Table 7.
Outcomes of linear regression models examining whether sleep changes were related to health behaviors.
| Predictors | Dietary Risk B (p-Value) |
Alcohol Misuse Scores B (p-Value) |
Physical Activity (METs/wk) B (p-Value) |
Sitting Time (min/d) B (p-Value) |
|---|---|---|---|---|
| (Constant) | 6.756 (<0.001) | 2.005 (<0.001) | 2517.932 (<0.001) | 368.664 (<0.001) |
| Sleep quality change | ||||
| Worse | 0.486 (0.001) | 0.310 (0.051) | −272.907 (0.123) | 19.909 (0.082) |
| Better | −0.323 (0.081) | 10.162 (0.413) | 18.737 (0.934) | −36.064 (0.010) |
| Did not change (reference) | - | - | - | - |
| Sleep duration change | ||||
| Less | 0.409 (0.040) | 0.473 (0.019) | −128.772 (0.582) | 4.436 (0.770) |
| More | 0.408 (0.008) | 0.459 (0.004) | −568.192 (0.002) | 22.841 (0.050 *) |
| Did not change (reference) | - | - | - | - |
n = 1877 for dietary risk, n = 1761 for alcohol misuse scores, n = 1736 for physical activity, and n = 1872 for sitting time. One model was built for each dependent variable, and the models controlled for age, BMI, gender, and countries. * The exact p value is 0.049880. Dietary risk was defined by score on the Starting the Conversation, a simplified food frequency questionnaire. 0 = best to 16 = worst dietary quality. Alcohol misuse was defined by score on the Alcohol Use Disorders Identification Test Consumption questionnaire. 0 = no alcohol use to 12 = highest alcohol use. Scores ≥ 3 in females and ≥4 in males were defined as misuse. Physical activity level and sitting time were reported on the International Physical Activity Questionnaire. METs = metabolic equivalents.