 ± 3.1 years) randomly assigned to 1 day, 2 days, or 3 days rest between the 1st and 2nd session. Sessions included bench press performed with 0° (flat bench), 30°, or 45° inclines, 4 sets/position, 8 RM load/set, reps to failure/set, and moderate (2 min) rest between sets. The results demonstrated that 2-3 days were necessary to perform similar total work (sets x reps) during the re-assessment of the flat or incline (30° or 45°) bench press. 1-day rest resulted in a significant decrease in total work, and 3 days of rest resulted in more total work than 2 days of rest. Additionally, blood lactate concentrations immediately post-exercise were compared following the first and second sessions. The findings demonstrated that 1 day of rest between sessions resulted in significantly higher blood lactate concentrations; however, 2-3 days of rest resulted in similar concentrations.
    • Miranda, H., Maia, M. F., Paz, G. A., de Souza, J. A. A. A., Simão, R., Farias, D. A., & Willardson, J. M. (2018). Repetition performance and blood lactate responses adopting different recovery periods between training sessions in trained men. The Journal of Strength & Conditioning Research, 32(12), 3340-3347.

Links to Strength and Performance Articles

Acute Variable Courses

© 2023 Brent Brookbush (B2C Fitness, LLC d.b.a. Brookbush Institute)

Comments, critiques, and questions are welcome!