Page 34 - JSOM Winter 2021

TABLE 1 Demographic Variables of the Control Group and the
Chronic Low Back Pain Group
CTRL LBP
No. 22 22
Age (y) 28.0 ± 3.9 28.7 ± 4.4
Mass (kg) 83.5 ± 7.8 84.2 ± 7.0
Height (m) 1.79 ± 0.05 1.77 ± 0.06
Body mass index 26.7 ± 2.1 27.4 ±1.8
FIGURE 3 Representative Time of service (y) 7.4 ± 3.9 8.2 ± 4.0
set up for isokinetic knee E4 = 1, E5 = 11, E4 = 2, E5 = 7,
quadriceps and hamstring Rank E6 = 2, E7 = 3, E6 = 9, E7 = 1,
strength on an isokinetic O3 = 3 E8 = 1, O3 = 2
dynamometer.
Days per week of 4.6 ± 1.7 3.8 ± 1.9
physical activity
Time of pain (y) N/A 3.9 ± 2.5
Approximate
number of healthcare N/A 25.4 ± 30.6
visits for LBP
CTRL = control group, LBP = low back pain group

in both groups reported completing some type of physical
training at least 3 days a week. The LBP group reported hav-
ing pain on average for 3.9 ± 2.5 years prior to participating
in this study and all indicated that while the low back pain
was unspecified in nature, it likely stemmed from one of three
places: (1) physical training, (2) load carriage, and/or (3)
occupational-related.

There were no differences between the LBP and the CTRL
FIGURE 4 Representative group in any strength or biomechanical variables (Table 2).
set up for isokinetic trunk
flexion and extension There also were no differences between groups in jump height
strength measured on an during the double-limb stop jump. Group comparisons identi-
isokinetic dynamometer. fied that the LBP group exhibited higher VGRF impulse on the
dominant limb compared to the nondominant limb (p = .036),
with the dominant limb experiencing a take-off VGRF impulse
of 0.26% BW/s, while the nondominant limb experienced less
of an impulse at 0.25% BW/s (Table 2). The between-limb dif-
ferences in VGRF impulse were small yet significant and had
clinical significance with an effect size of 0.512.
We identified relationships between strength and take-off bio-
mechanics variables to jump height via correlations (Table 3).
Shapiro-Wilks tests were run to determine normal distribu- The CTRL group exhibited significant positive relationships
tion of all variables. Independent sample t-tests were used for between jump height and both dominant and nondominant
normally distributed variables, while Mann-Whitney U tests quadriceps (dominant: r = 0.436, p = .043; nondominant:
were used for non-normally distributed data of between group r = 0.571, p = .006) and hamstring (dominant: r = 0.575, p =
analyses. For within-group differences comparing the domi- .005; nondominant: r = 0.654, p = .001) strength and trunk
nant and nondominant limbs, paired t-tests were used for nor- flexion (r = 0.501, p = .015) and extension (r = 0.663, p =
mally distributed data, while Wilcoxon signed rank tests were .001) strength. The relationship of strength to jump height
used for non-normally distributed data. Comparison of means was not present in the LBP group. In the LBP group there were
will be coupled with Cohen d effect sizes. Effect sizes (ES) significant positive relationships between jump height and
were interpreted as weak (≤ 0.50), moderate (0.50–0.79), or dominant limb knee work (r = 0.470, p = .027) and ankle
strong (≥ 0.80). Pearson’s correlation coefficients were used work (r = 0.447, p = .037). The relationship between domi-
17
to examine relationships between jump height, strength, and nant limb knee and ankle work was not present in the CTRL
biomechanical variables from the take-off phase for both the group. There were also no significant relationships between
low back pain (LBP) and CTRL group. A significance level of jump height and nondominant knee or ankle work, or VGRF
alpha = .05 was used. All data analyses were conducted using impulse on either limb, from either group.
SPSS (IBM, https://www.ibm.com/products/spss-statistics).
Discussion
Results
The purpose of this study was to determine if MARSOC per-
A total of 44 MARSOC personnel were recruited for this study. sonnel with low back pain exhibited reduced strength or ex-
Demographic variables are presented in Table 1. Individuals plosive performance compared to controls, and to examine

32 | JSOM Volume 21, Edition 4 / Winter 2021