Resistive vibration exercise during bed-rest reduces motor control changes in the lumbo-pelvic musculature

Electromyographic signal processing and parameter measurement. (A) Goniometer signal from a movement-trial at 125cyc/min. Of the total of five “regions” (three consecutive movement cycles) of the goniometer signal found to be of sufficient accuracy (see text for details), two regions are marked by the vertical lines. Measurements on the goniometer signal at this stage also provided movement accuracy data. (B) Corresponding EMG signal from the lumbar erector spinae after band-pass filtering (Butterworth) from 20 to 500Hz. The root-mean-square (RMS) was calculated from the EMG signal within each data region at this stage. The RMS data was also used to calculate EMG amplitude “activity-ratios” (ratio) between movement speeds. (C) The entire EMG signal was high-pass (Butterworth) filtered at 100Hz, full-wave rectified, and low-pass filtered (Bessel) at 10Hz to produce the linear-envelope. (D) Maxima and minima are detected in the linear-envelope subsections using a detection algorithm to enable calculation of amplitude-modulation (BTR). The movement frequency is determined from the goniometer signal sub-section from its amplitude spectrum. At this movement frequency, the phase-value (in radians) in the phase spectrum of the EMG linear-envelope sub-section is subtracted from the corresponding value of the goniometer signal sub-section phase spectrum. This generates the phase-lead/lag (PHZ) measurement. To determine lumbo-pelvic flexor and extensor co-contraction (CoCon), the area of overlap between the linear-envelope signals from the abdomen and spinal extensors was calculated. See text for further details on signal processing. Data on PHZ and RMS have been presented in online Supplementary material for the interested reader and are not discussed in detail here.

Changes in lumbo-pelvic extensor–flexor co-contraction (CoCon) and activity-ratios (Ratio) over the course of the study. Values are mean(SEM) percentage difference to baseline (BR1) values. ^∗p<0.05; ^†p<0.01; ^‡ p<0.001 and indicate significance of difference to baseline. BR=bed-rest, R+=recovery. CTRL: inactive control group; RVE: resistive vibration exercise group. Group×study-date effect on ANOVA reached significance for Ratio (p=0.011), but not for CoCon (p=0.058). Movement speed did not influence changes of CoCon over study-date (p=0.34), and effects on ratio was similar across all muscles (p⩾0.59) therefore results are pooled for muscle (ratio only) and movement speeds (ratio and CoCon). Some evidence (p=0.012) was apparent from ANOVA for a different response of ratio between groups over time across movement speeds; see Fig. 3.

Changes in activity-ratios (ratio) over the course of the study dependent upon movement speed. Values are mean(SEM) percentage difference to baseline (BR1) values. ^∗p<0.05; ^†p<0.01; ^‡ p<0.001 and indicate significance of difference to baseline. BR=bed-rest, R+=recovery. CTRL: inactive control group; RVE: resistive vibration exercise group. Data represent ratios of root-mean-square electromyographic amplitude between 75 and 50cyc/min (ratio_75/50), 100 and 75 cyc/min (ratio_100/75) and 125 and 100 cyc/min (ratio_125/100) movement speeds. Effects on ratio were similar across all muscles (p⩾0.59) therefore results are pooled across muscles. Analysis showed a difference in the response of the two subject groups over the course of the study dependent upon movement speed (p=0.012), with the differences between the groups for ratio_75/50 (p=0.0068) and ratio_125/100 (p=0.0036) reaching significance (otherwise p=0.26).

Changes in tonic muscle activity (BTR) in the lumbo-pelvic musculature. Values are mean(SEM) percentage difference to baseline (BR1) values. ^∗p<0.05; ^†p<0.01; ^‡ p<0.001 and indicate significance of difference to baseline. BR=bed-rest, R+=recovery. CTRL: inactive control group (left); RVE: resistive vibration exercise group (right). EO: external oblique, IO: internal oblique, IGM: inferior gluteus maximus, TES: thoracic erector spinae, LES: lumbar erector spinae. Movement speed did not influence changes of each muscle over study-date (p⩾0.11). Therefore results are averaged across movement speeds. ANOVA suggested a different response between groups over time for the different muscles (BTR: p=0.013). Subsequent analyses for each muscle showed a significant group×study-date interaction for BTR in IGM only (p=0.001; otherwise p⩾0.113).

Changes in trunk fat mass as measured by dual X-ray absorptiometry. Values are mean(SEM) percentage difference to baseline values. Baseline values were average of data collected 3-days prior to beginning of bed-rest (BDC-3) and on day-3 of bed-rest (BR2).^∗p<0.05; ^†p<0.01; ^‡p<0.001 and indicate significance of difference to baseline. BR=bed-rest, R+=recovery. CTRL: inactive control group; RVE: resistive vibration exercise group. Although the CTRL group showed stronger increases in trunk fat levels than the RVE group, this effect was not statistically significant on ANOVA (group×study-date: p=0.19).