Scand J Med Sci Sports 2014: ••: ••–•• doi: 10.1111/sms.12190

© 2014 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

Muscle functional MRI analysis of trunk muscle recruitment during extension exercises in asymptomatic individuals E. M. D. De Ridder1, J. O. Van Oosterwijck1, A. Vleeming2, G. G. Vanderstraeten1, L. A. Danneels1 1

Department of Rehabilitation Sciences and Physical Therapy, Faculty of Medicine and Health Sciences, Ghent University, Belgium, Department of Anatomy, University of New England College of Osteopathic Medicine, Biddeford, Maine, USA Corresponding author: Lieven A. Danneels, Ghent University Hospital, Department of Rehabilitation Sciences and Physical Therapy, De Pintelaan 185, 3B3, B-9000 Ghent, Belgium. Tel: +32 9 332 26 35, Fax: +32 9 332 38 11, E-mail: [email protected] 2

Accepted for publication 13 January 2014

The present study examined the activity levels of the thoracic and lumbar extensor muscles during different extension exercise modalities in healthy individuals. Therefore, 14 subjects performed four different types of extension exercises in prone position: dynamic trunk extension, dynamic–static trunk extension, dynamic leg extension, and dynamic–static leg extension. Pre- and post-exercise muscle functional magnetic resonance imaging scans from the latissimus dorsi, the thoracic and lumbar parts of the longissimus, iliocostalis, and multifidus were performed. Differences in water relaxation values (T2-relaxation) before and after exercise were calculated (T2-shift) as a measure of muscle activity and compared between extension modalities. Linear mixed-

model analysis revealed higher lumbar extensor activity during trunk extension compared with leg extension (T2shift of 5.01 ms and 3.55 ms, respectively) and during the dynamic–static exercise performance compared with the dynamic exercise performance (T2-shift of 4.77 ms and 3.55 ms, respectively). No significant differences in the thoracic extensor activity between the exercises could be demonstrated. During all extension exercises, the latissimus dorsi was the least activated compared with the paraspinal muscles. While all extension exercises are equivalent effective to train the thoracic muscles, trunk extension exercises performed in a dynamic–static way are the most appropriate to enhance lumbar muscle strength.

Up to 40% of the athlete population is affected by low back pain (LBP), which is the most common cause of lost playing time in professional sports (Bono, 2004). Sports-induced muscle imbalance within the trunk or hip muscles seems to be related to LBP (Renkawitz et al., 2006, 2008). Furthermore, there is considerable evidence suggesting that trunk muscle strength and endurance do not only play a key role in the prevention and treatment of LBP (Holmstrom et al., 1992; Luoto et al., 1995; Kuukkanen & Malkia, 1996; Moffroid, 1997; Holm & Dickinson, 2001; Mannion et al., 2001a, b; Kell & Asmundson, 2009), but are also related to sports performance (Smith et al., 2008; McGill, 2010). As decreased muscle performance implicates a lower fatigue threshold, the precision and control of movements is reduced, which results into a poorer sports performance (Durall et al., 2009). This implies that optimal functioning of the trunk extensors is beneficial for sports performance in athletes. Athletic training and LBP rehabilitation programs exist of different exercise regimes, which are performed to enhance trunk extensor muscle strength, endurance and spinal control, and will lead to decreased levels of pain and disability (Henchoz & So, 2008; Franca et al.,

2010, 2012; Henchoz et al., 2010). While spinal control is optimized during stabilization and mobilization exercises, muscle strength and endurance are often enhanced by extension exercises of the trunk and/or the legs. These extension exercises are performed in a dynamic or dynamic–static way, and specifically strengthen the thoracic and lumbar extensors. However, at present it is not clear to which extent the contraction modality (dynamic vs dynamic–static and trunk extension vs leg extension) influences the activation of the thoracic and lumbar extensors. This is due to the scarce literature regarding this topic and the lacunas in existing studies. The few studies that exist have either compared muscle activation patterns between dynamic and dynamic–static contractions or between trunk and bilateral leg extension exercises, and have reported conflicting results. Although these studies have used electromyography (EMG), more recently muscle functional magnetic resonance imaging (mfMRI) has been used to determine the amount of muscle activity during exercise. Its main advantage compared with surface EMG (sEMG) is its superior spatial resolution, imaging deep and superficial muscles simultaneously at multiple levels and at both sides of the spine (Adams et al., 1992; Mayer et al., 2005; Cagnie et al.,

1

De Ridder et al. 2009; Dickx et al., 2010a, b). Although fine-wire EMG can also be used to investigate the activity of deep muscles, it only provides an idea on electrical activity of a few motor units and is invasive in nature. Whereas EMG has been widely used to investigate thoracic and lumbar muscle activation, and lumbar muscle work has frequently been investigated during trunk extension, studies in which both thoracic and lumbar muscle activity during trunk and leg extension exercises is measured with mfMRI are nonexistent. Therefore, this study was the first to evaluate simultaneously the amount of activity of the thoracic and lumbar muscles during standardized extension exercises with mfMRI in healthy subjects. The present study examined: (a) the influence of different exercise modalities, i.e., trunk or leg extension, on the amount of the thoracic and lumbar extensor muscle activity by evaluating the T2-shift; and (b) whether the findings were influenced by the contraction modality of the exercise, i.e., dynamic or dynamic–static contraction. It was hypothesized that the thoracic extensors, which are conjoining the thorax with the pelvic via long tendons, will be recruited more during trunk extension compared with leg extension, whereas leg extension may generate more activity of the lumbar extensors, which directly attach onto the lumbar vertebrae. Regarding the contraction modality, it was hypothesized that both the thoracic and the lumbar extensors will show bigger T2-shifts during the dynamic–static exercise performance compared with the dynamic exercises.

and two during the third session. The exercise sequence was randomized and determined by lottery. MRI scanning was performed before and immediately after performing each exercise modality. To prevent the potential influence of muscle fatigue, a rest period of 40 min was provided between the two extension exercise modalities. There were at least 7 days between the second and third session. The study protocol, information leaflet, and informed consent were approved by the local ethics committee.

Extension exercises To perform the trunk extension exercises, subjects were installed in prone position on a variable angle chair, with their upper body in a 45° of trunk flexion. The superior border of the anterior iliac (SIAS) was positioned on the edge of the table and the ankles were strapped to the table. Hands were placed on the opposite shoulder (Fig. 1). The dynamic trunk extension implied that subjects raised their trunk to the horizontal in 2 s and returned to the start position in 2 s. During the dynamic–static trunk extension, the trunk was raised to the horizontal in 2 s, the horizontal position was maintained for 5 s, after which the subject returned to 45° flexion in 2 s. To perform the leg extension exercises, subjects were installed in prone position with their lower body positioned at 45° flexion. The upper body was strapped at the level of the angulus inferior of the scapulae, hands were positioned under the forehead (Fig. 2). The dynamic leg extension consisted of a 2-s raise of the legs till the horizontal, followed by a period of 2 s to return to the start position. During the dynamic–static leg extension, the legs were extended horizontally in 2 s, held in that position during 5 s, and lowered in 2 s. To reach the horizontal position, tactile feedback was given by a rope between the two vertical stands. A metronome (60 beats/ min) was used to ensure appropriate timing of the different movements. A set of 20 repetitions of each exercise modality was

Materials and methods Subjects Fourteen subjects, eight men and six women, participated in this study. Subjects were characterized by a mean age of 24.73 ± 3.19 years, mean height of 172.9 ± 6.4 cm, mean weight of 64.47 ± 12.5 kg, and body mass index (BMI) of 22.98 ± 3.1 kg/ m2, indicating normal weight.

Study design An observational study to evaluate the recruitment of thoracic and lumbar extensor during different extension exercises was conducted on 14 healthy individuals. Subjects were recruited through adverts, which were spread among personal contacts of the researchers, the staff of Ghent University and Ghent University Hospital. If subjects experienced back pain recently, had a medical consultation concerning LBP in the past year, reported previous back surgery or spinal deformities, or when MRI was contradicted, they were not eligible for study participation. Each subject attended three sessions. The first session included a consultation of the information leaflet and signing of the informed consent, followed by anthropometric measurements and determination of the one repetition maximum (1-RM) for each extension exercise. Four different modalities of extension exercises (i.e., dynamic trunk extension, dynamic–static trunk extension, dynamic leg extension, and dynamic–static leg extension) were performed during the second and third sessions. Two exercise modalities were performed during the second session

2

Fig. 1. Position trunk extension exercises.

mfMRI of recruitment of trunk muscles during extension Table 1. Individual load adjustments of the extension exercises at 60% 1-RM

Subjects

1 2 3 4 5 6 7 8 9 10 11 12 13 14 Mean (kg)

Dynamic trunk extension

Dynamic–static trunk extension

Dynamic leg extension

Dynamic–static leg extension

Load

Repetitions

Load

Repetitions

Load

Repetitions

Load

Repetitions

0 4 −1 14 1 1 8.5 10 −6 22.5 4 −1.5 9 5 5

28 35 26 49 29 29 42 43 16 58 35 25 40 34

−1.5 −1 −17 −2 −5 −5 5 4 −7 −2.5 0 −6.5 1 −3 −3

20 26 10 25 20 20 20 33 14 24 27 16 29 22

6 6.5 −0.5 5.5 5.5 3 7 11 −1 10.5 4.5 9 6 5.5 5.5

62 55 26 46 50 39 55 71 24 67 51 70 58 49

−1 −0.5 −3 1 3 −1.5 −0.5 3.5 −4 0.5 0.0 5.5 −2 0.5 0

21 25 20 31 40 22 25 43 10 30 27 54 22 29

The resistant (positive) or assistant (negative) weight, relative to the weight of the trunk or legs and the maximum number of repetitions achieved, necessary to perform each extension modality at an exercise load of 60% of one repetition maximum (1-RM) are presented per subject. Weights are expressed in kg and are accurate up to 0.5 kg. Bold indicates mean adjusted weights (kg) to adjust the exercise load to 60% 1-RM.

exercise weight. Each 1-RM test was executed in the same position, over the same range of motion, and with an identical timing as during the respective exercise modality. Afterwards, the exercise load (kg) corresponding to 60% of 1-RM, was estimated using the Holten diagram. This diagram describes the relation between the performed number of repetitions and the exercise intensity (Danneels et al., 2001b). Calculation of the individual exercise weight occurred identically as described by Dickx et al. (2010a), using the following formula: upper/lower body weight (kg) × exercise load (60% 1-RM)/ exercise load determined on testing day (Holten diagram). The weight of the upper body is calculated as 70% of the total body weight and of the lower body as 30% of the total body weight. The total body weight was determined on a body scale during the first session. To adjust the exercise weight, the body was assisted via a load pulley system or extra weights were added (Table 1). Extra weight was added to the trunk by holding weight pockets against the chest by crossing their arms. In order to adjust the leg extension, exercise weight cuffs were tied around both thighs.

Muscle functional MRI

Fig. 2. Position leg extension exercises.

performed continuously at an exercise intensity of 60% of 1-RM. Thus to complete the dynamic–static exercise, 180 s (20 repetitions × 9 s) were needed, while the dynamic exercise condition was finalized in 80 s (20 repetitions × 4 s).

Determination of the exercise intensity (60% 1-RM) Minimum of 3 days before the second session took place, the individual 1-RM was indirectly determined by registering the maximum number repetitions participants could perform of each exercise modality with the weight of their upper/lower body as the

A 3-Tesla Trio Trim scanner (Siemens Erlangen, Germany) was used to assess changes in the relaxation time of muscle water (T2-relaxation time) as a result of muscle work during the extension exercises. The amount of muscle activity can be assessed by quantifying shifts in T2-relaxation times and is expressed as the T2-shift. (Meyer & Prior, 2000). To ensure a neutral spine position, subjects were placed symmetrically in supine position, with their head first. Two coils were used to partly cover the thoracic and lumbar spine: ventral, a flexible six-element body matrix coil centered on the belly button, and dorsal, a standard phased-array spine coil, was positioned. Two transversal slices corresponding the lower endplate of T12 and the lower endplate of L4 (Danneels et al., 2001a) were positioned as horizontal as possible on a sagittal view. A spin-echo multi-contrast sequence (Semc) was used for the acquisition of T2-weighted images. The following parameters were applied: repetition time 1000 ms, 256 × 176 mm2 matrix, 256 mm field of view (FOV), slice thickness 5 mm. Total scan time was 6 min 15 s. The images were obtained after a period of 15 min of prone lying (rest

3

De Ridder et al. Statistical analysis

Fig. 3. Image at lower endplate level T12. 1, latissimus dorsi; 2, longissimus thoracis pars thoracic; 3, iliocostalis lumborum pars thoracic.

SPSS 19.0 (IBM Corporation, Somers, New York, USA) was used to carry out statistical analyses. At first, baseline and post-exercise T2-values of the thoracic and lumbar muscles of the left and the right side were averaged, due to the symmetry of the exercises and the lack of significant side differences in T2-values (P < 0.05). In addition, the symmetry of the exercise was monitored by sEMG, the results which are published elsewhere confirmed the lack of significant side differences (De Ridder et al., 2013). Subsequently, descriptive statistics [means and standard deviation (SD)] were calculated for the anthropometric group characteristics and T2-values. To investigate the T2-shift values (i.e., the difference in T2-values between post-exercise and baseline) of the back muscles between different exercise modalities, a linear mixedmodel analysis was conducted. The following main factors were used: muscle (LD, IT, IL, LT, LL, MF), extension modality (trunk vs leg extension), and contraction type (dynamic vs dynamic– static). In case one of the main factors was significant, separate mixed-models were conducted. Post-hoc comparisons were made when required and adjusted using a Bonferroni correction. Statistical significance for all tests was set at P ≤ 0.05 (CI 95%).

Results Recruitment of the trunk muscles

Fig. 4. Image at lower endplate level L4. 1, multifidus; 2, longissimus thoracis pars lumborum; 3, iliocostalis lumborum pars lumborum.

T2), and immediately following the exercises (exercise T2). The time span between the end of the exercise and the beginning of the scan ranged from 1 min 55 s to 2 min 30 s. Scanning was performed before and immediately after performance of every extension modality. Between two different exercise modalities, subjects rested in prone lying over a period of 40 min to ensure that the T2-values were able to decrease to baseline values. The MRI images were analyzed using Image J (Java-based version of the public domain NIH Image Software, Research Services Branch, National Institutes of Health, Washington, USA). Using the “MRI analysis calculator plug-in” a T2-value (in milliseconds per voxel) was calculated out of 16 echoes. Subsequently, the region of interest (ROI) was determined on all images by drawing the outlines of all muscles, avoiding visual fat, blood vessels, and connective tissue. This method has proven to be reliable in previous work (Danneels et al., 2000; Dickx et al., 2010b; D’Hooge et al., 2013). At T12, the latissimus dorsi (LD) and the thoracic parts of the longissimus (LT) and iliocostalis (IT) were analyzed bilaterally (Fig. 3). At L4 the multifidus (MF) and the lumbar parts of the longissimus (LL) and iliocostalis (IL) were analyzed. Figure 4 presents the MF, LL, and IL on a mfMRI image. Finally, the mean T2-value was derived for each ROI and used for further analysis.

4

The mixed-model analysis, which was used to examine the T2-shift in thoracic and lumbar muscles between and within the different extension exercise modalities, showed no significant interaction effects between the main factors, whereas the main factors muscle (P < 0.001) and extension modality (P = 0.045) had a significant effect on the T2-shift. No main effect for contraction type (P = 0.193) could be established. A post-hoc comparison between the different trunk extensors demonstrated that during all exercise modalities the LD was recruited significantly less compared with all other trunk extensors (Fig. 5). No differences between the other muscles could be established. Moreover, post-hoc data revealed that in general the mean T2-shift of all trunk muscles (mean of the sum of all T2-shift values) was significantly higher during trunk extension than during leg extension, regardless of the type of contraction (P = 0.045). Due to the significance of the factor muscle, reflecting possible anatomical and functional differences between the thoracic and lumbar muscles, two new mixed-models were conducted using the same factors, but one including the thoracic muscles (IT and LT) and one the lumbar muscles (IL, LL, and MF).

Recruitment of the thoracic muscles An analysis of the thoracic muscles only was performed and showed no. three-way or two-way interaction effects between the main factors. Furthermore, the T2-shift of the IT was not significantly different to the T2-shift of the LT during the extension exercises (muscle P = 0.574). Moreover, neither the extension modality (trunk or leg extension), nor the contraction type

Values in the table present the mean T2-value (in ms) and standard deviation (± SD) at rest (pre), after performing the exercise modality (post), and the difference between these two conditions (T2-shift) of the trunk muscles. IL, iliocostalis lumborum pars lumborum; IT, iliocostalis lumborum pars thoracis; LD, latissimus dorsi; LL, longissimus thoracis pars lumborum; LT, longissimus thoracis pars thoracic; MF, multifidus.

± ± ± ± ± ± −0.64 6.50 5.34 4.33 3.68 3.75 46.43 ± 7.18 52.95 ± 5.52 53.00 ± 3.44 52.73 ± 4.37 51.88 ± 2.72 52.96 ± 3.94 46.93 ± 5.70 47.09 ± 7.73 48.25 ± 7.16 49.00 ± 6.02 48.18 ± 6.35 49.81 ± 6.54 2.88 6.53 1.78 1.91 3.04 2.21 ± ± ± ± ± ± 0.75 5.61 6.53 3.75 1.73 4.14 46.06 ± 3.56 51.68 ± 4.14 54.48 ± 3.60 51.00 ± 3.59 50.01 ± 3.68 52.31 ± 4.82 45.27 ± 3.99 46.68 ± 6.29 47.95 ± 4.72 47.09 ± 3.13 48.30 ± 3.99 47.86 ± 4.39 1.86 4.60 6.61 2.11 1.04 1.96 ± ± ± ± ± ± 1.67 7.06 6.65 5.70 4.94 6.22 1.06 6.40 4.95 4.40 4.60 4.09

± ± ± ± ± ±

2.70 5.40 2.67 5.11 2.72 1.67

45.14 ±3.98 47.83 ± 4.79 48.33 ± 3.22 46.95 ± 2.33 46.30 ± 2.85 46.85 ± 2.77

46.12 ± 4.42 54.06 ± 2.66 55.80 ± 5.75 52.65 ± 2.85 51.24 ± 2.72 53.65 ± 2.73 46.60 ± 5.91 54.45 ± 4.81 55.35 ± 6.60 52.15 ± 3.58 51.63 ± 2.58 53.75 ± 3.76 45.24 ± 4.25 47.66 ± 7.13 49.79 ± 5.65 47.67 ± 3.01 47.03 ± 4.00 48.83 ± 4.63

T2-shift T2-shift

T2 post T2 pre T2 post T2 pre

The present study examined whether the amount of activity (estimated by the shift in T2-values) of the trunk extensor muscle extensors is influenced by different modalities of extension exercises, i.e., which

Dynamic–static trunk extension

Discussion

Dynamic trunk extension

The analysis, which included solely the lumbar muscles, showed no three-way or two-way interaction effect between the main factors; however, a clear significant main effect of extension modality and contraction type was demonstrated. Lumbar muscles were recruited at a higher degree during the trunk extension exercises (T2-shift of 5.01 ms) compared with the leg extension exercises (P ≤ 0.001; T2-shift of 3.55 ms). Furthermore, the dynamic–static extension exercises demanded more lumbar muscle work than the dynamic extension exercises (P ≤ 0.014; T2-shift 4.77 vs 3.78 ms). The mean T2-shift of the lumbar muscles during the different exercise modalities are displayed in Figure 5.

Muscles

Recruitment of the lumbar muscles

Table 2. Shift of the T2 values of the trunk extensor muscles in response to each extension exercise modality

(dynamic vs dynamic–static) had a significant effect on the shift in T2-values of the thoracic muscles (P = 0.902 and P = 0.591, respectively). The mean T2-values of the thoracic muscles during all exercise modalities are presented in Table 2.

Dynamic leg extension

Fig. 5. Mean T2-shift of the trunk muscles in response to each extension exercise modality. LD, latissimus dorsi; LT, longissimus thoracis pars thoracic; IT, iliocostalis lumborum pars thoracic; LL, longissimus thoracis pars lumborum; IL, iliocostalis lumborum pars lumborum; MF, multifidus; *, significant difference in mean T2-shift among the trunk muscles at P ≤ 0.05 level. (Bars: ± 1 standard deviation).

T2 pre

Extension exercise modality

T2 post

T2-shift

−5 Dynamic-static Dynamic Dynamic-static Dynamic leg trunk trunk leg

LD LT IT LL IL MF

0

T2 post

5

T2 pre

Dynamic–static leg extension

10

T2-shift

Muscle LD LT IT LL IL MF

15

Mean T2-shift

2.64 5.89 2.76 1.34 2.37 0.96

mfMRI of recruitment of trunk muscles during extension

5

De Ridder et al. body part is extended (trunk or legs) and in which way the extension exercise is performed (dynamic or dynamic–static). The difference in mean T2-shift of the thoracic and lumbar extensors, between the various exercise conditions, supports the hypothesis that the activity level of the back extensors is influenced by the manner in which an extension exercise is performed. These results implicate that although the exercise load of the different exercises was identical, the T2-shift of the lumbar muscles was higher during trunk extension exercises compared with leg extension exercises. The higher shift implies enhanced levels of (metabolic) activity within the lumbar muscles when performing a trunk extension compared with a leg extension exercise, which is the result of more activity of the lumbar muscles. Although previous studies already showed that the thoracic and lumbar extensors are activated during extension exercises (Mayer et al., 1999, 2002; Plamondon et al., 1999, 2002; Clark et al., 2002), to our knowledge this is the first study to compare differences in the activity level of thoracic and lumbar extensor muscles, and between different trunk and leg extension exercises. In addition, this is the first study on this topic using mfMRI. The results confirm our previous findings, which were obtained using sEMG and using an identical exercise protocol on the same study population (De Ridder et al., 2013). The increased lumbar muscle activity during trunk extension in the present study, is inconsistent with the results of Plamondon et al. (2002). Those authors reported greater levels of lumbar extensor spinae muscles (LES) during performance of a dynamic prone-leg extension compared with a dynamic pronetrunk extension. However, we need to consider that Plamondon et al. (2002) used a different technique to evaluate lumbar muscle usage. While they used sEMG measures, which reflect the real-time ‘neural’ muscle changes upon exercises, in the current study mfMRI was used, which displays the acute activityinduced prolongation of T2-relaxation times of muscle water. Furthermore, inequalities in starting angle exist between the two studies, which could have a significant effect on the muscle force production, due to differences in muscle length (Mannion & Dolan, 1996). The present study was unable to demonstrate differences among the lumbar muscles, which suggest a similar action of these muscles during the extension exercises. This finding is in line with Danneels et al. (2002), who described that the MF and IL have a similar function during trunk movements. However, Ng et al. (1997, 1999) found that the MF was more activated than the IL or LL during isometric trunk and leg holding, respectively. The discrepancy between these results could be explained by dissimilarities in exercise modal-

6

ity. Ng et al. (1997, 1999) studied isometric exercises, while the current study examined dynamic and dynamic–static contractions. As it has been demonstrated that the MF has a more stabilizing function compared with the IL and LL (Wilke et al., 1995), isometric contractions could be more sensitive to little variations in function. There are few other studies that used MRI to examine differences in lumbar muscle recruitment during trunk extension in a healthy population (Mayer et al., 2005; Dickx et al., 2010b). Mayer et al. (2005) demonstrated higher activity of the MF compared with the LES during dynamic trunk extension, and showed that there was a relationship between lumbar muscle contribution and exercise intensity. This relationship could explain previous described differences in study findings. More specifically, when a relative high-load exercise is performed, as was the case in the present study; the different muscles are recruited in a homogeneous way in order to obtain and maintain a relatively high force output. As other studies did not adjust or report the exact exercise intensity, we assume that these low-load exercises are more sensitive to objectify subtle differences in muscle activity. Further investigation in low-load conditions is recommended to verify this assumption. Besides the influence of the extended body part, we also showed clear differences in lumbar muscle usage between a dynamic and dynamic–static performance of the extension exercises. Although the exercise intensity was identical in both cases, performing the extension exercises in a dynamic–static way caused a higher T2-shift of the lumbar muscles compared with a dynamic contraction. The acute higher metabolic reactions in response to the dynamic–static performance could support the findings of Danneels et al. (2001a), who studied the long-term effects of two training programs on the cross-sectional area (CSA) of the paravertebral muscles. They demonstrated that an increase in the CSA of the paravertebral muscles occurred after dynamic–static muscle training, whereas dynamic training did not affect CSA. On the long term, the higher metabolic cost of dynamic–static exercises may have triggered the volume-growing effect within the lumbar muscles, which resulted in a higher CSA. Because the contraction type (and the duration) of both modalities (dynamic vs dynamic–static) differed, both conditions were tested separately in advance and the load (number of kilograms of resistance) was individually adapted to express 60% of 1-RM. Following this procedure the exercise load, which was applied in the dynamic–static condition, was systematically lower than the load used in the dynamic version. This ensured that the exercise intensity of both exercises was identical. But although the perceived intensity was identical, it is apparent that the nature of the muscle

mfMRI of recruitment of trunk muscles during extension contraction is different. We can assume that during a static muscle contraction, the arterial pressure is higher and the blood flow is lower compared with a dynamic muscle contraction (Masuda et al., 1999; Vanderthommen et al., 2003; Arimoto et al., 2005). In this way, the added static element during the dynamic– static exercise condition could have resulted in an increased lactic acid accumulation compared with the dynamic condition. However, future exercise studies, in which mfMRI is combined with lactate determination, are necessary to verify these assumptions. The higher lactate accumulation can support the higher T2-shift in this condition. Moreover, previous studies proved that changes in T2 depend on the duration of the exercise load (Jenner et al., 1994), resulting in a higher T2-value when the duration of the exercise is increased. We expected that the work of the thoracic muscles during trunk extension would have been higher compared with leg extension, and may be influenced by thecontraction type. The lack of difference in thoracicmuscle work between the exercise modalities in the present study did not support this hypothesis. The present study demonstrated that during all extension exercises the LD was less active compared with the other trunk muscles, which is in agreement with other studies and can be clarified by the main function of the LD, namely, arm movement. The lower activity levels of the LD are confirmed by our previous findings using sEMG in a similar protocol (De Ridder et al., 2013), and in line with the findings of Coorevits et al. (2008) who showed that the LD was the least fatigued muscle of the trunk and hip extensor muscles following isometric trunk extension. This study was limited to a small sample of healthy subjects. Although most MRI studies have a limited number of subjects, it would be useful to investigate a larger and more varied population. Furthermore, care should be taken with generalization of the recent study results. The current study did not investigate the trunk muscle activity patterns during a pure static exercise modality or exercises at lower intensities (i.e., stabilization exercises). It has been recommended that in the initial stage of spine-strengthening programs, subjects should be instructed to become aware of motor patterns and to recruit muscles in isolation at lower exercise intensities (Hibbs et al., 2008), but that an essential requirement is to progress to more functional and dynamic modalities (Akuthota & Nadler, 2004). Especially from an athletic perspective, extension exercises which comprise of a dynamic component and a high exercise load are usually preferred for strengthening the trunk muscles (Hibbs et al., 2008). Future studies may reveal which static exercise modality is the most appropriate to target the lumbar extensors during the early stages of LBP rehabilitation. Furthermore, to

examine whether LBP or excessive sports participation will cause alterations in the recruitment of the trunk extensor muscles, it would be useful for future studies to examine these populations. If impaired recruitment patterns exist, it would be desirable to examine the efficacy of various extension exercise modalities in order to optimize the function of the trunk extensors. Our results demonstrated that during extension exercises the level of activity (shift in T2) of the lumbar extensors is influenced by the modality of the extension exercise, whereas the thoracic extensor activity is not. The highest activity of the lumbar muscles was found during the dynamic–static trunk extension. Therefore, due to the need of a metabolic stimulus to enhance muscle strength, the dynamic–static exercise performance and the trunk extension exercises physiologically seem the most appropriate to train the lumbar muscles in clinical practice, although this will need to be established by future studies. In order to strengthen the thoracic muscles, none of the exercise conditions seems to be superior. During all extension exercises, the LD was less active then the other trunk extensor muscles. Thus in case it is desirable to strengthen this muscle in addition to the trunk extensors, the previous described extensions exercises need to be complemented by more appropriate exercise which target this muscle.

Perspective Optimal functioning of the trunk extensors is beneficial for sports performance in athletes and plays a key role in the prevention and treatment of LBP. In training programs, endurance and strength of the lumbar muscles are often enhanced using different extension exercise modalities. The present study examined how activity patterns of trunk muscles differ between several modalities, and which type of modality achieves the highest level of activation of the lumbar extensors. It was shown that the type of extension exercise did not influence the activity levels of the thoracic extensors, but does indeed determine the activity levels of the lumbar muscles. More specifically, it was shown that when trunk and leg extensions exercises comprise a dynamic and static component they will result in higher lumbar muscle activation than when performed pure dynamic. Furthermore, trunk extension exercises will result in higher levels of lumbar muscle activation compared with leg extension exercises. In conclusion, exercise programs which wish to efficiently optimize endurance and strength of the lumbar muscles should give preference to dynamic–static trunk extension exercises.

7

De Ridder et al. Key words: Imaging, trunk extensor muscles, prone extension exercises, posterior muscle chain, spine, rehabilitation.

Acknowledgements Jessica Van Oosterwijck is a postdoctoral research fellow funded by the Special Research Fund of Ghent University.

References Adams GR, Duvoisin MR, Dudley GA. Magnetic-resonance-imaging and electromyography as indexes of muscle function. J Appl Physiol 1992: 73: 1578–1589. Akuthota V, Nadler SF. Core strengthening. Arch Phys Med Rehabil 2004: 85: S86–S92. Arimoto M, Kijima A, Muramatsu S. Cardiorespiratory response to dynamic and static leg press exercise in humans. J Physiol Anthropol Appl Human Sci 2005: 24: 277–283. Bono CM. Low-back pain in athletes. J Bone Joint Surg Am 2004: 86-A: 382–396. Cagnie B, Barbe T, Vandemaele P, Achten E, Cambier D, Danneels L. MRI analysis of muscle/fat index of the superficial and deep neck muscles in an asymptomatic cohort. Eur Spine J 2009: 18: 704–709. Clark BC, Manini TM, Mayer JM, Ploutz-Snyder LL, Graves JE. Electromyographic activity of the lumbar and hip extensors during dynamic trunk extension exercise. Arch Phys Med Rehabil 2002: 83: 1547–1552. Coorevits P, Danneels L, Cambier D, Ramon H, Vanderstraeten G. Assessment of the validity of the Biering-Sorensen test for measuring back muscle fatigue based on EMG median frequency characteristics of back and hip muscles. J Electromyogr Kinesiol 2008: 18(6): 997–1005. Danneels LA, Cools AM, Vanderstraeten GG, Cambier DC, Witvrouw EE, Bourgois J, De Cuyper HJ. The effects of three different training modalities on the cross-sectional area of the paravertebral muscles. Scand J Med Sci Sports 2001a: 11: 335–341. Danneels LA, Coorevits PL, Cools AM, Vanderstraeten GG, Cambier DC, Witvrouw EE, De CH. Differences in electromyographic activity in the multifidus muscle and the iliocostalis lumborum between healthy subjects and patients with sub-acute and chronic low back pain. Eur Spine J 2002: 11 (1): 13–19. Danneels LA, Vanderstraeten GG, Cambier DC, Witrouw EE, De Cuyper HJ. CT imaging of trunk muscles in chronic low back pain

8

patients and healthy control subjects. Eur Spine J 2000: 9.4 (Aug): 266–272. Danneels LA, Vanderstraeten GG, Cambier DC, Witvrouw EE, Bourgois J, Dankaerts W, De Cuyper HJ. Effects of three different training modalities on the cross sectional area of the lumbar multifidus muscle in patients with chronic low back pain. Br J Sports Med 2001b: 35: 186–191. De Ridder EM, Van Oosterwijck JO, Vleeming A, Vanderstraeten GG, Danneels LA. Posterior muscle chain activity during various extension exercises: an observational study. BMC Musculoskelet Disord 2013: 14: 204. D’Hooge R, Cagnie B, Crombez G, Vanderstraeten G, Achten E, Danneels L. Lumbar muscle dysfunction during remission of unilateral recurrent nonspecific low-back pain: evaluation with muscle functional MRI. Clin J Pain 2013: 29: 187–194. Dickx N, Cagnie B, Achten E, Vandemaele P, Parlevliet T, Danneels L. Differentiation between deep and superficial fibers of the lumbar multifidus by magnetic resonance imaging. EurSpine J 2010a: 19: 122–128. Dickx N, D’hooge R, Cagnie B, Deschepper E, Verstraete K, Danneels L. Magnetic resonance imaging and electromyography to measure lumbar back muscle activity. Spine (Phila Pa 1976) 2010b: 35: E836–E842. Durall CJ, Udermann BE, Johansen DR, Gibson B, Reineke DM, Reuteman P. The effects of preseason trunk muscle training on low-back pain occurrence in women collegiate gymnasts. J Strength Cond Res 2009: 23: 86–92. Franca FR, Burke TN, Caffaro RR, Ramos LA, Marques AP. Effects of muscular stretching and segmental stabilization on functional disability and pain in patients with chronic low back pain: a randomized, controlled trial. J Manipulative Physiol Ther 2012: 35: 279–285. Franca FR, Burke TN, Hanada ES, Marques AP. Segmental stabilization and muscular strengthening in chronic

low back pain: a comparative study. Clinics (Sao Paulo) 2010: 65: 1013–1017. Henchoz Y, de Goumoëns P, Norberg M, Paillex R, So AK. Role of physical exercise in low back pain rehabilitation: a randomized controlled trial of a three-month exercise program in patients who have completed multidisciplinary rehabilitation. Spine (Phila Pa 1976) 2010: 35: 1192–1199. Henchoz Y, So AKL. Exercise and nonspecific low back pain: a literature review. Joint Bone Spine 2008: 75: 533–539. Hibbs AE, Thompson KG, French D, Wrigley A, Spears I. Optimizing performance by improving core stability and core strength. Sports Medicine 2008: 38: 995–1008. Holm SM, Dickinson AL. A comparison of two isometric back endurance tests and their predictability of first-time back pain: a pilot study. JNMS 2001: 9: 46–53. Holmstrom E, Moritz U, Andersson M. Trunk muscle strength and back muscle endurance in construction workers with and without low-back disorders. Scand J Rehabil Med 1992: 24: 3–10. Jenner G, Foley JM, Cooper TG, Potchen EJ, Meyer RA. Changes in magnetic resonance images of muscle depend on exercise intensity and duration, not work. J Appl Physiol (1985) 1994: 76: 2119–2124. Kell RT, Asmundson GJ. A comparison of two forms of periodized exercise rehabilitation programs in the management of chronic nonspecific low-back pain. J Strength Cond Res 2009: 23: 513–523. Kuukkanen T, Malkia E. Muscular performance after a 3 month progressive physical exercise program and 9 month follow-up in subjects with low back pain. A controlled study. Scand J Med Sci Sports 1996: 6: 112–121. Luoto S, Heliovaara MH, Hurri H, Alaranta H. Static back endurance and the risk of low-back-pain. Clin Biomech 1995: 10: 323–324. Mannion AF, Dolan P. The effects of muscle length and force output on the

mfMRI of recruitment of trunk muscles during extension EMG power spectrum of the erector spinae. J Electromyogr Kinesiol 1996: 6(3): 159–168. Mannion AF, Muntener M, Taimela S, Dvorak J. Comparison of three active therapies for chronic low back pain: results of a randomized clinical trial with one-year follow-up. Rheumatology 2001a: 40: 772–778. Mannion AF, Taimela S, Muntener M, Dvorak J. Active therapy for chronic low back pain Part 1. Effects on back muscle activation, fatigability, and strength. Spine 2001b: 26: 897–908. Masuda K, Masuda T, Sadoyama T, Inaki M, Katsuta S. Changes in surface EMG parameters during static and dynamic fatiguing contractions. J Electromyogr Kinesiol 1999: 9: 39–46. Mayer JM, Graves JE, Clark BC, Formikell M, Ploutz-Snyder LL. The use of magnetic resonance imaging to evaluate lumbar muscle activity during trunk extension exercise at varying intensities. Spine 2005: 30: 2556–2563. Mayer JM, Graves JE, Robertson VL, Pierra EA, Verna JL, Ploutz-Snyder LL. Electromyographic activity of the lumbar extensor muscles: effect of angle and hand position during Roman chair exercise. Arch Phys Med Rehabil 1999: 80: 751–755. Mayer JM, Verna JL, Manini TM, Mooney V, Graves JE. Electromyographic activity of the

trunk extensor muscles: effect of varying hip position and lumbar posture during Roman chair exercise. Arch Phys Med Rehabil 2002: 83: 1543–1546. McGill S. Core training: evidence translating to better performance and injury prevention. Strength Cond J 2010: 32: 33–46. Meyer RA, Prior BM. Functional magnetic resonance imaging of muscle. Exerc Sport Sci Rev 2000: 28 (2): 89–92. Moffroid MT. Endurance of trunk muscles in persons with chronic low back pain: assessment, performance, training. J Rehabil Res Dev 1997: 34: 440–447. Ng JKF, Chan CCH, Lui CYH, Sin AKY, Wong ISW, Yip JHH. Back and lower limb muscle activities in a prone leg holding exercise. J Back Musculoskelet 1999: 13: 101–106. Ng JKF, Richardson CA, Jull GA. Electromyographic amplitude and frequency changes in the iliocostalis lumborum and multifidus muscles during a trunk holding test. Phys Ther 1997: 77: 954–961. Plamondon A, Marceau C, Stainton S, Desjardins P. Toward a better prescription of the prone back extension exercise to strengthen the back muscles. Scand J Med Sci Sports 1999: 9: 226–232. Plamondon A, Serresse O, Boyd K, Ladouceur D, Desjardins P. Estimated

moments at L5/S1 level and muscular activation of back extensors for six prone back extension exercises in healthy individuals. Scand J Med Sci Sports 2002: 12: 81–89. Renkawitz T, Boluki D, Grifka J. The association of low back pain, neuromuscular imbalance, and trunk extension strength in athletes. Spine J 2006: 6: 673–683. Renkawitz T, Linhardt O, Grifka J. Electric efficiency of the erector spinae in high performance amateur tennis players. J Sports Med Phys Fitness 2008: 48: 409–416. Smith CE, Nyland J, Caudill P, Brosky J, Caborn DNM. Dynamic trunk stabilization: a conceptual back injury prevention program for volleyball athletes. J Orthop Sports Phys Ther 2008: 38: 703–720. Vanderthommen M, Duteil S, Wary C, Raynaud JS, Leroy-Willig A, Crielaard JM, Carlier PG. A comparison of voluntary and electrically induced contractions by interleaved 1H- and 31P-NMRS in humans. J Appl Physiol 2003: 94: 1012–1024. Wilke HJ, Wolf S, Claes LE, Arand M, Wiesend A. Stability increase of the lumbar spine with different muscle groups. A biomechanical in vitro study. Spine (Phila Pa 1976) 1995: 20(2): 192–198.

9