Lumbar Kinematic Pattern Variability during Gait and the Effects of Isolated Lumbar Extension Exercise in Chronic Low Back Pain Participants

4.2.1 Results

Results for ILEX strength, VAS, ODI standing ROM and lumbar ROM were already presented in this report. Thus, to avoid replication, only kinematic data and relationships examined between VAS, ODI, ILEX strength and kinematic variables are reported here.

4.2.1.1 Participants

Participant demographics, pain, disability and ILEX strength data are shown in Table 9 for groups. Comparison between groups revealed that most demographic variables at baseline did not significantly differ thus it was considered that randomisation had been successful. The only significantly different characteristic between groups was VAS score (t(22) = 2.420, p = 0.024).

Table 9. Participant group demographics (Study 2)

Training (n = 17) Control (n = 7) p

Age (years) 47+13 44+16 0.645

Stature (cm) 171.90+9.26 180.02+8.92 0.076

Body Mass (Kg) 75.00+15.49 82.92+19.37 0.324

BMI (Kg/m2_{) 25.12+3.10 25.33+4.36 0.899}

Symptom Duration (years) 14+11 12+10 0.800

VAS (mm) 47.26+24.09 23.00+16.62 0.024

ODI (pts) 34.71+12.69 27.15+7.65 0.158

Lumbar Extension Strength (Nm) 177.80+83.80 192.21+67.60 0.691

Gender Ratio (M:F) 8:9 5:2

Note: Results are mean +SD

4.2.1.2 Baseline Kinematic Data

Between group comparisons again revealed that the majority of kinematic variables did not significantly differ at baseline, only sagittal CVo (U = 23.000, Z = -2.318, p = 0.019),

and both transverse Winter’s CV and CVo (respectively; U = 17.000, Z = -2.699, p =

174 | P a g e Due to inclusion of novel methods of determining ensemble average variation in this study (CVp and CVo; O’Dwyer et al., 2009), compared with others use of Winter’s CV research

(Vogt et al., 2001), baseline data were pooled for all participants in order to compare

Winter’s CV, CVp, and CVo in this population of chronic LBP participants. Displacement and Winter’s CV were highest and similar in frontal and transverse planes. Contrastingly CVp and CVo were higher in the sagittal plane than in frontal and transverse planes which

were both also similar. Figure 15 presents a comparison of these pooled data showing

mean and SDs with Winter’s CV, and mean and SDs transformed to zero with both CVp

and CVo. It can be noted that, particularly for the sagittal plane, Winters CV does not offer

an accurate reflection of the variability across the ensembled waveform patterns (SD reflected by width of dotted lines) whereas CVp does.

175 | P a g e

Figure 15. Waveform patterns, Winter’s CV, CVp and CVo. The x axis of each graph reflects the gait cycle (0-100%) with the y axis reflecting angular displacement (movement amplitude in degrees). Waveform patterns, Winters CV, CVp and CVo. Left hand graphs from top to bottom show mean (+ standard deviation) for [a] frontal, [b] sagittal and [c] transverse lumbar spine kinematics and Winter’s CV respectively, with offsets shown by the horizontal axis, for pooled data (n = 24). Right hand graphs from top to bottom show mean (+ standard deviation) with raw signals transformed to zero mean removing offsets and CVp and CVo for the same data.

Spearman’s correlations revealed a significant moderate positive correlation between VAS

and only sagittal plane Winter’s CV (r = .411, p = 0.023). Significant moderate positive correlations were found between ODI and sagittal plane Winter’s CV (r = .457, p = 0.012), transverse plane Winter’s CV (r = .404, p = 0.025) and transverse plane CVp (r = .401, p

176 | P a g e 0.026). Significant moderate negative correlations were also found between ILEX strength and frontal plane CVo(r = -.370, p = 0.045), sagittal plane Winter’s CV (r = -.467, p

=0.014), transverse plane Winter’s CV (r = -.435, p = 0.021), transverse plane CVp (r = -

.411, p = 0.029), transverse plane CVo (r = -.378, p = 0.042) and a significant moderate

positive correlation with transverse plane displacement (r = .442, p =0.020).

4.2.1.3 Effects of Intervention upon Kinematic Variables

Table 10 shows pre and post data for displacement, Winter’s CV, CVp and CVo. Figure 16

presents an example participant’s data to demonstrate change in the waveform pattern variability. Wilcoxon Signed Ranks Exact test revealed significant changes from pre to post only for sagittal plane CVp (W(16), Z = -1.728, p = 0.044) in the training group only

suggesting improvement in stride to stride waveform pattern replication after the intervention. Effect size for saggital CVp in the training group was 0.48.

177 | P a g e

Table 10. Pre and Post Kinematic data

* Denotes significant changes from pre to post

Figure 16. Example participant data showing pre and post individual trial waveforms and pre and post waveform variability transformed to zero (CVp)

Displacement (degrees) CVp (%) CVo (%)

Frontal Sagittal Transverse Frontal Sagittal Transverse Frontal Sagittal Transverse

Training Pre 10.61+3.74 3.92+1.20 8.85+2.72 41.95+16.62 111.99+42.64 46.49+20.57 27.48+18.34 103.94+52.78 41.69+28.15 Post 10.80+2.88 4.31+1.37 9.41+3.26 39.35+12.72 91.09+28.27* 48.20+24.02 25.87+15.02 87.95+41.10 42.35+25.28 Control Pre 8.15+1.94 4.13+1.78 6.91+7.87 52.65+19.23 92.95+27.07 33.41+11.74 32.30+29.09 66.33+69.07 14.15+5.46 Post 7.25+2.31 3.80+1.54 8.86+2.32 56.45+11.82 89.51+26.63 40.25+20.83 44.59+46.13 85.91+39.78 31.66+27.27

178 | P a g e

4.2.2 Discussion

This study yields several novel findings: 1) sagittal plane lumbar kinematic waveform patterns appear to be considerably more variable than frontal or tranverse planes in chronic LBP, observed using novel methods of differentiating offset variability from pattern

variability in this population and in contrast to earlier studies using Winter’s CV, 2)

transverse plane lumbar pattern variability is significantly correlated with ILEX strength and ODI, and 3) a 12 week ILEX resistance training intervention significantly improves sagittal plane pattern variability during gait in chronic LBP participants. These findings potentially offer further understanding regarding relationships between chronic LBP, gait variability and lumbar extensor deconditioning.

Within this study the foremost interest was repeatability of lumbar movement patterns exhibited (intra-subject stride-to-stride variability) as, despite similar average movements occurring amplitudes, symptomatic participants appear less able to replicate these consistently (Vogt et al., 2001). Vogt et al. (2001) reported data using Winter’s CV suggesting lumbar movement variability during gait was significantly higher in chronic LBP participants compared with asymptomatic controls, and that sagittal and transverse plane variability was greater than frontal plane variability. For comparison with previous research Winter’s CV was calculated for the present study’s data. Results for Winter’s CV differed from those of Vogt et al. (2001) where sagittal plane variability appeared lowest (Vogt et al. 2001 – 26.93%; Present study – 6.73%), and both frontal and transverse plane variability was slightly higher (Vogt et al., 2001 – 14.87% and 26.45% frontal/transverse respectively; Present study – 34.74% and 38.66% frontal/transverse respectively).

This difference in sagittal plane Winter’s CV might be accounted for by the large mean offset in the data’s waveform. Vogt et al. (2001) calibrated their measurements to angles during the standing posture to zero their measurements whereas in the present study they were not. The sagittal plane data were instead similar to that reported by Lamoth et al.