Surface Electromyographic Activation Patterns and Elbow Joint Motion During a Pull-Up, Chin-Up, or Perfect-Pullup™ Rotational Exercise

Abstract

Youdas, JW, Amundson, CL, Cicero, KS, Hahn, JJ, Harezlak, DT, and Hollman, JH. Surface electromyographic activation patterns and elbow joint motion during a pull-up, chin-up, or Perfect-Pullup™ rotational exercise. J Strength Cond Res 24(12): 3404-3414, 2010-This study compared a conventional pull-up and chin-up with a rotational exercise using Perfect·Pullup™ twisting handles. Twenty-one men (24.9 ± 2.4 years) and 4 women (23.5 ± 1 years) volunteered to participate. Electromyographic (EMG) signals were collected with DE-3.1 double-differential surface electrodes at a sampling frequency of 1,000 Hz. The EMG signals were normalized to peak activity in the maximum voluntary isometric contraction (MVIC) trial and expressed as a percentage. Motion analysis data of the elbow were obtained using Vicon Nexus software. One-factor repeated measures analysis of variance examined the muscle activation patterns and kinematic differences between the 3 pull-up exercises. Average EMG muscle activation values (%MVIC) were as follows: latissimus dorsi (117-130%), biceps brachii (78-96%), infraspinatus (71-79%), lower trapezius (45-56%), pectoralis major (44-57%), erector spinae (39-41%), and external oblique (31-35%). The pectoralis major and biceps brachii had significantly higher EMG activation during the chin-up than during the pull-up, whereas the lower trapezius was significantly more active during the pull-up. No differences were detected between the Perfect·Pullup™ with twisting handles and the conventional pull-up and chin-up exercises. The mean absolute elbow joint range of motion was 93.4 ± 14.6°, 100.6 ± 14.5°, and 99.8 ± 11.7° for the pull-up, chin-up, and rotational exercise using the Perfect·Pullup™ twisting handles, respectively. For each exercise condition, the timing of peak muscle activation was expressed as a percentage of the complete pull-up cycle. A general pattern of sequential activation occurred suggesting that pull-ups and chin-ups were initiated by the lower trapezius and pectoralis major and completed with biceps brachii and latissimus dorsi recruitment. The Perfect·Pullup™ rotational device does not appear to enhance muscular recruitment when compared to the conventional pull-up or chin-up.

Introduction

The pull-up maneuver has been used as a performance measure to assess physical fitness and muscular strength and endurance of the shoulder (^25,30). The concentric or up-phase of the pull-up exercise begins when the performer grasps a horizontal overhead bar and then uses muscles of the shoulder-arm-forearm complex to lift his or her body mass so that the feet leave the floor and the nose or chin is elevated above the bar. Subsequently, during the eccentric or down phase, the performer lowers his or her body mass to the start position. The ability to successfully perform multiple repetitions of a pull-up is dependent upon a performer's body mass (^8,39), muscle strength, and movement style. Traditionally, the pull-up has been used to assess the degree of physical fitness in adolescents (³⁶) and in men and women who attend the US military service academies (^17,38,40).

Rehabilitation or fitness professionals are concerned with the level of balance between upper body pressing and pulling performance. Excessive emphasis on pressing movements in upper extremity resistance training may increase the likelihood of shoulder complex trauma such as rotator cuff injury or strain (^6,14). The strength ratio between upper body pulling and pressing strength can be determined clinically by using the 1 repetition maximum (1RM) bench press and pull-up. This ratio was reported to be about 100% (⁵). By periodically assessing the ratio of press-to-pull, rehabilitation professionals or coaches, can monitor an athlete's physical training to help preserve a more balanced shoulder complex and prevent injury. At first glance the conventional pull-up or chin-up appears straightforward to perform. However, the mechanical demands placed upon the muscles of the shoulder-arm-forearm complex and trunk could be readily modified by simply changing the position of the handgrip from one of a forward grasp in which the forearms are pronated and the palms are directed away from the face (pull-up) to a reverse grasp whereby the forearms are supinated and the hands are directed toward the face (chin-up). Despite the familiarity with the pull-up and chin-up exercise, movement scientists or fitness professionals have a paucity of information on muscle activation patterns and kinematics associated with the conventional pull-up and chin-up exercise.

Surprisingly only 1 report was found which described the muscle activation patterns of the shoulder and arm muscles during the performance of a pull-up and chin-up (³³). Four healthy volunteers (3 men and 1 woman) whose ages ranged from 21 to 53 years (mean ± SD = 33.5 ± 15 years) performed multiple repetitions of either a pull-up or chin-up. Surface electrodes were used to record electromyographic (EMG) signals from 7 muscles during the concentric or shortening phase of the pull-up and chin-up exercise. The EMG signals were filtered, full-wave-rectified, and integrated, but the activation patterns were not normalized to a maximum voluntary isometric contraction (MVIC). The investigators divided the EMG tracings into 3 equal parts to display the initial, middle, and final phase of the pull-up and chin-up. According to the investigators, the teres major, upper portion of the pectoralis major, biceps brachii, infraspinatus, and latissimus dorsi were active during the early phase of the exercise. During the middle phase, EMG activity of the latissimus dorsi and teres major was “pronounced,” whereas activation of the pectoralis major and biceps brachii declined. During the final phase, EMG activity of the latissimus dorsi and infraspinatus muscles remained high, whereas the other 5 muscles showed a marked decline. Despite this information, we presently do not know the relative activation of a given muscle during the pull-up and chin-up exercise when normalized to a MVIC. Furthermore, to our knowledge, no report has provided kinematic data regarding movement of the shoulder-arm-forearm complex during the performance of a pull-up and chin-up procedure.

Manufacturers of novel strength training devices often claim their product will produce improved results when compared to conventional strength training procedures. The new Perfect·Pullup™ twisting handles are designed to rotate 360° upon a stationary base. Recently, the originator of the Perfect·Pullup™ unit claimed the device's distinctive rotation handles move with the arm's natural motion to improve strength, balance, flexibility, and endurance, while reducing risk of muscle strain and injury. Presently, no one has investigated these claims. Therefore, the purpose of this study was to answer the following 3 questions. (a) Would the concentric phase of a combined pull-up and chin-up performed using the Perfect·Pullup™ twisting handles demonstrate significantly greater EMG activation of back, shoulder, arm, and abdominal musculature when compared to a conventional pull-up or chin-up exercise? (b) Would the concentric phase of a combined pull-up and chin-up performed using the Perfect·Pullup™ twisting handles demonstrate significantly greater absolute (end position-start position) elbow joint sagittal plane range of motion (ROM) when compared to a conventional pull-up or chin-up exercise? (c) Will there be a significant difference in the temporal sequence of peak back, shoulder, arm, and abdominal muscle activation expressed as a percentage of the complete pull-up and chin-up cycle among the 3 exercise conditions?

Methods

Experimental Approach to the Problem

Electromyography is used to record motor unit activity within a muscle or group of muscles. Researchers in health and wellness professions use EMG to observe differences in muscle activity between exercise conditions (^24,25). According to Distefano et al. (¹³), investigators have discovered higher EMG signals correspond with larger strength gains. The conventional pull-up is typically performed with a pronated forearm grip on the horizontal bar with the hands shoulder width apart. The chin-up is a variation whereby a supinated forearm grip on the horizontal bar is used in place of a pronated one. The Perfect·Pullup™ uses patented rotating handles to combine the conventional pull-up with the chin-up. The subject started by using a pronated forearm grip and subsequently supinated the forearm while advancing the body mass in an upward direction. We examined 7 muscles recruited during a conventional pull-up, chin-up, and Perfect·Pullup™ including the following: (a) lower trapezius, (b) latissimus dorsi, (c) infraspinatus, (d) erector spinae, (e) pectoralis major, (f) external oblique, and (g) biceps brachii. Electromyographic data were analyzed with Delsys software by processing the root mean square (RMS) algorithm and normalizing values to an MVIC. Motion analysis was used to describe sagittal plane upper extremity kinematic data of the elbow during the pull-up/chin-up exercises. Both EMG and motion analysis data were combined to determine a temporal relationship between muscle activation and elbow flexion ROM throughout the pull-up cycle. The independent variables examined were the 3 aforementioned pull-up exercises. The dependent variables were the normalized EMG values of the 7 muscles and the kinematic elbow data. The question that we sought to answer was as follows: “Do the Perfect·Pullup™ handles require increased EMG activation compared to the conventional pull-up and chin-up exercises?”

Subjects

Twenty-one men (24.9 ± 2.4 years) and 4 women (23.5 ± 1.0 years) volunteered to participate. The average height, weight, and body mass index (BMI) of the subjects were 180 ± 3.7 cm, 76 ± 13 kg, and 23 ± 3 kg·m⁻¹, respectively. All subjects demonstrated upper extremity ROM within normal limits and self-reported the ability to perform all 3 pull-up exercises. A sample size of 22 subjects was required to detect a mean difference in EMG recruitment of 10% MVIC (effect size = 0.20) between conditions with a statistical power (1 − β) equal to 0.80 at α = 0.05 (²⁰). At the time the study was completed, 64% (n = 16) of subjects engaged in an upper extremity weight training program that included some form of back exercises (pull-ups, lat pull-downs, or seated rows). In addition, 84% (n = 21) of subjects were actively strength training at least 2-3 times per week. Subjects self-reporting a history of the following upper extremity conditions were excluded from the study: (a) previous shoulder subluxation, dislocation, or fracture; (b) a history of joint instability, tendinitis, bursitis, impingement, adhesive capsulitis, neurovascular complications, or any condition that limited physical activity for greater than 2 days over the last 6 months; and (c) current complaints of neuromuscular pain, numbness, or tingling in the upper extremity, neck, or back during the pull-up screening. The study procedures were approved by the Mayo Institutional Review Board, Mayo Clinic, Rochester, MN, USA. Before enrollment in the study all subjects provided written informed consent and completed a survey that assessed their current level of physical activity.

Instrumentation

Raw EMG signals were collected with Bagnoli™ DE 3.1 double-differential surface EMG sensors (Delsys Inc., Boston, MA, USA). The sensor contacts were made from 99.9% pure silver bars 10 mm in length and spaced 10 mm apart and encased within preamplifier assemblies measuring 41 × 20 × 5 mm. The preamplifiers had a gain of 10 V/V. The combined preamplifier and main amplifier permitted a gain from 100 to 10,000. The common mode rejection ratio was 92 dB at 60 Hz, and input impedance was >1,015 Ω at 100 Hz. Data were collected at a sampling frequency of 1,000 Hz. Raw EMG signals were processed with EMG works® Data Acqusition and Analysis software (Delsys Inc.).

Three-dimensional motion analysis was performed with a computer-aided Vicon MX motion analysis system with 5 high-resolution MX20+ infrared digital cameras (Vicon Motion Systems, Oxford, United Kingdom). Kinematic data were sampled at 100 Hz. Spherical reflective markers were placed on specific anatomic landmarks to model each subject's torso and upper extremities. Cameras were positioned so that each marker was detected by a minimum of 2 cameras throughout the task. Vicon Nexus software was used to quantify upper extremity kinematics in the sagittal, frontal, and transverse planes of motion. Raw kinematic data were processed and smoothed with a Woltring quintic spline filter at a mean square error of 20 mm.

The manufacturer of the Perfect·Pullup™ device intended the apparatus be installed in a standard residential doorway; however, this was not possible because the doorway to our laboratory opened into a busy corridor. Instead a free-standing pull-up apparatus was constructed out of solid steel with the vertical bars measuring 1.9 m in length and the horizontal bar measuring 0.9 m in length and 0.03 m in diameter (Figure 1). The base of the apparatus measured 0.91 m × 0.99 m. The diameter of the horizontal bar was suitable for secure placement of the Perfect·Pullup™ handgrip devices.

Figure 1: The free-standing pull-up and chin-up apparatus was constructed out of solid steel with the vertical bars measuring 1.9 m in length and the horizontal bar measuring 0.9 m in length and 0.03 m in diameter. The base of the apparatus measured 0.91 m by 0.99 m. The diameter of the horizontal bar was suitable for secure placement of the Perfect·Pullup™ handgrip devices.

Procedures

Data were collected in a research laboratory at the Mayo Clinic in Rochester, MN, USA. All subjects wore appropriate attire to permit correct placement of EMG electrodes and motion analysis reflective markers. To allow optimal EMG reading, the subject's skin was abraded with an alcohol wipe until erythema was attained. Electrodes were then placed superficially parallel to the direction of the muscle fibers on the subject's hand-dominant side. A ground electrode was placed ipsilaterally on the medial malleolus. The gain on the Delsys EMG instrumentation was set individually for each muscle by having each subject perform an informal manual muscle break test for 2- to 3-second hold times. The MVICs of each muscle were then collected by using formal manual muscle testing techniques as described by Hislop and Montgomery (²¹). Palm, wrist, and elbow widths and shoulder-offset measurements were recorded with a hand-held caliper to satisfy anthropometric input values specified by the motion analysis software. Twenty-seven spherical motion analysis reflective markers were placed on the subject (Figure 2). Markers were placed bilaterally as follows: (a) dorsum of the hand on the middle of the third metacarpal, (b) ulnar and radial styloid processes, (c) middle lateral forearm, (d) lateral epicondyle, (e) deltoid tuberosity, (f) tip of the acromion process, (g) anterior superior iliac spine, and (h) posterior superior iliac spine. Markers were also placed on the (i) sternal notch, (j) xiphoid process, (k) spinous processes of C7 and T10, and (l) middle of the scapular spine of the dominant side. Subjects donned a head band with 4 equidistantly spaced markers. With the subject standing in the anatomic position, a 10-second static trial was recorded to establish baseline muscle activity. Education on appropriate cadence of the pull-up, chin-up, or rotational exercise using the Perfect·Pullup™ twisting handles using a metronome (50 b·min⁻¹) was provided. Subjects began each exercise in the anatomical position and performed 3 consecutive cycles of each of the 3 randomized pull-up exercises. The pull-up, chin-up, and rotational exercise using the Perfect·Pullup™ twisting handles are illustrated in Figures 3-5, respectively. A 2-minute rest period was provided for each subject between exercise conditions.

Figure 2: An anterior (A) and posterior (B) view of the electromyographic (EMG) electrode and motion analysis reflective marker placement locations. Anteriorly EMG electrodes (A) were placed on the biceps brachii, pectoralis major, and external oblique of the subject's dominant side. Posteriorly (B) EMG electrodes were placed on the lower trapezius, infraspinatus, latissimus dorsi, and erector spinae.

Figure 3: The pull-up (A) began when the subject grasped the horizontal bar with palms facing away from the subject and elbows in terminal extension. Next, the subject flexed his/her knees to about 90° so he or she was suspended from the pull-up/chin-up apparatus by the upper extremities and the lower extremities were non-weight-bearing. The subject began the up or concentric phase of the pull-up by lifting the torso until the nose was raised just superior to the horizontal bar (B), whereupon the down phase began as the torso was lowered to the start position. Subjects performed 3 consecutive repetitions of the pull-up.

Figure 4: The chin-up (A) began when the subject grasped the horizontal bar with palms facing the subject and elbows in terminal extension. Next the subject flexed his/her knees to about 90° so he or she was suspended from the apparatus by the upper extremities and the lower extremities were nonweight-bearing. The subject began the up or concentric phase of the chin-up by lifting the torso until the nose was raised just superior to the horizontal bar (B), whereupon the down phase began as the torso was lowered to the start position. Subjects performed 3 consecutive repetitions of the chin-up.

Figure 5: The Perfect·Pullup™ handles consisted of a soft cell foam handle with a width of 14.5 cm and a height of 24 cm. Perfect·Pullup™ twisting handles were designed to rotate 360° upon a stationary base. The diameter of the horizontal bar of the pull-up and chin-up apparatus safely secured the twisting handles while the subject performed a combined pull-up/chin-up. The combined pull-up and chin-up (A) began when the subject grasped the Perfect·Pullup's™ twisting handles with palms facing away from the subject and elbows in terminal extension. Next, the subject flexed his or her knees to about 90° so he or she was suspended from the pull-up and chin-up apparatus by the upper extremities and the lower extremities were non-weight bearing. The subject began the up or concentric phase of the combined pull-up and chin-up by lifting the torso and supinating the forearms until the elbows reached terminal flexion (B). The down phase began as the torso was lowered to the start position accompanied by forearm pronation. Subjects performed 3 consecutive repetitions of the combined pull-up and chin-up.

Statistical Analyses

An intraclass correlation coefficient (ICC_3,1) was used to estimate the test-retest reliability of the EMG recordings. A 1-factor repeated measures analysis of variance (ANOVA) with Bonferroni corrections addressed question 1 whereby we examined differences in EMG activation (%MVIC) for a specific muscle group among the 3 exercise conditions. Similarly, a 1-factor repeated-measures ANOVA with Bonferroni corrections also addressed question 2 where we compared the absolute elbow joint ROM across the 3 exercise conditions. Lastly, question 3 explored the temporal sequence of muscle activation for each of the 3 exercise conditions and was tested with a repeated-measures (3 conditions × 7 muscles) ANOVA. Post hoc analysis of the condition × muscle interaction on timing of peak muscle activation as a percentage of the complete pull-up cycle was performed. Simple effects tests with Bonferroni correction were used to handle multiple comparisons. For all 3 ANOVAs, we used a level of significance of p ≤ 0.05. All data were analyzed with SPSS 15.0 for Windows software (SPSS Inc, Chicago, IL, USA).

Results

Reliability

During the pull-up test-retest procedure, EMG recordings from 8 subjects were separated by 2 weeks. The ICCs for the 7 muscles were as follows: (a) infraspinatus-0.77; (b) biceps brachii-0.64; (c) lower trapezius-0.64; (d) external oblique-0.57; (e) erector spinae-0.48; (f) latissimus dorsi-0.35; and (f) pectoralis major-0.35.

Electromyographic Exercise Data

During the concentric phase, statistically significant differences existed between the conventional pull-up and the chin-up exercises as follows: (a) the lower trapezius had greater activation (p = 0.006) in the conventional pull-up when compared to the chin-up (mean difference = 10.8% MVIC, 95% confidence interval [CI] = 2.8-18.8% MVIC, effect size = 0.51); (b) the pectoralis major had greater activation (p = 0.009) in the chin-up when compared to the pull-up (mean difference = 13.6% MVIC, 95% CI = 3.0-24.3% MVIC, effect size = 0.3); and (c) the biceps brachii had greater activation (p = 0.03) in the chin-up when compared to the pull-up (mean difference = 17.9% MVIC, 95% CI = 1.5-34.2% MVIC, effect size = 0.54). Values of mean peak activation (percentage) data for the 3 exercise conditions are contained in Figure 6.

Figure 6: Mean (±SD) normalized electromyographic (EMG) activation of back, shoulder, arm, and abdominal musculature expressed as a percentage of maximum voluntary isometric contraction (MVIC) during a pull-up, chin-up, or rotational exercise with Perfect·Pullup™ twisting handles. External oblique (EO); erector spinae (ES); pectoralis major (PM); lower trapezius (LT); infraspinatus (ISP); biceps brachii (BB); and latissimus dorsi (LD). ^* p < 0.05; ^** p < 0.01.

Elbow Joint Range of Motion Data

During the concentric phase, we observed a statistically significant difference in absolute elbow joint sagittal plane ROM among the 3 pull-up/chin-up exercises (F_2,30 = 4.13, p = 0.03); however, Bonferroni tests for multiple comparisons did not detect significant pairwise differences (Figure 7). Mean absolute elbow joint ROM was 93.4 ± 14.6°, 100.6 ± 14.5°, and 99.8 ± 11.7° for the pull-up, chin-up, and combined pull-up/chin-up performed using the Perfect·Pullup™ twisting handles, respectively.

Figure 7: Sequential timing of mean (±SE) peak muscle activation of the pull-up is expressed as a percentage of the complete exercise cycle. Brackets indicate the results of statistically significant multiple comparisons. External oblique (EO); erector spinae (ES); pectoralis major (PM); lower trapezius (LT); infraspinatus (ISP); biceps brachii (BB); and latissimus dorsi (LD). ^* p < 0.05; ^** p < 0.01.

Muscle Recruitment during the 3 Exercises

The temporal order of peak motor unit recruitment was visualized from RMS processed EMG data to determine the temporal sequence of selected back, shoulder, arm, and abdominal muscles during the complete pull-up cycle for each of the 3 conditions. Timing of peak EMG activation was expressed as a percentage of a complete pull-up, chin-up, or rotational exercise cycle using the Perfect·Pullup™ twisting handles. A repeated measures (3 conditions × 7 muscles) ANOVA detected a statistically significant muscle × conditions interaction (F_2,276 = 2.83, p = 0.021), which indicated the presence of differences in the order of peak activation among the pull-up conditions. Sequential muscle activation expressed as a percentage of the complete pull-up cycle and the results of multiple comparisons are displayed in Figures 7-9 for the pull-up, chin-up, and combined pull-up/chin-up using the Perfect·Pullup™ twisting handles, respectively. On average (Figure 10) a complete cycle for each of the 3 exercise conditions lasted about 2,500 milliseconds. The midpoint of each exercise cycle occurred at the point of termination of elbow flexion and signified the completion of the concentric or up-phase of the cycle. Additionally, the timing at which peak activation of a muscle occurred ranged from 12% to 44% of the cycle.

Figure 8: Sequential timing of mean (±SE) peak muscle activation of the chin-up is expressed as a percentage of the complete exercise cycle. Brackets indicate the results of statistically significant multiple comparisons. External oblique (EO); erector spinae (ES); pectoralis major (PM); lower trapezius (LT); infraspinatus (ISP); biceps brachii (BB); and latissimus dorsi (LD). ^* p < 0.05; ^** p < 0.01.

Figure 9: Sequential timing of mean (±SE) peak muscle activation of the rotational exercise with Perfect·Pullup™ twisting handles is expressed as a percentage of the complete exercise cycle. Brackets indicate the results of statistically significant multiple comparisons. Abbreviations: external oblique (EO); erector spinae (ES); pectoralis major (PM); lower trapezius (LT); infraspinatus (ISP); biceps brachii (BB); and latissimus dorsi (LD). ^* p < 0.05; ^** p < 0.01.

Figure 10: The left column expresses raw electromyographic (EMG) activity for each muscle studied from a representative subject during the performance rotational exercise of a with Perfect·Pullup™ twisting handles. The right column displays the timing at which peak activation of a muscle occurred expressed as a percentage of the complete exercise cycle for the pull-up, chin-up, and rotational exercise with Perfect·Pullup™ twisting handles.

Discussion

To our knowledge, no one has normalized EMG activity for the pull-up/chin-up exercise to % MVIC, so the primary purpose of this study was to compare the magnitude of action of the 7 muscles across the 3 exercise conditions to determine if the pull-up/chin-up mode of exercise influenced muscle recruitment. On the basis of EMG activation data, DiGiovine et al. (¹²) classified relative muscular demand into 4 categories: low (<20% MVIC), moderate (20-40% MVIC), high (41-60% MVIC), and very high (>60% MVIC). Moreover, the scheme proposed by DiGiovine et al. agreed with other investigators who contended the threshold value for promoting muscle strength gains during therapeutic exercise required muscle activation greater than 50-60% MVIC (^1,4,28). Therefore, we chose to use the EMG signal amplitude (%MVIC) as a method to discuss the relative exercise intensity of the back, shoulder, arm, and trunk muscles when comparing a pull-up, chin-up, and rotational exercise using the Perfect·Pullup™ twisting handles.

The external oblique and erector spinae muscles provide muscular-based or “core stability” to the trunk under destabilizing external forces (³). Both the external oblique and erector spinae are classified as extrinsic muscular stabilizers because they attach partially or totally to the ribs and pelvis, which lie outside the vertebral column (²⁹). During the concentric phase the magnitude of EMG activity for the external oblique ranged from 31 ± 24% MVIC for the pull-up to 35 ± 24% MVIC for the chin-up, whereas the magnitude of EMG activity for the erector spinae ranged from 39 ± 31% MVIC for the pull-up to 41 ± 24% for the chin-up. Both external oblique and erector spinae muscles demonstrated moderate muscle activity (30-40% MVIC), but at a level of recruitment that would not be considered appropriate for strength training (^1,4,28). During the pull-up and chin-up exercise the external oblique and erector spinae assisted with stabilization of individual spinal segments within the vertebral column and provided a firm base for the superficial back muscles to move the upper extremities (²⁹). Therefore, on the basis of moderate muscle EMG signal amplitude we would consider the pull-up and chin-up exercise to be useful for promoting endurance training of these extrinsic muscular stabilizers of the trunk (¹⁵). Similar to this study results, other investigators have reported EMG signal amplitudes less than 40% MVIC and appropriate for endurance training when healthy subjects perform a variety of core rehabilitation exercises (abdominal crunch, supine bridge, lunge, bent-knee sit-up, active hip abduction in side lying, and quadruped arm and lower extremity lift) that activate the external oblique and erector spinae muscles (^15,18,19).

The 5 muscles of the shoulder and arm examined in the present study-pectoralis major, lower trapezius, infraspinatus, latissimus dorsi, and biceps brachii,-displayed high to very high muscle activation at a level that would be appropriate for strength training (^1,4,28). Activation of the pectoralis major ranged from 44 ± 27% MVIC for the pull-up to 57 ± 36% MVIC for the chin-up. Compared to values from this study, other investigators have recorded larger magnitudes of pectoralis major muscle activation during upper extremity weight-bearing exercise. For example, Townsend et al. (³⁷) found pectoralis major EMG signal amplitudes of 64 ± 63% MVIC and 84 ± 42% MVIC for a conventional push-up and seated press-up, respectively, whereas Cogley et al. (⁹) reported mean activation of 94 ± 8% MVIC during the concentric phase of a conventional push-up with hands at shoulder width. Similar to the pectoralis major, the lower fibers of the trapezius in this study displayed high muscle activation ranging from 45 ± 22% MVIC for the chin-up to 56 ± 21% MVIC for the pull-up. Other investigators have reported higher values for the lower trapezius using non-weight-bearing exercises. Moseley et al. (²⁷) and Ekstrom et al. (¹⁶) both recorded relatively high EMG activity (60 ± 22% MVIC and 63 ± 41% MVIC, respectively) during elevation of the arm in the scapular plane above 120° using dumbbell weights. Myers et al. (²⁸) reported lower trapezius EMG activity of 88 ± 51% MVIC during external rotation of the shoulder against rubber-tubing resistance with both the arm abducted and elbow simultaneously flexed to 90°. Lastly, Ekstrom et al. (¹⁶) recorded peak lower trapezius muscle activation (97 ± 16% MVIC) using a dumbbell load during prone arm raise overhead in line with the muscle fibers.

In this study, the infraspinatus muscle developed very high EMG signals that varied from 71 ± 52% MVIC for the rotational exercise using the Perfect·Pullup™ twisting handles to 79 ± 56% MVIC for the pull-up. Reinhold et al. (³²) studied common shoulder external rotation exercises and found side-lying external rotation at 0° abduction against dumbbell resistance produced the peak amount of activity (62 ± 13% MVIC) for the infraspinatus muscle. Earlier Townsend et al. (³⁷) reported infraspinatus muscle activation of 88 ± 25% MVIC when healthy subjects performed prone horizontal shoulder abduction in external rotation with the elbow fully extended against dumbbell resistance. Using a weight-bearing upper extremity exercise, Decker et al. (¹¹) discovered that a push-up with a plus activated the infraspinatus muscle at a level of 104 ± 54% MVIC. The plus movement occurred when the performer pushed his or her upper back toward the ceiling at the conclusion of the traditional up-phase of the push-up. The plus movement has been credited for additional recruitment of motor units within the serratus anterior and rotator cuff muscles (³⁷).

The biceps brachii produced very high EMG signals that varied from 78 ± 32% MVIC for the pull-up to 96 ± 34% MVIC for the chin-up. The chin-up exercise generated greater activation of the biceps brachii when compared to the pull-up because the forearm was supinated-palms facing the subject-when gripping the horizontal bar. Movement scientists have reported the biceps brachii muscle develops its maximal EMG signal (⁷) and peak torque production (³¹) with simultaneous elbow flexion and supination. Dynamic upper extremity activities requiring elbow flexion have shown a range of EMG activation signals from the biceps brachii. For example, Koukoubis et al. (²³). reported EMG signals were 33 ± 12% MVIC when experienced rock climbers performed finger-tip pull-ups with the forearms pronated. DiGiovine et al. (¹²) documented that skilled baseball pitchers activated the biceps brachii at a value of 44 ± 32% MVIC during the arm acceleration phase of throwing, whereas Ryu et al. (³⁴) recorded biceps brachii activation of 86% MVIC during the acceleration phase of the tennis forehand.

Of the 7 muscles examined in this study during the pull-up and chin-up procedures, the latissimus dorsi clearly developed the greatest EMG signals ranging from 117 ± 46% MVIC for the chin-up to 130 ± 53% MVIC for the rotational exercise using the Perfect·Pullup™ twisting handles. In comparison, during the arm acceleration phase of baseball pitching (¹²) the latissimus dorsi elicited 88 ± 53% MVIC activation. The push-up plus (¹¹) and press-up (³⁷), both traditional upper extremity weight-bearing exercises, were reported to generate 49 ± 25% MVIC and 55 ± 27% MVIC activation of the latissimus dorsi, respectively. Furthermore, Signorile et al. (³⁵) discovered that a wide grip anterior lat pull-down exercise normalized to a percentage of each subject's 10RM resulted in muscle activation of 108 ± 8% MVIC.

Data from this study revealed a performer of a pull-up, chin-up, or rotational exercise with the Perfect·Pullup™ twisting handles required about 100° of absolute elbow joint sagittal plane ROM to successfully complete the exercise. Using data supplied by Antinori et al. (²), we estimated 9 healthy subjects (5 men and 4 women) required about 120° of absolute elbow sagittal plane motion to execute a pull-up. The discrepancy between kinematic elbow data from the present study and those of Antinori et al. could be attributed to differences in methodology. Our data were captured with a computer-aided Vicon MX motion analysis system, whereas Antorini et al. used a videocamera connected to a videorecorder. Meanwhile, Morrey et al. (²⁶) reported a healthy person required a 100° “functional arc” (30-130°) to accomplish routine activities of daily living. Neither a pull-up nor chin-up would be considered a routine activity of daily living, but performance of the exercise is well within the expected normal sagittal plane elbow joint ROM.

We proposed the following rationale to justify a muscle's peak activation during the temporal sequence of the pull-up and chin-up cycle. All 7 muscles studied displayed EMG activity throughout the duration of the cycle for each exercise condition particularly during the concentric phase. The lower trapezius was activated early in the 3 pull-up/chin-up exercises at an average of 14.7% of the pull-up cycle and a magnitude of recruitment ranging from 45 ± 22% MVIC for the chin-up to 56 ± 21% MVIC for the pull-up. At the start of the concentric phase of the cycle, the scapula is positioned close to maximum upward rotation. With the subject's hand gripping the horizontal bar and the torso suspended from the pull-up and chin-up apparatus, the scapula is relatively fixed when compared to the trunk thus reversing the action of lower trapezius. The oblique inferomedial-directed muscle fiber orientation allows the lower trapezius to assist with elevation of the trunk. Judging by visual observation trunk elevation initiated the exercise cycle in preparation for shoulder adduction.

Overall, the second muscle recruited was the pectoralis major with peak activation occurring at 18.0% of the pull-up cycle and a magnitude of recruitment that ranged from 44 ± 27% MVIC for the pull-up to 57 ± 36% MVIC for the chin-up. According to Jenkins (²²), the sternocostal portion of the pectoralis major muscle extends the shoulder at the glenohumeral joint when the arm is in a flexed position, and, therefore, it is reasonable to assume why the pectoralis major is active early in the pull-up or chin-up cycle. This finding is in agreement with the report of Ricci et al. (³³) whereby the lower fibers (sternocostal) of the pectoralis major are activated early in the initial thrust of the pull-up cycle.

The erector spinae was recruited at 23.9% of the pull-up cycle with a magnitude of muscle activation that ranged from 39 ± 31% MVIC for the pull-up to 41 ± 24% for the chin-up. The erector spinae served 2 primary purposes during the exercise cycle. As the subject approached the bar during the concentric phase of the pull-up, back extension was necessary to permit the subject's head to clear the horizontal bar. Moreover, the erector spinae also assisted with stabilization of the pelvis which provided the latissimus dorsi a stable fixation site to pull from later in the exercise cycle.

The fourth muscle recruited was the infraspinatus at 29.6% of the cycle with a magnitude of recruitment that varied from 71 ± 52% MVIC for the rotational exercise using the Perfect·Pullup™ twisting handles to 79 ± 56% MVIC for the pull-up. Neumann (²⁹) commented that during a pull-up exercise the infraspinatus functioned as an arm adductor and extensor. Furthermore, the infraspinatus also provided dynamic shoulder stability to centralize the humeral head against the rigidity of the anterior structures of the shoulder (²⁹).

The biceps brachii was the fifth muscle recruited at 32.4% of the pull-up cycle and a magnitude of recruitment that ranged from 78 ± 32% MVIC for the pull-up to 96 ± 34% MVIC for the chin-up. With the forearm complex stabilized as the subject grasped the bar, the biceps brachii flexed the elbow with reverse action by pulling the humerus toward the fixed forearm. In contrast to data from this study, Ricci et al. (³³) commented that the biceps brachii was active during the initial thrust of the pull-up exercise, although the muscle was less active during the remainder of the pull-up cycle.

The penultimate muscle to reach peak muscle activation was the external oblique at 36.4% of the pull-up cycle and a magnitude of activation that varied from 31 ± 24% MVIC for the pull-up to 35 ± 24% MVIC for the chin-up. This muscle primarily functioned as a core stabilizer throughout all 3 exercises. We postulated the late activation displayed in the chin-up was a result of relative trunk flexion when subjects elevated their nose above the horizontal bar near the completion of the concentric phase of the exercise cycle.

The last muscle to reach peak activation was the latissimus dorsi at 37.3% of the pull-up cycle with a magnitude of activation ranging from 117 ± 46% MVIC for the chin-up to 130 ± 53% MVIC for the rotational exercise using the Perfect·Pullup™ twisting handles. This muscle functioned to complete trunk elevation through reverse activation, which is made possible by pelvic stabilization from the erector spinae and external oblique. The latissimus dorsi also completed adduction and hyperextension of the shoulder to finish the concentric phase of the exercise cycle. Results from this study are consistent with the findings by Ricci et al. (³³) whereby the most pronounced activity of the latissimus dorsi is at the completion of the concentric phase of the pull-up or chin-up cycle.

We identified an assortment of limitations in our study. As with all motion analysis, reflective markers are designed to measure skeletal kinematics; however, they are indirectly measuring joint angles via skin movement. Soft tissue artifact was inevitable because the only way to truly measure skeletal kinematics is through the use of needle penetration into bone. In addition, as with all EMG systems there is the potential for skin artifact and crosstalk. We minimized this by using standard EMG electrode placement as described by Cram and Kassman (¹⁰). The skin was wiped with alcohol, and shaved if needed, to clean the skin of oil and debris. Tape reinforced the EMG electrodes preventing the sensor from falling off the desired muscle motor point. Additionally, we used double-differential electrodes vs. single-differential electrodes which served to minimize the potential for crosstalk. To increase generalizability, future research should include subjects with varying age, activity level, BMI, and gender. Future research should examine joint forces to confirm the manufacturer's claim that joint contact forces are minimized because of the Perfect·Pullup's™ twisting handles allowing the arm to move in a natural motion to improve strength, balance, flexibility, and endurance, while reducing the risk of muscle strain and injury.

Practical Applications

The pull-up or chin-up exercise can be used to train muscles used regularly in activities of daily living that involve pulling motions (shoulder extension combined with elbow flexion) of the upper extremity. Back, shoulder, arm, and trunk muscle activation patterns and elbow joint kinematic information are of interest to physical therapists and other fitness professionals when prescribing variations of the pull-up and chin-up exercise to their clients. Excessive emphasis placed on pressing movements in the upper extremity may increase the likelihood of shoulder complex trauma such as rotator cuff injury or strain. To prevent this, pulling movements should be incorporated in rehabilitation programs to achieve a balance between pressing and pulling performance. The 7 muscles of the back, shoulder, arm, and trunk examined in the present study during the pull-up and chin-up displayed high to very high muscle activation at a level that would be appropriate for strength training (^1,4,28). Based upon normalized EMG activity alone, the rotational exercise using Perfect·Pullup™ twisting handgrips is not preferable to a conventional pull-up or chin-up when the goal is to produce muscle strengthening of back, shoulder, arm, and trunk muscles.

Acknowledgments

Funding for this project came from the Mayo Program in Physical Therapy, College of Medicine, Mayo Clinic, Rochester, MN, USA. The authors disclose they have no professional relationship with the company that manufactures the Perfect·Pullup™ device. The results of this study do not constitute endorsement of the Perfect·Pullup™ device by the authors or the National Strength and Conditioning Association. The authors wish to thank Mr. Mike Ilse for constructing the free-standing pull-up and chin-up apparatus.