Table 2.
Regression results showing the association between health behaviors and frequency of social media use by middle and high school studentsa,b,c.
| Social media use frequency | % marginal effect (95% CI) | |||||
| Vigorous physical activity (n=19,543) | Sleeping 7 hours/day (n=19,596) | |||||
| Never/seldom | Sometimes | Every day | Never/seldom | Sometimes | Every day | |
| Never | 26.4 (23.4-29.4) | 38.1 (34.8-41.3) | 35.5 (32.6-38.5) | 16.6 (14.2-18.9) | 36.9 (33.7-40.0) | 46.6 (43.4-49.7) |
| A few times a year | 24.4 (19.8-28.9) | 37.8 (32.4-43.2) | 37.8 (32.5-43.1) | 21.4 (16.9-25.9) | 37.9 (32.6-43.3) | 40.7 (35.4-46.0) |
| 1-2 times a month | 20.4 (16.9-24.0) | 42.0 (37.6-46.3) | 37.6 (33.6-41,6) | 19.5 (16.2-22.9) | 36.4 (32.2-40.6) | 44.0 (39.9-48.2) |
| Once a week | 18.6 (16.4-20.8) | 36.5 (33.9-39.1) | 44.9 (42.3-47.5) | 17.4 (15.4-19.3) | 40.1 (37.5-42.8) | 42.5 (40.0-45.0) |
| Every day | 15.8 (15.1-16.4) | 34.1 (33.2-35.0) | 50.1 (49.2-51.0) | 18.3 (17.6-19.0) | 40.4 (39.5-41.4) | 41.3 (40.4-42.1) |
aGeneralized ordered logistic regression models are used for the estimation. Marginal effect indicates predicted percentage of students with different intensities of health behaviors and social media use, holding all the other covariates at their observed values.
bAll results are statistically significant at P<.001.
cData are from the Monitoring the Future survey (2014 and 2015) with students from 8th, 10th, and 12th grades.