Four events (Figure 4-5, E1-E4) were defined by force data and the position of the marker placed on the sacrum. The first, onset (E1), was defined as the point at which the sitting subject starts the first movement by leaning the trunk forward and causing
second, seat-off (E2), was defined as the point at which the subject changes from a sitting to a standing position, defined as the point in time when the vertical force under the stool becomes equal to the weight of the stool. The third event, standing (E3), was defined as the point at which a standing position is reached, with the marker on the sacrum at its highest position. The fourth event, end of task (E4), represents the instant when a stable standing phase is achieved with no fluctuations greater than 5% of body weight.
Thus, the duration of the SitTS task was defined as the elapsed time from E1 to E4, subdivided into three phases: preparation (E1 to E2), ascending (E2 to E3), and
stabilization (E3 to E4). Figure 4-6 shows data collected from a representative subject.
Vertical ground-reaction force was measured to calculate weight-bearing
asymmetry during the task. The index of asymmetry was defined as follows, Asymmetry index =Vna−Vperfect symmetry
Vperfect symmetry
where Vna is the vertical GRF measured from the non-affected side, and Vperfect symmetry
equals to 50% of the sum of the total vertical GRF under the feet [16]. The unit of index was expressed as a percentage, with 0 representing perfect symmetry, −100%
indicating the load was entirely on the affected side, and +100% corresponding to loading entirely on the non-affected side.
To better understand the fluctuations of vertical GRF during the SitTS task. The vertical GRF under both legs was acquired to calculate the leg load discrepancy during the SitTS task (Figure 4-7). The value of the Leg Load Discrepancy (LLD)
was defined according to the following formula, Load discrepancy = ∫ (Vna
T
0 −Va)dt T
∫ (VT na 0 -Va)dt
T
where T is the duration of the phase and Va is the vertical GRF measured from the affected side [17].
Figure 4-5 Events of sit to stand task.
Figure 4-6 Bilateral leg load in FspHk and FabHc position.
Figure 4-7 Asymmetrical leg load compared to FspHk and FabHc.
4.3.2 Statistical analysis
Descriptive statistics (means) were calculated for duration, asymmetry index and leg load discrepancy. Results were analyzed to determine the effect of foot and hand positions during the SitTS task for the duration and LLD were assessed for each phase. The asymmetry index was calculated at each event. Because some subjects failed to perform few difficult postural configurations (such as Faf), it led to data missing for further statistical analysis. Therefore, we used the Generalized Estimating Equations (GEE) method to evaluate the effect of foot and hand position on duration, asymmetry index, and leg load discrepancy, instead of using repeated-measure ANOVA. Where results of GEE were significant only in main effect of foot or hand, pair-wise comparisons were tested to determine significant differences for each foot or hand components. If any complex interactions of foot and hand (foot x hand), a further simple effect analysis with pairwise contrasts (a Bonferroni adjustment) to determine where the differences were, such as fixed one independent variable (e.g., foot position) on each of the other positions (e.g., for Hc, Hk, and Ha)
Results
The effect of foot and hand position on the duration of SitTS (Figure 4-8).
An interaction between foot and hand positioning and the duration of preparation and ascending phases was observed. The FabHc position led to the shortest preparation phase duration (0.51 s) and the FspHa position led to the longest (0.59 s). In the Fab
position, Ha position led to a longer preparation phase duration than Hc or Hk
positions [the difference of duration between configurations (Δ), Δ (FabHa vs. FabHc)
= 0.071 s, P = 0.031 and Δ (FabHa vs. FabHk) = 0.044 s, P = 0.012]. In the Faf
position, the Hc position was found to lead to a shorter duration than Ha or Hk positions (Δ (FafHc vs. FafHa) = 0.070 s, P = 0.044 and Δ (FafHc vs. FafHk) = 0.060
s, P = 0.004). No significant differences were observed between other foot and hand positions. For the ascending phase, the FspHc position led to the shortest movement duration (1.21 s) and the FabHk position led to the longest (1.41 s). An effect of foot position and hand position on the duration of the stabilization phase was observed.
The mean duration of the stabilization phase significantly differed among different foot (Wald X2 = 10.03, P = 0.018) and and positions and (Wald X2 = 32.56, P <
0.001). Therefore, further pairwise tests were performed demonstrating the duration in the Fab position was longer than Faf (Fab > Faf, P = 0.01) or Fsp (Fab > Fsp, P = 0.004) positions, and the duration in the Hc position was longer than in Ha or Hk
positions (P < 0.001 for Hc > Ha, Hk).
Figure 4-8 The durations of sit to stand by phase: (A) preparation phase. (B) ascending phase. (C) stabilization phase. (D) total sit to stand. The values were plot by mean with standard deviation.
Fab, Affected foot backward; Faf, Affected foot forward; Fsp, Foot spontaneous; Fs, Foot Symmetry; Hk, Hands on knees; Hc, Hands clasped; Ha, Hands aside.
E1, Onset; E2, Seat-off; E3, Standing; and E4, End of Task.
No interaction (foot x hand) was observed between the total duration of the SitTS task, but for the main effect of foot (Wald X2 = 9.77, P = 0.021) and effect of hand (Wald X2 = 19.42, P < 0.001) were significant. The affected foot placed backward (Fab) position led to a longer duration than Faf (Fab > Faf, P = 0.027) or Fsp (Faf >
Fsp, P = 0.002) positions. The Hc position led to a longer than duration than Hk or Ha (Hc > Hk, Ha, P < 0.001) positions. The shortest mean duration was 2.57 s (FspHk) and the longest was 3.36 s (FabHc). Compared to the FspHk (2.57 s) position, the normal posture of patients, subjects in the affected foot backward and hands clasped position needed 30% longer (3.36 s) to accomplish the SitTS task.
The effects of hand and foot position on the asymmetry index at each event in the SitTS task (Figure 4-8). In general, no interactions of foot x hand position on the asymmetry index were observed for any events. At standing (E3), no significant difference between the mean asymmetry index and foot and hand position was found.
A significant different was found among foot position on the asymmetry index at E4 only (Wald X2 = 27.79, P < 0.001).
At onset (E1), the mean asymmetry indices were all negative, indicating that the affected leg took more load than the non-affected leg. At this event, the Fs position
had increased asymmetry than Fsp (Fs > Fsp, P = 0.047) and the Hc position led to decreased asymmetry than Hk (Hc >Hk, P = 0.047).
At seat-off (E2), pairwise comparisons revealed asymmetry in the Fab position was significantly lower than in the Faf, Fsp, or Fs positions (P < 0.001,Figure 4-8), and significantly lower in the Hc position than in the Hk position (P = 0.004, Figure 4-9 ).
Similar results were found for foot positions at the end of the task (E4). Although asymmetry in the Hc position was slightly higher than in other hand positions, it did not reach statistical significance.
(A) (B)
(C) (D)
Figure 4-9 The asymmetry Index of events according to Foot and Hand Positions:
(A) onset. (B) seat off. (C) standing. (D) end of task. All values are represented as mean (standard deviation).
Fab, Affected foot backward; Faf, Affected foot forward; Fsp, Foot spontaneous; Fs, Foot symmetry; Hk, Hands on knees; Hc, Hands clasped; Ha, Hands aside.
The effect of hand and foot position on leg load discrepancy during the SitTS task (Figure 4-10). Leg load discrepancy was used as an index for the difference in leg load between the two sides during the task. Significant effects were found only for foot positions, in each phase, throughout the entire task duration, but no such effects were found for hand positions (Table 4-1). The results showed that Faf had the
greatest LLD in each phase and total sitTS. Despite of hand position, the placement of affected foot backward (Fab) was less LLD than Fsp or Fs during the sitTS task.
Figure 4-7 shows a visual presentation of the GRF percentages of each leg in the FspHk and FabHc positions during SitTS task.
(A)
(B)
(C) (D)
Figure 4-10 Leg Load Discrepancy of sit to stand by phase: (A) preparation phase. (B) ascending phase. (C) stabilization phase. (D) total sit to stand. All values are
represented as mean (standard deviation).
Fab, Affected foot backward; Faf, Affected foot forward; Fsp, Foot spontaneous; Fs, Foot symmetry; Hk, Hands on knees; Hc, Hands clasped; Ha, Hands aside.
E1, Onset; E2, Seat-off; E3, Standing; and E4, End of Task.
Table 4-1 Leg load discrepancy by phase.
Factors Wald X2 P value Pairwise test
Preparation phase (E1–E2)
Foot effect 13.36 0.004 Faf > Fs, Fsp, Fab
Fs > Fab
Hand effect 2.95 0.229
Foot x Hand interaction 5.11 0.530
Ascending phase (E2–E3)
Foot effect 25.05 0.003 Faf > Fsp, Fs, Fab
Fs, Fsp > Fab
Hand effect 0.91 0.956
Foot x Hand interaction 3.77 0.708
Stabilization phase (E3–E4)
Foot effect 13.87 <0.001 Faf > Fsp, Fs, Fab
Fsp > Fs, Fab
Hand effect 0.38 0.826
Foot x Hand interaction 2.47 0.872
Total SitTS
Foot effect 22.87 <0.001 Faf > Fsp, Fs, Fab
Fsp > Fab
Hand effect 0.55 0.758
Foot x Hand interaction 5.94 0.430
Discussion
Most studies of the SitTS movement have included constraints on the use of the hands to simplify experimental procedures. However, we believe the use of the hands in performing the SitTS task is common among elderly and hemiparetic patients and likely to significantly influence foot placement. Therefore, the effects of foot and hand placement during the SitTS task were the main foci of this study. Our results demonstrated that a number of events and phases during the SitTS process were influenced by foot and hand position configurations.