Table 4.
Changes in indicators of abdominal obesity as classified according to commuting mode transition (n = 4322)
| Crude model | P values | Multivariable model 1a | P values | Multivariable model 2b | P values | |
|---|---|---|---|---|---|---|
| Waist circumference (cm) | ||||||
| Maintained inactive | 0.7 (0.4, 1.0) | Reference | 0.9 (0.6, 1.1) | Reference | 0.9 (0.6, 1.1) | Reference |
| Switching to inactive | 1.2 (0.6, 1.9) | 0.20 | 0.9 (0.5, 1.2) | 0.15 | 0.8 (0.4, 1.2) | 0.10 |
| Switching to active or public transportation | −0.4 (−1.1, 0.4) | 0.007 | 0.6 (0.4, 0.8) | 0.052 | 0.5 (0.3, 0.7) | 0.17 |
| Maintained active or transportation | 0.7 (0.4, 1.1) | 0.98 | 0.7 (0.4, 1.0) | 0.85 | 0.7 (0.4, 1.0) | 0.27 |
| Visceral fat areas (cm2) | ||||||
| Maintained inactive | 5.3 (3.7, 6.9) | Reference | 5.4 (4.1, 6.6) | Reference | 5.6 (4.3, 6.9) | Reference |
| Switching to inactive | 6.4 (2.4, 10.3) | 0.58 | 5.2 (3.1, 7.3) | 0.63 | 4.4 (2.3, 6.5) | 0.20 |
| Switching to active or public transportation | −0.8 (−5.4, 3.9) | 0.007 | 2.2 (1.1, 3.4) | 0.11 | 1.5 (0.3, 2.7) | 0.25 |
| Maintained active or public transportation | 4.6 (2.8, 6.4) | 0.64 | 4.2 (2.5, 5.9) | 0.42 | 4.2 (2.5, 5.9) | 0.59 |
| Subcutaneous fat areas (cm2) | ||||||
| Maintained inactive | 5.1 (3.7, 6.5) | Reference | 5.0 (3.9, 6.2) | Reference | 5.4 (4.2, 6.5) | Reference |
| Switching to inactive | 5.7 (2.0, 9.5) | 0.73 | 5.5 (3.6, 7.3) | 0.55 | 4.2 (2.4, 6.0) | 0.19 |
| Switching to active or public transportation | 2.0 (−1.5, 5.5) | 0.11 | 3.2 (1.8, 4.7) | 0.16 | 3.0 (2.0, 4.0) | 0.51 |
| Maintained active or public transportation | 3.5 (1.9, 5.0) | 0.13 | 3.7 (2.8, 4.7) | 0.13 | 3.2 (1.7, 4.7) | 0.80 |
Data are shown as estimated mean with 95% confidence intervals of changes in obesity outcome. Maintained inactive group, n = 2129; switching from active or public transportation commuting to inactive group, n = 328; switching from inactive to active or public transportation commuting group, n = 322; maintained active or public transportation commuting group, n = 1532
aModel 1: Adjusted for age and sex at baseline. In addition, baseline variables of each outcome, treated as continuous data, were adjusted for (e.g., baseline waist circumference when outcome is changes in waist circumference)
bModel 2: Further adjusted for other variables