Manual Therapy 20 (2015) 412e419
Contents lists available at ScienceDirect
Manual Therapy journal homepage: www.elsevier.com/math
Original article
Effects of external pelvic compression on electromyographic activity of the hamstring muscles during unipedal stance in sportsmen with and without hamstring injuries Ashokan Arumugam a, *, Stephan Milosavljevic b, Stephanie Woodley c, Gisela Sole a a b c
Centre for Health, Activity and Rehabilitation Research, School of Physiotherapy, University of Otago, New Zealand School of Physical Therapy, University of Saskatchewan, Canada Department of Anatomy, University of Otago, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 February 2014 Received in revised form 9 July 2014 Accepted 16 October 2014
There is some evidence that hamstring function can be influenced by interventions focusing on the pelvis via an anatomic and neurophysiologic link between these two segments. Previous research demonstrated increased electromyographic activity from injured hamstrings during transition from bipedal to unipedal stance (BUS). The aim of this study was to investigate the effects of a pelvic compression belt (PCB) on electromyographic activity of selected muscles during BUS in sportsmen with and without hamstring injury. Electromyographic amplitudes (normalised to maximum voluntary isometric contraction [MVIC]) of the hamstrings, gluteus maximus, gluteus medius and lumbar multifidus were obtained during BUS from 20 hamstring-injured participants (both sides) and 30 healthy participants (one side, randomly selected). There was an increase in biceps femoris (by 1.23 ± 2.87 %MVIC; p ¼ 0.027) and gluteus maximus (by 0.63 ± 1.13 %MVIC; p ¼ 0.023) electromyographic activity for the hamstringinjured side but no significant differences other than a decrease in multifidus activity (by 1.36 ± 2.92 %MVIC; p ¼ 0.023) were evident for healthy participants while wearing the PCB. However, the effect sizes for these findings were small. Wearing the PCB did not significantly change electromyographic activity of other muscles in either participant group (p > 0.050). Moreover, the magnitude of change induced by the PCB was not significantly different between groups (p > 0.050) for the investigated muscles. Thus, application of a PCB to decrease electromyographic activity of injured hamstrings during BUS is likely to have little effect. Similar research is warranted in participants with acute hamstring injury. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Electromyography Hamstring injury Pelvic compression belt Unipedal stance
1. Introduction Hamstring injuries are common in sports that involve highvelocity running or extensive lengthening of the hamstring muscles (Askling et al., 2012) and it has been suggested that loss of optimal neuromotor control may contribute towards occurrence and recurrence of these injuries (Cameron et al., 2003; Sole et al., 2008). An inherent anatomic link exists between the hamstrings and the sacroiliac joint (SIJ) via the sacrotuberous ligament (Vleeming et al., 1989; Woodley and Mercer, 2005) and it is through this structural pathway that biomechanical and neuromotor alterations of lumbopelvic function is thought to influence hamstring function (Mason et al., 2007; Panayi, 2010). Clinical examination of * Corresponding author. Tel.: þ64 3 479 9619 ; fax: þ64 3 479 8367. E-mail address: [email protected] (A. Arumugam). http://dx.doi.org/10.1016/j.math.2014.10.011 1356-689X/© 2014 Elsevier Ltd. All rights reserved.
the SIJ often includes assessment of transition from bipedal to unipedal stance (BUS) using Gillet's test (Potter and Rothstein, 1985; Sturesson et al., 2000). Variations of this technique have also been implemented in the assessment of neuromotor control of lumbopelvic and/or lower limb muscles in patients with SIJ, pelvic, groin and lower limb injuries (Hungerford et al., 2003; van Deun et al., 2007; Morrissey et al., 2012; Sole et al., 2012; Jung et al., 2013). Transitions from BUS are fundamental for the initiation of gait (Rogers and Pai, 1993) and to perform sporting techniques such as kicking, shooting or passing a ball (Paillard et al., 2006). Investigations of BUS are useful for examining motor control of the lumbopelvic and hamstring muscles as this task requires pelvic stabilisation and is functionally relevant to walking and climbing stairs (Morrissey et al., 2012). A recent systematic review on the effects of external pelvic compression reported moderate evidence to support the role of
A. Arumugam et al. / Manual Therapy 20 (2015) 412e419
pelvic compression in altering electromyographic (EMG) activity of muscles associated with the pelvis (Arumugam et al., 2012b). More recently, Jung et al. (2013) reported that wearing a pelvic compression belt (PCB) decreases biceps femoris (BF) activity and increases gluteus maximus (GMa) activity during BUS in participants with and without SIJ pain. A decrease in BF EMG has also been demonstrated during standing (Snijders et al., 1998) and walking (Hu et al., 2010) with application of a PCB in healthy individuals. In addition, Hu et al. (2010) documented an increase in GMa activity during walking in healthy women, which might indicate that the need for the BF to extend the hip could be compensated by increased recruitment of the GMa with the PCB. Previous studies have reported abnormal recruitment patterns, evident by earlier EMG onset and/or increased amplitudes, of the BF on the affected side during BUS in patients with unilateral SIJ pain (Hungerford et al., 2003) and with hamstring injuries (Sole, 2008; Sole et al., 2012). Increased (aberrant) recruitment of injured hamstrings might predispose to further (re)injury (Sole et al., 2008). It is unknown whether application of a PCB has an effect on EMG of injured hamstrings; however, various hypothetical mechanisms underpinning the effects of a PCB on hamstring function have been proposed (Arumugam et al., 2012a). We have recently reported that application of the PCB increased isokinetic eccentric muscle strength in the outer range of movement for participants with hamstring injuries, suggesting that application of such a belt had effects on neuromotor control of the hamstrings (Arumugam et al., 2014). The aim of the current study was to investigate changes in recruitment patterns of the hamstrings, glutei and lumbar multifidi (MF) with application of a PCB during BUS in the same group of sportsmen with and without recent hamstring injuries. We hypothesised that application of the PCB would reduce EMG activity of the hamstrings in both study groups (Arumugam et al., 2012b). 2. Methods 2.1. Study design and participants A cross-over design was used in which the order of belt conditions (PCB vs. no PCB) was randomised. Ethical approval was received from the University of Otago Human Ethics Committee (Reference e 11/115). All participants provided written informed consent before participating in the study. Data collection was conducted at the Biomechanics laboratory, School of Physiotherapy, University of Otago, Dunedin, New Zealand.
413
Participants were recruited from the University, local sports clubs and physiotherapy clinics using posters, flyers and emails. The eligibility criteria (Sole et al., 2012) are summarised in Table 1. The same group of sportsmen also participated in a study investigating the effects of the application of the PCB during isokinetic strength testing (Arumugam et al., 2014) and walking. Data for weightbearing tasks (BUS followed by walking) were collected during the first session while the isokinetic task was performed during a second session within one week (Arumugam, 2014). Participants underwent musculoskeletal screening to confirm their eligibility. Anthropometric measurements were recorded and footedness was determined based on self-declared leg preference when kicking a ball (Teixeira and Teixeira, 2008). Bilateral hamstring flexibility was assessed simultaneously using the sit-and-reach test (Liemohn et al., 1994). 2.2. Procedures Data were collected from one side (left/right) for the healthy participants and both sides for the hamstring-injured participants. The choice of leg to be tested for the healthy participants and the order of testing for the hamstring-injured participants were randomised using computer-generated numbers list. Standard guidelines recommended by the Surface Electromyography for the Non-Invasive Assessment of Muscles (SENIAM) committee for skin preparation and electrode placement were followed (Hermens et al., 1999, 2000). Two Ag/AgCl surface electrodes were placed on the targeted site on each muscle at an interelectrode distance of 2 cm. The ground electrode was placed on the L2 spinous process. Data were collected only when skin impedance, measured with a multimeter (Tequipment.NET™, NJ), was less than 10 KU (Konrad, 2005) and negligible crosstalk was observed during voluntary muscle contraction. A 16 channel Noraxon Telemyo™ 2400 T G2 system (Noraxon Inc., AZ) with MyoResearch XP MasterEdition software™ V1.06.54 was used to record and process EMG data. Conventional manual muscle testing positions (Daniels et al., 2007) were adopted to record maximum voluntary of isometric contraction (MVIC) of the MF, GMa, gluteus medius, and hamstrings. A total of three MVIC trials were performed for each muscle and participants were encouraged verbally to produce a maximal contraction. Each trial lasted for 5 s and a rest period of 1 min was allowed between trials. EMG data were recorded for 3 s of each trial after the EMG signals reached a steady state following the first second.
Table 1 Eligibility criteria for recruiting participants. Hamstring-injured groupa
Healthy group
Inclusion criteria - An onset of pain in the posterior thigh while playing a sport (nonimpact) within - No previous history of hamstring injury that was diagnosed and treated by a the previous 12 months, but not less than a month health professional - The injury was severe enough to necessitate intervention from a health professional or prevent participation in at least one match or competition (Bennell et al., 1998; Orchard, 1998), and one weekb of regular sports training (Brockett et al., 2004), within the previous 12 months - History of unilateral or bilateral, first-time or recurrent hamstring injuries Exclusion criteria - Trauma or pathology in the knee joint or lumbopelvic region within the last six months that was treated by a health professional or prevented involvement in at least one week of training sessions, a competition or match - Evidence of any neuromotor or musculoskeletal abnormality of the lumbopelvic region and/or the lower limb during clinical examination (Laslett et al., 2005; Laslett, 2008; Petty, 2011) - Ongoing musculoskeletal (lumbopelvic and/or lower limb), neurological, cardiorespiratory, inflammatory or other systemic disorder a
Sportsmen were included based on their self-reported history of hamstring injury within the past 12 months as it is reported to be reliable (Gabbe et al., 2003). Based on the period of absence from sports participation, injury severity was classified as minor (7 days), moderate (8e21 days) or severe (>21 days) (Arnason et al., 2008). b
414
A. Arumugam et al. / Manual Therapy 20 (2015) 412e419
TeleMyo™ 2400T-G2 transmitter. EMG of only the weight-bearing leg was registered during each 5 s trial. Simultaneously, the ground reaction force of each leg and the light signal were recorded using Cortex™ software-V2.0.2.917. A minimum of five trials were performed for each condition (PCB vs. no PCB). 2.3. Data processing
Fig. 1. Location of external pelvic compression e pelvic compression belt applied below the anterior superior iliac spines.
Participants stood comfortably with each foot on one of two force plates (BP2436 and OR6-5, Advanced Medical Technologies, MA), wearing their sports/leisure shoes. The PCB (SI-brace neoprene-ADL-anatomisch, 3200202; Rafys, The Netherlands) was fitted below the anterior superior iliac spines, applied with maximal manual tension without causing any discomfort to the participant (Fig. 1.). A custom-made stand with two bulbs (one green and one red) was placed in front of the participant. The choice of leg to be lifted was indicated by the light colour: red signalled the left leg and green the right leg (Sole et al., 2012). As soon as either light was randomly switched on by the researcher (AA), the participant raised his foot to a height level with that of the top of a wooden block (measuring 18.5 cm in height), by flexing his hip and knee as quickly as possible. Each trial lasted for 5 s and the limb was lowered to the floor when the light bulb was turned off. The sequencing of the light signals was randomly determined, therefore, participants were unaware of subsequent moves. A minimum of four practice trials was undertaken to familiarise participants with the task. Participants walked (Damen et al., 2002) for 5 min (Jung et al., 2013) between the belt conditions to minimise any carryover effects of the PCB which could potentially confound the outcome variables. EMG signal recording was instigated using a manual trigger via a wireless sync trigger unit (232 Transmitter, Noraxon). The synchronisation pulse was sent to a wireless sync receiver unit (Model234 Inline-wireless-sync-receiver, Noraxon) which was positioned in front of the participant's chest after being connected to a
Onset of movement transition in BUS was defined by an increase in the vertical ground reaction force obtained from force plate data of the flexing leg by more than three standard deviations from the baseline value for more than 50 ms (Sims and Brauer, 2000). EMG data were processed with a fourth order Butterworth band-pass filter of 10e500 Hz (MyoResearch XP-V1.06.54). A 50 ms epoch RMS approach was used to process the filtered EMG signal using MATLAB® software (V12.0.0.58851, The Mathworks, Inc., MA). The mean EMG RMS value of 3 s of MVIC for each muscle was calculated and used as the anchor (denominator) to normalise the EMG amplitudes obtained during the transition phase of BUS (500 ms following the start of movement) and expressed as a percentage (Fig. 2). 2.4. Data analyses Data were inspected for normal distribution using the ShapiroeWilk test and histograms, as well as homogeneity of variances using Levene's test. If data were skewed, their log values were used for further analysis provided that the log-transformed data were normally distributed. Within-group comparisons (PCB vs. no PCB) using paired t-tests and between-group comparisons (hamstringinjured vs. healthy group) using independent t-tests were performed. Effect sizes (d) were determined using Cohen's formula and interpreted as small (0.20 d 0.50), medium (0.50 d 0.80) or large ( 0.8) (Cohen, 1988; Kinnear and Gray, 2006). In addition, Spearman's r correlation coefficient was used to correlate the change scores (with belt e no belt) of injured hamstring muscles and respective time since (recent) hamstring injury of the corresponding limb for all hamstring-injured participants. IBM-SPSS Version 20 (IBM, NY) was used for statistical analyses. 3. Results After screening, 20 (out of 37) volunteers with hamstring injuries and 30 (out of 36) healthy volunteers who met the eligibility criteria were included in the study (Table 2). In the hamstring-
Fig. 2. Maximum voluntary isometric contraction (MVIC) normalised electromyographic (EMG) root mean square (RMS) values (%) of the lumbar multifidus (MF), gluteus maximus (GMa), gluteus medius (GMe), biceps femoris (BF), and medial hamstring (MH) muscles of the weight-bearing limb of one participant during a trial of transition from bipedal to unipedal stance. T0 indicates the start of motion (anticipatory postural adjustment).
A. Arumugam et al. / Manual Therapy 20 (2015) 412e419
of the PCB on the injured side (p ¼ 0.068). However, no significant differences were found between the test conditions for normalised EMG RMS values of the medial hamstring, and gluteus medius during BUS in this group (Table 3, Fig. 3.). For the healthy participants, there was a significant decrease of 15% in MF activity (p ¼ 0.023) while wearing the PCB but the corresponding effect size was small (d ¼ 0.29). No other muscles showed a significant change in EMG with the PCB in the healthy group (Table 3, Fig. 4.). Between-group comparisons did not reveal any significant difference in EMG activity of any of the investigated muscles either for the baseline values recorded during trials without the PCB (Table 4) or for the magnitude of change induced by the PCB (Table 3). Further, no significant correlation was noted for injured hamstrings and the time since injury (Spearman's r 0.25).
Table 2 Participant characteristics. Variable
Hamstring-injured group (n ¼ 20)
Control group (n ¼ 30)
Age (years), mean (SD) 22.00 ± 1.45 Anthropometric measurements, mean (SD) Mass (kg) 85.52 ± 14.40 Height (m) 1.81 ± 0.08 2 BMI (kg/m ) 25.89 ± 3.38 b Body fat (%) 23.32 ± 3.42 Flexibility, mean (SD) Sit-and-reach (cm) 23.10 ± 6.46 Sports participation, (n) Soccer/Football 8 Rugby 9 Hockey 1 Long distance running 0 Sprinting 1 Triathlon 0 Ice hockey 0 Weight-lifting 0 Racquet sports 1 Basket ball 0 Cricket 0
p Value 0.046a
23.53 ± 3.68 70.86 ± 11.01 1.76 ± 0.08 22.92 ± 2.68 23.40 ± 4.53
Contents lists available at ScienceDirect
Manual Therapy journal homepage: www.elsevier.com/math
Original article
Effects of external pelvic compression on electromyographic activity of the hamstring muscles during unipedal stance in sportsmen with and without hamstring injuries Ashokan Arumugam a, *, Stephan Milosavljevic b, Stephanie Woodley c, Gisela Sole a a b c
Centre for Health, Activity and Rehabilitation Research, School of Physiotherapy, University of Otago, New Zealand School of Physical Therapy, University of Saskatchewan, Canada Department of Anatomy, University of Otago, New Zealand
a r t i c l e i n f o
a b s t r a c t
Article history: Received 14 February 2014 Received in revised form 9 July 2014 Accepted 16 October 2014
There is some evidence that hamstring function can be influenced by interventions focusing on the pelvis via an anatomic and neurophysiologic link between these two segments. Previous research demonstrated increased electromyographic activity from injured hamstrings during transition from bipedal to unipedal stance (BUS). The aim of this study was to investigate the effects of a pelvic compression belt (PCB) on electromyographic activity of selected muscles during BUS in sportsmen with and without hamstring injury. Electromyographic amplitudes (normalised to maximum voluntary isometric contraction [MVIC]) of the hamstrings, gluteus maximus, gluteus medius and lumbar multifidus were obtained during BUS from 20 hamstring-injured participants (both sides) and 30 healthy participants (one side, randomly selected). There was an increase in biceps femoris (by 1.23 ± 2.87 %MVIC; p ¼ 0.027) and gluteus maximus (by 0.63 ± 1.13 %MVIC; p ¼ 0.023) electromyographic activity for the hamstringinjured side but no significant differences other than a decrease in multifidus activity (by 1.36 ± 2.92 %MVIC; p ¼ 0.023) were evident for healthy participants while wearing the PCB. However, the effect sizes for these findings were small. Wearing the PCB did not significantly change electromyographic activity of other muscles in either participant group (p > 0.050). Moreover, the magnitude of change induced by the PCB was not significantly different between groups (p > 0.050) for the investigated muscles. Thus, application of a PCB to decrease electromyographic activity of injured hamstrings during BUS is likely to have little effect. Similar research is warranted in participants with acute hamstring injury. © 2014 Elsevier Ltd. All rights reserved.
Keywords: Electromyography Hamstring injury Pelvic compression belt Unipedal stance
1. Introduction Hamstring injuries are common in sports that involve highvelocity running or extensive lengthening of the hamstring muscles (Askling et al., 2012) and it has been suggested that loss of optimal neuromotor control may contribute towards occurrence and recurrence of these injuries (Cameron et al., 2003; Sole et al., 2008). An inherent anatomic link exists between the hamstrings and the sacroiliac joint (SIJ) via the sacrotuberous ligament (Vleeming et al., 1989; Woodley and Mercer, 2005) and it is through this structural pathway that biomechanical and neuromotor alterations of lumbopelvic function is thought to influence hamstring function (Mason et al., 2007; Panayi, 2010). Clinical examination of * Corresponding author. Tel.: þ64 3 479 9619 ; fax: þ64 3 479 8367. E-mail address: [email protected] (A. Arumugam). http://dx.doi.org/10.1016/j.math.2014.10.011 1356-689X/© 2014 Elsevier Ltd. All rights reserved.
the SIJ often includes assessment of transition from bipedal to unipedal stance (BUS) using Gillet's test (Potter and Rothstein, 1985; Sturesson et al., 2000). Variations of this technique have also been implemented in the assessment of neuromotor control of lumbopelvic and/or lower limb muscles in patients with SIJ, pelvic, groin and lower limb injuries (Hungerford et al., 2003; van Deun et al., 2007; Morrissey et al., 2012; Sole et al., 2012; Jung et al., 2013). Transitions from BUS are fundamental for the initiation of gait (Rogers and Pai, 1993) and to perform sporting techniques such as kicking, shooting or passing a ball (Paillard et al., 2006). Investigations of BUS are useful for examining motor control of the lumbopelvic and hamstring muscles as this task requires pelvic stabilisation and is functionally relevant to walking and climbing stairs (Morrissey et al., 2012). A recent systematic review on the effects of external pelvic compression reported moderate evidence to support the role of
A. Arumugam et al. / Manual Therapy 20 (2015) 412e419
pelvic compression in altering electromyographic (EMG) activity of muscles associated with the pelvis (Arumugam et al., 2012b). More recently, Jung et al. (2013) reported that wearing a pelvic compression belt (PCB) decreases biceps femoris (BF) activity and increases gluteus maximus (GMa) activity during BUS in participants with and without SIJ pain. A decrease in BF EMG has also been demonstrated during standing (Snijders et al., 1998) and walking (Hu et al., 2010) with application of a PCB in healthy individuals. In addition, Hu et al. (2010) documented an increase in GMa activity during walking in healthy women, which might indicate that the need for the BF to extend the hip could be compensated by increased recruitment of the GMa with the PCB. Previous studies have reported abnormal recruitment patterns, evident by earlier EMG onset and/or increased amplitudes, of the BF on the affected side during BUS in patients with unilateral SIJ pain (Hungerford et al., 2003) and with hamstring injuries (Sole, 2008; Sole et al., 2012). Increased (aberrant) recruitment of injured hamstrings might predispose to further (re)injury (Sole et al., 2008). It is unknown whether application of a PCB has an effect on EMG of injured hamstrings; however, various hypothetical mechanisms underpinning the effects of a PCB on hamstring function have been proposed (Arumugam et al., 2012a). We have recently reported that application of the PCB increased isokinetic eccentric muscle strength in the outer range of movement for participants with hamstring injuries, suggesting that application of such a belt had effects on neuromotor control of the hamstrings (Arumugam et al., 2014). The aim of the current study was to investigate changes in recruitment patterns of the hamstrings, glutei and lumbar multifidi (MF) with application of a PCB during BUS in the same group of sportsmen with and without recent hamstring injuries. We hypothesised that application of the PCB would reduce EMG activity of the hamstrings in both study groups (Arumugam et al., 2012b). 2. Methods 2.1. Study design and participants A cross-over design was used in which the order of belt conditions (PCB vs. no PCB) was randomised. Ethical approval was received from the University of Otago Human Ethics Committee (Reference e 11/115). All participants provided written informed consent before participating in the study. Data collection was conducted at the Biomechanics laboratory, School of Physiotherapy, University of Otago, Dunedin, New Zealand.
413
Participants were recruited from the University, local sports clubs and physiotherapy clinics using posters, flyers and emails. The eligibility criteria (Sole et al., 2012) are summarised in Table 1. The same group of sportsmen also participated in a study investigating the effects of the application of the PCB during isokinetic strength testing (Arumugam et al., 2014) and walking. Data for weightbearing tasks (BUS followed by walking) were collected during the first session while the isokinetic task was performed during a second session within one week (Arumugam, 2014). Participants underwent musculoskeletal screening to confirm their eligibility. Anthropometric measurements were recorded and footedness was determined based on self-declared leg preference when kicking a ball (Teixeira and Teixeira, 2008). Bilateral hamstring flexibility was assessed simultaneously using the sit-and-reach test (Liemohn et al., 1994). 2.2. Procedures Data were collected from one side (left/right) for the healthy participants and both sides for the hamstring-injured participants. The choice of leg to be tested for the healthy participants and the order of testing for the hamstring-injured participants were randomised using computer-generated numbers list. Standard guidelines recommended by the Surface Electromyography for the Non-Invasive Assessment of Muscles (SENIAM) committee for skin preparation and electrode placement were followed (Hermens et al., 1999, 2000). Two Ag/AgCl surface electrodes were placed on the targeted site on each muscle at an interelectrode distance of 2 cm. The ground electrode was placed on the L2 spinous process. Data were collected only when skin impedance, measured with a multimeter (Tequipment.NET™, NJ), was less than 10 KU (Konrad, 2005) and negligible crosstalk was observed during voluntary muscle contraction. A 16 channel Noraxon Telemyo™ 2400 T G2 system (Noraxon Inc., AZ) with MyoResearch XP MasterEdition software™ V1.06.54 was used to record and process EMG data. Conventional manual muscle testing positions (Daniels et al., 2007) were adopted to record maximum voluntary of isometric contraction (MVIC) of the MF, GMa, gluteus medius, and hamstrings. A total of three MVIC trials were performed for each muscle and participants were encouraged verbally to produce a maximal contraction. Each trial lasted for 5 s and a rest period of 1 min was allowed between trials. EMG data were recorded for 3 s of each trial after the EMG signals reached a steady state following the first second.
Table 1 Eligibility criteria for recruiting participants. Hamstring-injured groupa
Healthy group
Inclusion criteria - An onset of pain in the posterior thigh while playing a sport (nonimpact) within - No previous history of hamstring injury that was diagnosed and treated by a the previous 12 months, but not less than a month health professional - The injury was severe enough to necessitate intervention from a health professional or prevent participation in at least one match or competition (Bennell et al., 1998; Orchard, 1998), and one weekb of regular sports training (Brockett et al., 2004), within the previous 12 months - History of unilateral or bilateral, first-time or recurrent hamstring injuries Exclusion criteria - Trauma or pathology in the knee joint or lumbopelvic region within the last six months that was treated by a health professional or prevented involvement in at least one week of training sessions, a competition or match - Evidence of any neuromotor or musculoskeletal abnormality of the lumbopelvic region and/or the lower limb during clinical examination (Laslett et al., 2005; Laslett, 2008; Petty, 2011) - Ongoing musculoskeletal (lumbopelvic and/or lower limb), neurological, cardiorespiratory, inflammatory or other systemic disorder a
Sportsmen were included based on their self-reported history of hamstring injury within the past 12 months as it is reported to be reliable (Gabbe et al., 2003). Based on the period of absence from sports participation, injury severity was classified as minor (7 days), moderate (8e21 days) or severe (>21 days) (Arnason et al., 2008). b
414
A. Arumugam et al. / Manual Therapy 20 (2015) 412e419
TeleMyo™ 2400T-G2 transmitter. EMG of only the weight-bearing leg was registered during each 5 s trial. Simultaneously, the ground reaction force of each leg and the light signal were recorded using Cortex™ software-V2.0.2.917. A minimum of five trials were performed for each condition (PCB vs. no PCB). 2.3. Data processing
Fig. 1. Location of external pelvic compression e pelvic compression belt applied below the anterior superior iliac spines.
Participants stood comfortably with each foot on one of two force plates (BP2436 and OR6-5, Advanced Medical Technologies, MA), wearing their sports/leisure shoes. The PCB (SI-brace neoprene-ADL-anatomisch, 3200202; Rafys, The Netherlands) was fitted below the anterior superior iliac spines, applied with maximal manual tension without causing any discomfort to the participant (Fig. 1.). A custom-made stand with two bulbs (one green and one red) was placed in front of the participant. The choice of leg to be lifted was indicated by the light colour: red signalled the left leg and green the right leg (Sole et al., 2012). As soon as either light was randomly switched on by the researcher (AA), the participant raised his foot to a height level with that of the top of a wooden block (measuring 18.5 cm in height), by flexing his hip and knee as quickly as possible. Each trial lasted for 5 s and the limb was lowered to the floor when the light bulb was turned off. The sequencing of the light signals was randomly determined, therefore, participants were unaware of subsequent moves. A minimum of four practice trials was undertaken to familiarise participants with the task. Participants walked (Damen et al., 2002) for 5 min (Jung et al., 2013) between the belt conditions to minimise any carryover effects of the PCB which could potentially confound the outcome variables. EMG signal recording was instigated using a manual trigger via a wireless sync trigger unit (232 Transmitter, Noraxon). The synchronisation pulse was sent to a wireless sync receiver unit (Model234 Inline-wireless-sync-receiver, Noraxon) which was positioned in front of the participant's chest after being connected to a
Onset of movement transition in BUS was defined by an increase in the vertical ground reaction force obtained from force plate data of the flexing leg by more than three standard deviations from the baseline value for more than 50 ms (Sims and Brauer, 2000). EMG data were processed with a fourth order Butterworth band-pass filter of 10e500 Hz (MyoResearch XP-V1.06.54). A 50 ms epoch RMS approach was used to process the filtered EMG signal using MATLAB® software (V12.0.0.58851, The Mathworks, Inc., MA). The mean EMG RMS value of 3 s of MVIC for each muscle was calculated and used as the anchor (denominator) to normalise the EMG amplitudes obtained during the transition phase of BUS (500 ms following the start of movement) and expressed as a percentage (Fig. 2). 2.4. Data analyses Data were inspected for normal distribution using the ShapiroeWilk test and histograms, as well as homogeneity of variances using Levene's test. If data were skewed, their log values were used for further analysis provided that the log-transformed data were normally distributed. Within-group comparisons (PCB vs. no PCB) using paired t-tests and between-group comparisons (hamstringinjured vs. healthy group) using independent t-tests were performed. Effect sizes (d) were determined using Cohen's formula and interpreted as small (0.20 d 0.50), medium (0.50 d 0.80) or large ( 0.8) (Cohen, 1988; Kinnear and Gray, 2006). In addition, Spearman's r correlation coefficient was used to correlate the change scores (with belt e no belt) of injured hamstring muscles and respective time since (recent) hamstring injury of the corresponding limb for all hamstring-injured participants. IBM-SPSS Version 20 (IBM, NY) was used for statistical analyses. 3. Results After screening, 20 (out of 37) volunteers with hamstring injuries and 30 (out of 36) healthy volunteers who met the eligibility criteria were included in the study (Table 2). In the hamstring-
Fig. 2. Maximum voluntary isometric contraction (MVIC) normalised electromyographic (EMG) root mean square (RMS) values (%) of the lumbar multifidus (MF), gluteus maximus (GMa), gluteus medius (GMe), biceps femoris (BF), and medial hamstring (MH) muscles of the weight-bearing limb of one participant during a trial of transition from bipedal to unipedal stance. T0 indicates the start of motion (anticipatory postural adjustment).
A. Arumugam et al. / Manual Therapy 20 (2015) 412e419
of the PCB on the injured side (p ¼ 0.068). However, no significant differences were found between the test conditions for normalised EMG RMS values of the medial hamstring, and gluteus medius during BUS in this group (Table 3, Fig. 3.). For the healthy participants, there was a significant decrease of 15% in MF activity (p ¼ 0.023) while wearing the PCB but the corresponding effect size was small (d ¼ 0.29). No other muscles showed a significant change in EMG with the PCB in the healthy group (Table 3, Fig. 4.). Between-group comparisons did not reveal any significant difference in EMG activity of any of the investigated muscles either for the baseline values recorded during trials without the PCB (Table 4) or for the magnitude of change induced by the PCB (Table 3). Further, no significant correlation was noted for injured hamstrings and the time since injury (Spearman's r 0.25).
Table 2 Participant characteristics. Variable
Hamstring-injured group (n ¼ 20)
Control group (n ¼ 30)
Age (years), mean (SD) 22.00 ± 1.45 Anthropometric measurements, mean (SD) Mass (kg) 85.52 ± 14.40 Height (m) 1.81 ± 0.08 2 BMI (kg/m ) 25.89 ± 3.38 b Body fat (%) 23.32 ± 3.42 Flexibility, mean (SD) Sit-and-reach (cm) 23.10 ± 6.46 Sports participation, (n) Soccer/Football 8 Rugby 9 Hockey 1 Long distance running 0 Sprinting 1 Triathlon 0 Ice hockey 0 Weight-lifting 0 Racquet sports 1 Basket ball 0 Cricket 0
p Value 0.046a
23.53 ± 3.68 70.86 ± 11.01 1.76 ± 0.08 22.92 ± 2.68 23.40 ± 4.53
