3.1. Demographic characteristics of subjects with postural abnormalities
As shown in Table 1, 22 of the 1,003 students were eventually diagnosed with scoliosis; the Cobb angle was 14.96 ± 4.55° and the prevalence of scoliosis was 2.2%.
Table 1 Summary of screening results
Mark
|
All
|
Male
|
Famal
|
|
N(%);
|
N(%);
|
N(%);
|
Positive
|
mean (95% CI)
|
mean (95% CI)
|
mean (95% CI)
|
N (%)
|
N(n)
|
1 003
|
607 (60.5)
|
396(39.5)
|
22(2.2)
|
Age(y)
|
20.38(20.31-20.45)
|
20.50(20.41-20.58)
|
20.20(20.10-20.30)
|
|
Weight(kg)
|
65.88(64.85-66.97)
|
70.50(69.40-71.56)
|
53.71(52.49-54.91)
|
|
Height(m)
|
1.77(1.73-1.84)
|
1.82(1.77-1.92)
|
1.64(1.63-1.65)
|
|
Disciplinary majors (n)
|
284(28.3)
|
129 (45.4)
|
155 (54.6)
|
12(54.6)
|
Technical disciplines (n)
|
719(71.7)
|
478 (66.5)
|
241 (33.5)
|
10(45.5)
|
BMI (kg/m2)
|
21.39(21.23-21.56)
|
22.33(22.12-22.53)
|
19.98(19.76-20.19)**
|
|
ATR≥5°(n)
|
236(23.6)
|
136(57.6)
|
100(42.4)
|
14(63.6)
|
ATI≥7˚(n)
|
5.09(4.98-5.21)
|
4.76(4.62-4.90)
|
5.61(5.42-5.80)**
|
19(86.4)
|
Thoracic kyphosis (˚)
|
32.17(31.65-32.72)
|
33.99(33.32-34.66)
|
29.39 (28.58-30.21)**
|
|
Foot length (cm)
|
24.33(24.21-24.45)
|
24.91(24.77-25.06)
|
23.45(23.35-23.64)
|
|
AHI
|
0.31(0.31-0.31)
|
0.31(0.30-0.31)
|
0.31(0.31-0.32)*
|
|
Percentage difference in AHI (%)
|
3.91(3.53-4.37)
|
3.93(3.54-4.38)
|
3.89(3.51-4.16)
|
|
Handgrip strength (kg)
|
36.77(35.92-36.63)
|
44.00(42.98-45.02)
|
25.69(25.15-26.23)**
|
|
Percentage difference in grip strength (%)
|
12.16(11.19-13.14)
|
14.21(12.63-15.78)
|
10.11(9.01-11.22)**
|
|
Higher left shoulder (n)
|
189(18.8)
|
111(58.7)
|
78(42.3)
|
17(77.3)
|
Higher right shoulder(n)
|
179(17.9)
|
93(60.0)
|
86(48.0)
|
Anterior pelvis tilt (n)
|
274(24.6)
|
187(68.3)
|
87(31.8)
|
8(36.4)
|
Posterior pelvic tilt (n)
|
19(1.9)
|
11(57.9)
|
8(42.1)
|
Higher left pelvis (n)
|
113(11.3)
|
76(67.3)
|
37(32.7)
|
10(45.5)
|
Higher right pelvis (n)
|
34(3.4)
|
19(55.9)
|
15(44.1)
|
“O” leg(n)
|
232(23.1)
|
190(81.9)
|
42(18.1)
|
6(27.3)
|
Foot inversion (n)
|
63(6.3)
|
39(61.9)
|
24(38.1)
|
12(54.6)
|
Foot eversion (n)
|
188(18.7)
|
109(58.0)
|
79(42.0)
|
FMS SM
|
2.50(2.44-2.56)
|
2.23(2.14-2.34)
|
2.76(2.69-2.82)**
|
|
FMS DS
|
2.76(2.72-2.80)
|
2.78(2.73-2.84)
|
2.74(2.68-2.81)
|
|
FMS RS
|
2.76(2.72-2.80)
|
2.73(2.67-2.78)
|
2.79(2.73-2.84)
|
|
FMSTSPp
|
2.74(2.69-2.79)
|
2.96(2.94-2.96)
|
2.51(2.43-2.60)**
|
|
X Positive (n)
|
22(2.2)
|
9(40.9)
|
13(59.1)
|
|
P.S.: Comparison with male values *P<0.05; **: P<0.01
23.6% of the students had abnormal rotation angles ≥5°, 36.7% had high and low shoulders, 29.2% had anterior and posterior pelvic tilt, 14.7% had pelvic tilt, and 23.4% had X/O legs (3 students had "X" legs); 25.0% of the students had foot inversion and eversion, with more eversion than inversion. As can be seen in Table 1, the proportion of boys with abnormalities was greater than that of girls in all items, but the number of positives was 13 more girls than 9 boys. ATR, ATI, and high and low shoulders had the highest proportion of confirmed diagnoses (63.6%, 86.4%, and 77.3%).
As shown in Table 1, there were significant differences between men and women in BMI (P < 0.001), ATI (P < 0.001), grip strength (P < 0.001), thoracic kyphosis (P < 0.001), grip strength difference (P < 0.001), AHI (left: P = 0.02; right: P = 0.03), FMS SM (P < 0.001), and FMSTSPp (P < 0.001).
3.2. Factors affecting scoliosis
3.2.1. Analysis of factors influencing scoliosis
As shown in Figure 1, pelvic tilt, high and low shoulders, O-legs, foot inversion and eversion, and AHI interacted with each other in a relatively close relationship (orange-red area). FMS shoulder mobility scores were weakly and negatively correlated with BMI (r = -0.34), pelvis (r = -0.26), and grip strength (r = -0.33). FMS trunk stability push-up scores were weakly negatively correlated with discipline category specialization (r = -0.34) and weakly positively correlated with training period (r = 0.22), BMI (r = 0.24), and grip strength (r = 0.39).
3.2.2. Risk factors for scoliosis
Based on the above correlation results, the items with significant differences were selected, and the Chi-square test and independent sample t-test were conducted from the measured and count data, respectively, to find out the identified influencing factors. As shown in Table 2, the incidence of scoliosis in disciplines (non-athletic students) was 12 cases with a prevalence rate of 4.2%, which was significantly higher than that of students in surgical disciplines (10 cases, 1.4%), and the difference was statistically significant (χ²=7.65, P=0.01); the incidence of scoliosis in students with high and low shoulders was 18 cases, with a prevalence rate of 4.9%, which was significantly higher than that in students without high and low shoulders (4 cases, 0.6%), a statistically significant difference (χ² = 8.16, P < 0.01); students with pelvic tilt had 15 cases of scoliosis, with a prevalence rate of 5.1%, which was significantly higher than those without pelvic tilt (7 cases, 1.0%), a statistically significant difference (χ² = 9.58, P < 0.01). BMI, thoracic kyphosis angle, grip strength, and AHI difference were significantly different in scoliosis patients (all P < 0.05, Table 2).
Table 2 Independent risk factors with Chi-square test and Independent sample t-test results for scoliosis
|
χ test
|
|
|
Independent samples t-test (mean±SD)
|
|
Factors
|
χ2
|
P
|
Factors
|
Without scoliosis
|
Scoliosis
|
P
|
Specialty
|
7.65
|
0.01
|
BMI
|
21.43±2.71
|
19.66±1.66
|
0.00
|
High and low shoulders
|
8.16
|
0.00
|
Thoracic kyphosis
angle (°)
|
5.03±1.79
|
8.57±2.44
|
0.00
|
pelvic tilt
|
9.58
|
0.00
|
Grip strength(kg)
|
36.95±13.82
|
28.93±11.01
|
0.01
|
ATR≥5°
|
8.62
|
0.00
|
AHI difference(%)
|
3.93±4.23
|
2.26±1.14
|
0.01
|
Gender
|
3.64
|
0.06
|
|
|
|
|
O-legs
|
0.31
|
0.58
|
|
|
|
|
Foot inversion and eversion
|
0.22
|
0.64
|
|
|
|
|
Several other risk factors for scoliosis were further determined by binary logistic regression. As shown in Table 3, the chi-square value of the model was 44.28, p = 0.00, and the logistic model obtained was statistically significant. Excluding tilt angle, a well-recognized indicator of scoliosis screening, binary logistic regression identified specialty, ATR, and BMI as independent variables in the model and possible independent risk factors. Disciplinary majors students had 2.99 times the risk of getting scoliosis than technical disciplines students. The risk of getting scoliosis with an ATR ≥5° was 3.53 times higher than with an ATR ˂5°. For every 1 increase in BMI, the risk of scoliosis increased 1.19 times. Sensitivity was 0, specificity was 100%, and overall prediction was 97.8%.
Table 3 Logistic regression analysis results of scoliosis
Mark
|
|
OR
|
95% CI
|
P
|
Upper-bound
|
Lower-bound
|
Specialty
|
Technical disciplines
|
1
|
--
|
--
|
--
|
|
Disciplinary majors
|
2.99
|
7.11
|
1.26
|
0.01
|
ATR
|
<5°
|
1
|
--
|
--
|
--
|
|
≥5°
|
3.53
|
8.40
|
1.49
|
0.00
|
BMI
|
--
|
1.19
|
1.34
|
1.06
|
0.01
|
3.3. Predictive thresholds for body mass index
ROC curve analysis is commonly used in clinical medicine and epidemiologic studies to evaluate diagnostic accuracy as well as to determine the cut-off point. Continuing with BMI as a variable, the AUC(Area Under Curve) value under the ROC curve was calculated to obtain the optimal cutoff value for the BMI risk factor prediction of scoliosis.
ROC curve analysis is commonly used in clinical medicine and epidemiologic studies to evaluate diagnostic accuracy as well as to determine the cut-off point. Continuing with the body mass index as a variable, the AUC value under the ROC curve was calculated to obtain the optimal cut-off value for the BMI risk factor prediction of scoliosis.
As shown in Figure 2, the AUC area under ROC was 0.71 (95% CI: 0.63-0.80), as the value >0.7 indicates that BMI is a good model for predicting scoliosis. The BMI cut-off value was 20.69, which is greater than this value to predict a high likelihood of scoliosis-free college students.