Surgical results of patients with LSS
Figure 2 shows the breakdown of preoperative and postoperative GLFS-25 scores and CDL stages 3 months postoperatively. Lumbar spinal surgery improved the mean GFLS-25 score from 48.8 preoperatively to 30 postoperatively (Fig. 2a). At 3 months postoperatively, the rates of CDL stages 0, 1, 2, and 3 were 7%,18.3%, 19.7%, and 55%, respectively. In total, the proportion of patients with improved CDL stage was 45.1% (32/71) (Fig. 2b).
Comparison of preoperative status between the two groups
Patients in the improvement group were significantly younger than those in the non-improvement group. In addition, the LSA score was significantly higher in the improvement group than in the non-improvement group (Table 2). The TUG times in the improvement group were significantly shorter than that in the non-improvement group. The two groups had no significant differences in limb skeletal muscle, trunk skeletal muscle, PNI, or VAS scores.
Table 2
Comparison between the two groups
| Improvement group (N = 32) | Non-improvement group (N = 39) | P |
| Median | Range (min-max) | Median | Range (min-max) | |
Age (years) | 75 | 65–81 | 78 | 67–91 | 0.008 |
Sex (Male/Female) | 14/18 | | 13/26 | | 0.4 |
BMI (kg/m2) | 22.8 | 15.4–32.1 | 24.2 | 17.7–31.2 | 0.3 |
Limb skeletal muscle mass (kg) | 14.9 | 9.8–23.3 | 13.9 | 8.9–21.9 | 0.3 |
Trunk skeletal muscle mass (kg) | 16.8 | 12.4–23.3 | 16.5 | 11.7–23.7 | 0.3 |
Hand grip strength (kg) | 23.3 | 12.8–44 | 20.6 | 8.1–43 | 0.1 |
TUG (s) | 9.6 | 5.4–21.5 | 14.5 | 6.4–38.4 | 0.0002 |
LSA | 54 | 4.5–120 | 37.5 | 0-110 | 0.02 |
PNI | 52.4 | 36.9–64.9 | 51.2 | 36.2–60.5 | 0.2 |
Low back pain (VAS) | 51 | 8-100 | 67 | 0-100 | 0.2 |
Lower limb pain (VAS) | 65 | 18–100 | 69.5 | 0-100 | 0.8 |
Lower limb numbness (VAS) | 51 | 0-100 | 65.5 | 0-100 | 0.4 |
GLFS-25 | 40 | 24–65 | 58 | 25–87 | 0.002 |
Abbreviations: BMI; body mass index, TUG; Timed Up and Go test, LSA; life-space assessment, PNI; prognostic nutritional index, VAS; visual analog scale, GLFS-25; the 25-Question Geriatric Locomotive Function Scale |
* indicates statistical significance |
[insert Table 2]
With the exception of pain and lumbar function in the non-improvement group, lumbar spinal surgery improved the JOABPEQ scores in both the groups. Similarly, lumbar spinal surgery improved the VAS scores for low back pain, lower limb pain, and lower limb numbness in both groups (Additional file 1).
Cox regression analysis for factors related to improvement in LS
Multivariate regression analysis was performed with the following variables related to improvements in LS: age, sex, BMI, LSA, PNI, handgrip strength, and TUG time.
The TUG time and age were significantly associated with improvement in LS (p = 0.0017 and p = 0.031) (Table 3). LSA, BMI, PNI, and handgrip strength were not significantly associated with improvements in LS (p = 0.38, p = 0.45, p = 0.61, p = 0.65).
Table 3: Multivariate analysis for the factors associated with the improvement of locomotive syndrome
Factors
|
Multivariate analysis for the improvement of locomotive syndrome
(95% Confidence interval, P-value)
|
TUG
|
1.22
(1.07-1.47, 0.0017)
|
Age
|
1.15
(1.01-1.32, 0.031)
|
LSA
|
1.01
(0.99-1.04, 0.38)
|
BMI
|
1.07
(0.90-1.28, 0.45)
|
PNI
|
1.04
(0.90-1.21, 0.61)
|
Hand Grip strength
|
1.02
(0.94-1.12, 0.65)
|
Abbreviations: TUG; Timed Up and Go test, LSA; life-space assessment, BMI; body mass index, PNI; prognostic nutritional index,
* indicates statistical significance
Decision tree algorithm for factors related to improvements in LS
A decision tree algorithm was used for variables related to improvements in LS, including age, sex, BMI, LSA, PNI, handgrip strength, and TUG time. In the decision tree analysis, the TUG time was selected as the first divergence variable and LSA was selected as the second divergence variable. Among patients with TUG time <12.4 s, CDL stage 3 was improved in 66.7%, and LSA was the second divergence variable in patients with TUG ≥12.4. Patients with a TUG time ≧12.4 s and LSA ≥40 showed improvement from CDL stage 3 by 46.2%. Among patients with reduced physical function, CDL stage 3 was improved in 46.2%, with an LSA score of 40 or higher (Figure 3).
Relationship between the breakdown of CDL stage and preoperative TUG 3 months postoperatively
Figure 4 shows the relationship between the breakdown of the CDL stage and preoperative TUG time 3 months after lumbar spinal surgery. In patients with CDL stage 3 LS, the preoperative TUG time 3 months after the procedure was significantly longer than that for those in CDL stages 0 and 1 (Figure 4). The results showed that patients who had low preoperative TUG times improved to either CDL stage 0 or 1 from stage 3 following the surgical procedure.
Relationship between Δ GFLS-25 and Δ pain scales
Δ GLFS-25 was significantly positively correlated with Δ lower limb pain, Δ lower limb numbness, and Δ low back pain (Table 4).
Table 4: Relationship between the Δ GLFS-25S and other Δ VAS scores.
|
Δ Lower limb pain
|
Δ Lower limb numbness
|
Δ Low back pain
|
Δ GLFS-25
|
Δ Lower limb pain
|
1.0000
|
|
|
|
Δ Lower limb numbness
|
0.5684*
|
1.0000
|
|
|
Δ Low back pain
|
0.4479*
|
0.3315*
|
1.0000
|
|
Δ GLFS-25
|
0.3615*
|
0.3723*
|
0.4103*
|
1.0000
|
Abbreviations: GLFS-25; the 25-Question Geriatric Locomotive Function Scale; VAS, visual analog scale
The numbers in the table indicate correlation coeffect.
Comparison of Δ GLFS-25 and Δ pain scales between groups
Lower limb pain and GFLS-25 scores were significantly better in the improvement group than in the non-improvement group (p=0.0107, p=0.002). Low back pain and lower limb numbness in the improvement group were not significantly higher than those in the non-improvement group (p=0.0953, p=0,1041) (Figure 5).