TABLE 1.
Risk factors at baseline (Phase 1) of low leisure-time physical activity (LTPA) at follow-up (Phase 3) for men (n=4291) and women (n=1794).
| Variables at Phase1 | Men | Women | Sex interactionc |
||
|---|---|---|---|---|---|
| % with LTPA at Phase 3 (Total participants) |
Prevalence ratio (95% CI) |
% with LTPA at Phase 3 (Total participants) |
Prevalence ratio (95% CI) |
||
| N | 44.7 (4291) | 66.2 (1794) | |||
| Age (mean (SDa)) | 43.6 (5.8) | 1.01 (100-1.01) | 44.8 (5.8) | 1.01 (1.00-1.08) | 0.74 |
| P for linear trend b | 0.08 | 0.07 | |||
| Ethnicity | 0.14 | ||||
| Non-white | 57.4 (296) | 1.00 | 72.2 (212) | 1.00 | |
| White | 43.8 (3995) | 0.76 (0.65-0.89) | 65.4 (1582) | 0.91 (0.76-1.07) | |
| P for linear trend b | 0.001 | 0.26 | |||
| Employment grade | 0.015 | ||||
| High | 42.9 (1696) | 1.00 | 70.1 (244) | 1.00 | |
| Middle | 44.1 (2287) | 1.03 (0.93-1.13) | 63.7 (793) | 0.91 (0.76-1.08) | |
| Low | 58.8 (308) | 1.37 (1.16-1.61) | 67.6 (757) | 0.97 (0.81-1.15) | |
| P for linear trend b | 0.006 | 0.96 | |||
| Self-rated health | 0.20 | ||||
| Good | 42.5 (3430) | 1.00 | 64.3 (1234) | 1.00 | |
| Average | 52.8 (740) | 1.24 (1.11-1.39) | 68.9 (450) | 1.07 (0.94-1.22) | |
| Poor | 57.0 (121) | 1.34 (1.05-1.71) | 77.3 (110) | 1.20 (0.96-1.50) | |
| P for linear trend b | <0.001 | 0.09 | |||
| Body Mass Index (kg/m2) | 0.88 | ||||
| Underweight (<18.5) | 48.8 (215) | 1.13 (0.93-1.38) | 66.5 (161) | 1.04 (0.85-1.28) | |
| Healthy weight (≥18.5 & <25) | 43.2 (2468) | 1.00 | 63.6 (1000) | 1.00 | |
| Overweight (≥ 25 but <30) | 46.5 (1418) | 1.08 (0.98-1.19) | 69.1 (473) | 1.09 (0.95-1.24) | |
| Obese (≥ 30) | 46.3 (190) | 1.07 (0.86-1.33) | 738 (160) | 1.16 (0.95-1.41) | |
| P for linear trend b | 0.12 | 0.09 | |||
| LTPA | <0.0001 | ||||
| At least recommended level | 28.7 (2601) | 1.00 | 47.7 (727) | 1.00 | |
| Below recommended level | 69.3 (1690) | 2.41 (2.20-2.64) | 78.8 (1067) | 1.65 (1.46-1.87) | |
| P for linear trend b | <0.001 | <0.001 | |||
a
SD=standard deviation
b
P-value for linear trend across the levels of the explanatory variable
c
Probability that the model including the interaction of sex with the explanatory variable fits the data significantly better than the model without the interaction.