A sample of 30 subjects with mean age 27.6±4.9 years, mass151.9±30.3 lbs., height 5.4±0.4 feet, BMI 25.0 ±3.9 kg/m2 participated in this study. Fifty-seven percent of the subjects were males (n=17). The distribution of all quantitative variables was approximately normal. There was no significant difference in subjects’ characteristics by study group (p>0.05). Subjects' characteristics are summarized in Table 2.
Primary Analysis
Figure 1. shows the differences in lumbosacral kinematics between groups at the beginning and the end of 1-hour sitting. There was a significant difference in ST and L3 angle among the three study groups at the lowest level of pain, which was at beginning of the sitting period, (p=0.029, h2= 0.23 for ST and p<0.001, h2= 0.44 for L3 angle). Bonferroni post hoc comparisons showed that the difference in ST was only significant between the FP and AEP subgroups (p=0.031), namely, the FP subgroup had slight posterior sacral tilt (-0.45°±2.02°), while the AEP subgroup had an increased anterior sacral tilt (3.35°±2.91°). In addition, the difference in L3 angle was only significant between FP subgroup and healthy controls (p=0.004), and between FP and AEP subgroups (p=0.001). Specifically, FP subgroup had greater L3 flexion (-3.11°±2.53°) compared to healthy controls (0.53°±2.48°) and AEP subgroup (1.19°±1.64°) who demonstrated slight L3 extension.
In contrast, all lumbosacral angles differed significantly among the three study groups at the highest level of pain, which was at minute 60 of sitting period (p<0.001, h2= 0.74 for ST, p<0.001, h2= 0.72 for L3 angle, and p=0.013, h2= 0.36 for RLLA). Bonferroni post hoc comparisons revealed that the difference in ST was only significant between FP subgroup and healthy controls (p<0.001), and between FP and AEP subgroups (p<0.001). Namely, FP had greater posterior sacral tilt (-6.59°±2.95°) compared to healthy controls (1.04°±2.97°) and AEP subgroup (3.45°±2.02°) who exhibited anterior sacral tilt. Furthermore, the difference in L3 angle was significant among all study groups (p<0.01). The FP subgroup had greater L3 flexion compared to healthy controls (-6.82°±2.05° vs. 0.15°±4.24°, p<0.001) and AEP subgroup who demonstrated L3 extension (-6.82°±2.05° vs. 4.97°±2.94°, p<0.001). Also, the AEP had greater L3 extension compared to healthy controls (4.98°±2.94° vs. 0.15°±4.24°, p=0.007). Moreover, the difference in RLLA was only significant between FP subgroup and healthy controls (p=0.020), and between FP and AEP subgroups (p=0.048). The FP subgroup had greater lumbar kyphosis (-7.55°±10.0°) compared to healthy controls (1.20°±5.45°) and AEP subgroup (0.09°±1.56°) who assumed slightly lumbar lordosis.
Secondary Analysis
Figure 2. shows the mean of lumbosacral kinematics of all study groups over the 1-hour sitting. There was a significant difference in mean lumbosacral angles (ST and L3 angle) among the three study groups over the entire 1-hour sitting (p<0.001, h2= 0.43, p<0.001, h2= 0.70, respectively). Bonferroni post hoc comparisons showed that the difference in mean ST angle was only significant between the FP and AEP subgroups (p<0.001). Specifically, the FP subgroup displayed, at large, a posterior sacral tilt presentation (-1.65°±1.16°), whereas the AEP subgroup exhibited an increased anterior sacral tilt (3.90°±2.70°). In addition, the difference in mean L3 was significant among all study groups (p<0.01). The FP subgroup had greater L3 flexion compared to healthy controls (-5.40°±1.60° vs. -0.47°±3.31°, p<0.001) and AEP subgroup who demonstrated L3 extension (-5.40°±1.60° vs. 2.92°±1.70°, p<0.001). In addition, the AEP had greater L3 extension compared to healthy controls (2.92°±1.70°, vs. -0.47°±3.31°, °, p=0.009). However, there was no significant difference in mean RLLA among all study groups (p=0.412).
The results of the analysis of time (min), expressed as a % of the total 1-hour of sitting period, spent in the available ranges of the studied lumbosacral angles showed that the FP subgroup spent more time sitting with posteriorly tilted pelvis (73.3%), whereas the AEP subgroup spent almost all of their sitting time in anterior pelvic tilt (91.1%). In contrast to the pain subgroups, the healthy controls spent 66.7% of their sitting time in anterior pelvic tilt compared to only 33.4% in the posterior direction. In addition, the FP subgroup sat with flexed L3 for 96.7% of the entire sitting period, in contrast, AEP subgroup spent 93.7% of the total sitting time in L3 extension. In comparison to pain subgroups, the healthy controls spent 65.1% of their sitting time in L3 flexion compared to 34.9% in extension. Furthermore, the FP subgroup spent, on average, 53.7% of their sitting time in lower lumbar kyphosis, while the AEP subgroup spent 71.1% of the sitting time in lordosis. Similar to the FP subgroup, healthy controls postured themselves in kyphosis for 53.3% of the sitting time compared to 46.7% in lordosis for the FP subgroup. Refer to Figure 3.
Third Analysis
Figure 4. shows the average pain scores of all groups over the 1-hour sitting. The mixed factorial analysis showed a significant group by time interaction effect for pain (p<0.001, h2=0.47). Results of the one-way ANOVA indicated that the difference in the amount of change from baseline was significant among the three groups at all time periods (p<0.001). Specifically, Bonferroni post hoc comparisons revealed that both pain subgroups significantly differed from healthy controls at all time periods (p<0.05). However, during the first 30 minutes of sitting, pain subgroups did not differ from each other, whereby, during the last 30 minutes, both FP and AEP subgroups were significantly different (p<0.01). Namely, the FP subgroup reported a significant increase in pain scores compared to the AEP subgroup at minute 40 (38.10±15.03 vs. 20.00±9.53, p=0.002), minute 50 (45.50±18.29 vs. 24.50±16.09, p=0.007), and minute 60 (49.20±16.82 vs. 27.40±19.67, p=0.009). Similar results were found when adding pain at baseline as a covariate. The level of pain reported by the pain subgroups increased significantly over time (p<0.001, h2=0.80 for FP and h2=0.44 for AEP), whereby the pain peaked towards the end of the sitting period and increased significantly from baseline after 20 minutes of the sitting period (FP, p<0.01 and AEP, p<0.05).
There was a significant group by time interaction effect for ST after controlling for baseline ST angle (p<0.001, h2=0.45). Results of the one-way ANOVA indicated that the difference in the amount of ST change from baseline was significant among the three groups at minute 20 until the end of the 60-minute sitting period (p<0.05). To narrow the results, the amount of ST change from baseline was only reported for the late phase of sitting (at minute 60). Bonferroni post hoc comparisons revealed that the FP subgroup significantly differed from the AEP subgroup (P<0.001), and healthy controls (p=0.001), but no significant difference was found between AEP subgroup and healthy controls. The degree of ST displayed by the pain subgroups increased significantly over time (FP; p<0.001, h2=0.57; AEP; p<0.001, h2=0.36). Specifically, the FP subgroup showed a significant increase in posterior sacral tilt from baseline at only minute 60 (the end of the sitting period) (-0.45°±2.02° vs. -6.59°±2.95°, p=0.002), but this change was not statistically significant during the first 50 minutes of sitting. In contrast, the AEP subgroup showed a significant increase in anterior sacral tilt from baseline only at minute 40 (3.35°±2.91° vs. 5.98°±2.03°, p=0.005), which then followed by a significant decreased at minute 50 (5.98°±2.03° vs. 3.79°±2.25°, p<0.001) and minute 60 (5.98°±2.03° vs. 3.45°±2.03°, p=0.001). However, there was no any significant change in ST over time in healthy controls (p=0.153, h2=0.16). Refer to Figure 5.a.
In addition, there was a significant group by time interaction effect for L3 after controlling for baseline L3 angle (p<0.001, h2=0.34). Results of the one-way ANOVA indicated that the difference in the amount of L3 change from baseline was significant among the three groups at all 6 time periods (p<0.05). Bonferroni post hoc comparisons revealed that at minute 60, all groups significantly differed from each other (FP vs. controls, p=0.005; AEP vs. controls, p<0.001; and FP vs. AEP, p<0.001). The degree of L3 displayed by the pain subgroups increased significantly over time (FP; p=0.017, h2=0.24; AEP; p<0.001, h2=0.55). Specifically, the FP subgroup showed a significant increase in L3 flexion from baseline to minute 60 (-3.12°±2.53° vs. -6.81°±2.05°, p<0.001). In contrast, the AEP subgroup showed a significant increase in L3 extension from baseline to minute 60 (1.20°±1.64° vs. 5.01°±2.94°, p=0.001). However, the control group did not show any significant change in L3 over time and remained relatively close to the neutral range (p=0.294, h2=0.12). Refer to Figure 5.b.
Furthermore, there was a significant group by time interaction effect for RLLA after controlling for baseline RLLA angle (p<0.001, h2=0.36). Results of the one-way ANOVA showed that the difference in the amount of RLLA change from baseline was significant among the three groups at minute 20 and 40 (p<0.5). Bonferroni post hoc comparisons revealed that at minute 20, the FP subgroup significantly differed from healthy controls (p=0.002), but no significant difference was found between both pain subgroups or between the AEP subgroup and healthy controls. However, at minute 40, only the AEP subgroup significantly differed from healthy controls (p=0.025). The degree of RLLA exhibited by the pain subgroups increased significantly over time (FP; p<0.001, h2=0.45; AEP; p=0.001, h2=0.33). Specifically, the FP subgroup showed a significant increase in lower lumbar kyphosis from minute 20 to minute 60 (4.49°±6.31° vs. -7.55°±10.00°, p=0.038). In contrast, the AEP subgroup showed a significant increase in lower lumbar lordosis from baseline to minute 40 (2.06°±3.04° vs. 5.98°±0.88°, p=0.006), which was followed by a significant decreased at minute 50 (5.98°±0.88°, vs. 1.30°±1.41°, p<0.001) and minute 60 (5.98°±0.88°, vs. 0.09°±1.56°, p<0.001). However, healthy controls did not show any significant change in RLLA over time (p=0.288, h2=0.12). Refer to Figure 5.c.