Trunk muscle co-activation of several muscles is considered necessary in achieving adequate spinal stability to prevent and treat low back injury [1]. Common exercise recommendations from health professionals include trunk exercises to prevent and treat low back injuries. Knowing the trunk muscle activation levels during exercises is important in the prescription and design of exercise programs that aim to increase the training intensity over time (progressive resistance model). Previous research has documented trunk muscle EMG during various exercises designed to train the trunk musculature and during functional activities [2–7]. Ng et al [7] found that abdominal and trunk muscles not only produce torque but also maintain spinal posture and stability during axial rotation exertions. Vera-Garcia et al [8] showed that performing curl-ups on a labile (moveable) surface changes the muscle activity amplitude required to perform the movement. Increases were greatest in the external oblique muscles. Mori [9] documented the trunk muscle activity during a variety of trunk muscle exercises on a Swiss ball. However, comparisons in muscle activity were not made with ground based exercises (no Swiss ball present), therefore, the influence of a Swiss ball on trunk muscle activity compared with ground based bridging is not known.

The importance of trunk muscles in providing adequate spine stability is well established and the role of trunk muscles during a variety of tasks has been well documented. Swiss balls are a common addition to trunk muscle exercises. In fitness centres and rehabilitation centres, Swiss balls are often touted as being superior to ground based exercises in their ability to recruit trunk muscles (rectus abdominis, external oblique, internal oblique, erector spinae). Considering the ubiquity of Swiss balls, one research question was posed: Does the addition of a Swiss ball to bridging exercises influence trunk muscle activity?

The implications of this study are twofold: 1. Modifying trunk muscle activity could be important in the safety and efficacy of rehabilitation exercises when a low level of trunk muscle activity is desired; and 2. Identifying exercises which maximally activate the trunk muscles may make it possible to develop an efficient and less time consuming general strength program that conditions the trunk muscles.

Table 1 depicts the muscle activation levels across exercises. The addition of an exercise ball did not influence the muscle activity in the Internal Oblique (Figure 6) in both bridging exercises. During the prone bridge the addition of an exercise ball resulted in increased myoelectric activity in the rectus abdominis and external oblique (Figure 7 and Figure 8). The exercise ball did not influence the Rectus Abominis or the External Oblique muscle activity during a supine bridge. The addition of an exercise ball did not influence the Erector Spinae (Figure 9) activity during the supine bridge or the prone bridge.

Exercise	Pr Br Floor	Pr Br Ball	Side Bridge	Su Br Ball	Su Br Floor
Column #	1	2	3	4	5
IO Avg	29.5	39.8	42.5	19.7	12.3
Stdev	18.8	23.9	25.2	15.8	9.5
Different From*	-	4,5	4,5	2,3	2,3
RA Avg	26.6	55.9	24.4	6.05	5.84
Stdev	11.1	28.8	11.7	1.3	1.1
Different From*	2,4,5	1,3,4,5	2,4,5	1,2,3	1,2,3
EO Avg	44.6	62.5	46.1	10.6	7.8
Stdev	14.8	26.3	15.4	5.7	6.3
Different From*	2,4,5	1,3,4,5	2,4,5	1,2,3	1,2,3
ES Avg	4.98	5.00	25.7	27.4	25.01
Stdev	1.05	1.46	11.3	7.56	9.02
Different From*	3,4,5	3,4,5	1,2	1,2	1,2

* This row indicates which other exercise the myoelectric signal for the respective column is statistically different (p < .05) from. (Avg = Average muscle activity in % MVC, Stdev = Standard deviation, IO = Internal Oblique, RA = Rectus Abdominis, EO = External Oblique, ES = Erector Spinae. Pr Br = Prone Bridge, Su Br = Supine Bridge.

The side bridge produced the highest myoelectric activity in both the Internal Oblique and Erector Spinae. The prone bridge with arms on a Swiss ball produced the highest myoelectric activity in both the Rectus Abdominis and External Oblique.

The primary aim of this study was to determine if performing bridging exercises on a Swiss ball rather than the ground resulted in increases in trunk muscle activity. A blanket statement that a more labile surface (the Swiss ball) increases trunk muscle activity cannot be made. The influence of surface stability on muscle activity appears to be muscle and exercise dependent. For example, during the prone bridge the primary mover (the rectus abdominis resisting trunk extension) was the most influenced by the addition of a Swiss ball. Conversely, during a supine bridge, one of the primary movers, the erector spinae, was not influenced by surface stability (the other primary mover, the Gluteus Maximus was not studied). It may be argued that the increase in activation levels of the external oblique and the rectus abominis during prone bridging appear to be caused by decreases in surface stability and not different biomechanical demands due to the body's position relative to gravity. This finding agrees with the Vera-Garcia et al [8] study that investigated trunk curl up exercises. While there were differences in the body's position relative to gravity between the ground exercise and the ball exercise during prone bridging, performing the bridge on a ball finds the participant in a more vertical position. This suggests there is less force creating a trunk extension movement (i.e. gravity attempts to increase lordosis which is resisted by muscle activity) due to the fact that the centre of mass of the trunk and head segment would be closer to the axis for trunk extension. Therefore less muscle activity may have been generated to resist this torque (compared with the ground based bridge) and more muscle activity may have been required to produce secondary spinal stabilization due to the labile surface.

An important observation from all exercise tasks was the large variability in muscle activity between subjects that can greatly influence the interpretation of these results. Figure 10 illustrates an example of this variability. Figure 6 shows the average activity in the internal oblique muscle during prone bridging on and off a Swiss ball for one repetition from each subject. This indicates that some subjects showed large changes in muscle activity while others showed minimal changes when modifications to the exercise tasks were made. It is possible that some subjects volitionally contracted their trunk muscles to provide stability while some others may have not. It is possible that individuals may be able to influence their trunk muscle activity either through verbal encouragement, or feedback produced by electromyography. Additionally, the variability may have been due to slight variations in participant posture or task performance. While exercise standardization was sought through verbal correction of form, it is possible that differences in task performance between the subjects still occurred. Further research may wish to determine the influence of electromyographic feedback on influencing the trunk activation levels during resistance exercise. This may decrease the variability between subjects.

This study is limited because it only measured the trunk muscle activity during the various exercises. No measurements were made nor a biomechanical model constructed to determine the compressive or shear loading on the spine during the tasks. This type of kinematic, and subsequently force data, is optimal when determining the safety and tissue loading properties of various movements. Also, this study did not quantitatively measure spinal posture. This may influence the muscle activation levels. While a consistent spinal posture was encouraged and monitored by the experimenters, it is also possible that minor differences in spine posture did occur. Monitoring spinal posture via a kinematic measurement system (eg. Electromagnetic tracking) may be important in future work.

While increased trunk muscle activation can result in higher compressive loads on the spine [12], is this amount of trunk muscle activity necessarily increasing the risk of injury? We are unable to say if increases in activity levels are due to biomechanical demands, or if they are due to motor control decisions that permit enhancements in spine stability that may decrease the risk of injury. Conversely, if an exercise modification results in decreases in muscle activity, is this always beneficial in terms of injury prevention? Is it possible that subjects who lower their muscle activation levels during tasks predispose themselves to a "buckling" type injury because sufficient spinal stability is not created with the current amount of trunk muscle activity [13]? It is important to note that this study measured trunk muscle activity in an athletic young homogenous population. Sedentary individuals or those with trunk or lower leg injuries may show different results.

Differences in trunk muscle activity are seen with the addition of a Swiss ball to bridging exercises. It cannot be concluded that these differences are solely due to changes in surface stability due to the different biomechanical demands of the exercises. Future research should control for exercise posture to determine how surface stability influences muscle activity.