Table 2.
Associations between diabetes and physical activity (levels and trajectories)
| Outcome: physical activity | ||
|---|---|---|
| b (95% CI) | p | |
| Fixed effects | ||
| Intercept | 3.32 (3.30; 3.34) | <.001 |
| Age (10-year follow-up) | −.30 (−.32; −.29) | <.001 |
| Age2 (10-year follow-up) | −.11 (−.12; −.10) | <.001 |
| Diabetes (ref. no diabetes) | ||
| Having diabetes | −.13 (−.15; −.11) | <.001 |
| Age × diabetes (ref. no diabetes) | ||
| Age × having diabetes | −.04 (−.06; −.02) | <.001 |
| Age2 × diabetes (ref. no diabetes) | ||
| Age2 × having diabetes | −.01 (−.02; .01) | .222 |
| Random effects | ||
| Participants | ||
| Intercept | .32324 | |
| Age (10-year follow-up) | .22670 | |
| Age2 (10-year follow-up) | .01993 | |
| Covariance | ||
| Intercept − Age (10-year follow-up) | .39 | |
| Intercept − Age2 (10-year follow-up) | −.43 | |
| Age (10-year follow-up) − Age2 (10-year follow-up) | .66 | |
| Residuals | .58263 | |
Note. All models were adjusted for gender (male, female), body mass index (BMI), birth cohort, country of residence, education, household’s ability to make ends meet, and participants’ attrition. “Age (10-year follow-up)” and “Age (10-year follow-up) squared” estimated the linear and quadratic changes in the engagement in physical activity over a 10-year period.