Skip to main content
NCBI home page
Pain Medicine: The Official Journal of the American Academy of Pain Medicine logoLink to Pain Medicine: The Official Journal of the American Academy of Pain Medicine
. 2015 Nov 25;17(8):1436–1446. doi: 10.1093/pm/pnv023

Trunk Muscle Size and Composition Assessment in Older Adults with Chronic Low Back Pain: An Intra-Examiner and Inter-Examiner Reliability Study

Jaclyn Megan Sions *,✉, Andrew Craig Smith †, Gregory Evan Hicks *, James Matthew Elliott †
PMCID: PMC4975013  PMID: 26814258

Abstract

Objective. To evaluate intra- and inter-examiner reliability for the assessment of relative cross-sectional area, muscle-to-fat infiltration indices, and relative muscle cross-sectional area, i.e., total cross-sectional area minus intramuscular fat, from T1-weighted magnetic resonance images obtained in older adults with chronic low back pain.

Design. Reliability study.

Subjects. n = 13 (69.3 ± 8.2 years old)

Methods. After lumbar magnetic resonance imaging, two examiners produced relative cross-sectional area measurements of multifidi, erector spinae, psoas, and quadratus lumborum by tracing regions of interest just inside fascial borders. Pixel-intensity summaries were used to determine muscle-to-fat infiltration indices; relative muscle cross-sectional area was calculated. Intraclass correlation coefficients were used to estimate intra- and inter-examiner reliability; standard error of measurement was calculated.

Results. Intra-examiner intraclass correlation coefficient point estimates for relative cross-sectional area, muscle-to-fat infiltration indices, and relative muscle cross-sectional area were excellent for multifidi and erector spinae across levels L2-L5 (ICC = 0.77–0.99). At L3, intra-examiner reliability was excellent for relative cross-sectional area, muscle-to-fat infiltration indices, and relative muscle cross-sectional area for both psoas and quadratus lumborum (ICC = 0.81–0.99). Inter-examiner intraclass correlation coefficients ranged from poor to excellent for relative cross-sectional area, muscle-to-fat infiltration indices, and relative muscle cross-sectional area.

Conclusions. Assessment of relative cross-sectional area, muscle-to-fat infiltration indices, and relative muscle cross-sectional area in older adults with chronic low back pain can be reliably determined by one examiner from T1-weighted images. Such assessments provide valuable information, as muscle-to-fat infiltration indices and relative muscle cross-sectional area indicate that a substantial amount of relative cross-sectional area may be magnetic resonance-visible intramuscular fat in older adults with chronic low back pain.

Keywords:  Intramuscular Fat, Low Back Pain, Magnetic Resonance Imaging, Multifidi, Muscle

Introduction

Lumbar spinal stability is maintained by interactions among the following three subsystems: active, passive, and neuromuscular control [1,2]. The active control subsystem is comprised of contractile tissues, i.e., muscles and their tendons. Critical lumbar spine stabilizers include the multifidi [3,4], the erector spinae [5], the psoas [6,7], and the quadratus lumborum [8]. The multifidi provide two-thirds of active spine stability [4]. The erector spinae, comprised of the spinalis, longissimus, and iliocostalis, assist with maintaining static trunk equilibrium by resisting flexion moments imposed by gravity and loads anterior to the spinal column [5]. The psoas muscle, originating on the transverse processes of the lumbar vertebrae and inserting on the lesser trochanter of the femur [6], is architecturally equipped to be both a hip flexor and active spinal stabilizer [7]. The quadratus lumborum, a spinal extensor and lateral flexor, may also play a critical role in spinal stabilization owing to its attachments to the lumbar transverse processes [8].

Age- and low back pain (LBP)-related morphological changes in the trunk muscles may include reduced cross-sectional area (CSA) and increased intramuscular fat [9–13]. Such morphological changes may impact muscle function, i.e., spinal stability, and physical performance [14–16], although there is little research in this area among older adults with LBP. Further, the quantity of intramuscular fat is a distinguishing factor between individuals with acute, remitting pain and chronic LBP [17], as well as individuals with low versus high LBP intensity [18]. Thus, establishing a simple, reliable, and valid analytical procedure for evaluating trunk muscle morphology may enable further exploration of the role of trunk muscle morphology in clinical presentations and recovery trajectories of older patients with LBP. Ultimately, improved understanding of the impact of trunk muscle morphology on muscle function and physical performance may support treatments targeting trunk muscle morphology and behavior, such as exercises that are utilized in the management of younger adults with chronic LBP [19]. Finally, simple analytical procedures that allow accurate quantification of trunk muscle size and structure among older adults with LBP may enable researchers to determine whether trunk-targeting treatments result in morphological changes that are related to improvements in muscle function and physical performance.

Imaging options currently used for evaluating trunk muscle size and structure include ultrasound imaging (USI) and magnetic resonance imaging (MRI). USI is a reliable and valid clinical tool for muscle CSA assessment, e.g., multifidi, in younger adults [20,21]. However, USI is largely unable to accurately delineate muscle from noncontractile tissue, such as intramuscular fat [22]. With aging comes an increase in intramuscular fat [13,23]. Consequently, when evaluating older adults using USI, muscle CSA may be artificially inflated by the presence of intramuscular fat. MRI affords a gold-standard measure for accurately delineating fat from muscle [24–26] and can be used to quantify the muscle-to-fat index (MFI) as well as relative muscle cross-sectional area (rmCSA), i.e., total CSA minus intramuscular fat CSA [27].

Currently, there are a number of fat-suppression MRI techniques that allow for quantification of fat in a given tissue, including fat saturation, inversion recovery, and opposed-phase imaging [28,29]. One of the major limitations of fat saturation is that its reliability decreases at lower field strengths [28,30]. While the availability of higher-field-strength MRI magnets is improving, accessibility is not always guaranteed, particularly in rural and lower socioeconomic regions. So, while emerging imaging technologies may be desirable, magnet availability may dictate measurement technique. Inversion recovery, although appropriate for use with low-field-strength magnets, is not specific to fat [28], and thus fat content must be assumed [31]. Opposed-phase imaging is best for structures with only small amounts of fat [28]; multifidi intramuscular fat among younger adults with LBP has been reported to be 14–23% [32–34], while multifidi and erector spinae intramuscular fat among older adults may be 50% or more [9]. So, while opposed-phase imaging may be used among younger adults with LBP to assess muscle composition, it may not be the best choice for older adults with LBP. Further, fat-suppression techniques are not typically used in a standard lumbar MRI series.

Chronic LBP prevalence among older adults has been reported to be greater than among younger adults, i.e., 12.3% versus 6.5%, with greater symptom duration and disability reported among older adults [35]. Many adults with chronic LBP receive MRI procedures to rule out serious pathology. While current data with regard to the frequency of obtaining MRIs in older adults with LBP is largely unknown, a trend toward increased frequency of advanced imaging, i.e., MRI and computed tomography, in the United States among Medicare recipients has been recently reported [36]. A standard lumbar MRI series utilizes T1-weighted parameters, which provide images with sufficient contrast to delineate fat from muscle [27,37,38] and, therefore, may provide additional patient data, i.e., MFI and rmCSA. While MRI reports do not typically report such data, extraction of these quantitative data is possible. However, we contend that for widespread adoption, the analytical procedure for obtaining MFI and rmCSA should be not only reliable and valid but simple. Establishing inter-examiner reliability could allow assessments of MFI and rmCSA to be delegated to trained personnel.

The primary purpose of this study was to establish measurement reliability and obtain the standard error of measurement (SEM) for an analytical procedure for determining MFI and rmCSA from T1-weighted images of the lumbar spine obtained in older adults with chronic LBP. We hypothesized that two examiners could attain similar relative CSA (rCSA), MFI, and rmCSA measurements for the multifidi, erector spinae, psoas, and quadratus lumborum using ImageJ software (National Institutes of Health, Bethesda, MD). When compared to inter-examiner reliability, we hypothesized that intra-examiner reliability would be higher, resulting in lower SEMs. We also hypothesized that combined measurements of the multifidus and erector spinae in older adults with chronic LBP would be more reliable than isolated measurements of the multifidus and erector spinae as examiners would not need to distinguish between the multifidi and erector spinae, which can be difficult in the presence of increased intramuscular fat from both aging and chronic LBP. Further, we hypothesized that averaging the right and left sides would be as reliable as assessing a single side, despite increased variability introduced by incorporating measurements from both sides.

Methods

Data Collections

Thirteen community-dwelling adults, ages 60–85 years, with self-reported chronic LBP (i.e., LBP of at least 3 months’ duration) of ≥ 3/10 pain intensity, underwent lumbar spine MRI. Exclusion criteria included history of low back surgery, receipt of services for LBP within the past 6 months (e.g., injections, chiropractic services, physical therapy, acupuncture), recent traumatic events, neurological disorders, and terminal illnesses. This study was approved by the Institutional Review Board for Human Subjects Research at the University of Delaware.

Participants completed the informed consent, underwent body anthropometric measurements, i.e., height and weight for calculation of body mass index, and participated in a MRI safety screen with a MRI technician before supine MRI [39]. Two-dimensional images were obtained on a 1.5 Tesla scanner (MAGNETOM, Siemens, Erlangen, Germany) with a flexible spine coil. T1-weighted, spin-echo images were produced in the axial plane (repetition time/echo time = 879/13 ms; field of view = 230 × 230 mm; encoding matrix = 480 × 640; phase encoding direction = anterior to posterior; bandwidth = 180; flip angle = 150 degrees; slice thickness = 5 mm with 1.5 mm gap; acquisition time ∼ 8 minutes).

Data Analysis

Sagittal scouts were used to plan the anatomical placement of the axial slices at levels L2–L5. Slices were included only if the entire sagittal image scout line was anterior to posterior through the vertebral body and did not cross the disc space. After scaling images using the known distance of a 4 cm line, right and left rCSA measurements of the multifidi, the erector spinae (longissimus and iliocostalis), the psoas, and the quadratus lumborum were taken by tracing just inside the fascial borders using ImageJ software (Figure 1); a combined CSA measurement of the multifidi and erector spinae was also obtained. Examiners were blinded to measurement outputs until all CSA measurements on a given image were taken. Using the ImageJ histogram feature, a pixel intensity summary for the muscle was obtained for each rCSA measurement [27,40]. Fat pixel intensity summaries were also obtained for 0.5 × 0.5 cm areas of extramuscular fat lateral to the erector spinae (for comparison to multifidi, erector spinae, and combined multifidi and erector spinae) and to the psoas (for comparison to psoas and quadratus lumborum). MFIs were computed as (mean rCSA pixel intensity)/(mean extramuscular fat pixel intensity). rmCSA was computed as (1 − MFI)×rCSA, effectively removing the T1-weighted fat portion of the muscle from the rCSA [27]. Examiner 1 completed this analytical procedure on all participants, waited 5 months to decrease recall bias, and repeated the procedure. Examiners 1 and 2 trained together 3 hours; sessions included review of procedures and cross-sectional anatomy and measurements on a sample set of T1-weighted images from older adults without LBP. Examiners were blinded to measurements generated by one another.

Figure 1.

Figure 1

Open in a new tab

Relative cross-sectional area measurements and extramuscular fat regions-of-interest on a magnetic resonance image at the L3 level from an older adult with chronic low back pain. A = multifidi; B = erector spinae; C = quadratus lumborum; D = psoas; E = extramuscular fat lateral to erector spinae;  F = extramuscular fat lateral to psoas.

Statistical analyses were performed using IBM SPSS Statistics 22 (SPSS, Inc., Armonk, NY). Two-way, intraclass correlation coefficients (ICCs) with 95% confidence intervals (CIs) were used to estimate intra-examiner and inter-examiner reliability (model 3,k). Based on ICC cutoffs proposed by Fleiss, ICC > 0.75 was considered excellent, while ICCs in the range 0.40–0.75 were considered fair to good [41]. The SEM, which provides information on measurement precision, was calculated using the formula [SEM = SD √1 − r)], where SD is the standard deviation calculated from the measurements and “r” is the reliability coefficient for that measurement, in this case the calculated ICC (3,k) value [42].

Results

Participant characteristics are provided in Table 1. Of the sample, seven participants were male. Mean age was 69.3 ± 8.2 years and mean body mass index was 29.1 ± 3.4 kg/m2. Means and standard deviations for the right side and the right-left average for rCSA, MFI, and rmCSA, as obtained by examiner 1, are provided in Table 2. Mean right-left average rCSA for the multifidi ranged from 3.47 to 9.33 cm2, with a progressive increase in rCSA from L2 through L5. Mean right-left average rCSA for the erector spinae ranged from 8.36 to 18.89 cm2, with a progressive decrease in rCSA from L2 through L5. Mean MFI for the multifidi and erector spinae ranged from 0.46 to 0.56. Psoas right-left average rCSA ranged from 5.46 to 10.98 cm2 with a MFI of 0.36–0.40. Average right-left quadratus lumborum rCSA ranged from 3.53 to 4.55cm2 for levels L3 and L4; average right-left quadratus MFI ranged from 0.51 to 0.52. After removing visible fat from the rCSA, the average right-left rmCSA ranged from 1.76 to 4.34 cm2 for multifidi, 3.62–10.10 cm2 for the erector spinae, and 3.54–6.93 cm2 for the psoas. Average right-left rmCSA for the quadratus lumborum at levels L3 and L4 were 1.63 and 2.19 cm2, respectively.

Table 1.

Participant characteristics (N = 13)

Sex, male 7
Age, y* 69.3 ± 8.2
Height, m* 1.6 ± 0.1
Weight, kg* 80.5 ± 14.8
BMI, kg/m2* 29.1 ± 3.4
Open in a new tab

BMI = body mass index.

*Values are presented as means ± SD.

Table 2.

Intra-examiner reliability results

Right
 Right-left average
L2 Mean (SD) ICC (95% CI) SEM Mean (SD) ICC (95% CI) SEM
Relative CSA, cm2
Multifidus 3.44 (0.94) 0.93 (0.80, 0.98) 0.24 3.47 (0.85) 0.98 (0.94, 0.99) 0.12
ES 18.76 (4.46) 0.99 (0.98, 0.99) 0.44 18.89 (4.15) 0.99 (0.98, 0.99) 0.41
Combined multifidus & ES 22.49 (5.00) 0.99 (0.98, 0.99) 0.50 22.93 (4.85) 0.99 (0.99, 0.99) 0.48
Psoas 5.04 (2.14) 0.99 (0.98, 0.99) 0.21 5.46 (2.37) 0.90 (0.67, 0.97) 0.74
Quadratus 2.53 (1.14) 0.98 (0.95, 0.99) 0.16 3.07 (1.89) 0.11 (−2.01, 0.73) 1.78
Muscle-to-fat index
Multifidus 0.46 (0.13) 0.98 (0.95, 0.99) 0.01 0.47 (0.09) 0.97 (0.85, 0.98) 0.01
ES 0.45 (0.13) 0.99 (0.97, 0.99) 0.01 0.46 (0.10) 0.98 (0.94, 0.99) 0.01
Combined multifidus & ES 0.45 (0.13) 0.99 (0.96, 0.99) 0.01 0.46 (0.10) 0.97 (0.92, 0.99) 0.01
Psoas 0.28 (0.10) 0.64 (−0.18, 0.89) 0.06 0.35 (0.03) 0.44 (−0.78, 0.83) 0.02
Quadratus 0.53 (0.07) 0.80 (0.37, 0.93) 0.03 0.53 (0.07) 0.84 (0.50, 0.95) 0.02
Relative muscle CSA, cm2
Multifidus 1.79 (.66) 0.99 (0.97, 0.99) 0.06 1.76 (.57) 0.98 (0.89, 0.99) 0.08
ES 10.18 (3.86) 0.99 (0.98, 0.99) 0.38 10.10 (3.08) 0.98 (0.89, 0.98) 0.43
Combined multifidus & ES 12.06 (4.35) 0.99 (0.97, 0.99) 0.43 12.14 (3.57) 0.98 (0.94, 0.99) 0.50
Psoas 3.69 (1.54) 0.99 (0.97, 0.99) 0.15 3.54 (1.52) 0.92 (0.73, 0.97) 0.42
Quadratus 1.20 (0.63) 0.97 (.092, 0.99) 0.10 1.56 (1.58) 0.04 (−2.23, 0.71) 1.54
L3 Mean (SD) ICC (95% CI) SEM Mean (SD) ICC (95% CI) SEM
Relative CSA, cm2
Multifidus 5.07 (2.02) 0.98 (0.96, 0.99) 0.28 5.29 (1.91) 0.98 (0.95, 0.99) 0.27
ES 17.63 (4.00) 0.94 (0.81, 0.98) 0.97 17.51 (3.54) 0.96 (0.89, 0.99) 0.70
Combined multifidus & ES 22.79 (4.79) 0.99 (0.97, 0.99) 0.47 23.10 (4.35) 0.99 (0.98, 0.99) 0.43
Psoas 8.22 (3.60) 0.97 (0.93, 0.99) 0.62 8.52 (3.86) 0.99 (0.97, 0.99) 0.38
Quadratus 3.07 (1.58) 0.99 (0.97, 0.99) 0.15 3.53 (1.57) 0.99 (0.97, 0.99) 0.15
Muscle-to-fat index
Multifidus 0.48 (0.13) 0.94 (0.83, 0.98) 0.03 0.50 (0.10) 0.91 (0.72, 0.97) 0.03
ES 0.49 (0.14) 0.96 (0.88, 0.99) 0.02 0.48 (0.11) 0.94 (0.83, 0.98) 0.04
Combined multifidus & ES 0.48 (0.12) 0.91 (0.69, 0.97) 0.03 0.48 (0.10) 0.90 (0.68, 0.97) 0.03
Psoas 0.39 (0.03) 0.87 (0.61, 0.96) 0.01 0.40 (0.04) 0.83 (0.45, 0.95) 0.01
Quadratus 0.52 (0.10) 0.81 (0.39, 0.94) 0.04 0.52 (0.05) 0.88 (0.63, 0.96) 0.01
Relative muscle CSA, cm2
Multifidus 2.54 (1.13) 0.98 (0.95, 0.99) 0.15 2.57 (1.07) 0.97 (0.91, 0.99) 0.18
ES 8.97 (3.55) 0.96 (0.85, 0.98) 0.71 8.94 (2.45) 0.93 (0.77, 0.98) 0.64
Combined multifidus & ES 11.83 (4.49) 0.96 (0.87, 0.98) 0.89 11.74 (3.15) 0.94 (0.81, 0.98) 0.77
Psoas 5.03 (2.23) 0.98 (0.94, 0.99) 0.31 5.07 (2.27) 0.98 (0.95, 0.99) 0.32
Quadratus 1.44 (0.73) 0.99 (0.96, 0.99) 0.07 1.63 (0.70) 0.96 (0.90, 0.99) 0.14
L4 Mean (SD) ICC (95% CI) SEM Mean (SD) ICC (95% CI) SEM
Relative CSA, cm2
Multifidus 8.76 (3.02) 0.98 (0.92, 0.99) 0.42 8.82 (3.02) 0.99 (0.96, 0.99) 0.30
ES 13.51 (2.00) 0.89 (0.66, 0.96) 0.66 13.44 (2.18) 0.95 (0.86, 0.98) 0.48
Combined multifidus & ES 22.73 (3.85) 0.96 (0.88, 0.98) 0.77 22.50 (12.36) 0.98 (0.95, 0.99) 1.74
Psoas 10.52 (3.79) 0.99 (0.98, 0.99) 0.37 10.98 (4.30) 0.99 (0.99, 0.99) 0.43
Quadratus 4.26 (1.80) 0.98 (0.95, 0.99) 0.25 4.55 (1.88) 0.97 (0.91, 0.99) 0.32
Muscle-to-fat index
Multifidus 0.57 (0.14) 0.94 (0.78, 0.98) 0.03 0.56 (0.08) 0.86 (0.53, 0.95) 0.02
ES 0.51 (0.13) 0.96 (0.85, 0.99) 0.02 0.50 (0.07) 0.90 (0.63, 0.97) 0.02
Combined multifidus & ES 0.53 (0.13) 0.93 (0.79, 0.98) 0.04 0.52 (0.08) 0.87 (0.60, 0.96) 0.02
Psoas 0.34 (0.05) 0.58 (−0.15, 0.86) 0.03 0.36 (0.04) 0.85 (0.32, 0.96) 0.01
Quadratus 0.53 (0.07) 0.97 (0.90, 0.99) 0.01 0.51 (0.05) 0.96 (0.86, 0.98) 0.01
Relative muscle CSA, cm2
Multifidus 3.82 (2.02) 0.96 (0.87, 0.99) 0.40 3.88 (1.62) 0.95 (0.83, 0.98) 0.36
ES 6.51 (1.94) 0.94 (0.77, 0.98) 0.47 6.63 (1.27) 0.89 (0.65, 0.96) 0.42
Combined multifidus & ES 10.66 (3.82) 0.94 (0.83, 0.98) 0.93 10.69 (2.49) 0.91 (0.72, 0.97) 0.74
Psoas 6.93 (2.61) 0.98 (0.90, 0.99) 0.36 6.93 (2.76) 0.99 (0.96, 0.99) 0.27
Quadratus 1.97 (0.92) 0.99 (0.96, 0.99) 0.09 2.19 (0.91) 0.99 (0.97, 0.99) 0.09
L5 Mean (SD) ICC (95% CI) SEM Mean (SD) ICC (95% CI) SEM
Relative CSA, cm2
Multifidus 9.35 (1.83) 0.94 (0.82, 0.98) 0.44 9.33 (1.82) 0.96 (0.89, 0.99) 0.36
ES 8.46 (3.08) 0.84 (0.50, 0.95) 1.23 8.36 (2.82) 0.83 (0.46, 0.94) 1.16
Combined multifidus & ES 17.73 (4.08) 0.95 (0.84, 0.98) 0.91 17.43 (3.74) 0.96 (0.90, 0.99) 0.74
Psoas 10.16 (3.87) 0.85 (0.50, 0.95) 1.49 10.13 (3.96) 0.86 (0.56, 0.95) 1.48
Muscle-to-fat index
Multifidus 0.52 (0.10) 0.91 (0.72, 0.97) 0.03 0.53 (0.08) 0.94 (0.83, 0.98) 0.01
ES 0.54 (0.09) 0.90 (0.68, 0.97) 0.02 0.53 (0.07) 0.92 (0.76, 0.97) 0.01
Combined multifidus & ES 0.53 (0.30) 0.93 (0.77, 0.97) 0.07 0.53 (0.07) 0.94 (0.81, 0.98) 0.01
Psoas 0.38 (0.04) 0.87 (0.60, 0.96) 0.01 0.39 (0.04) 0.79 (0.28, 0.93) 0.01
Relative muscle CSA, cm2
Multifidus 4.40 (1.29) 0.95 (0.85, 0.98) 0.28 4.34 (1.13) 0.97 (0.90, 0.99) 0.19
ES 3.61 (1.19) 0.85 (0.52, 0.95) 0.46 3.62 (1.08) 0.77 (0.25, 0.93) 0.51
Combined multifidus & ES 8.15 (2.00) 0.90 (0.68, 0.97) 0.63 8.01 (1.62) 0.93 (0.77, 0.98) 0.42
Psoas 6.42 (2.15) 0.97 (0.85, 0.99) 0.37 6.39 (2.14) 0.98 (0.93, 0.99) 0.30
Open in a new tab

CSA = Cross-sectional area; cm = centimeters; ES = erector spinae; SD = standard deviation; ICC = intraclass correlation coefficient; %=percentage; CI = confidence interval; SEM = standard error of measurement. Bold indicates ICC point estimate > 0.75 with CI that does not cross 0.

Intra-Examiner Reliability

ICCs and SEMs for rCSA, MFI, and rmCSA for the right side and the average of the right and left sides for repeat measurements by a single examiner are provided in Table 2. Assessments by one examiner of rCSA, MFI, and rmCSA had excellent reliability point estimates for the multifidi and the erector spinae muscles across levels L2–L5 (ICC = 0.77–0.99); CIs are generally narrow, with the exception of assessment of ES at level L5. Overlap of CIs suggests that reliability is similar for isolated measurements of multifidus or erector spinae when compared to the combined measurement of multifidus plus erector spinae for levels L2–L5. Intra-examiner ICC point estimates and CIs for the multifidi and the erector spinae as assessed on a single side versus the average of the right and left sides are similar. Intra-examiner reliability point estimates are excellent for psoas at levels L3 and L5; overlapping CIs suggest similar reliability regardless of whether a single side or the average of the right and left sides is assessed. Quadratus rCSA, MFI, and rmCSA may be reliably assessed at levels L3 and L4 (ICC = 0.81–0.99), with the largest 95% CIs for MFI at L3 (ICC = 0.39–0.96); reliability is similar for the assessment of a single side when compared to the average of the right and left sides. At L4 and L5, SEMs are smaller when multifidus measurements are obtained by averaging the right and left sides when compared to measurements obtained from a single side.

Inter-Examiner Reliability

Table 3 provides ICCs and SEMs for rCSA, MFI, and rmCSA for the right side and the average of the right and left sides combined when measurements by examiner 1 were compared to those of examiner 2. Overall, inter-examiner reliability ICC point estimates for rCSA of the multifidi, erector spinae, and combined measurements of multifidi and erector spinae are good to excellent for levels L2–L4 (ICC = 0.76–0.98); however, wide CIs are noted (95% CIs = 0.24–0.99). Inter-examiner reliability point estimates for psoas rCSA are excellent for the L2 and L3 levels (ICC = 0.93–0.96), while those for the quadratus are generally poor based on 95% CIs that cross 0. When compared to rCSA, lower reliability and wider CIs are noted for MFI and rmCSA, which is not unexpected since MFI and rmCSA inherently have greater variability, due in part to multiple measurements rather than a single measurement. Based on ICC point estimates, CIs, and large SEMs, inter-examiner reliability for measurements at L5 is generally poor. Finally, overall reliability for this analytical procedure for the determination of rCSA, MFI, and rmCSA for the multifidi, erector spinae, psoas, and quadratus lumborum was better when repeat measurements were taken by one examiner rather than two examiners; further, intra-examiner SEMs were lower.

Table 3.

Inter-examiner reliability results

Right
 Right-left average
L2 Mean (SD) ICC (95% CI) SEM Mean (SD) ICC (95% CI) SEM
Relative CSA, cm2
Multifidus 3.72 (1.00) 0.79 (0.32, 0.93) 0.45 3.68 (0.89) 0.83 (0.36, 0.95) 0.36
ES 18.65 (4.49) 0.98 (0.96, 0.99) 0.63 19.07 (4.20) 0.98 (0.95, 0.99) 0.59
Combined multifidus & ES 22.58 (5.05) 0.97 (0.90, 0.99) 0.87 22.94 (4.82) 0.97 (0.91, 0.99) 0.83
Psoas 5.39 (2.12) 0.93 (0.56, 0.98) 0.56 5.89 (2.42) 0.93 (0.79, 0.98) 0.64
Quadratus 2.94 (1.17) 0.87 (−0.10, 0.97) 0.42 3.10 (1.08) 0.86 (−0.10, 0.97) 0.40
Muscle-to-fat index
Multifidus 0.45 (0.12) 0.90 (0.69, 0.97) 0.03 0.46 (0.09) 0.77 (.029, 0.93) 0.04
ES 0.45 (0.10) 0.72 (0.08, 0.91) 0.05 0.47 (0.14) −.09 (−3.32, 0.68) –
Combined Multifidus & ES 0.45 (0.10) 0.76 (0.19, 0.92) 0.04 0.45 (0.09) 0.58 (-.036, 0.87) 0.05
Psoas 0.30 (0.14) 0.59 (−0.26, 0.88) 0.08 0.36 (0.03) 0.64 (−0.25, 0.90) 0.01
Quadratus 0.54 (0.07) 0.70 (0.04, 0.91) 0.03 0.54 (0.06) 0.70 (0.11, 0.90) 0.03
Relative muscle CSA, cm2
Multifidus 1.96 (0.70) 0.89 (0.29, 0.97) 0.23 1.90 (0.60) 0.87 (0.10, 0.96) 0.21
ES 10.27 (3.63) 0.92 (0.75, 0.97) 1.02 9.89 (3.26) −0.15 (−3.56, 0.66) –
Combined multifidus & ES 12.35 (4.21) 0.92 (0.76, 0.97) 1.19 12.51 (3.65) 0.89 (0.68, 0.96) 1.21
Psoas 3.77 (1.38) 0.90 (0.66, 0.97) 0.43 3.72 (1.47) 0.94 (0.80, 0.98) 0.36
Quadratus 1.34 (0.61) 0.96 (.14, .99) 0.12 1.40 (0.55) 0.94 (0.07, 0.98) 0.13
L3 Mean (SD) ICC (95% CI) SEM Mean (SD) ICC (95% CI) SEM
Relative CSA, cm2
Multifidus 5.47 (2.00) 0.80 (0.34, 0.94) 0.89 5.52 (1.86) 0.90 (0.65, 0.97) 0.58
ES 17.84 (4.10) 0.94 (0.82, 0.98) 1.00 17.78 (3.52) 0.94 (0.80, 0.98) 0.86
Combined multifidus & ES 23.21 (5.03) 0.98 (.89, 0.99) 0.71 23.30 (4.36) 0.97 (0.90, 0.99) 0.75
Psoas 8.80 (3.52) 0.94 (0.71, 0.98) 0.86 9.16 (3.86) 0.96 (0.41, 0.99) 0.77
Quadratus 3.72 (1.75) 0.87 (−0.12, 0.97) 0.63 4.09 (1.67) 0.88 (−0.11, 0.97) 0.57
Muscle-to-fat index
Multifidus 0.47 (0.11) 0.85 (0.53, 0.95) 0.04 0.48 (0.08) 0.61 (−0.14, 0.87) 0.04
ES 0.48 (0.12) 0.73 (0.07, 0.91) 0.06 0.47 (0.09) 0.64 (−0.23, 0.89) 0.05
Combined multifidus & ES 0.47 (0.10) 0.83 (0.48, 0.95) 0.04 0.47 (0.07) 0.68 (−0.10, 0.90) 0.03
Psoas 0.40 (0.04) 0.78 (0.21, 0.93) 0.01 0.41 (0.04) 0.80 (0.30, 0.94) 0.01
Quadratus 0.55 (0.03) 0.53 (−0.26, 0.85) 0.02 0.54 (0.04) 0.69 (−0.09, 0.91) 0.02
Relative muscle CSA, cm2
Multifidus 2.86 (1.17) 0.87 (0.16, 0.96) 0.42 2.84 (1.04) 0.90 (0.14, 0.97) 0.32
ES 9.27 (3.36) 0.90 (0.67, 0.97) 1.06 9.35 (2.49) 0.88 (0.61, 0.96) 0.86
Combined multifidus & ES 12.25 (4.25) 0.91 (0.72, 0.97) 1.27 12.28 (3.15) 0.89 (0.65, 0.96) 1.04
Psoas 5.25 (2.17) 0.95 (0.85, 0.98) 0.48 5.38 (2.27) 0.96 (0.81, 0.99) 0.45
Quadratus 1.64 (0.79) 0.92 (−0.07, 0.98) 0.22 1.84 (0.74) 0.92 (−0.06, 0.98) 0.2
L4 Mean (SD) ICC (95% CI) SEM Mean (SD) ICC (95% CI) SEM
Relative CSA, cm2
Multifidus 9.42 (2.90) 0.91 (0.45, 0.97) 0.87 9.37 (2.84) 0.94 (0.62, 0.98) 0.69
ES 13.87 (2.39) 0.76 (0.24, 0.93) 1.17 13.70 (2.56) 0.81 (0.41, 0.94) 1.11
Combined multifidus & ES 23.36 (4.09) 0.87 (0.58, 0.96) 1.47 23.14 (93.95) 0.89 (0.64, 0.97) 1.31
Psoas 11.35 (3.75) 0.94 (−0.03, 0.99) 0.91 11.71 (4.24) 0.95 (0.17, 0.99) 0.94
Quadratus 4.91 (1.95) 0.84 (0.14, 0.96) 0.78 5.11 (1.88) 0.90 (0.00, 0.98) 0.59
Muscle-to-fat index
Multifidus 0.53 (0.10) 0.74 (0.21, 0.91) 0.05 0.53 (0.08) 0.61 (−0.10, 0.87) 0.04
ES 0.49 (0.10) 0.78 (0.29, 0.93) 0.04 0.48 (0.07) 0.79 (0.33, 0.93) 0.03
Combined multifidus & ES 0.51 (0.11) 0.75 (0.24, 0.92) 0.05 0.50 (0.07) 0.69 (0.05, 0.90) 0.03
Psoas 0.38 (0.04) 0.76 (0.00, 0.93) 0.01 0.39 (0.04) 0.80 (0.32, 0.94) 0.01
Quadratus 0.55 (0.05) 0.84 (0.11, 0.96) 0.02 0.51 (0.11) 0.12 (−1.19, 0.71) 0.10
Relative muscle CSA, cm2
Multifidus 4.46 (1.87) 0.84 (0.35, 0.95) 0.74 4.46 (1.58) 0.83 (0.27, 0.95) 0.65
ES 6.98 (1.96) 0.66 (−0.10, 0.89) 1.14 7.10 (1.66) 0.64 (− 0.07, 0.89) 0.99
Combined multifidus & ES 11.47 (3.50) 0.75 (0.24, 0.92) 1.75 11.58 (2.84) 0.70 (0.11, 0.90) 1.55
Psoas 7.04 (2.56) 0.97 (0.62, 0.99) 0.44 7.16 (2.75) 0.97 (0.49, 0.99) 0.47
Quadratus 2.16 (0.90) 0.89 (0.63, 0.96) 0.29 2.21 (0.97) 0.88 (0.55, 0.96) 0.33
L5 Mean (SD) ICC (95% CI) SEM Mean (SD) ICC (95% CI) SEM
Relative CSA, cm2
Multifidus 9.99 (1.86) 0.54 (−0.22, 0.85) 1.36 9.96 (1.78) 0.53 (−0.23, 0.84) 1.22
ES 8.33 (2.59) 0.78 (0.33, 0.93) 1.21 8.32 (2.49) 0.82 (0.44, 0.94) 1.05
Combined multifidus & ES 18.03 (3.79) 0.65 (−0.08, 0.89) 2.24 17.85 (3.42) 0.56 (−0.39, 0.86) 2.26
Psoas 10.84 (3.96) 0.75 (0.02, 0.93) 1.98 10.77 (4.04) 0.81 (0.12, 0.95) 1.76
Muscle-to-Fat Index
Multifidus 0.51 (0.08) 0.64 (−0.16, 0.89) 0.04 0.51 (0.07) 0.64 (−0.06, 0.88) 0.04
ES 0.52 (0.24) 0.72 (0.17, 0.91) 0.12 0.50 (0.09) 0.63 (−0.08, 0.88) 0.05
Combined Multifidus & ES 0.52 (0.08) 0.69 (0.03, 0.90) 0.04 0.51 (0.26) 0.58 (−0.18, 0.86) 0.16
Psoas 0.39 (0.04) 0.88 (0.52, 0.96) 0.01 0.39 (0.04) 0.71 (0.13, 0.91) 0.02
Relative muscle CSA, cm2
Multifidus 4.84 (1.12) 0.62 (0-.18, 0.88) 0.69 4.83 (1.03) 0.50 (−0.26, 0.83) 0.72
ES 3.74 (1.04) 0.89 (0.67, 0.96) 0.34 3.79 (1.17) 0.73 (0.10, 0.92) 0.60
Combined Multifidus & ES 8.60 (1.77) 0.74 (0.14, 0.92) 0.90 8.39 (1.88) 0.69 (0.08, 0.90) 1.04
Psoas 6.70 (2.14) 0.90 (0.34, 0.97) 0.67 6.68 (2.17) 0.93 (0.48, 0.98) 0.57
Open in a new tab

CSA = cross-sectional area; cm = centimeters; ES = erector spinae; SD = standard deviation; ICC = intraclass correlation coefficient; %=percentage; CI = confidence interval; SEM = standard error of measurement. Bold indicates ICC point estimate > 0.75 with CI that does not cross 0.

Discussion

With respect to the assessment of MFI and rmCSA among adults with LBP, few studies targeting adults over 60 years old have been conducted. This study is the first step toward establishing an analytical procedure for extracting data related to MFI and rmCSA from MRIs obtained in older adults with LBP that considers the availability of equipment and images and age-related challenges. Our procedure, shown to be reliable when performed by a single examiner, is performed on T1-weighted images. Clinically, T1-weighted images are typically obtained with routine clinical protocols for patients with chronic LBP, and the ImageJ software, used for analysis, is free to the public, resulting in no additional cost to the patient or examiner. Further, we have adapted previously established procedures for the determination of MFI and rmCSA based on challenges noted when analyzing trunk muscles of older adults, and we have shown that the methods proposed in this paper can be reliable in the assessment of multifidi and erector spinae at levels L2–L5, psoas at levels L3 and L5, and quadratus lumborum at levels L3 and L4. A single examiner can reliably perform a combined measurement of the lumbar (L2–L5 levels) multifidi and erector spinae rCSA, MFI, and rmCSA. An advantage of the combined measurement may be decreased data processing time, but this advantage should be weighed against the loss of independent multifidus data. Further, this article provides SEMs that may be valuable in determining whether changes in MFI and rmCSA over time exceed measurement error. For example, when a single examiner conducts measurements, changes in MFI that exceed 0.07, regardless of the muscle or vertebral level, may indicate a “true change.”

Among younger adults with LBP, and similar to other musculoskeletal conditions (e.g., rotator cuff injuries) [43–45], previous investigators utilized visual grading systems for the evaluation of trunk muscle intramuscular fat [12,46,47]. Recently, Battaglia and colleagues demonstrated inter-examiner reliability for both a visual grading system and quantitative analytical procedure for determining multifidus intramuscular fat using T1-weighted MRI scans in adults (mean age: 55.84 years) [47]. The following ordinal scale was used for the visual grading system: grade 0: normal muscle; grade 1: fatty streaks within muscle; grade 2: fat less than muscle; grade 3: fat and muscle equal; and grade 4: fat greater than muscle; their average grade was 1.90 [47]. Among our participants, the average right-left MFI for multifidi, erector spinae, and quadratus lumborum indicated that 43–53% of the rCSA was fat. Therefore, we question whether a similar reliability for this visual grading system could have been obtained in our sample as we believe it would be more difficult to differentiate between grades 2 and 4 than grades 1 and 2.

For the quantitative procedure, Battaglia et al. used minimal and maximal pixel intensity values obtained via ImageJ to determine intramuscular fat pixel intensities [47] rather than a ratio, as in our study. Furthermore, Battaglia et al. [47] utilized a single MR slice when determining values related to intramuscular fat compared to our multiple-slice protocol. Despite methodological differences, our ICCs were similar to those of Battaglia et al. [47]. Further, the MFI and rmCSA procedures employed in our study were previously used by Elliott et al. to quantify muscle composition in adults with chronic traumatic neck pain [27], with one exception—the MFI in our study was derived from a mean extramuscular fat pixel intensity rather than intermuscular fat pixel intensity. In our older adult population, during pilot testing, we were unable to consistently determine a region of intra- or intermuscular fat, void of muscle tissue, for the muscles of interest. Instead, we opted for a region of extramuscular fat adjacent to the muscle. The use of extramuscular fat pixel intensity profiles, however, could have resulted in greater MFI and smaller rmCSA values for our participants, which could be due to coil proximity. Nonetheless, our MFI and rmCSA results for multifidi and erector spinae are similar to those published by Fortin et al. among older adults who used minimal and maximal pixel intensities of lean muscle tissue to determine MFI and rmCSA [9]. Validation of our analytical procedure for the assessment of MFI and rmCSA against a criterion standard, such as proton magnetic resonance spectroscopy, proton-density fat fraction, or chemical shift Dixon techniques [24], is needed.

While, clinically speaking, USI may be used as a valid means of assessing multifidus muscle CSA, i.e., rCSA [21], USI is largely unable to differentiate fat from muscle. Given that intramuscular fat appears to be approximately 50% of the rCSA among older adults with chronic LBP, USI assessments may be significantly overestimating “true” multifidus muscle CSA in these individuals. Similarly, our findings show that erector spinae, quadratus lumborum, and psoas USI CSA assessment among older adults with chronic LBP should be cautiously interpreted as a significant portion of the CSA may be fat.

Study Limitations

Images were obtained on a single 1.5 Tesla magnet. Magnet strength, parameter selection, patient positioning, patient anthropometrics, or movement artifact may have impacted our reliability results. Prior to the start of the study, examiner 1’s experience was predominantly in lumbar spine image analysis, while examiner 2’s experience was in the cervical spine; differing examiner experience may help explain our reliability results. Sample size was intentionally kept small to determine whether inter-examiner reliability could be established with minimal examiner training and practice; inter-examiner reliability may be improved with increased training or practice. While a small clinical trial could utilize a single examiner for image analysis, larger clinical trials utilizing this analytical technique (or other advanced, available, and possibly time-efficient measures [24]) may find it more feasible to attempt to improve inter-examiner reliability through increased training and practice. We believe a larger sample size would likely improve reliability owing to inherent variability in measurements, particularly for MFI and rmCSA. Further, despite advancements in MRI technology, such analyses may be limited to clinical trial research given that current economic costs associated with advanced imaging in healthcare may preclude the referral for and performance of serial MRIs in the vast majority of patients.

Conclusions

Assessment of rCSA, MFI, and rmCSA in older adults with chronic LBP can be reliably determined by one examiner from T1-weighted images. Proposed methods are procedurally simple, providing valuable information given that rCSA appears to significantly overestimate muscle tissue in these individuals. It appears that a substantial amount of the rCSA of trunk muscles among older adults with chronic LBP may be magnetic resonance-visible MFI. Future studies with larger samples utilizing advanced, but available, MRI techniques to confirm these findings are warranted.

Funding sources: The work of Dr. Sions is supported in part by the Foundation for Physical Therapy PODS I/II, the Fellowship for Geriatric Research Award from the Academy of Geriatric Physical Therapy, R21 HD057274 (NICHD), and 1R01AG041202-01(NIA). The work of Andrew C. Smith is supported by T32 HD057845 (NIH) and in part by the Foundation for Physical Therapy PODS I/II. The work of Dr. Hicks is supported in part by R21 HD057274 (NICHD) and R01AG041202-01(NIA).

Disclosure and conflicts of interest: Dr. Elliott has financial activities outside this work that include an ownership/investment interest in a medical consulting start-up, Pain ID, LLC. The authors have no conflicts of interest related to this work.

References

  • 1.Panjabi MM. The stabilizing system of the spine. Part I. Function, dysfunction, daptation, and enhancement. J Spinal Disord 1992;5:383–9. [DOI] [PubMed] [Google Scholar]
  • 2.Panjabi MM. The stabilizing system of the spine. Part II. Neutral zone and instability hypothesis. J Spinal Disord 1992;5:390–6. [DOI] [PubMed] [Google Scholar]
  • 3.Briggs AM, Greig AM, Wark JD, Fazzalari NL, Bennell KL. A review of anatomical and mechanical factors affecting vertebral bodintegrity. Int J Med Sci 2004;1:170–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Wilke HJ, Wolf S, Claes LE, Arand M, Wiesend A. Stability increase of the lumbar spine with different muscle groups. A biomehanical in vitro study. Spine 1995;20:192–8. [DOI] [PubMed] [Google Scholar]
  • 5.Moseley GL, Hodges PW, Gandevia SC. Deep and superficial fibers of the lumbar multifidus muscle are differentially active during voluntary arm movements. Spine 2002;27:E29–36. [DOI] [PubMed] [Google Scholar]
  • 6.Philippon MJ, Devitt BM, Campbell KJ, et al. Anatomic variance of the ilipsoas tendon. Am J Sports Med 2014;42:807–11. [DOI] [PubMed] [Google Scholar]
  • 7.Regev GJ, Kim CW, Tomiya A, et al. Psoas muscle architectural design, in vivo sarcomere length range, and passive tensile properties support its role as a lumbar spine stabilizer. Spine 2011;36:E1666–74. [DOI] [PubMed] [Google Scholar]
  • 8.McGill S, Juker D, Kropf P. Quantative intramuscular myoelectric activity of quadratus lumborum during a wide variety of tasks. Clin Biomech 1996;11:170–2. [DOI] [PubMed] [Google Scholar]
  • 9.Fortin M, Macedo LG. Multifidus and paraspinal muscle group cross-sectional areas of patients with low back pain and control patients: A systematic review with a focus on blinding. Phys Ther 2013;93:873–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Hadar H, Gadoth N, Heifetz M. Fatty replacement of lower paraspinal muscles: Normal and neuromuscular disorders. AJR A J Roentgenol 1983;141:895–8. [DOI] [PubMed] [Google Scholar]
  • 11.Kamaz M, Kiresi D, Oguz H, Emlik D, Levendoglu F. CT measurement of trunk muscle area in patients with chronic low back pain. Diagn Interv Radiol 2007;13:144–8. [PubMed] [Google Scholar]
  • 12.Kjaer P, Bendix T, Sorensen JS, Korsholm L, Leboeuf-Yde C. Are MRI-defined fat infiltrations in the multifidus muscles associated with low back pain? BMC Med 2007;5:2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Sions JM, Tyrell CM, Knarr BA, Jancosko A, Binder-Macleod SA. Age- and stroke-related skeletal muscle changes: A review for the geriatric clinician. J Geriatr Phys Ther 2012;35:155–61. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Chen YY, Pao JL, Liaw CK, Hsu WL, Yang RS. Image changes of paraspinal muscles and clinical correlations in patients with unilateral lumbar spinal stenosis. Eur Spine J 2014;23:999–1006. [DOI] [PubMed] [Google Scholar]
  • 15.Goodpaster BH, Carlson CL, Visser M, et al. Attenuation of skeletal muscle and strength in the elderly: The Health ABC Study. J Appl Physiol 2001;90:2157–65. [DOI] [PubMed] [Google Scholar]
  • 16.Hicks GE, Simonsick EM, Harris TB, et al. Trunk muscle composition as a predictor of reduced functional capacity in the health, aging and body composition study: The moderating role of back pain. J Gerontol a Biol Sci Med Sci 2005;60:1420–4. [DOI] [PubMed] [Google Scholar]
  • 17.Hultman G, Nortin M, Saraste H, Ohlsen H. Body composition, endurance, strength, cross-sectional area, and density of MM erector spinae in men with and without low back pain. J Spinal Disord 1993;6:114–23. [PubMed] [Google Scholar]
  • 18.Hicks GE, Simonsick EM, Harris TM, et al. Cross-sectional associations between trunk muscle composition, back pain, and physical function in the health aving and body composition study. J Gerontol a Biol Sci Med Sci 2005;60:882–7. [DOI] [PubMed] [Google Scholar]
  • 19.Hayden JA, van Tulder MV, Malmivaara A, Koes BW. Exercise therapy for treatment of non-specific low back pain. Cochrane Database Syst Rev 2005;20:CD000335. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Pressler JF, Heiss DG, Buford JA, Chidley JV. Between-day repeatability and symmetry of multifidus cross-sectional area measured using ultrasound imaging. J Orthop Sports Phys Ther 2006;36:10–8. [DOI] [PubMed] [Google Scholar]
  • 21.Hides JA, Richardson CA, Jull GA. Magnetic resonance imaging and ultrasonography of the lumbar multifidus muscle. Comparison of two different modalities. Spine (Phila Pa 1976) 1995;20:54–8. [DOI] [PubMed] [Google Scholar]
  • 22.Sions JM, Velasco TO, Teyhen DS, Hicks GE. Ultrasound imaging: Intraexaminer and interexaminer reliability for multifidus muscle thickness assessment in adults aged 60 to 85 years versus younger adults. J Orthop Sports Phys Ther 2014;44:425–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Fortin M, Videman T, Gibbons LE, Battie MC. Paraspinal muscle morphology and composition: A 15-yr longitudinal magnetic resonance imaging study. Med Sci Sports Exerc. 2014;46:893–901. [DOI] [PubMed] [Google Scholar]
  • 24.Reeder SB, Hu HH, Sirlin CB. Proton density fat-fraction: A standardized MR-based biomarker of tissue fat concentration. J Magn Reson Imaging 2012;36:1011–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Elliott JM, Walton DM, Rademaker A, Parrish TB. Quantification of cervical spine muscle fat: A comparison between T1-weighted and multi-echo gradient echo imaging using a variable projection algorithm (VARPRO). BMC Med Imaging 2013;13:30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Smith AC, Parrish TB, Abbott R, et al. Muscle-fat MRI: 1.5 tesla and 3.0 tesla versus histology. Muscle Nerve 2014;50:170–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Elliott JM, Pedler AR, Jull GA, et al. Differential changes in muscle composition exist in traumatic and nontraumatic neck pain. Spine (Phila Pa 1976) 2014;39:39–47. [DOI] [PubMed] [Google Scholar]
  • 28.Delfaut EM, Beltran J, Johnson G, et al. Fat suppression in MR imaging: Techniques and pitfalls. Radiographics 1999;19:373–82. [DOI] [PubMed] [Google Scholar]
  • 29.Paalanne N, Niinimaki J, Karppinen J, et al. Assessment of association between low back pain and paraspinal muscle atrophy using opposed-phase magnetic resonance imaging: A population-based study among young adults. Spine (Phila Pa 1976) 2011;36:1961–8. [DOI] [PubMed] [Google Scholar]
  • 30.de Kerviler E, Leroy-Willig A, Clement O, Frija J. Fat suppression techniques in MRI: An update. Biomed Pharmacother 1998;52:69–75. [DOI] [PubMed] [Google Scholar]
  • 31.Elliott JM, Kerry R, Flynn T, Parrish TB. Content not quantity is a better measure of muscle degeneration in whiplash. Man Ther 2013;18:578–82. [DOI] [PubMed] [Google Scholar]
  • 32.Le Cara EC, Marcus RL, Dempsey AR, Hoffman MD, Hebert JJ. Morphology versus function: The relationship between lumbar multifidus intramuscular adipose tissue and muuscle function among patients with low back pain. Arch Phys Med Rehabil 2014;95:1846–985. [DOI] [PubMed] [Google Scholar]
  • 33.Mengiardi B, Schmid MR, Boos N, et al. Fat content of lumbar paraspinal muscles in patients with chronic low back pain and in asymptomatic volunteers: Quantification with MR spectroscopy. Radiology 2006;240:786–92. [DOI] [PubMed] [Google Scholar]
  • 34.D'Hooge R, Cagnie B, Crombez Get al. Increased intramuscular fatty infiltration without differences in lumbar muscle cross-sectional area during remission of unilateral recurrent low back pain. Man Ther 2012;17:584–8. [DOI] [PubMed] [Google Scholar]
  • 35.Knauer SR, Freburger JK, Carey TS. Chronic low back pain among older adults: A population-based perspective. J Aging Health 2010;22:1213–34. [DOI] [PubMed] [Google Scholar]
  • 36.Weiner DK, Kim YS, Bonino P, Wang T. Low back pain in older adults: Are we utilizing healthcare resources wisely? Pain Med 2006;7:143–50. [DOI] [PubMed] [Google Scholar]
  • 37.Murphy WA, Totty WG, Carroll JE. MRI of normal and pathologic skeletal muscle. AJR Am J Roentgenol 1986;146:565–74. [DOI] [PubMed] [Google Scholar]
  • 38.Elliott J, Jull G, Noteboom JT, Galloway G. MRI study of the cross-sectional area for the cervical extensor musculature in patients with persistent whiplash associated disorders (WAD). Man Ther 2008;13:258–65. [DOI] [PubMed] [Google Scholar]
  • 39.Willemink MJ, van Es HW, Helmhout PH, et al. The effects of dynamic isolated lumbar extensor training on lumbar multifidus functional cross-sectional area and functional status of patients with chronic nonspecific low back pain. Spine 2012;37:E1651–8. [DOI] [PubMed] [Google Scholar]
  • 40.Elliott J, Jull G, Noteboom JT, et al. Fatty infiltration in the cervical extensor muscles in persistent whiplash-associated disorders: A magnetic resonance imaging analysis. Spine 2006;31:E847–55. [DOI] [PubMed] [Google Scholar]
  • 41.Fleiss JL. The Design and Analysis of Clinical Experiments. Wiley Classics Library Edition New York: Wiley; 1999. [Google Scholar]
  • 42.Arab AM, Rasouli O, Amiri M, Tahan N. Reliability of utlrasound measurement of automatic activity of the abdominal muscle in participants with and without chronic low back pain. Chiropr Man Therap 2013;21:37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Gerber C, Schneeberger AG, Hoppeler H, Meyer DC. Correlation of atrophy and fatty infiltration on strength and integrity of rotator cuff repairs: A study in thirteen patients. J Shoulder Elbow Surg 2007;16:691–696. [DOI] [PubMed] [Google Scholar]
  • 44.Gladstone JN, Bishop JY, Lo IK, Flatow EL. Fatty infiltration and atrophy of the rotator cuff do not improve after rotator cuff repair and correlate with poor functional outcome. Am J Sports Med 2007;35:719–28. [DOI] [PubMed] [Google Scholar]
  • 45.Meyer DC, Pirkl C, Pfirrmann CW, Zanetti M, Gerber C. Asymmetric atrophy of the supraspinatus muscle following tendon tear. J Orthop Res 2005;23:254–8. [DOI] [PubMed] [Google Scholar]
  • 46.Parkkola R, Rytokoski U, Kormano M. Magnetic resonance imaging of the discs and trunk muscles in patients with chronic low back pain and healthy control subjects. Spine 1993;18:830–6. [DOI] [PubMed] [Google Scholar]
  • 47.Battaglia PJ, Maeda Y, Welk A, Hough B, Kettner N. Reliability of the Goutallier classification in quantifying muscle fatty degeneration in the lumbar multifidus using magnetic resonance imaging. J Manipulative Physiol Ther 2014;37:190–7. [DOI] [PubMed] [Google Scholar]

Articles from Pain Medicine: The Official Journal of the American Academy of Pain Medicine are provided here courtesy of Oxford University Press

RESOURCES