Table 1: The coefficient of the classical OLS regression models (n = 35).
|
|
Coefficients for positional OSAS |
Coefficients for non-positional OSAS |
||||
|
|
Training set |
Training set |
||||
|
Variables |
β |
Standard Error |
p |
β |
Standard Error |
p |
|
Constant |
123.96 |
99.08 |
0.226 |
-216.80 |
226.32 |
0.392 |
|
BMI (X1) |
-4.10 |
3.29 |
0.229 |
9.01 |
9.24 |
0.385 |
|
REM (X2) |
0.14 |
0.21 |
0.520 |
-1.33 |
1.32 |
0.370 |
|
ODI (X3) |
0.85 |
0.11 |
0.001* |
0.82 |
0.71 |
0.027* |
|
AROUSAL (X4) |
-0.13 |
0.08 |
0.128 |
0.01 |
0.22 |
0.939 |
|
AH (X5) |
0.05 |
0.02 |
0.082 |
-0.04 |
0.33 |
0.912 |
|
Model (Enter) |
|
R2 = 0.781 |
< 0.001* |
|
R2 = 0.889 |
0.001* |
|
|
|
MSE = 36.27 |
|
|
MSE = 34.53 |
|
|
|
Test set |
Test set |
||||
|
Constant |
107.89 |
308.65 |
0.744 |
6.11 |
62.60 |
0.923 |
|
BMI (X1) |
-3.17 |
10.41 |
0.776 |
0.13 |
2.32 |
0.956 |
|
REM (X2) |
-0.31 |
0.41 |
0.494 |
-0.41 |
0.26 |
0.133 |
|
ODI (X3) |
0.54 |
0.28 |
0.122 |
0.84 |
0.11 |
< 0.001* |
|
AROUSAL (X4) |
-0.24 |
0.28 |
0.447 |
0.01 |
0.09 |
0.985 |
|
AH (X5) |
0.06 |
0.07 |
0.412 |
0.02 |
0.03 |
0.629 |
|
Model (Enter) |
|
R2 = 0.679 |
0.123 |
|
R2 = 0.719 |
< 0.001* |
|
|
|
MSE = 28.69 |
|
|
MSE = 6.23 |
|
*Significant at 0.05 level; AHI: Apnea-Hypopnea Index (event/h); BMI: Body Mass Index (kg/m2); REM: Rapid Eye Movement Sleep (%); ODI: Oxygen Desaturation Index (event/h); AROUSAL: Stimulating in sleep (event/h); AH: Apnea-Hypopnea duration (h); R2: Coefficient of determination; MSE: Mean Squared Error; RMS: Root Mean Squared Error.