Table 3.
Odds ratios and confidence intervals for sedentary TV viewing and computer use related to the children’s body mass index groupsa stratified by weekly leisure-time exercise duration (n = 10,228).
Risk for Underweightb | Risk for Overweightb | |||||
---|---|---|---|---|---|---|
OR | 95% CI | p valuec | OR | 95% CI | p valuec | |
Model for children in the Low exercise group (≤5 hours/week), n = 3,703 | ||||||
TV viewingd | ||||||
Light | 1 | 1 | ||||
Medium | 0.79 | 0.61–1.02 | 0.069 | 1.28 | 1.01–1.62 | 0.039 |
Heavy | 0.70 | 0.54–0.92 | 0.010 | 1.46 | 1.15–1.85 | 0.002 |
Computer usee | ||||||
Light | 1 | 1 | ||||
Medium | 1.15 | 0.89–1.48 | 0.286 | 1.19 | 0.95–1.50 | 0.137 |
Heavy | 1.10 | 0.83–1.47 | 0.503 | 1.37 | 1.08–1.76 | 0.010 |
Model for children in the Medium exercise group (6-8 hours/week), n = 3,154 | ||||||
TV viewingd | ||||||
Light | 1 | 1 | ||||
Medium | 0.97 | 0.73–1.29 | 0.836 | 1.29 | 0.97–1.73 | 0.082 |
Heavy | 0.83 | 0.60–1.13 | 0.233 | 1.74 | 1.30–2.33 | <0.001 |
Computer usee | ||||||
Light | 1 | 1 | ||||
Medium | 0.88 | 0.66–1.17 | 0.388 | 1.00 | 0.76–1.32 | 0.994 |
Heavy | 1.00 | 0.72–1.38 | 0.988 | 1.33 | 0.99–1.78 | 0.056 |
Model for children in the High exercise group (9-10 hours/week), n = 3,371 | ||||||
TV viewingd | ||||||
Light | 1 | 1 | ||||
Medium | 0.69 | 0.51–0.92 | 0.012 | 1.37 | 1.02–1.83 | 0.037 |
Heavy | 0.80 | 0.58–1.10 | 0.168 | 1.55 | 1.14–2.12 | 0.005 |
Computer usee | ||||||
Light | 1 | 1 | ||||
Medium | 1.02 | 0.76–1.35 | 0.919 | 1.00 | 0.75– 1.34 | 0.998 |
Heavy | 1.02 | 0.72–1.44 | 0.927 | 1.62 | 1.19–2.21 | 0.002 |
Likelihood ratio test to evaluate models with and without interaction between exercise and TV viewing: p = 0.012.
Likelihood ratio test to evaluate models with and without interaction between exercise and computer use: p = 0.010.
aCategorization based on the Cole and Lobstein Classification45.
bCompared to Normal Weight.
cMain effects from multinomial logistic regression. Variables in the model: age, sex, language, sleep duration on school nights, TV viewing and computer use.
dCategorization based on the distribution of responses for two questions on recreational TV viewing during school days and weekends/days off.
eCategorization based on the distribution of responses for two questions on recreational computer use during school days and weekends/days off.
OR, odds ratio; CI, confidence interval.