Table 3. Regression coefficients (mL/min per 1.73 m2) and 95% confidence intervals for the associations between 5-year changes in physical activity and cystatin C-based estimated glomerular filtration rate at the subsequent round, adjusted for attained time-varying covariates.
| 5-year changes in physical activity | |||||
|---|---|---|---|---|---|
| Becoming inactive | Staying inactive | Staying moderately (in)active | Staying active | Becoming active | |
| Model 1 | -0.40 (-1.14,0.34) | -3.46 (-7.01,0.09) | -0.28 (-1.09,0.53) | 1.41 (0.57,2.26) | Reference |
| Model 2 | -0.52 (-1.25,0.20) | -2.30 (-5.64,1.05) | -0.01 (-0.78,0.76) | 0.53 (-0.26,1.32) | Reference |
| Model 3 | -0.28 (-1.02,0.45) | -1.22 (-4.63,2.19) | -0.09 (-0.86,0.68) | 0.63 (-0.15,1.42) | Reference |
| Model 4 | -0.27 (-1.01,0.47) | -1.10 (-4.50,2.30) | -0.05 (-0.83,0.72) | 0.68 (-0.12,1.47) | Reference |
Model 1: crude; Model 2: adjusted for age and sex; Model 3: model 2 and highest attained level of education and time-dependent smoking, alcohol consumption, body mass index and animal protein; Model 4: model 3 and time-dependent diabetes, hypertension, hypercholesterolemia and cardiovascular disease. Percentage of people in the various 5-year change groups: 17.05% becoming inactive, 1.54% staying inactive, 31.07% staying moderately (in)active, 35.90% staying active, 14.45% becoming active.