Activation of latissimus dorsi muscle in patients with sacroiliac joint dysfunction when lifting a load

Abstract

BACKGROUND:

Sacroiliac Joint Dysfunction (SIJD) may be observed in 13% to 30% people with idiopathic low back pain (LBP). Latissimus dorsi (LD) muscle works by stabilizing the SIJ, providing a pathway for force transmission through the thoraco lumbar fascia. Literature has shown that muscles can change their activation pattern in response to pain, altering motor control. However, to date, there have been no studies evaluating the EMG activity of LD in people with SIJD while lifting a load, this could guide for a better understanding about how muscle activation occurs in this group of patients.

AIM:

To evaluate and compare activation of LD in people with LBP, SIJD and without LBP in load lifting.

METHODS:

One hundred fourteen people were evaluated and divided into 3 groups: LBP, SIJD and Control group. EMG signals were recorded from LD while the subjects lifted a load in a symmetrical posture. Subjects started in an upright position, grasped the box from the floor and returned to the initial position with flexed elbows. Root Mean Square (RMS) amplitude and latency were calculated. The Kruskal-Wallis and the post hoc Dunn’s tests were used to compare groups.

RESULTS:

Results showed that left LD in SIJD group is activated 26.21% more than in the control group and 23.98% than the LBP group (p = 0.02). Besides, right LD has a delayed onset in SIJD by 0.68 ms compare with the control group and 0.29 ms with LBP group (p = 0.03).

CONCLUSION:

In a specific group of individuals with SIJD, alterations in LD muscle activation, could be evidenced by an increased RMS amplitude which is accompanied with a delay in activation in the opposite side during lifting a load in a symmetrical posture.

Keywords

Low back pain sacroiliac joint dysfunction electromyography latissimus dorsi

1 Background

Sacroiliac joint dysfunction (SIJD) is a pathomechanical disturbance of the sacroiliac joint (SIJ), which is characterized by the combination of pelvic rotation, joint block, hypo or hypermobility, and muscle imbalance [1]. The pain caused by SIJD is located below the fifth lumbar vertebra and can be localized in the joint and/or radiated to several regions [2].

It has been established that the SIJD has an important impact in quality of life in a similar magnitude as it can be observed in patients with debilitating orthopedic conditions, such as in hip osteoarthritis and spinal stenosis, and its impact is higher than many related cardiovascular medical conditions [3]. In addition, it has been suggested that SIJD may cause between 13% to 30% of non-specific low back pain (LBP) [4].

SIJ is a relatively flat joint, aligned close to the vertical plane therefore its stability largely depends on the surrounding myofascial structures [1]. Anatomical and biomechanical studies have outlined the importance of latissimus dorsi (LD) muscle in SIJ stability [5, 6], which through its synchronous interaction with opposite gluteus maximus muscle, provides a pathway for the forces transmission through the posterior layer of the thoraco lumbar fascia (TLF) [7]. During daily activities, alterations in coordinated work between LD and gluteus maximus affects SIJ stability disrupting the pelvic girdle capacity to transmit and dissipate forces generated from lower limbs (LL) to the lumbar spine and vice versa [2].

Previous studies have reported that muscle activation patterns can change in response to pain, which can lead to altered motor control [2, 8]. It has been found that people with SIJD shown delays in the LD onset on the affected side, when rising from a seated position task. This result suggests muscle deficiencies in the movement anticipation during functional activities [9] which relies not only on muscle activation but also on muscle deactivation as well. There are few studies about LD activation in people with SIJD, and, to our knowledge, its EMG behavior has not been evaluated during other important functional activities such as lifting loads, which is an essential part of human daily living activities.

To study LD muscle activation is necessary to understand the motor control changes that can occur in people with SIJD and which can contribute to symptoms perpetuation in this population. Besides, this knowledge could help to establish suitable treatment guides that take in consideration muscular activation alterations, allowing to correct these impairments, improving muscular performance, functionality, and quality of life in people with SIJD. Therefore, this study aimed to evaluate the EMG activity of LD when lifting a load in people with LBP, with and without SIJD and people without LBP. We hypothesized that persons with SIJD exhibit alterations in muscle activity compared with other participants of study.

2 Methods

A cross-sectional study was performed. Subjects were selected from a convenience sampling, using Snowball sampling as a recruitment technique. The sample was composed by 114 individuals (men and women), aged between 18 and 40 years, with a residence in the city of Bucaramanga (Colombia) and its metropolitan area. They were allocated into three study groups: SIJD Group –composed by participants with LBP and a positive diagnosis of SIJD, which was established by a trained physical therapist through the Laslett's multitest regimen [11]; LBP Group - composed by participants with LBP and a negative result for the diagnosis of SIJ dysfunction; Control Group –composed by participants without LBP or SIJD with the same age range as the participants of the LBP and SIJ dysfunction groups.

Inclusion criteria for the LBP and SIJ dysfunction groups were at least one episode of continuous or intermittent LBP over the last 6 months, located between L1 and the gluteal fold, with pain intensity less than or equal to 4 according to the visual analog scale at rest and during SIJ palpation. The exclusion criteria for the three groups were: fibromyalgia, diagnosed intervertebral disc diseases, diagnosed sacroiliac joint disease, pregnant women and women who were in the luteal phase of their menstrual cycle. For the latter criterion if participant met other selection criteria, the assessment was scheduled at a different phase of the menstrual cycle.

All procedures were in accordance with the guidelines of the Declaration of Helsinki and Resolution 008430 of 1993 of the Ministry of Health of Colombia in which the scientific, technical, and administrative regulations for health research are enacted [18]. This research was classified in the category of minimum risk and was approved by the Scientific Research Ethics Committee of Industrial University of Santander. All participants signed a consent form prior to participation.

2.1 Procedure

Electromyographic recording: Surface electromyography (sEMG) was recorded bilaterally during the functional task of lifting a load. The test was performed using a mobile device (Biometrics, Ltd, UK) with a 14-bit analog-to-digital (A/D) conversion board, amplifier with gain range of 1000 fold, analog frequency filter from 20 to 450 Hz, common mode rejection ratio (CMRR) of 96 dB or more, signal-to-noise ratio less than 5 microvolts and impedance greater than 10 [14] Ohms. The EMG was recorded using 10-mm circular active surface electrodes (Biometrics, Ltd, UK), Ag/AgCl and inter-electrode distance of 10 mm [11].

Before placing the electrodes, the skin was first shaved and cleaned after with alcohol [12]. The reference electrode was located on the spiny process of the C7 vertebra, and the recording electrodes were placed 3 cm laterally at the scapula’s lower angle [13]. Subsequently, the submaximal voluntary contraction (SVC) was recorded during a 5-second isometric contraction [14], while the participant was in prone position and asked to place the arm in adduction, internal rotation and shoulder extension while the elbow remained extended. The evaluator stabilized the pelvis and applied resistance on the lower third of the forearm in opposite movements. This procedure was performed 3 times with a rest period of 1 minute between each trial to prevent muscle fatigue.

The functional activity of lifting a symmetrical load was selected because it is frequently used in daily activities. Moreover, the movement required during the task has shown to generate a lower peak of compressive force over L5/S1, producing less tension on back ligaments and minimizing overall force requirements [15, 16], which decrease the injury of risk.

Inter-rater reproducibility of EMG activity for LD muscle was evaluated in a pilot test. The intraclass correlation coefficient (ICC) for latency and for RMS amplitude were 0.77 and 0.82, respectively.

The task of lifting the load was divided into 3 phases:

Phase 1: Resting phase. Three seconds in bipedal upright position, legs hip-width apart and upper limbs down both sides of the body.

Phase 2: Flexion phase. Since the moment in which the participants leaned to lift the box (with knees, hips, trunk and shoulders in flexion and elbows in extension to grasp the box), until a load cell located under the box recorded when the box was lifted.

Phase 3: Extension phase. Since the moment that the box detaches from the load cell until the participant returned to the initial bipedal upright position (with extended shoulders and bent elbows) and located the upper edge of the box at the level of the iliac crests.

This activity was performed 3 times. The weight to be lifted was determined based on the NIOSH (National Institute for Occupational Safety and Health) work practice guide for manual lifting [17], which stablish the recommended weight limit for the task, without placing an excessive strain on the back.

All sEMG signals were processed using Matlab (version 8.0, The Mathworks Inc., Natick, MA, USA). They were corrected for offset and band-pass filtered using a fourth order zero-lag Butterworth filter in the 20–450 Hz band. Moving windows of 100 ms. and 99% of overlap were used to convert signals into RMS values. The EMG signals recorded during the task were normalized to the mean value of the three SVC’s. The normalized RMS signals were used to identify the start of the activity, as the point at which the baseline amplitude increased more than 2 standard deviations, considering a window of 100 ms., the variables obtained from the EMG recording were:

Latency: It was the time elapsed between the lifting of the load determined by a load cell under it and the beginning of the muscular activation in ms. To obtain this data, the time elapsed from the beginning of the recording, to picked up the box, was subtracted from the total time for carrying out the activity.

Root Mean Square Amplitude: mean RMS was calculated for both total task and phase-wise task, representing the amount of muscular energy required to perform the action.

SIJD diagnosis: It was established through the multitest regimen described by Laslett. This regimen consists in the application of 6 pain provocation tests for the SIJ and the diagnosis was established if 3/6 positive tests are obtained. This regimen has a sensitivity of 91% and a specificity of 78% for the diagnosis of SIJD. In this study, the diagnosis of SIJD was made after sEMG recording to ensure that possible residual pain after pain provocation tests did not interfere with participants’ performance in the functional activity [10].

There were no losses as the study was based on a single evaluation. All procedures were performed by a physiotherapist with 20 years of experience, properly trained in evaluation procedures. In addition, the standardization of procedures and the pilot test for the evaluation of the sEMG reproducibility were carried out.

2.2 Data analysis

The database was double typed in Microsoft Excel and then validated using the statistical software STATA IC.14. For the data analysis, the data distribution was evaluated through Shapiro-Wilk test, because they not-normally distributed, quantitative variables are shown as median and 25–75 interquartile ranges (IQR).

Bivariate analysis was carried out to evaluate the differences in muscle activation among the study groups, for which the post hoc Dunn’s test was applied after the Kruskal-Wallis test to determine which group provided the difference in the results. Tests with a P≤0.05 were considered statistically significant.

3 Results

Hundred twenty-eight participants were evaluated to establish a sample of 38 people for each study group. The final population consisted of 114 people (76 men and 38 women) between the ages of 18 and 40 years. The descriptive statistics for the personal data of the subjects of the 3 groups are presented in Table 1. The SIJD affected mainly the right side (47.8% right, 31.2% left and 21% bilateral).

Table 1
Gender and age of study groups

Variable Control n = 38 LBP n = 38 SIJD n = 38 p

Female gender 5 (13.2) 11 (28.9) 22 (57.9) <0.001^*

n (%)

Age - years 22.5 (21–33) 21.5 (19–29) 31 (24–37) 0.01^*

Me (IQR)

Variable	Control n = 38	LBP n = 38	SIJD n = 38	p
Female gender	5 (13.2)	11 (28.9)	22 (57.9)	<0.001^*
n (%)
Age - years	22.5 (21–33)	21.5 (19–29)	31 (24–37)	0.01^*
Me (IQR)

Me: Median, IQR: 25–75 interquartile range, ^*Significant difference in SIJD group compare with control and LBP.

In the EMG variables, the SIJD group had a delayed onset of the right LD and a higher RMS amplitude of the left LD, both in the overall activity and in each of the activity phases, compared whit LBP and control groups (Table 2).

Table 2

Electromyographic variables obtained for total task and the three phases, compared by group. Data are presented as median (IQR)

Phase	Variable	Control n = 38	LBP n = 38	SIJd n = 38	p
Me (IQR)	Me (IQR)	Me (IQR)
Total task	RLD LAT (ms)	–1.88	–1.49	–1.2	0.02^*
		(–3.03; –1.14)	(–1.88; –0.9)	(–2.31; –0.91)
	LLD LAT (ms)	–1.48	–1.48	–1.18	0.32
		(–2.25; –0.92)	(–1.92; –0.71)	(–2.07; –0.69)
	RLD RMS (% SVC)	29.76	29.66	45.10	0.23
		(18.4–52.4)	(21.9–55.1)	(26.2–63.3)
	LLD RMS (% SVC)	45.07	47.42	71.28	0.03^*
		(28.9–79.5)	(32.1–71.2)	(46.9–87.0)
Phase 1	RLD RMS (% SVC)	4.73	4.56	7.78	0.08
		(3.5–7.1)	(3.2–5.9)	(3.3–12.7)
	LLD RMS (% SVC)	7.32	6.3	9.86	0.003^*
		(4.09–12.3)	(4.48–9.5)	(7.47–15.59)
Phase 2	RLD RMS (% SVC)	12.1	8.3	15.3	0.06
		(6.62–20.95)	(6.96–14.8)	(7.34–24.39)
	LLD RMS (% SVC)	14.3	11.5	20.89	0.003^*
		(9.01–23.96)	(8.36–18.09)	(16.22–34.73)
Phase 3	RLD RMS (% SVC)	44.9	44.6	59.5	0.2
		(26.68–66.25)	(29.82–75.24)	(36.52–84.16)
	LLD RMS (% SVC)	45.1	47.4	73.2	0.02^*
		(28.87–81.47)	(30.79–76.44)	(46.99–113.7)

Me: median, IQR: 25–75 interquartile range, Lat: latency, RLD: right latissimus dorsi, LLD: left latissimus dorsi, SVC: Submaximal voluntary contraction, RMS: root mean square. ^*Significant difference in SIJD group compared with control and LBP groups.

4 Discussion

This study examined the EMG activity of LD when lifting a load in people with LBP with and without SIJD, and people without LBP. The SIJD patients showed a delay in the activation of right LD when they performed the activity of lifting the load. This finding is relevant, considering that the LD muscle is part of the appendicular skeleton that forms the posterior layer of the TLF and together with the other muscles that form it, enhances the stability of the lumbar spine and increase the force closure of the SIJ [19, 20]. Additionally, the interaction between the LD and the contralateral GM in order to improve the stability of the SIJ to transferring loads through the trunk and the LL [6] has been described in the literature.

Based on the above, the EMG delay of the LD suggests an alteration in the motor control of people with SIJD, decreasing the efficiency of this muscle as a stabilizer of the SIJ while performing functional activities. Therefore, the EMG delay of LD in people with SIJD identified in this study suggests failures in the force closure of the SIJ and thus in muscular system effectiveness to give an adequate and anticipated response to unexpected changes, largely dependent on the control of the central nervous system which continuously interprets the state of stability and movement of the body [8, 21]. Similar results were found by Capobianco et al. [9], who evaluated muscle activation in sit to stand activity, finding that people with SIJD had delays in the activation of LD. According to the authors, this suggests muscle deficiencies in the anticipation of movement during functional activities.

Another relevant finding of this study was that the left LD increased its RMS amplitude in the SIJD group, both in the total activity and in each of the phases covered by the activity. Considering that the SIJD mainly affected the right SIJ (47.8% right and 21% bilateral), the increase in the RMS amplitude can be considered a compensatory mechanism of the left LD for getting a better stability of the SIJ while lifting the load, acting together with the right GM to favor the force closure of the SIJ on the affected side.

These findings are similar to those of Solomonow et al. [22, 23], who found that muscles show increased activation in people with LBP while lifting loads in order to prevent further damage caused by joint instability. Similarly, Marras et al. [24] found that people with LBP had greater activation of the LD and abdominals, probably for the purpose of improving spinal stability during load lifting.

On the other hand, the increase in RMS amplitude in the LD muscle recorded in our study can also be explained from the painful experience of those suffering from SIJD [25]. In this regard, the new theory of adaptation proposed by Hodges et al., states that in the presence of pain, there is a redistribution of muscle activity within the painful muscles and among the muscles surrounding the pain focus, as a protective mechanism against the harmful threat [25]. Therefore, the increase in RMS amplitude in the LD could be the result of redistribution of muscle activity to muscles related to SIJ stability. Hodges’s postulate supports the results of the Palsson et al. study, in which after experimental pelvic pain was generated through injection of hypertonic saline over the posterior sacroiliac ligament, was observed a bilateral increase in the RMS amplitude of the LD during the active straight leg elevation test, followed by a regular EMG pattern once pain decreased [26].

The EMG findings on LD have clinical relevance because during functional activities, people with SIJD presented changes in the activation of this muscle that participates in the force closure of the SIJ. Therefore, future studies should evaluate the effect of exercise on LD in people with LBP caused by SIJD through the performance of specific training for this muscle, which would improve the reorganization in the neuronal network of the motor cortex and thereby improve its activation [27].

Another relevant result of this study was the high prevalence of SIJD in people over 24 years. Some authors have described that the likelihood of having SIJ-originated pain increases from age 30 onwards by 1.6 times for every 5 years of age [28]. This may be due to the fact that upon reaching adulthood, the SIJ progressively becomes more rigid and towards the third decade of life, adaptive changes are evident such as the increase in size and number of elevations and depressions in joint surfaces [29]. These changes can generate excessive joint stiffness with consequent deficits in joint biomechanics so it may favor the appearance of the dysfunction.

In addition, the higher prevalence of SIJD observed in women agrees with previous studies [30 –32] that have described the influence of hormonal factors on the higher prevalence of SIJ pain in women. In this regard, the peak in the production of estrogen and relaxin during the luteal phase of the menstrual cycle is considered to decrease the intrinsic strength and rigidity of collagen, which makes ligaments become laxer [34], altering their stabilizing capacity in joints [30], predisposing to the appearance of joint symptoms such as SIJD. In our study, the women were evaluated on a specific day other than the luteal phase of their menstrual cycle, to reduce the hormonal effect at the time of application of the pain provocation tests, since it has been shown that this could produce false positives during the diagnosis of SIJD [9].

Another factor that may predispose to SIJD in women is the history of previous pregnancies. According to Capobianco et al., 80% of pregnant women have SIJ dysfunction and out of these, about 20% persist with SIJ dysfunction after childbirth [34], which has been related to weight gain, exaggerated lordosis posture, and trauma during childbirth [19, 35]. However, this study did not consider women’s parity, which should be included in future research.

Additionally, the anatomical differences between the female and male pelvis make women more susceptible to SIJD because the female pelvis is lower and wider with larger iliac crests and smaller sacroiliac joints, favoring greater freedom of movement (2.8° for women, 1.2° for men) [19]. Moreover, the female sacrum is generally wider, less curved, and more inclined backward than the male one [36], favoring SIJ instability.

All the above mentioned factors can generate failures in the transfer of forces between the trunk and the lower limbs [37], altering the SIJ ability to reduce tension on the L5-S1 discs and to increase shear forces on the lumbar discs during walking [38], leading to overload the lumbar spine and favoring the presence of pain in this region.

One of the main limitations of this study was that we worked with a convenience sampling, so the potential selection bias was always present. Although in our study the participants were evaluated on a day when they were not in the luteal phase of their cycle, this calculation was made considering the first day of the last menstrual period, but there was no blood test that could ensure this condition. In addition, measurements of EMG activation could be complemented with kinematic measurements to allow a better explaining of EMG activation findings. Therefore, the results of this study should be interpreted with caution and could not be generalized to other types of population.

Despite these limitations, this study was conducted with a high level of methodological rigor and instruments with adequate psychometric properties. The reproducibility analysis of tests used in the study was evaluated before in the pilot test.

5 Conclusion

People with SIJD have higher activation of left LD and delayed onset of this muscle on the opposite side when lifted a load in a symmetrical posture compared to controls and for people with LBP. This situation can cause alterations in motor control and affect the force closure of SIJ. These findings also suggest the need to evaluate treatment plans that include specific handling for this muscle as a complement to current management, to support the rehabilitation of people with SIJD. In addition, future studies should evaluate the relationship between changes in EMG activation and affected side by SIJD.

Footnotes

Acknowledgments

Financial support for the development of this study was provided by Industrial University of Santander.

Conflict of interest

The authors report no conflicts of interest.

References

Cusi

Paradigm for assessment and treatment of SIJ mechanical dysfunction, J Bodyw Mov Ther (2010); 14(2): 152–61.

Hungerford

, Gilleard

, Hodges

Evidence of altered lumbopelvic muscle recruitment in the pr...: Spine. Exerc Physiol Phys Exam (2003); 28(14): 1593–600.

Cher

, Polly

, Berven

Sacroiliac joint pain: Burden of disease. Med Devices 8 (2014); 7 73–81. doi: 10.2147/MDER.S59437.

Chou

, Slipman

, Bhagia

, Tsaur

, Bhat

, Isaac

,et al. Inciting events initiating injection-proven sacroiliac joint syndrome, Pain Med (2004); 5(1): 26–32.

Barker

, Hapuarachchi

, Ross

, Sambaiew

, Ranger

, Briggs

Anatomy and biomechanics of gluteus maximus and the thoracolumbar fascia at the sacroiliac joint, Clin Anat (2014); 27(2): 234–40.

Carvalhais VO do

, Ocarino J de

, Araújo

, Souza

, Silva

PLP

, Fonseca

Myofascial force transmission between the latissimus dorsi and gluteus maximus muscles: An in vivo experiment, J Biomech (2013); 46(5): 1003–7.

Willard

, Vleeming

, Schuenke

, Danneels

, Schleip

The thoracolumbar fascia: Anatomy, function and clinical considerations, J Anat (2012); 221(6): 507–36.

Hodges

, Moseley

Pain and motor control of the lumbopelvic region: Effect and possible mechanisms, J Electromyogr Kinesiol (2003); 13(4): 361–70.

Capobianco

, Feeney

, Jeffers

, Nelson-wong

, Morreale

, Grabowski

, et al. Patients with sacroiliac joint dysfunction exhibit altered movement strategies when performing a sit-to-stand task, Spine J (2018); 18(8): 1434–40.

10.

Laslett

, Aprill

, McDonald

, Young

Diagnosis of Sacroiliac Joint Pain: Validity of individual provocation tests and composites of tests. Man Ther (2005); 10(3): 207–18.

11.

BIOMEC. Manual de usuario Biometrics MiniDatalog MWX8.

12.

Hermens

, Freriks

, Disselhorst-Klug

, Rau

Development of recommendations for SEMG sensors and sensor placement procedures, J Electromyogr Kinesiol (2000); 10(5): 361–74.

13.

Beaudette

, Unni

, Brown

SHM

Electromyographic assessment of isometric and dynamic activation characteristics of the latissimus dorsi muscle, J Electromyogr Kinesiol (2014); 24(3): 430–6.

14.

Boettcher

, Ginn

, Cathers

Standard maximum isometric voluntary contraction tests for normalizing shoulder muscle EMG, J Orthop Res (2008); 26(12): 1591–7.

15.

Pinto

, Córdova

Técnica de levantamiento manual de carga: Actualización de algunos conceptos biomecánicos y fisiológicos, Cienc Trab (2009); 11(34): 193–6.

16.

Gil

, Macuacé M del

, Orozco

, Rojas

Análisis biomecánico de tres posiciones diferentes para agacharse a recoger un objeto Escuela Nacional del Deporte; 2010.

17.

NIOSH Subject. NIOSH Document: Applications Manual for the Revised NIOSH Lifting Equation. 94–110. 1991, p. 35.

18.

República de Colombia Ministerio de salud. Resolucion 30 (4de Octubre de Normas científicas, técnicas yadministrativas para la investigación en salud Bogotá.

19.

Cohen

Sacroiliac joint pain: A comprehensive review of anatomy, diagnosis and treatment. Anesth Analg (2005); 101(5): 1440–53.

20.

Foley

, Buschbancher

Sacroiliac joint pain Anatomy, biomechanics, diagnosis and treatment, Am J Phys Med Rhabil (2006); 85(12): 997–1006.

21.

Hodges

, Tucker

Moving differently in pain: A new theory to explain the adaptation to pain. Pain (2011); 152(SUPPL.3): S90–8.

22.

Solomonow

, Hatipkarasulu

, Zhou

, Baratta

, Aghazadeh

Biomechanics and electromyography of a common idiopathic low back disorder, Spine (Phila Pa (2003); 28(12): 1235–48.

23.

Solomonow

, Zhou

, Baratta

, Burger

Biomechanics and electromyography of a cumulative lumbar disorder: Response to static flexion, Clin Biomech (2003); 18(10): 890–8.

24.

Marras

, Ferguson

, Burr

, Davis

, Gupta

Spine loading in patients with low back pain during asymmetric lifting exertions (2004); 4, 64–75.

25.

Hodges PW

Pain and motor control: From the laboratory to rehabilitation, J Electromyogr Kinesiol (2011); 21(2): 220–8.

26.

Palsson

, Hirata

, Graven-Nielsen

Experimental pelvic pain impairs the performance during the active straight leg raise test and causes excessive muscle stabilization, Clin J Pain (2015); 31(7): 642–51.

27.

Crow

, Pizzari

, Buttifant

Physical Therapy in Sport Muscle onset can be improved by therapeutic exercise: A systematic review, Phys Ther Sport (2011); 12(4): 199–209.

28.

Depalma

, Ketchum

, Trussell

, Saullo

, Slipman

Does the location of low back pain predict its source? PM&R (2011); 3(1): 33–9.

29.

Brolinson

, Kozar

, Cibor

Sacroiliac joint dysfunction in athletes, Curr Sports Med Re (2003); 2(1): 47–56.

30.

Depalma

, Ketchum

, Saullo

Multivariable analyses of the relationships between age, gender, and body mass index and the source of chronic low back pain, Pain Med (2012); 13, 498–506.

31.

Mejía

, Arias

, Valdez

, Carrillo

, Infante

Dolor de laarticulación sacroilíaca Anatomía, diagnóstico ytratamiento, Rev la Soc Esp del Dolor (2008); 15(3): 170–80.

32.

O’Sullivan

, Beales

Diagnosis and classification of pelvic girdle pain disorders-Part A mechanism based approach within a biopsychosocial framework, Man Ther (2007); 12(2): 86–97.

33.

Seoane

Influence of menstrual cycle in synchronized swimming flexibility, Agon Int J Sport Sci (2013); 3(2): 53–9.

34.

Capobianco

, Cher

Safety and effectiveness of minimally invasive sacroiliac joint fusion in women with persistent post-partum posterior pelvic girdle pain: 12- month outcomes from a prospective, multi-center trial. Springerplus (2015); 4(1): 1–12.

35.

Albert

, Godskesen

, Westergaard

Prognosis in four syndromes of pregnancy-related pelvic pain, Acta Obstet Gynecol Scand (2001); 80(6): 505–10.

36.

Joukar

, Shah

, Kiapour

, Vosoughi

, Duhon

, Agarwal

,et al. Gender specific sacroiliac joint biomechanics during standing upright. Spine (Phila Pa 1976). 2018;1.

37.

Cohen

, Chen

, Neufeld

Sacroiliac joint pain: A comprehensive reviewof epidemiology, diagnosis and treatment. Expert Rev Neurother (2013); 13(1): 99–116.

38.

Ramírez

, Ayala

, Marcela

, Pinzón

Disfunciónde la articulación sacro ilíaca: Causa potencial de dolorlumbar, Salud UIS (2007); 39, 43–53.