Sequencing of superficial trunk muscle activation during range-of-motion tasks

Human Movement Science 43 (2015) 67–77

Contents lists available at ScienceDirect

Human Movement Science journal homepage: www.elsevier.com/locate/humov

Sequencing of superficial trunk muscle activation during range-of-motion tasks Alison Schinkel-Ivy 1, Janessa D.M. Drake ⇑ School of Kinesiology & Health Science, York University, 4700 Keele Street, Toronto, ON M3J 1P3, Canada

a r t i c l e

i n f o

Article history: Received 8 March 2015 Revised 14 July 2015 Accepted 14 July 2015 Available online 25 July 2015 Keywords: Trunk musculature Electromyography Muscle activation sequences Cross-correlation

a b s t r a c t Altered lumbo-pelvic activation sequences have been identified in individuals with low back pain. However, an analysis of activation sequences within different levels of the trunk musculature has yet to be conducted. This study identified the activation sequences characteristic of the trunk musculature during upright standing and range-of-motion tasks. Surface electromyography was recorded for eight trunk muscles bilaterally during trunk range-of-motion movement tasks in 30 participants. Cross-correlation was performed on 48 pairings of muscles, consisting of one lower- and one mid-level muscle, or one mid-level and one upper muscle. Time lags of the maximum cross-correlation coefficient were extracted and defined as a top-down or bottom-up activation sequence, or similar activation timing. Pairings that demonstrated a specific activation sequence in 50% or more of participants were then identified. Similar activation timing was consistently identified between muscle pairings for upright standing. Top-down sequences and similar timing were identified for abdominal – mid-level pairings in maximum flexion and slumped standing, respectively, while both tasks were characterized by bottom-up sequences when considering the lumbar and lower-thoracic erector spinae. Sequences were more variable across muscle pairings for lateral bend and axial twist tasks. These results provide insight into the synergy of the trunk musculature for movements in the three planes of motion. These findings can be used for comparison to low back pain populations, as altered activation sequences in these individuals may contribute to maladaptive loading patterns and consequently the development or exacerbation of low back pain. Ó 2015 Elsevier B.V. All rights reserved.

1. Introduction Studies employing surface electromyography (EMG) to evaluate trunk muscle function are numerous throughout the literature. Various techniques may be employed during data collection and processing to examine different characteristics of the EMG signal, such as activation timing and sequencing (Nelson-Wong et al., 2013), changes with fatigue (Beneck, Baker, & Kulig, 2013), and the relationships to muscle forces and spine loading (Cholewicki & McGill, 1996; Granata & Marras, 2000). Further, it has been established that EMG variables can be used to distinguish between healthy individuals and individuals with low back pain (Nelson-Wong & Callaghan, 2010; Watson, Booker, Main, & Chen, 1997). For example, different activation

⇑ Corresponding author at: 2030 Sherman Health Science Research Centre, School of Kinesiology and Health Science, York University, 4700 Keele Street, Toronto, Ontario M3J 1P3, Canada. E-mail address: [email protected] (J.D.M. Drake). 1 Present address: Toronto Rehabilitation Institute, University Health Network, 550 University Avenue, Toronto, ON M5A 2G2, Canada. http://dx.doi.org/10.1016/j.humov.2015.07.003 0167-9457/Ó 2015 Elsevier B.V. All rights reserved.

68

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

sequences have been identified in individuals with low back pain, compared to healthy individuals (Nelson-Wong, Alex, Csepe, Lancaster, & Callaghan, 2012; Nelson-Wong et al., 2013). As a result of the insight provided by these characteristics of the EMG signal into various physiological processes (De Luca, 1997), and the potential relationships to injury in the lumbar spine, the lumbar musculature has been extensively researched with respect to these characteristics. With the traditional focus on the lumbar musculature in spine biomechanics research, there is a relative paucity of work relating to muscle activation patterns and levels in the thoracic spine, as well as the interactions of the musculature amongst the different levels of the spine. While the musculature of the thoracic spine has been somewhat neglected due to the lower incidence of injury relative to the lumbar spine, the median prevalence of thoracic spine pain has been documented to be approximately 30% in the general working population (Briggs, Bragge, Smith, Govil, & Straker, 2009). These findings suggest that thoracic spine pain is a substantial occupational health problem, and therefore a detailed examination of muscle activation within this region is warranted. Further, the thoracic spine may exert substantial influence on postures and motion patterns in the whole spine and in the cervical and lumbar spine regions (Edmondston & Singer, 1997). Lumbar spine posture has also been shown to influence the motion that can subsequently be achieved in the thoracic spine (Nairn & Drake, 2014). Given the mutual influence of the lumbar and thoracic spine regions on each other with respect to posture and movement, a detailed examination of the patterns by which the muscles activate along the spine during trunk movements is warranted to better understand neuromuscular control of the trunk. Past work has documented activation levels in several muscles in the thoracic spine, such as the lower-thoracic erector spinae (ES) (Drake, Fischer, Brown, & Callaghan, 2006; McGill, 1991; Nelson-Wong & Callaghan, 2010) and latissimus dorsi (Drake et al., 2006; McGill, 1991), both characterized at the T9 level. These muscles are often examined in relation to lumbar spine mechanics. Conversely, for studies with a focus on the mechanics of the cervical or cervico-thoracic spine, activation levels have been recorded for the thoracic ES musculature at the T4 level (Burnett et al., 2009; Caneiro et al., 2010; Edmondston, Sharp, Symes, Alhabib, & Allison, 2011), and for the trapezius muscle between the C7 spinous process and the acromion (Burnett et al., 2009; Caneiro et al., 2010). The majority of studies incorporating the thoracic musculature have focused on one to two of these four muscles, with little work done to integrate multiple muscles within the thoracic musculature itself, or across the thoracic and lumbar spine regions. Nairn, Azar, and Drake (2013), Nairn, Chisholm, and Drake (2013), and Schinkel-Ivy, Nairn, and Drake (2013) presented three such studies, investigating activation levels of the surface trunk musculature during short-duration slumped sitting (Nairn, Chisholm, et al., 2013) and prolonged sitting (Nairn, Azar, et al., 2013), and co-contraction between the various muscles of the trunk during prolonged sitting (Schinkel-Ivy et al., 2013). In these studies, the bilateral upper-thoracic ES (T4), lower-thoracic ES (T9), and latissimus dorsi (T9) activation levels were measured along with the lumbar ES (L3), rectus abdominis, external oblique, and internal oblique muscles. However, an examination of the interactions between these muscles during trunk range-of-motion (ROM) tasks has yet to be undertaken. Nelson-Wong and Callaghan (2010), Nelson-Wong, Gregory, Winter, and Callaghan (2008), and Nelson-Wong et al. (2012, 2013) have conducted a series of studies investigating the sequences of muscle activation within the lumbo-pelvic region, focusing on the relationships between selected trunk muscles and the gluteus medius (Nelson-Wong & Callaghan, 2010), and between the bilateral gluteus medius (Nelson-Wong et al., 2008) during prolonged standing; between the thoracic ES, lumbar ES, and gluteus maximus during trunk flexion (Nelson-Wong et al., 2012); and between selected trunk muscles and either rectus femoris or gluteus maximus during two commonly used clinical assessments for lumbo-pelvic control (Nelson-Wong et al., 2013). These works used cross-correlation to identify the sequencing of activation between trunk and pelvic muscles, for the purposes of better understanding lumbo-pelvic control. However, the understanding of activation sequences and patterns within the muscles of the trunk specifically during trunk motion is lacking; work of this nature would provide insight into the means by which trunk movement is accomplished in healthy individuals. Therefore, the purpose of this study was to identify the activation sequences characteristic of the trunk musculature during upright standing and trunk ROM tasks. 2. Methods 2.1. Participants Thirty individuals (15 male/15 female) participated in the study, with mean (SD) age, height, and weight, respectively, of 25.0 years (3.8), 1.80 m (0.05), and 79.64 kg (8.75) for the males and 22.8 years (2.7), 1.66 m (0.05), and 59.12 kg (6.38) for the females. All participants were right-hand dominant and asymptomatic for low back pain in that none had sought treatment for back pain, nor missed any days of school or work due to back pain, for twelve months prior to collection. All procedures were approved by York University’s Office of Research Ethics, and written informed consent was obtained from all participants prior to collection. Data were collected as part of a larger study investigating trunk muscle activation and three-dimensional motion. 2.2. Instrumentation Pairs of Ag/Ag–Cl electrodes (AmbuÒ Blue Sensor N, Ambu A/S, Denmark) were applied over the bellies of the muscles of interest (Fig. 1): external oblique (McGill, 1991; Mirka & Marras, 1993), internal oblique (McGill, 1991), latissimus

69

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

(a)

(b) Upper trapezius Upperthoracic ES Lowerthoracic ES

Rectus abdominis Internal oblique

External oblique

Latissimus dorsi Lumbar ES

Fig. 1. Electrode placements for (a) the abdominal musculature and (b) the back musculature. ES: erector spinae.

dorsi (McGill, 1991), lumbar erector spinae (lumbar ES) (Mirka & Marras, 1993), lower-thoracic erector spinae (lower-thoracic ES) (McGill, 1991), rectus abdominis (Mirka & Marras, 1993), upper trapezius (Jensen, Vasseljen, & Westgaard, 1993), and upper-thoracic erector spinae (upper-thoracic ES) (Burnett et al., 2009 (adapted from Solomonow, Baratta, Banks, Freudenberger, & Zhou, 2003); Zipp, 1982). Levels of the ES muscles were determined based on the literature. Further, kinematic analyses suggested that thoracic and lumbar spine regions could be partitioned into three segments (C7–T6, T6–T12, T12–L5; Schinkel-Ivy & Drake, 2015), and therefore one site was chosen at approximately the midpoints of each of these segments. EMG signals were differentially amplified (frequency response 10–1000 Hz, common mode rejection 115 dB at 60 Hz, input impedance 10 GX; model AMT-8, Bortec, Calgary, Canada) and sampled at 2400 Hz (Vicon MX, Vicon Systems Ltd., Oxford, UK). Three-dimensional kinematic data were also collected using a seven-camera Vicon MX motion capture system (Vicon MX, Vicon Systems Ltd., Oxford, UK). Further information regarding kinematics instrumentation, data collection procedures, and data processing can be found in Schinkel-Ivy and Drake (2015). 2.3. Procedures Following shaving and swabbing of electrode sites with rubbing alcohol, isometric maximum voluntary contraction (MVC) protocols were performed to elicit the maximum activation level of each muscle. All MVCs were performed against manual resistance provided by an investigator. For the trunk flexors, participants sat in a slightly reclined, bent-knee sit-up position at the edge of a therapy table with the arms crossed over the chest and performed maximal isometric trunk flexion, lateral bending, and axial rotation against resistance (McGill, 1991; McGill, 1992). The trunk extensor MVCs entailed participants lying prone on a therapy table with the upper body cantilevered over the edge and feet restrained, and attempting to extend the trunk against resistance applied to the shoulders (McGill, 1991; McGill, 1992). For the latissimus dorsi MVC, participants abducted the arm to 90° and externally rotated so the forearm was in an almost-vertical position. Participants then pulled their elbow downwards and backwards against resistance applied under the elbow (Arlotta, LoVasco, & McLean, 2011). To elicit the upper trapezius MVC, participants abducted the arm to 90° and attempted to continue abducting the arm past this angle against resistance applied to the lateral aspect of the elbow (McLean, 2005). MVC trials lasted 3–5 s, with verbal encouragement provided by an investigator. Three trials were performed for each protocol, with a minimum of 3 min of rest between trials to minimize the effect of fatigue. The maximum value of the three trials was designated the MVC for that muscle (see Section 2.4, Data Processing and Analysis). Participants then performed ten trials of upright standing and four movement tasks: maximum trunk flexion, lateral bend, and axial twist, and slumped standing. Upright standing trials were each 10 s long, in which participants stood with the head looking forwards and arms at their sides. For the maximum range-of-motion trials, participants crossed the arms over the chest. The trials consisted of head movement followed by trunk movement in a smooth, continuous motion. The slumped standing trials entailed participants ‘rounding out’ the shoulders and spine (Callaghan & Dunk, 2002; Dankaerts, O’Sullivan, Burnett, & Straker, 2006), with the arms at the sides and the head remaining forward. Bending and twisting trials were performed to the right side, and trials were presented in a random order. All movement trials lasted approximately 10 s, during which participants moved to the position in a controlled manner, held the position for 3 s, and moved back to upright standing. Participants received full instructions and time to practice the movement tasks prior to beginning the protocol, as well as brief prompts prior to each trial.

70

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

2.4. Data processing and analysis Data were processed using Visual3D v.4 (C-Motion, Inc., Germantown, USA). Data were high-pass filtered using a fourth-order, dual-pass Butterworth filter (cutoff frequency: 30 Hz (Drake & Callaghan, 2006)), full-wave rectified, and low-pass filtered using a fourth-order, dual-pass Butterworth filter (cutoff frequency: 2.5 Hz (Brereton & McGill, 1998)). The maximum value of any of the three MVC trials (see Section 2.3, Procedures) was designated the MVC for that muscle (Knutson, Soderberg, Ballantyne, & Clarke, 1994; Nairn & Drake, 2014; Schinkel-Ivy et al., 2013; Sparto & Parnianpour, 2001). The signals from subsequent trials (i.e. the movement tasks) were then normalized to the MVC value by dividing all samples in the signal by that value, resulting in EMG data expressed as %MVC. The eight muscles were divided into three groups: lower-level muscles (external and internal oblique, rectus abdominis, and lumbar ES), mid-level muscles (latissimus dorsi, lower-thoracic ES), and upper-level muscles (upper trapezius, upper-thoracic ES) (Fig. 2). Cross-correlation analyses were conducted between every possible pairing (48 in total) of one lower and one mid-level muscle, or one mid-level and one upper muscle (Table 1). The time series activation data were used as inputs into the analysis to examine the activation sequences of the selected muscle pairings over the full trial durations, using a custom program written in Matlab v.R2012a (The MathWorks, Inc., Natick, USA). Cross-correlation quantifies the extent to which two time-varying data sets are correlated (Shum, Crosbie, & Lee, 2007), to assess spatial and temporal similarities between the signals (Nelson-Wong, Howarth, Winter, & Callaghan, 2009). The time lags at which the maximum cross-correlation coefficient occurred (within a window of ±500 ms) were extracted (Johnson, Cacciatore, Hamill, & Van Emmerik, 2010; Nelson-Wong & Callaghan, 2010; Nelson-Wong et al., 2008). For each task, the time lags were then averaged

Upper Muscles Mid-Level Muscles Lower Muscles

UTES TR LTES LD EO IO RA LES

Fig. 2. The division of trunk muscles into the lower, mid-, and upper levels.

Table 1 The 48 pairings of muscles included in the analysis, with all possible combinations with one lower and one mid-level muscle, and one mid-level and one upper muscle. L/R: left/right; EO: external oblique; IO: internal oblique; RA: rectus abdominis; LES: lumbar erector spinae; LD: latissimus dorsi; LTES: lower-thoracic erector spinae; TR: upper trapezius; UTES: upper-thoracic erector spinae. Lower – mid-level muscle pairings

Mid-level – upper muscle pairings

LEO

LLD LLTES RLD RLTES

LRA

LLD LLTES RLD RLTES

LLD

LTR LUTES RTR RUTES

LIO

LLD LLTES RLD RLTES

RRA

LLD LLTES RLD RLTES

LLTES

LTR LUTES RTR RUTES

REO

LLD LLTES RLD RLTES

LLES

LLD LLTES RLD RLTES

RLD

LTR LUTES RTR RUTES

RIO

LLD LLTES RLD RLTES

RLES

LLD LLTES RLD RLTES

RLTES

LTR LUTES RTR RUTES

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

71

across trials within each participant. Based on the direction of the time lag at which the maximum cross-correlation coefficient was identified, each time lag was determined to be either a top-down or bottom-up sequence (based on the order in which the muscles were entered into the cross-correlation), or similar timing (maximum coefficient at zero time lag). Time lag values are presented with positive and negative values representing top-down and bottom-up sequences, respectively. Muscle pairings for which 15 or more participants (P50% of participants) demonstrated the same sequence were identified and compiled to obtain the predominant muscle sequencing pattern for each muscle pairing in each movement task. 3. Results 3.1. Upright standing In upright standing, all muscle pairings exhibited maximum cross-correlation coefficients at zero time lag (Tables 2 and 3). For all pairings, a time lag of zero was identified in at least 26 of 30 participants. 3.2. Maximum flexion In maximum flexion, pairings involving the oblique muscles and the lower-thoracic ES tended to demonstrate a top-down activation sequence, with time lags ranging from 0.029 s to 0.055 s. The lumbar ES demonstrated a similar sequence when paired with latissimus dorsi, but a bottom-up sequence when paired with lower-thoracic ES (time lags ranging from 0.093 s to 0.119 s). A bottom-up activation sequence was generally evident for the mid-level – upper muscle pairings (time lags ranging from 0.070 s to 0.014 s). For the 28 pairings in which a consistent sequence was identified (i.e. a sequence that was identified in the majority of participants), the sequence was demonstrated by between 15 and 21 participants, depending on the pairing. 3.3. Slumped standing For slumped standing, 11 of the 24 pairings involving the abdominal and mid-level muscles activated with similar timing (peak cross-correlation occurring at zero time lag). For these 11 pairings, between 16 and 23 participants showed this pattern. Conversely, the lumbar ES and latissimus dorsi muscles activated in a bottom-up sequence relative to the mid-level and upper muscles, respectively (15–26 participants; time lags ranging from 0.234 s to 0.017 s). When the lower-thoracic ES and upper-level muscles were examined together, the lower-thoracic ES activated prior to the upper-thoracic ES, but following the upper trapezius muscles (between 15 and 20 participants; time lags ranging from 0.052 s to 0.054 s). 3.4. Maximum lateral bend For maximum lateral bend to the right side, consistent trends were identified in the left-side oblique muscles (Table 4). When paired with the left-side, mid-level muscles, a bottom-up activation sequence was observed (time lags ranging from 0.209 to 0.013 s), while the opposite was found when the left-side obliques were paired with the right-side mid-level muscles (time lags ranging from 0.027 s to 0.117 s). Conversely, the rectus abdominis muscles generally activated with similar timing to the mid-level muscles, in between 15 and 23 participants (depending on the pairing). The lumbar ES muscles also demonstrated differential activation sequences based on the side of the body, with the left lumbar ES yielding top-down and bottom-up sequences with the latissimus dorsi and lower-thoracic ES, respectively, while the right lumbar ES was involved in top-down activation sequences or similar timing to the mid-level muscles. The pairings of the mid-level and upper muscles resulted in a combination of top-down and bottom-up sequences, as well as instances of similar activation timing, with no consistent trends. 3.5. Maximum twist During maximum twisting to the right side, different sequences were observed for the oblique muscles depending on the side of the body. The right side tended to activate in a bottom-up sequence with the mid-level muscles, while the left side activated in a top-down sequence. The rectus abdominis muscles activated with similar timing to the mid-level muscles. The lower – mid-level back pairings involving the lumbar ES muscles, as well as the mid-level – upper pairings, revealed a combination of top-down and bottom-up activation sequences, with no consistent trends across pairings. Time lags for the lower – mid-level and mid-level – upper pairings ranged from 0.148 s to 0.152 s, and from 0.168 s to 0.082 s, respectively. 4. Discussion The purpose of the present study was to identify the activation sequences characteristic of the trunk musculature during trunk ROM tasks, in order to understand how the trunk muscles function together to produce motion. Overall, the sequences

72

Pairing

Upright standing

Maximum flexion

Direction

N

Direction

N

Lag (s)

Slumped standing

LLD–LTR

0

29

–

–

LLD–LUTES

0

29

–

LLD–RTR LLD–RUTES

0 0

28 28

RLD–LTR

0

29

RLD–LUTES RLD–RTR RLD–RUTES

0 0 0

Direction

N

Lag (s)

–

–

–

–

–

–

15 21

0.018 (0.137) 0.040 (0.070)

–

28 29 29

–

Maximum lateral bend

Maximum axial twist

Direction

Direction

N

Lag (s)

–

17

0.050 (0.111)

N

Lag (s)

18

–

–

19

0.085 (0.123)

0.042 (0.086)

19

–

–

– 16

– 0.052 (0.101)

– –

– –

– –

15 23

0.016 (0.071) 0.054 (0.092)

–

–

–

–

–

–

–

18

0.053 (0.127)

18 15 17

0.034 (0.082) 0.022 (0.158) 0.033 (0.073)

–

17 – 20

0.031 (0.055) – 0.039 (0.067)

0 0

16 16 21

0.027 (0.052) – –

– –

– – 28

– – 0.168 (0.142)

0

LLTES–LTR

0

27

18

0.053 (0.149)

18

0.050 (0.113)

16

0.012 (0.152)

–

–

–

LLTES–LUTES

0

29

20

0.014 (0.061)

15

0.015 (0.071)

21

0.071 (0.118)

0

23

–

LLTES–RTR LLTES–RUTES

0 0

26 27

18 20

0.041 (0.150) 0.032 (0.068)

18 20

0.047 (0.117) 0.040 (0.083)

20 –

0.045 (0.100) –

–

–

– 22

– 0.050 (0.143)

RLTES–LTR

0

28

20

0.070 (0.119)

17

0.054 (0.116)

–

–

–

19

0.082 (0.193)

RLTES–LUTES RLTES–RTR

0 0

29 27

17 19

0.015 (0.078) 0.053 (0.164)

19 17

0.030 (0.086) 0.051 (0.128)

–

– 16

– 0.025 (0.127)

15 –

0.021 (0.075) –

RLTES–RUTES

0

29

19

0.021 (0.048)

18

0.035 (0.082)

0

21

–

23

0.064 (0.086)

–

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

Table 2 Direction of predominant activation sequence, mean (SD) time lag, and number of participants (N; out of 30) exhibiting that sequence for mid-level – upper pairings in all tasks. ": bottom-up activation sequence (negative lag); ;: top-down activation sequence (positive lag); 0: zero lag. Small, medium, and large bolded arrows represent time lags of ±0.002 s–0.049 s, ±0.050 s–0.149 s, and P0.150 s (absolute value), respectively. Dashes indicate pairings with no predominant sequence, or with no time lag. L/R: left/right; LD: latissimus dorsi; LTES: lower-thoracic erector spinae; TR: upper trapezius; UTES: upper-thoracic erector spinae.

73

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

Table 3 Direction of predominant activation sequence, mean (SD) time lag, and number of participants (N; out of 30) exhibiting that sequence for lower – mid-level pairings in upright standing, maximum flexion, and slumped standing. ": bottom-up activation sequence (negative lag); ;: top-down activation sequence (positive lag); 0: zero lag. Small, medium, and large bolded arrows represent time lags of ±0.002 s–0.049 s, ±0.050 s–0.149 s, and P0.150 s (absolute value), respectively. Dashes indicate pairings with no predominant sequence, or with no time lag. L/R: left/right; EO: external oblique; IO: internal oblique; RA: rectus abdominis; LES: lumbar erector spinae; LD: latissimus dorsi; LTES: lower-thoracic erector spinae. Pairing

Upright standing

Maximum flexion

Slumped standing

Direction

N

Direction

N

Lag (s)

Direction

LEO–LLD LEO–LLTES LEO–RLD LEO–RLTES REO–LLD REO–LLTES REO–RLD REO–RLTES

0 0 0 0 0 0 0 0

30 29 30 29 30 29 30 28

– – –

– – – 15 – 18 – 16

N

Lag (s)

– – – 0.036 (0.089) – 0.029 (0.090) – 0.050 (0.086)

0 – 0 – 0 – 0 –

22 – 21 – 21 – 22 –

– – – – – – – –

LIO–LLD LIO–LLTES

0 0

28 28

–

– 19

– 0.055 (0.143)

0 –

19 –

– –

LIO–RLD LIO–RLTES

0 0

29 29

–

– 16

– 0.078 (0.166)

0 –

18 –

– –

RIO–LLD RIO–LLTES

0 0

30 29

–

– 19

– 0.055 (0.124)

0 –

23 –

– –

RIO–RLD RIO–RLTES

0 0

30 29

–

– 20

– 0.065 (0.126)

0 –

19 –

– –

LRA–LLD LRA–LLTES LRA–RLD LRA–RLTES RRA–LLD RRA–LLTES RRA–RLD RRA–RLTES

0 0 0 0 0 0 0 0

29 28 30 29 30 28 30 28

– – – – – – – –

– – – – – – – –

– – – – – – – –

0 – – – 0 – 0

16 – – – 16 – 18 15

– – – – – – – 0.082 (0.133)

LLES–LLD

0

28

19

0.119 (0.200)

–

–

–

LLES–LLTES

0

29

19

0.090 (0.194)

25

0.231 (0.182)

LLES–RLD

0

29

18

0.087 (0.226)

–

–

LLES–RLTES

0

28

17

0.052 (0.197)

26

0.234 (0.182)

RLES–LLD

0

30

19

0.109 (0.204)

–

–

RLES–LLTES

0

28

20

0.093 (0.202)

25

0.232 (0.193)

RLES–RLD

0

30

17

0.100 (0.227)

15

0.017 (0.167)

RLES–RLTES

0

29

18

0.038 (0.188)

26

0.225 (0.186)

– –

–

–

appeared to be dependent on both the movement task and the level of the musculature (i.e. either lower – mid-level or mid-level – upper muscles), with consistent sequences across participants in some movement tasks and variation in the sequences used for other tasks. These results provide insight into the synergy of the trunk musculature that produces trunk movements in the three planes of motion, and have potential to be used as a baseline for the study of altered neuromuscular control in individuals with low back pain. The cross-correlation technique has been employed previously in electromyography studies to examine neuromuscular control and movement coordination (Nelson-Wong et al., 2012; Nelson-Wong et al., 2013), and to quantify co-activation between two muscles (Nelson-Wong & Callaghan, 2010; Nelson-Wong et al., 2008), due to its ability to quantify the extent of the spatial and temporal similarities between two signals. The time lags at which the cross-correlation function was maximal were within the range of values reported previously (Nelson-Wong et al., 2012; Nelson-Wong et al., 2013; Shan et al., 2014). In upright standing, participants (at least 26 of 30) consistently demonstrated time lags of zero for all muscle pairings. These results differ from those previously identified for prolonged standing by Nelson-Wong and Callaghan (2010), who found that participants tended to exhibit a bottom-up control strategy. The discrepancies between the findings of the two studies can potentially be attributed to the muscles of interest. That is, the focus of Nelson-Wong and Callaghan (2010) was the lumbo-pelvic musculature, specifically the relationship of gluteus medius to the trunk muscles, while the focus

74

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

Table 4 Direction of predominant activation sequence, mean (SD) time lag, and number of participants (N; out of 30) exhibiting that sequence for lower – mid-level pairings in maximum lateral bend and axial twist. ": bottom-up activation sequence (negative lag); ;: top-down activation sequence (positive lag); 0: zero lag. Small, medium, and large bolded arrows represent time lags of ±0.002 s–0.049 s, ±0.050 s–0.149 s, and P0.150 s (absolute value), respectively. Dashes indicate pairings with no predominant sequence, or with no time lag. L/R: left/right; EO: external oblique; IO: internal oblique; RA: rectus abdominis; LES: lumbar erector spinae; LD: latissimus dorsi; LTES: lower-thoracic erector spinae. Pairing

Maximum lateral bend Direction

Maximum axial twist

N

Lag (s)

N

Lag (s)

LEO–LLD

19

0.013 (0.123)

16

0.053 (0.111)

LEO–LLTES

25

0.209 (0.230)

17

0.050 (0.203)

LEO–RLD

22

0.117 (0.182)

25

0.152 (0.149)

17

0.033 (0.256)

20

0.058 (0.117)

–

–

21

0.055 (0.100)

16

0.019 (0.145)

17

0.091 (0.164)

15 15

– 0.072 (0.135)

16 25

0.019 (0.115) 0.148 (0.154)

LIO–LLD LIO–LLTES

19 24

0.035 (0.090) 0.153 (0.170)

16 –

0.029 (0.057) –

LIO–RLD

15

0.066 (0.124)

18

0.072 (0.121)

LIO–RLTES

16

0.027 (0.199)

21

0.095 (0.134)

LEO–RLTES REO–LLD

–

REO–LLTES REO–RLD REO–RLTES

0

Direction

–

16

0.078 (0.151)

17

0.044 (0.085)

RIO–LLTES

–

–

–

16

0.058 (0.144)

RIO–RLD RIO–RLTES

0 0

20 15

– –

15 25

0.015 (0.110) 0.136 (0.130)

LRA–LLD LRA–LLTES

–

– 19

– 0.080 (0.139)

0 0

19 24

– –

LRA–RLD LRA–RLTES RRA–LLD RRA–LLTES RRA–RLD RRA–RLTES LLES–LLD

0 – 0 0 0 0

18 – 19 15 23 20 28

– – – – – – 0.129 (0.133)

0 – 0 0 –

15 – 20 23 – 18 –

– – – – – 0.046 (0.114) –

LLES–LLTES LLES–RLD

20 19

0.023 (0.146) 0.062 (0.161)

–

– 18

– 0.018 (0.091)

LLES–RLTES

17

0.090 (0.248)

15

0.036 (0.165)

RLES–LLD

20

0.146 (0.185)

17

0.019 (0.069)

RLES–LLTES

27

0.155 (0.137)

15

0.029 (0.062)

20 18

– –

– 23

– 0.103 (0.129)

RIO–LLD

RLES–RLD RLES–RLTES

0 0

–

–

of the present study was relationships between muscles at different levels within the trunk. Further, upright standing trials in the present study consisted of 10 s of quiet standing, while the protocol of Nelson-Wong and Callaghan (2010) involved 2 h of prolonged standing. Therefore, activation strategies may alter over time with prolonged exposure to standing, or different strategies may be selected based on knowledge of the length of exposure. Time lag values for maximum flexion were within ranges previously reported for this task. Shan et al. (2014) reported time lags between the bilateral ES muscles ranging from 0.11 s to 0.16 s and 0.02 s to 0.91 s for trunk flexion and extension, respectively, while Nelson-Wong et al. (2012) observed time lags of 0.01 s, 0.20 s, and 0.10 s between the thoracic ES and lumbar ES, lumbar ES and gluteus maximus, and thoracic ES and gluteus maximus, respectively, during trunk flexion–extension in individuals without back pain. Additionally, the lumbar and lower-thoracic ES muscles were found in the present study to activate with a bottom-up sequence, further agreeing with the results of Nelson-Wong et al. (2012). The earlier activation of the lumbar ES may represent the need for stabilization of the lumbar spine, or preparation for the control of the movement of this region against gravity during flexion. The remaining lower – mid-level muscle pairings for which consistent activation sequences were observed displayed a top-down sequence, while a bottom-up sequence was consistently observed for the mid-level – upper muscle pairings. Taken together, these results generally indicate that the

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

75

mid-level muscles were activating prior to the muscles at the other levels. This may have been due to the design of the trunk ROM tasks, in that the head was moved prior to the trunk. Potentially, the mid-level muscles activated in a bottom-up sequence with the upper muscles to control the motion of the head, with the remaining muscles activating prior to the commencement of trunk movement. In slumped standing, time lags of zero were identified for the abdominals relative to the mid-level muscles, while a bottom-up sequence was observed for the lumbar ES relative to the mid-level muscles. The latter trend was present in at least 25 of 30 participants for pairings involving the lumbar and lower-thoracic ES muscles. As with maximum flexion, the earlier activation of the lumbar ES was likely due to the requirement for stabilization or movement control. Latissimus dorsi also activated prior to the upper muscles, while the remaining muscles demonstrated variable patterns. These variable patterns may have also been due to the control of head movement. In addition, although the instructions provided to participants regarding slumped standing were based on those previously reported (Callaghan & Dunk, 2002; Dankaerts et al., 2006), these were more subjective than those of the remaining tasks, possibly contributing to the variable activation sequences identified for slumped standing. The differences in activation patterns observed between maximum flexion and slumped standing, although both performed primarily in the sagittal plane, may be accounted for by the differences in movement patterns whereby the maximum flexion motion tended to occur in a superior to inferior direction along the spine, while slumped standing tended to elicit motion along the whole spine with approximately the same timing. For the movement tasks out of the sagittal plane (maximum lateral bend, maximum twist), activation sequences appeared to be more variable across muscle pairings, but showed approximately similar consistency across participants as the sagittal plane tasks. The lower – mid-level muscle pairings tended to display more consistent activation sequences across participants than the mid-level – upper muscle pairings. Specifically, lower – mid-level pairings with consistent activation sequences more often involved the abdominal muscles, as opposed to the lumbar ES. These findings may suggest the importance of the abdominal musculature in generating motion and the associated required spinal stiffness and stability during these movement tasks, complementing the results of Thelen, Schultz, and Ashton-Miller (1995) that increased abdominal co-contraction was identified during trunk lateral bend and axial twist exertions compared to flexion and extension exertions. Specific activation sequences involving the abdominals may be required for these movement tasks, as opposed to other movements in which there are a greater number of possible combinations of activation patterns to generate the movement, and therefore more variability in observed activation sequences. The examination of activation sequences in the lumbo-pelvic musculature has revealed important differences in activation sequences between individuals who do and do not develop transient back pain in response to a prolonged static exposure (Nelson-Wong & Callaghan, 2010; Nelson-Wong et al., 2012), as well as individuals with and without low back pain (Nelson-Wong et al., 2013). The aforementioned works have focused on the lumbo-pelvic musculature specifically, whereas the present study examined relationships between muscles at three different levels within the trunk. While the present study included only participants free of back pain, the results from individuals with low back pain may be compared to the present study, in order to identify differences in how individuals with low back pain activate their muscles relative to healthy individuals. Altered activation sequences observed in individuals with low back pain may contribute to maladaptive loading patterns along the length of the spine, thereby potentially contributing to the development or exacerbation of back pain (Nelson-Wong et al., 2012). There were several methodological concerns in the present study. The participant sample consisted of young adults asymptomatic for back pain. As pain status affects EMG measures (Nelson-Wong & Callaghan, 2010; Nelson-Wong et al., 2012; Nelson-Wong et al., 2013; Schinkel-Ivy et al., 2013; Watson et al., 1997), it cannot necessarily be assumed that individuals with low back pain would show similar activation sequences. The cross-correlation analyses were performed such that the entire task was analyzed as a single phase. However, additional information may be provided by dividing the task into smaller phases (for example, trunk flexion followed by return to upright standing), which may constitute a direction for future work. Further, the cross-correlation function is such that only two signals can be included, which was limiting in that a global indication of trunk muscle activation across all levels could not be determined concurrently. Principle component analysis could aid in addressing this limitation, and may be a direction for further study. The cut-off of 15 participants (i.e. at least 50% of participants) was selected to identify when consistent activation sequencing was present in a muscle pairing. As the aim of this study was to quantify the sequences most common to each muscle pairing and task, the presence of the sequence in the majority of participants was deemed sufficient. However, it may be prudent for future studies to focus on the sequences that were identified consistently across many of the participants (for example, if comparing healthy participants and those with low back pain, or if activation sequences are used to identify low back pain). Finally, the head and arm positions for the tasks of the present study were selected to promote a relatively consistent movement pattern. However, these standardized positions may have influenced the results of the analysis. Future work should seek to determine whether these results generalize to more functional activities with multiple movement strategies and greater involvement from both the upper and lower extremities. 5. Conclusion In conclusion, this study aimed to improve the understanding of the behavior of the trunk musculature through an in-depth investigation of activation sequences observed across different levels of the musculature. Activation sequences

76

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

were dependent on both the movement task and the level of the musculature. These results provide insight into the synergy of the trunk musculature that functions to produce trunk movements in the three planes of motion. Further, these findings are representative of activation sequencing in healthy individuals, and could potentially be used for comparison to individuals with low back pain. This would enable the identification of altered activation sequencing in this clinical population, which may be important in the study of low back pain as altered activation sequences in these individuals may contribute to maladaptive loading patterns and consequently the development or exacerbation of low back pain. Conflict of interest None. Acknowledgements Natural Sciences and Engineering Research Council (NSERC) of Canada and York University for funding. The funding sources had no role in the study design; in the collection, analysis, or interpretation of data; in the writing of the manuscript; or in the decision to submit the manuscript for publication. References Arlotta, M., LoVasco, G., & McLean, L. (2011). Selective recruitment of the lower fibers of the trapezius muscle. Journal of Electromyography and Kinesiology, 21, 403–410. Beneck, G. J., Baker, L. L., & Kulig, K. (2013). Spectral analysis of EMG using intramuscular electrodes reveals non-linear fatigability characteristics in persons with chronic low back pain. Journal of Electromyography and Kinesiology, 23, 70–77. Brereton, L. C., & McGill, S. M. (1998). Frequency response of spine extensors during rapid isometric contractions: Effects of muscle length and tension. Journal of Electromyography and Kinesiology, 8, 227–232. Briggs, A. M., Bragge, P., Smith, A. J., Govil, D., & Straker, L. M. (2009). Prevalence and associated factors for thoracic spine pain in the adult working population: A literature review. Journal of Occupational Health, 51, 177–192. Burnett, A., O’Sullivan, P. B., Caneiro, J. P., Krug, R., Bochmann, F., & Helgestad, G. W. (2009). An examination of the flexion–relaxation phenomenon in the cervical spine in lumbo-pelvic sitting. Journal of Electromyography and Kinesiology, 19(4), 229–236. Callaghan, J. P., & Dunk, N. M. (2002). Examination of the flexion relaxation phenomenon in erector spinae muscles during short duration slumped sitting. Clinical Biomechanics, 17, 353–360. Caneiro, J. P., O’Sullivan, P., Burnett, A., Barach, A., O’Neil, D., Tveit, O., et al (2010). The influence of different sitting postures on head/neck posture and muscle activity. Manual Therapy, 15, 54–60. Cholewicki, J., & McGill, S. M. (1996). Mechanical stability of the in vivo lumbar spine: Implications for injury and chronic low back pain. Clinical Biomechanics, 11(1), 1–15. Dankaerts, W., O’Sullivan, P., Burnett, A., & Straker, L. (2006). Altered patterns of superficial trunk muscle activation during sitting in nonspecific chronic low back pain patients: Importance of subclassification. Spine, 31(17), 2017–2023. De Luca, C. J. (1997). The use of surface electromyography in biomechanics. Journal of Applied Biomechanics, 13, 135–163. Drake, J. D. M., & Callaghan, J. P. (2006). Elimination of electrocardiogram contamination from electromyogram signals: An evaluation of currently used removal techniques. Journal of Electromyography and Kinesiology, 16, 175–187. Drake, J. D. M., Fischer, S. L., Brown, S. H., & Callaghan, J. P. (2006). Do exercise balls provide a training advantage for trunk extensor exercises? A biomechanical evaluation. Journal of Manipulative and Physiological Therapeutics, 29(5), 354–362. Edmondston, S. J., Sharp, M., Symes, A., Alhabib, N., & Allison, G. T. (2011). Changes in mechanical load and extensor muscle activity in the cervico-thoracic spine induced by sitting posture modification. Ergonomics, 54(2), 179–186. Edmondston, S. J., & Singer, K. P. (1997). Thoracic spine: Anatomical and biomechanical considerations for manual therapy. Manual Therapy, 2(3), 132–143. Granata, K. P., & Marras, W. S. (2000). Cost-benefit of muscle cocontraction in protecting against spinal instability. Spine, 25(11), 1398–1404. Jensen, C., Vasseljen, O., & Westgaard, R. H. (1993). The influence of electrode position on bipolar surface electromyogram recordings of the upper trapezius muscle. European Journal of Applied Physiology, 67, 266–273. Johnson, M. B., Cacciatore, T. W., Hamill, J., & Van Emmerik, R. E. A. (2010). Multi-segmental torso coordination during the transition from sitting to standing. Clinical Biomechanics, 25, 199–205. Knutson, L. M., Soderberg, G. L., Ballantyne, B. T., & Clarke, W. R. (1994). A study of various normalization procedures for within day electromyographic data. Journal of Electromyography and Kinesiology, 4, 47–59. McGill, S. M. (1991). Electromyographic activity of the abdominal and low back musculature during the generation of isometric and dynamic axial trunk torque: Implications for lumbar mechanics. Journal of Orthopaedic Research, 9(1), 91–103. McGill, S. M. (1992). A myoelectrically based dynamic three-dimensional model to predict loads on lumbar spine tissues during lateral bending. Journal of Biomechanics, 25(4), 395–414. McLean, L. (2005). The effect of postural correction on muscle activation amplitudes recorded from the cervicobrachial region. Journal of Electromyography and Kinesiology, 15(6), 527–535. Mirka, G. A., & Marras, W. S. (1993). A stochastic model of trunk muscle coactivation during trunk bending. Spine, 18(11), 1396–1409. Nairn, B. C., Azar, N. R., & Drake, J. D. M. (2013a). Transient pain developers show increased abdominal muscle activity during prolonged sitting. Journal of Electromyography and Kinesiology, 23, 1421–1427. Nairn, B. C., Chisholm, S. R., & Drake, J. D. M. (2013b). What is slumped sitting? A kinematic and electromyographical evaluation. Manual Therapy, 18, 498–505. Nairn, B. C., & Drake, J. D. M. (2014). Impact of lumbar spine posture on thoracic spine motion and muscle activation patterns. Human Movement Science, 37, 1–11. Nelson-Wong, E., Alex, B., Csepe, D., Lancaster, D., & Callaghan, J. P. (2012). Altered muscle recruitment during extension from trunk flexion in low back pain developers. Clinical Biomechanics, 27, 994–998. Nelson-Wong, E., & Callaghan, J. P. (2010). Is muscle co-activation a predisposing factor for low back pain development during standing? A multifactorial approach for early identification of at-risk individuals. Journal of Electromyography and Kinesiology, 20, 256–263. Nelson-Wong, E., Gregory, D. E., Winter, D. A., & Callaghan, J. P. (2008). Gluteus medius muscle activation patterns as a predictor of low back pain during standing. Clinical Biomechanics, 23, 545–553. Nelson-Wong, E., Howarth, S., Winter, D. A., & Callaghan, J. P. (2009). Application of autocorrelation and cross-correlation analyses in human movement and rehabilitation research. Journal of Orthopaedic & Sports Physical Therapy, 39(4), 287–295.

A. Schinkel-Ivy, J.D.M. Drake / Human Movement Science 43 (2015) 67–77

77

Nelson-Wong, E., Poupore, K., Ingvalson, S., Dehmer, K., Piatte, A., Alexander, S., et al (2013). Neuromuscular strategies for lumbopelvic control during frontal and sagittal plane movement challenges differ between people with and without low back pain. Journal of Electromyography and Kinesiology, 23, 1317–1324. Schinkel-Ivy, A., & Drake, J. D. M. (2015). Which motion segments are required to sufficiently characterize the kinematic behaviour of the trunk? Journal of Electromyography and Kinesiology, 25, 239–246. Schinkel-Ivy, A., Nairn, B. C., & Drake, J. D. M. (2013). Investigation of trunk muscle co-contraction and its association with low back pain development during prolonged sitting. Journal of Electromyography and Kinesiology, 23, 778–786. Shan, X., Wei, Y., Chen, Z., Fan, L., Shi, W., & Yang, S. (2014). Effect of leg support on muscle cross-correlation of bilateral erector spinae during trunk flexion– extension performance. Gait and Posture, 39, 161–165. Shum, G. L. K., Crosbie, J., & Lee, R. Y. W. (2007). Movement coordination of the lumbar spine and hip during a picking up activity in low back pain subjects. European Spine Journal, 16, 749–758. Solomonow, M., Baratta, R. V., Banks, A., Freudenberger, C., & Zhou, B. H. (2003). Flexion–relaxation response to static lumbar flexion in males and females. Clinical Biomechanics, 18, 273–279. Sparto, P. J., & Parnianpour, M. (2001). Generalizability of trunk muscle EMG and spinal forces. IEEE Engineering in Medicine and Biology Magazine, 20(6), 72–81. Thelen, D. G., Schultz, A. B., & Ashton-Miller, J. A. (1995). Co-contraction of lumbar muscles during the development of time-varying triaxial moments. Journal of Orthopaedic Research, 13, 390–398. Watson, P. J., Booker, C. K., Main, C. J., & Chen, A. C. N. (1997). Surface electromyography in the identification of chronic low back pain patients: The development of the flexion relaxation ratio. Clinical Biomechanics, 12(3), 165–171. Zipp, P. (1982). Recommendations for the standardization of lead positions in surface electromyography. European Journal of Applied Physiology, 50, 41–54.