Secondary Logo

Journal Logo

Original Research

Lumbopelvic-Hip Complex and Scapular Stabilizing Muscle Activations During Full-Body Exercises With and Without Resistance Bands

Wasserberger, Kyle W.; Downs, Jessica L.; Barfield, Jeff W.; Williams, Teasie K.; Oliver, Gretchen D.

Author Information
Journal of Strength and Conditioning Research: October 2020 - Volume 34 - Issue 10 - p 2840-2848
doi: 10.1519/JSC.0000000000002842
  • Free

Abstract

Introduction

Interdependent segments working in a proximal to distal manner are commonly known as the kinetic chain (3). Proper sequencing of the kinetic chain is imperative when performing overhead throwing tasks (3). Essentially, when performing an overhead throwing task, the lower extremity and lumbopelvic-hip complex (LPHC) must not only be able to exhibit efficient movement but also adequate stability for applicable energy transfer to the upper extremity of the shoulder, elbow, wrist, hand, and ball (3,12,15). The importance of effectively using the body as a kinetic chain has been associated with not only maximizing overhead athletic performance, but also reducing injury susceptibility (3,12–14,26,27). Shoulder rehabilitation programs have begun to implement the kinetic chain theory for a broader approach versus sole focus on isolated structures (15). Specifically examining shoulder rehabilitation protocols, resistance tubing/bands are a popular modality.

In the worlds of baseball and softball, resistance tubing/bands are a standard within strength and conditioning programs for prethrowing exercise. Traditionally, resistance band use has focused on the rotator cuff and surrounding shoulder musculature. However, isolating resistance band work at the shoulder does not adhere to the established theory of proximal stability for distal mobility (7,11,13). Therefore, including exercises that incorporate the interactions between proximal and distal portions of the kinetic chain may prove beneficial (8,17,22). Thus, it was the purpose of this study to examine activations of LPHC and scapular stabilizing musculature during 4 full-body exercises with and without the use of a resistance band. Specifically, the examination of the LPHC and scapular stabilizing musculature (bilateral external abdominal oblique [EO] and gluteus medius [GM], dominant side multifidus [MT], latissimus dorsi [LD], lower trapezius [LT], and upper trapezius [UT]) during 4 total-body exercises with and without resistance bands. It was hypothesized that those total-body exercises with the use of the resistant band would yield greater muscle activation than those without the use of the resistant band.

Methods

Experimental Approach to the Problem

The aim of this study was to quantitatively examine muscle activations of the LPHC and scapular stabilizing musculature (bilateral EO and GM, dominant side MT, LD, LT, and UT) during 4 total-body exercises with and without resistance bands. Muscle activation data were collected through surface electromyography (EMG). All EMG data were normalized as a percent of the subject's maximum voluntary isometric contraction (%MVIC) and normalized all subjects to be right-hand/right-leg dominant (only one subject was left-hand dominant).

Subjects

Twenty healthy active individuals (mean ± SD: 174.39 ± 1.58 cm; 74.10 ± 1.75 kg; 21.85 ± 1.13 years), regardless of sex, volunteered. Healthy was determined by having no history of upper- or lower-extremity injury within the past 6 months. Active was defined as 30 minutes of physical activity most days of the week. The Institutional Review Board of Auburn University approved all testing protocols. Before data collection, all testing procedures were explained to each subject, and written informed consent was obtained.

Procedures

On the day of testing, subjects reported to the Sports Medicine and Movement Laboratory before engaging in any vigorous physical activity or resistance training. Electromyography electrode placement was identified through palpation of the muscle belly. Before electrode placement, all musculature (EO, GM, MT, LD, LT, and UT) were abraded and cleaned using 70% isopropyl alcohol. Silver chloride 4-cm surface electrodes (Bio Protech Inc., Tustin, CA, USA) were placed over the muscle bellies parallel to the muscle fibers, with an interelectrode distance of 10 mm using previously established standards (1,4,20). The use of surface electrodes was chosen because they have been deemed to be a noninvasive technique that is able to reliably detect surface muscle activity (1).

External oblique electrode placement was approximated to be midway between the umbilicus and the anterior superior iliac spine (4). Electrode placement for the GM was identified as the proximal third of the distance from the iliac crest and greater trochanter (4). Multifidus electrode placement was 2 cm lateral to the lumbosacral junction (6). Latissimus dorsi electrode placement was oblique upward and laterally, below the inferior tip of the scapula, approximately half the distance between the spine and lateral torso (1). Lower trapezius electrode placement was obliquely upward and laterally midway between the spine of the scapula and vertebral border of the scapula and seventh thoracic spinous process (1). Upper trapezius electrode placement was identified as the midpoint of the seventh cervical vertebra and the acromion process (4).

After electrode placement, manual muscle testing (MMT) was performed in attempt to determine baseline MVIC to which all EMG data were normalized (10). Although the exercises under investigation were not purely isometric, normalization using MVIC was chosen due the nonballistic actions of the abdominal and scapular musculature during the selected exercises. Given the dynamic, yet not ballistic, nature of the selected exercises, normalization using MVIC provided the best compromise between specificity, repeatability, and protocol simplicity (19,24,25). Three MMTs lasting 5 seconds each were performed for each muscle. In efforts to maximize reliability, all electrode placements and MMTs were performed by one investigator. Manual muscle testing and EMG data were collected using a Noraxon Myopac 1,400 L 8-channel amplifier (Noraxon USA, Inc., Scottsdale, AZ, USA).

Subjects performed 4 exercises with and without the use of resistance bands (Jaeger Sports, Los Angeles, CA, USA) for a total of 8 exercises. The chosen exercises were included because of their full-body nature and were as follows: airplane (AP), overhead squat with “Y” hold (OH), lunge with “W” hold (LW), and single-leg Romanian deadlift (RDL) with horizontal row and external rotation. Exercise order for all 8 exercises was randomized for each subject. The anchor height of the resistance band was standardized to the subject's elbow height. Subjects were verbally cued to maintain band tension throughout each exercise. Before each trial, an investigator explained and demonstrated the exercise. Verbal cues were standardized for all subjects and each subject was given time to practice, if needed. All exercises were performed at a self-selected pace and a significant rest period between exercises was given to minimize the effects of fatigue.

Exercise Selection

For all exercises using the resistance bands, the subject stood at a minimum distance from the band anchor to eliminate slack from the bands but not produce any further tension. While performing the exercise using the bands, band tension was maintained throughout the duration of the exercise. Five repetitions were performed for the OH, LW, and RDL, whereas 3 repetitions were performed bilaterally for the AP, for both conditions.

Airplane

Subjects stood facing the resistance anchor with shoulders forward flexed and elbows extended. Next, they performed 90° shoulder horizontal abduction and then a reverse lunge with the leg ipsilateral to the dominant arm. The lunge was limited to 75% of subject's height. Subjects were verbally cued to maintain the horizontally abducted position throughout the exercise and to not let the ipsilateral knee touch the ground. Once the subject was in the appropriate lunge position, they performed alternating lateral trunk flexion while maintaining a straight spine. Subjects were verbally cued to not “break” their spine and to only go through the range of motion that a straight spine allowed (Figure 1).

Figure 1.
Figure 1.:
A= start of AP; B= reverse lunge of AP; C= lateral trunk flexion of AP; D= end of AP; AP = airplane.

Overhead Squat With Y

Subjects stood facing the resistance anchor with shoulders forward flexed and elbows extended. From this position, subjects performed shoulder abduction and shoulder horizontal abduction to form the traditional overhead “Y” position. Subjects were verbally cued to form a “Y” overhead at approximately 45°, where 0° was parallel to the ground and 90° was perpendicular to the ground, and to keep elbows fully extended. The subject then performed an overhead squat while maintaining the “Y” position. Subjects were verbally cued to reach maximum depth possible without compromising upper-body position. After reaching peak voluntary depth, subjects performed hip and knee extension to return to the standing overhead “Y” position (Figure 2).

Figure 2.
Figure 2.:
A= start of the overhead squat with Y exercise; B= end of the overhead squat with Y exercise.

Lunge With W

Subjects stood facing the resistance anchor with shoulders flexed and elbows extended. Subjects then performed elbow flexion, scapular retraction, scapular depression, and humeral external rotation to form a “W” with their arms. Once in the W position, they performed a reverse lunge (limited to 75% of height) with the leg ipsilateral to the dominant arm. Subjects were verbally cued to maintain the “W” position in their upper body throughout the lunge and to not let the ipsilateral knee touch the ground. After maximum contralateral knee flexion was achieved, subjects extended their contralateral knee while pushing off the ground with the ipsilateral leg to return to the starting position (Figure 3).

Figure 3.
Figure 3.:
A= start of LW; B= end of LW; LW = lunge with “W” hold.

Single Leg Romanian Deadlift With Horizontal Row and External Rotation

Subjects stood on the leg opposite the dominant arm with the torso flexed and ipsilateral leg extended. Subjects then performed torso extension, shoulder horizontal abduction, elbow flexion, and humeral external rotation while keeping the ipsilateral leg off the ground. Subjects were verbally cued to move in one fluid motion from the parallel to upright position. Once the torso was fully upright and throwing arm achieved peak external rotation, the subject was instructed to eccentrically control the return to the starting parallel position (Figure 4).

Figure 4.
Figure 4.:
A= start of RDL; B= finish of RDL; RDL = Romanian deadlift.

Statistical Analyses

All statistical analyses were conducted using IBM SPSS Statistics 22 software (IBM Corp., Armonk, NY, USA) with an alpha level set a priori at α = 0.05. Before analysis, Shapiro-Wilk tests of normality were run and determined the data were non-normal. The data were then logarithmically transformed and revealed to be normal. All subsequent statistical tests were conducted on the transformed data. Analyses were performed with 20 people for the EO, MT, LD, LT, and UT. Due to equipment malfunction, analyses were performed with 19 people for the GM. A within-subjects multivariate analysis of variance (MANOVA) was used to determine whether muscle activation varied by the interaction between band usage and exercise choice. To follow-up the within-subjects MANOVA, repeated-measures ANOVAs were used. The Bonferroni inequality was applied for all follow-up tests to adjust for multiple comparisons, setting test-wise error at α = 0.006. Because the assumption of sphericity was not met for the left GM (W < 0.01, p < 0.001) and MT (W = 0.21, p < 0.001), the Greenhouse-Geisser correction was applied to these variables. Bonferroni pairwise comparisons were used to determine how significant muscle activation differences varied across band usage and exercise choice. Ranges for muscle activations were defined as low (<20% MVIC), moderate (21–40% MVIC), high (41–60% MVIC), and very high (>61% MVIC) (5,7,18,21).

Results

We found a significant difference on muscle activation based on the interaction between band usage and exercise choice (Λ = 0.276, F24, 136.92 = 3.19, p < 0.001). About 37% of the variance in muscle activation was influenced by the interaction between band usage and exercise choice (ω2 = 0.37).

There was significant difference in right EO (F3, 54 = 4.96, p = 0.004) and UT activation (F3, 54 = 10.82, p < 0.001). About 22% of the variance in right EO activation (η2 = 0.22) and 38% of the variance in UT activation (η2 = 0.38) was explained by the interaction between band usage and exercise choice.

In terms of band usage, the UT (p < 0.001) differed between band and no band conditions, whereas the right EO exhibited no significant differences between band conditions (p = 0.185). For exercises, significance was found between the AP and LW (p < 0.001), LW and OH (p < 0.001), and RDL and OH (p = 0.001) for the UT. Mean values and SEMs for muscle activity for each exercise with and without resistance bands are presented in Figures 5–8.

Figure 5.
Figure 5.:
Mean muscle activation (airplane). MVIC = maximum voluntary isometric contraction; EO = external abdominal oblique; GM = gluteus medius; MT= multifidus; LD = latissimus dorsi; LT = lower trapezius; UT = upper trapezius. <20% MVIC = low; 21-40% MVIC = moderate; 41-60% MVIC = high; >60% MVIC = very high. *indicated significant difference between band and no band condition (p < 0.05).
Figure 6.
Figure 6.:
Mean muscle activation (W lunge). MVIC = maximum voluntary isometric contraction; EO = external abdominal oblique; GM = gluteus medius; MT= multifidus; LD = latissimus dorsi; LT = lower trapezius; UT = upper trapezius. <20% MVIC = low; 21-40% MVIC = moderate; 41-60% MVIC = high; >60% MVIC = very high. *indicated significant difference between band and no band condition (p < 0.05).
Figure 7.
Figure 7.:
Mean muscle activation (RDL). RDL = Romanian deadlift; MVIC = maximum voluntary isometric contraction; EO = external abdominal oblique; GM = gluteus medius; MT= multifidus; LD = latissimus dorsi; LT = lower trapezius; UT = upper trapezius. <20% MVIC = low; 21-40% MVIC = moderate; 41-60% MVIC = high; >60% MVIC = very high. *indicated significant difference between band and no band condition (p < 0.05).
Figure 8.
Figure 8.:
Mean muscle activation (overhead squat). MVIC = maximum voluntary isometric contraction; EO = external abdominal oblique; GM = gluteus medius; MT= multifidus; LD = latissimus dorsi; LT = lower trapezius; UT = upper trapezius. <20% MVIC = low; 21-40% MVIC = moderate; 41-60% MVIC = high; >60% MVIC = very high. *indicated significant difference between band and no band condition (p<0.05).

Discussion

It was the purpose of the current study to use surface EMG in the examination of scapular and LPHC muscle activations during 4 full-body exercises with and without the use of a resistance band. The results of this study partially support the initial hypothesis that adding a resistance band would increase mean muscle activation in scapular and LPHC stabilizing musculature. A combination of resistance band usage with exercise choice significantly affected muscle activation in the current study.

Resistance bands have gained popularity in the baseball and softball communities as a part of prethrowing routines and strength and conditioning programs. Traditional protocols have included the IYT series and rotator cuff strengthening exercises such as internal and external rotation drills at 0 and 90° of shoulder abduction. Commercially, these bands have been marketed to target musculature specific to the shoulder. However, it is well documented that shoulder function during overhead throwing also requires strength and coordination of the LPHC and lower extremities (3,7,12,13,19–22,26,27,29). Consequently, current practices that involve the use of resistance bands that target musculature local to the throwing extremity do not train the entire kinetic chain or promote proximal to distal sequencing. To achieve full benefit of scapular and LPHC stabilizer muscle activation, one should use a combination of resistance band with dynamic, full-body exercises.

For our selected exercises, adding a resistance band was particularly effective at increasing activation in the scapular stabilizing musculature. Specifically, UT activation increased with the introduction of a resistance band during every exercise. All exercises, regardless of resistance band usage, exhibited moderate to very high muscle activation among the scapular stabilizers. When the selected exercises were accompanied with the band, the LT showed moderate activation in the AP, LW, and RDL, whereas exhibiting high activation with the OH. Because the LT is known to be an important scapular stabilizer, previous LT engagement before any overhand throwing activity could help prevent scapular winging (16,27).

Special attention has been given in past studies regarding the relative balance between LT and UT activation levels during overhead movements and its relationship with pathological conditions of the throwing shoulder (2,7,26,28). Specifically, high UT and low LT activation levels have been linked to several pathological conditions in overhead athletes (2,28). A main cause of shoulder impingement, which account for upward of 50% of shoulder complaints during doctor visits, would be muscle imbalance between the scapular stabilizers (23). The current study revealed significantly higher UT activation while performing the AP during the no band condition and the OH during the band condition. Thus, the authors recommend that extra consideration be used before prescribing these exercises during shoulder rehabilitation programs because they may exacerbate problematic activation patterns in pathological populations such as those with scapular dyskinesis or shoulder impingement.

When considering the LD, which provides indirect scapular stabilization through its actions at the glenohumeral joint, all banded exercises exhibited greater activation when compared with their nonbanded counterpart. The banded AP and LW provided high muscle activation for the LD, whereas all other banded exercises indicated moderate LD activation. Previous research has noted the LD involvement in the arm acceleration phase of throwing (9,16). Although LD injuries are rare among overhand pitchers, activating the LD before it undergoes the stress of throwing could play a role in maintaining a healthy kinetic chain, causing the large LD to accept the stabilizing stresses placed on proximal upper extremity during the overhand throw.

Conversely, adding a resistance band in combination with the selected exercises for this study was only effective at increasing LPHC muscle activation at the right EO. Specifically, adding a resistance band failed to produce significant increases in bilateral GM and left EO activity during the AP, RDL, and OH exercises. Although limited increases in LPHC muscle activation were found with resistance band usage, all exercises exhibited moderate to very high muscle activation for the left GM, whereas the OH and RDL exercises showed moderate activation for the right GM. Yamanouchi (29) and Burkhart et al. (2) state the importance of hip abductor strength when trying to provide stabilization when throwing, along with how weakness in these muscles can cause breaks within the kinetic chain leading to increased stress accumulation up the chain, thus increasing one's risk of injury. Consequently, upper-body resistance band exercises that also elicit substantial GM activity may provide additional training benefits to overhead throwing athletes. The strength and conditioning professional would be well advised to consider using resistance bands in combination with dynamic exercises to increase both scapular and LPHC stabilizing muscle activation.

In summary, exercises that train proper energy transfer through the kinetic chain are imperative for overhead throwing athletes. Inefficiencies in proximal segments of the kinetic chain are often concurrent with pathomechanic adaptations such as scapular dyskinesis (2,7,13,14,27). These adaptations may cause distal segments to overwork in efforts to maintain optimal performance. Chronic overworking of more distal segments may increase the risk of injury to the musculature and connective tissue surrounding those segments over time (2,3,8,12,13). Furthermore, given that pathomechanic adaptations are often associated with muscle activation abnormalities (13,14,28), exercises that can maximize the ability of the kinetic chain to efficiently transfer forces from proximal to distal segments are essential for athletes and coaches who wish to enhance their overall performance and current level of play (2,3,8,12,14,15,27). It is, therefore, the authors' suggestion that those in charge of training overhead throwing athletes use a full-body, kinetic chain–oriented training approach.

It should be noted that there are several limitations of this study. First, as a result of the utilization of surface EMG, movement artifact may have occurred due to the dynamic nature of the exercises involved in this study. Moreover, although the authors believe that normalization using MVIC was appropriate for the current study, it should be noted that the method of using isometric contractions for normalization of muscle activity during movement has been contended. In addition, although care was taken for proper electrode placement, muscle crosstalk may have occurred. A further limitation was the use of healthy participants. Thus, muscle activations observed in this healthy population may not be representative of other clinical populations. Future studies should involve subjects from various pathological clinical populations in an attempt to generalize these findings to subjects with dysfunctions common to overhead throwing athletes. Also, because neither EO reached moderate activation in either condition, future studies should investigate exercises that produce moderate activation of the EO due to the key role the core plays in the kinetic chain (12).

Practical Applications

The results of this study add to the growing body of descriptive EMG data specific to overhead throwing exercises. Through continuing quantification, practitioners and coaches will be able to make increasingly informed decisions when designing training programs unique to each throwing athlete. By using the current findings in conjunction with other athlete-specific training considerations, the trained exercise or coaching professional will be able to increase performance and reduce risk of injury through a full-body training approach. Given that it elicited the highest LD activation in concert with being the only exercise to achieve higher LT than UT activity, the authors especially recommend the LW for throwing athletes. Coaches will be able to take these exercises and easily transfer them to the field as part of their prethrowing routine.

Acknowledgments

The authors have no conflicts of interest to disclose.

References

1. Basmajian JV, Deluca CJ. Apparatus, detection, and recording techniques. In: Muscles Alive, Their Functions Revealed by Electromyography. Baltimore, MD: Williams and Wilkins, 1985.
2. Burkhart SS, Morgan CD, Kibler WB. The disabled throwing shoudler: Spectrum of pathology. Part III: The SICK scapula, scapular dyskinesis, the kinetic chain, and rehabilitation. Arthroscopy 19: 641–661, 2003.
3. Chu SK, Jayabalan P, Kibler WB, Press J. The kinetic chain revisited: New concepts on throwing mechanics and injury. PM R 8: S69–S77, 2016.
4. Cram JR, Kasman GS, Holtz J. Electrode placement. In: Introduction to Surface Electromyography. Gaithersburg, MD: Aspen Publishers, 1998.
5. Digiovine NM, Jobe FW, Pink M, Perry J. An electromyographic analysis of the upper extremity in pitching. J Shoulder Elbow Surg 1: 15–25, 1992.
6. Ekstrom R, Soderberg G, Donatelli R. Normalization procedures using maximum voluntary isometric contraction for the serratus anterior and trapezius muscles during surface EMG analysis. J Electromyogr Kinesiol 15: 418–428, 2005.
7. Escamilla RF, Andrews JR. Shoulder muscle recruitment patterns and related biomechanics during upper extremity sports. Sports Med 39: 569–590, 2009.
8. Gilmer GG, Washington JK, Dugas J, Andrews J, Oliver GD. The role of lumbopelvic-hip complex stability in softball throwing mechanics. J Sport Rehabil 28: 196–204, 2019.
9. Jobe FW, Moynes DR, Tibone JE, Perry J. An EMG analysis of the shoulder in pitching. A second report. Am J Sports Med 12: 218–220, 1984.
10. Kendall F, McCreary EK, Provance PG, Rodgers MM, Romani W. Muscles: Testing and Function. Baltimore, MD: Lippincott Williams & Wilkins, 1993.
11. Kendall FP, McCreary EK, Provance PG, Rodger MM, Romani WA. Fundamental concepts. In: Muscles: Testing and Function With Posture and Pain. Baltimore, MD: Lippincott Williams and Wilkins, 2005. p. 31.
12. Kibler WB, Press J, Sciascia A. The role of core stability in athletic function. Sports Med 36: 189–198, 2006.
13. Kibler WB, Wilkes T, Sciascia A. Mechanics and pathomechanics in the overhead athlete. Clin Sports Med 32: 637–651, 2013.
14. Lintner D, Noonan TJ, Kibler WB. Injury patterns and biomechanics of the athlete's shoulder. Clin Sports Med 27: 527–551, 2008.
15. McMullen J, Uhl TL. A kinetic chain approach for shoulder rehabilitation. J Athl Train 35: 329–337, 2000.
16. Mehdi SK, Frangiamore SJ, Schickendantz MS. Latissimus dorsi and teres major injuries in major league baseball pitchers: A systematic review. Am J Orthop 45: 163–167, 2016.
17. Oliver GD. Relationship between gluteal muscle activation and upper extremity kinematics and kinetics in softball position players. Med Biol Eng Comput 52: 265–270, 2014.
18. Oliver GD, Plummer HA, Gascon SS. Electromyographic analysis of traditional and kinetic chain exercises for dynamic shoulder movements. J Strength Cond Res 30: 3146–3154, 2016.
19. Oliver GD, Sola M, Dougherty C, Huddleston S. Quantitative examination of upper and lower extremity muscle activation during common shoulder rehabilitation exercises using the bodyblade. J Strength Cond Res 27: 2509–2517, 2013.
20. Oliver GD, Stone AJ, Plummer H. Electromyographic examination of selected muscle activation during isometric core exercises. Clin J Sport Med 20: 452–457, 2010.
21. Oliver GD, Washington JK, Barfield JW, Gascon SS, Gilmer GG. Quantitative analysis of proximal and distal kinetic chain musculature during dynamic exercises. J Strength Cond Res 32: 1545–1553, 2018.
22. Oliver GD, Weimar WH, Plummer HA. Gluteus medius and scapula muscle activations in youth baseball pitchers. J Strength Cond Res 29: 1494–1499, 2015.
23. Page P. Shoulder muscle imbalance and subacromial impingement syndrome in overhead athletes. Int J Sports Phys Ther 6: 51–58, 2011.
24. Remaley DT, Fincham B, McCullough B, Davis K, Nofsinger C, Armstrong C, et al. Surface electromyography of the forearm musculature during the windmill softball pitch. Orthop J Sports Med 3: 2325967114566796, 2015.
25. Rojas IL, Provencher MT, Bhatia S, Foucher KC, Bach BR Jr, Romeo AA, et al. Biceps activity during windmill softball pitching: Injury implications and comparison with overhand throwing. Am J Sports Med 37: 558–565, 2009.
26. Sciascia A, Cromwell R. Kinetic chain rehabilitation: A theoretical framework. Rehabil Res Pract 853037: 1–9, 2012.
27. Seroyer ST, Nho SJ, Bach BR, Bush-Joseph CA. The kinetic chain in overhand pitching: Its potential role for performance enhancement and injury prevention. Sports Health 2: 135–146, 2010.
28. Wilk K, Williams RA, Dugas JR, Cain EL, Andrews JR. Current concepts in the assessment and rehabilitation of the thrower's shoulder. Oper Tech Sports Med 24: 170–180, 2016.
29. Yamanouchi T. EMG analysis of the lower extremities during pitching in high-school baseball. Kurume Med J 45: 21–25, 1998.
Keywords:

baseball; EMG; kinetic chain; shoulder; softball

© 2018 National Strength and Conditioning Association