Facebook Pixel
Brookbush Institute Logo

Tuesday, June 6, 2023

Multifidus Recovery Is Not Automatic Following Low Back Pain

Brent Brookbush

Brent Brookbush


Research Review: Multifidus Recovery Is Not Automatic Following Low Back Pain

By Jinny McGivern PT, DPT, Certified Yoga Instructor

Edited by Brent Brookbush DPT, PT, MS, PES, CES, CSCS, ACSM H/FS

Original Citation: Hides, J. A., Richardson, C. A., & Jull, G. A. (1996). Multifidus Muscle Recovery Is Not Automatic After Resolution of Acute, First‐Episode Low Back Pain.Spine, 21(23), 2763-2769. ARTICLE

Image courtesy of http://www.coreconcepts.com.sg/mcr/multifidus-smallest-yet-most-powerful-muscle/

Why is this relevant?: At some point in life, most individuals will grapple with at least one episode of low back pain (LBP). Hoy et. al (2012) reports a prevalence of 12-33% of individuals suffering from a 1st time episode of LBP, and 22-65% of individuals suffering from an episode of LBP each year. The authors of this study, Hides et. al (1996), report that LBP typically resolves of its own within 2-4 weeks. However there is a very high recurrence rate, 60-80%, within a year of the 1st episode (Hides et. al, 1996). This indicates that although pain reduces and function improves, impairments persist in critical areas of the individual's movement system that likely predispose them to subsequent episodes of pain. This article provides critical information on the behavior of the multifidi  (component of the intrinsic stabilization subsystem ) following injury, and how this behavior may play a role in the recurrence rate of LBP.

Study Summary

Study Design Prospective Randomized Controlled Trial (RCT)
Level of Evidence Level II - Evidence from a well designed RCT
Subject Demographics

39 individuals with an acute initial episode of low back pain (LBP) were randomly assigned to either a control group (CON - 19 subjects) which received medical management of LBP or an exercise group which received multifidi retraining (MRG - 20 subjects) activities plus medical management. All subjects underwent full neurological exam & lumbar X-ray prior to participation.

  • Age: 31 yrs +/- 7.9 (CON); 30.9 yrs +/- 6.5 (MRG)
  • Gender: 9 men, 10 women (CON); 7 men, 13 women (MRG)
  • Characteristics: Young, healthy, 1st episode of LBP
  • Inclusion Criteria: Aged 18-45 yrs; 1st episode of unilateral, mechanical LBP; less than 3 weeks duration of pain; pain location between T12 and gluteal fold (with or without radiation into the lower extremity); lumbar range of motion (ROM) restricted by pain; asymmetry of lumbar multifidi activation of  greater than 11%.
  • Exclusion Criteria: Previous history of LBP, injury or lumbar surgery; spinal abnormalities visible on X-ray (i.e. pars defect, lumbarization, scoliosis, spina bifida occulta); neuromuscular or joint disease; reflex or motor signs of nerve root or cauda equina compression; evidence of systemic/organ disease or carcinoma; pregnancy; any sports or fitness training for the lower back within prior 3 months.
Outcome MeasuresAssessments were performed by 2 independent blinded examiners at baseline, then again at the end of week 1, week 2, week 3, and week 4. Outcome measures included:
  • Pain - Pain location, quality, intensity via the McGill Pain Questionnaire (MPQ), Visual Analog Scale (VAS), and daily pain/pain medication use diaries.
  • Disability - via the Roland Morris Disability Index
  • Lumbar ROM (flexion, extension, side-bending, straight leg raise (SLR)) assessed with 2 inclinometers
  • Habitual Activity Level (work, sport, leisure) via questionnaire
  • Multifidus Muscle Cross Sectional Area (CSA) via ultrasound for spinal levels L2 to S1

If patients were low pain to pain-free during all functional activities and demonstrated full lumbar range of motion (ROM) at the end of week 4, they were included in a week 10 follow-up where the following was assessed:

  • Habitual Activity Level (work, sport, leisure) via questionnaire
  • Multifidus Muscle Cross Sectional Area (CSA) via ultrasound for spinal levels L2 to S1

If patients continued to report pain during activity or pain during lumbar movements, they were removed from the study and provided with a complete course of physical therapy treatment.

Results Description of Sample at Baseline:
  • Mean age, height, weight, pre-morbid activity levels, duration of pain (CON: 9.16 days, MRG: 8.10 days; p = .67), pain  scales, lumbar ROM, location of pain & medication use were not significantly different between groups.
  • Multifidus Cross Sectional Area (CSA) - Ultrasound imaging demonstrated that all subjects, except one, demonstrated multifidus size asymmetry on the painful side at a single spinal level. The exception showed changes at 2 adjacent levels (L5, S1).  The most common level to show atrophy was L5 (34 subjects), followed by S1 (4 subjects), then L4 (1 subject) in a distant 2nd and 3rd.  The mean difference between sides was represented as a percentage of the size of the unaffected side. The mean of both groups at the affected level was 24.03% +/- 8.67%.  The asymmetry at the affected level was significantly different from other spinal segments.
  • There was a statistically significant difference in the Roland Morris Disability Scale scores (CON: 13.6; MRG: 10.3; p = .05); however, the authors note that this may not be clinically significant because the average score on this scale for individuals with LBP is 11.4 (as reported by Roland & Morris).

Outcomes for Weeks 1 - 4

  • Pain - Pain decreased significantly in both groups from baseline to week 4.  Full recovery from pain occurred in all but 4 patients who were members of the the MRG group. There was no significant difference in pain reduction between groups at any point from weeks 1 - 4.
  • Disability - By week 4 both groups reported significantly reduced disability (mean score CON: 2.3; MRG: 0). There was no significant difference between groups.
  • Lumbar ROM - ROM increased in all directions by week 4 for both groups.  There was no significant difference between groups at week 4.
  • Multifidus CSA - There was significantly greater muscle recovery in the MRG over the CON group  at each week (p=.0001) and by week 4 (p=.0001).

Correlational Analysis

  • The authors found that pain was correlated with disability in both groups.  Both outcome measures decreased together in both groups.
  • In the CON group, pain and disability were not correlated with muscle size recovery.

At 10 week Follow up

• There was no significant change in multifidus size at the affected level between week 4 measurements and week 10 measurements (p=.02) in either group.  At week 4 the CON demonstrated a 16.84% (+/- 9.26%) difference between sides, and at week 10 it was 14.02% (+/- 6.31%).  At week 4 the MRG demonstrated a .71% (+/- 2.49%) difference between sides, and at week 10 it was .24% (+/- 3.29%).

• There was no significant difference between baseline and week 10 in the scores on the habitual activity questionnaire, indicating that the subjects resumed their regular activities.

ConclusionsMeasures of pain and ability to perform functional activities without pain are not sufficient criteria to deem the rehabilitation of individuals with low back pain complete.  Assessment of the muscular stabilization system is essential in order to fully restore stabilizing muscle function for optimal performance of functional activities.
Conclusions of the ResearchersMultifidus size does not recover automatically with the cessation of pain and resumption of full functional activities.  This may predispose individuals to subsequent episodes of low back pain.  Localized, non-resisted isometric training restored multifidi CSA, therefore it may be beneficial to focus on this specific training before transitioning to a more general stabilization program.

Image courtesy of http://www.keala.org/sc/anato/Palpation_List_Final.html

Review & Commentary:

This research had many strong components to its methodology. It examined the effects of two different approaches to the treatment of LBP, often a multi-faceted and complex condition. The authors were particularly interested in how these two approaches impacted a very specific component of the pathoanatomy of the lumbar spine - multifidi atrophy, as well as how this might have implications for the recurrence of LBP. This narrow research question helps to provide very specific information on a component of a complex condition. The medical management strategy included advice on rest and prescription of mild analgesics/anti-inflammatories. The exercise approach included the medical strategy plus a multifidus re-training activity. This exercise consisted of isometric facilitation of multifidus contraction in standing with an abdominal co-contraction. Real-time ultrasound imaging was utilized to ensure that multifidus was engaging. The use of imaging to confirm multifidi activation demonstrated that the exercise was actually activating the intended muscles, removing a potential source of error.

In terms of design, participants were randomly assigned to CON or MRG groups thus reducing the possibility of a selection bias. All assessments were performed by blinded examiners, lending additional objectivity to the data collection process. The authors provided a very clear, yet complete, clinical picture of the LBP presentation that this research observed with respect to area of pain, duration, ROM limitations and disability. This allows clinicians to have a better understanding of which patients/clients this information is most specific to, although it may also be applied more broadly. The use of a varied array of outcome measures examining everything from pain to disability to ROM to multididi size covers all the essential components of the Nagi Disablement Model, which describes the connection between pathology, impairment, function and disability of the individual (Jette, 2006). As human movement professionals, it is essential that we understand the connection between these processes.

While this research has many strengths, it also has its limitations. Because of the short follow up period of 6 weeks after the end of the intervention, it was not possible to determine if the changes observed at week 10 were maintained over time from this paper. However, the authors of this work published a follow up study in 2001 in the journal Spine where they report the results of 1 year and 3 year telephone follow ups. Without giving too much away (future research review on the follow up coming soon), at 1 year follow up alone, the MRG had a 30% recurrence rate whereas the CON group reported an 84% recurrence rate. This is quite a striking difference.

Another limitation of this research is that the authors did not indicate whether subjects had an insidious onset of low back pain versus a sudden or traumatic onset (although exclusion criteria included low back injury). It is possible that these sub-groups might respond to multifidi re-training differently. It also would have been beneficial if the authors provided more information about the dosage of exercise that was performed by the MRG. It is unclear as to whether or not subjects attended training sessions in the clinic or if they were prescribed a home program and what the intensity/frequency of this program was. While the use of real time ultrasound was helpful in confirming multifidi activation, it is not readily available to most human movement professionals in the clinic or gym. Future research should compare the validity of palpation and other clinical tests with confirmation via real time ultrasound to ensure multifidi activation. Finally, multifidi activation was only assessed during static standing, not during any functional activities such as walking, sitting, transferring sit to stand, etc. Future research should also focus on assessing the behavior of multifidi following low back pain during functional activities, including sport.

Why is this study important?

The authors of this study report that previous research has found that poor multifidi function and poor functional outcomes go hand in hand. This research demonstrates that multifidi do not automatically recover following LBP, even when pain is absent and normal activities resumed. This highlights that human movement professionals must look beyond pain and the ability to perform functional tasks, and look to more sophisticated and comprehensive assessments of movement quality and recruitment strategies. Because the multifidi do not return unless there is a specific challenge to facilitate their activation, re-training should be a component of any rehabilitation or performance program intended for the individual suffering from low back pain.

How does it affect practice?

From a practice standpoint, an important takeaway from this article is that although your client/patient may not be painful, dysfunction may still be present due to previous injury. It is essential to collect a history of prior injury (and more information depending on your scope of practice) prior to starting a new program. This will help to spotlight certain body systems or tendencies that may need to be addressed, and will help to guide your movement analysis. Second, analyze movement to determine if appropriate stabilization is occurring at the appropriate time. Even if the goal of a program is one of general lumbar stabilization, it is essential to confirm that the multifidi is functioning optimally prior to progressing to more global stabilization activities.

How does it relate to Brookbush Institute Content?

The findings of this research support the Brookbush Institute's description of the relationship between core subsystems  and models of postural dysfunction . The multifidi are a member of the Intrinsic Stabilization Subsystem (ISS) of the spine, along with the transverse abdominis , rotatores, interspinalis, intertransversarii , diaphragm and pelvic floor. The Brookbush Institute describes this system as having a tendency toward under-activity in all models of postural dysfunction: Lower Leg , Lumbo Pelvic Hip Complex, Sacroiliac Joint and Upper Body Dysfunction. Because the role of the ISS  is so critical to segmental control of the spine and trunk, it may be inevitable that dysfunction in this system will lead to compensation at distal segments. The Brookbush Institute advocates activation of the ISS  during the "Isolated Activation" or "Core Activation" portions of a rehabilitation program or integrated warm-up. Below are several videos that describe isolated activation techniques for the ISS , as well as reactive activation techniques that place time dependent demands on the ISS . The ability of the ISS to activate in a time appropriate fashion is key to provide the stability needed for other muscles & subsystems to function optimally. Reactive activation techniques incorporate that time component which is so essential for functional activities that range from pushing open a door to swinging a baseball bat.

Review of Core Subsystems

Transverse Abdominis TVA Isolated Activation

TVA and Gluteus Maximus Activation and Progression

Hardest Quadriped Progression Ever

Modified Mountain Climbers for Core Reactive Stabilization

Core Reactive Integration Crunch and Catch


Hoy, D., Bain, C., Williams, G., March, L., Brooks, P., Blyth, F., … & Buchbinder, R. (2012). A systematic review of the global prevalence of low back pain. Arthritis & Rheumatism, 64(6), 2028-2037.

Jette, A. M. (2006). Toward a common language for function, disability, and health. Physical therapy, 86(5), 726-734.

© 2015 Brent Brookbush

Questions, comments, and criticisms are welcomed and encouraged -