Table 3.
The associations [OR (95% CI)] of total screen time/online-course time with the risk of asthenopiaa.
| Quartiles of exposure (hours) | P for linear trendb | Random effects [σ (SE)]c | Residual ICC (95% CI) d | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Q1 (lowest) | Q2 | Q3 | Q4 (highest) | Level 3 | Level 2 | Level 3 | Level 2 | 3 | ||
| Total screen time (hours) | ≤ 84.0 | 86.8–175.0 | 176.4–350.0 | ≥357.0 | NA | NA | NA | NA | NA |
| No. of participants | 6,344 | 4,879 | 5,442 | 5,301 | NA | NA | NA | NA | NA |
| Events, No. (%) | 539 (8.50) | 363 (7.44) | 651 (11.96) | 1,094 (20.94) | NA | NA | NA | NA | NA |
| Model 1e | Reference | 1.05 (0.90, 1.23) |
1.24 (1.07, 1.43) |
1.67 (1.44, 1.94) |
<0.001 | 0.143 (0.079) | 0.154 (0.042) |
0.006 (0.001, 0.051) |
0.013 (0.005, 0.034) |
| Model 2f | Reference | 1.00 (0.86, 1.18) |
1.15 (0.99, 1.33) |
1.48 (1.27, 1.72) |
<0.001 | 0.412 (0.155) |
0.147 (0.043) |
0.049 (0.011, 0.184) |
0.055 (0.016, 0.176) |
| Model 3g | Reference | 1.02 (0.87, 1.20) |
1.16 (0.999, 1.35) |
1.44 (1.23, 1.68) |
<0.001 | 0.382 (0.152) |
0.162 (0.043) |
0.042 (0.009, 0.173) |
0.050 (0.014, 0.161) |
| Total online-course time (hours) | ≤ 84.0 | 85.8–140.0 | 142.8–280.0 | ≥ 283.5 | NA | NA | NA | NA | NA |
| No of participants | 6,732 | 4,468 | 5,372 | 5,394 | NA | NA | NA | NA | NA |
| Events, No. (%) | 411 (6.11) | 487 (10.90) | 737 (13.72) | 1,012 (18.76) | NA | NA | NA | NA | NA |
| Model 1e | Reference | 1.13 (0.97, 1.32) |
1.30 (1.11, 1.52) |
1.45 (1.23, 1.71) |
<0.001 | 0.118 (0.079) |
0.160 (0.043) |
0.004 (0.000, 0.054) |
0.012 (0.005, 0.029) |
| Model 2f | Reference | 1.13 (0.97, 1.31) |
1.27 (1.09, 1.49) |
1.42 (1.20, 1.68) |
<0.001 | 0.407 (0.154) |
0.147 (0.043) |
0.048 (0.011, 0.181) |
0.054 (0.015, 0.173) |
| Model 3g | Reference | 1.15 (0.99, 1.34) |
1.29 (1.10, 1.51) |
1.40 (1.18, 1.66) |
<0.001 | 0.376 (0.149) |
0.159 (0.044) |
0.041 (0.009, 0.168) |
0.048 (0.014, 0.157) |
CI, confidence interval; ICC, interclass correlation; NA, not applicable; OR, odds ratio; Q, quartile; SE, standard error.
All the models were constructed by using mixed-effects logistic regression with individuals (level 1) nested in school (level 2) and province (level 3).
P for linear trend was calculated by using the exposure of interest as a continuous variable.
From the models with exposure of interest in quartiles.
Model 1: adjusted age, gender, administrative district, physical activity, sleep time, myopia, astigmatism, and glasses-wearing status.
Model 2: adjusted variables in model 1, lying down or lying on the stomach while watching a screen, and distance from eyes to screen.
Model 3: adjusted variables in model 2, rest time between classes, eye exercise, and eye drops for foreign body sensation, dry or fatigued eyes.