Skip to main content
. 2020 Jun 5;11:100610. doi: 10.1016/j.ssmph.2020.100610

Table 2.

Summary of fixed effects logistic regression model predicting self-rated health (German Ageing Survey (DEAS) 1996–2014).

Predictor Model 1:
Sport activity
Model 2:
Perception of aging
Model 3:
Interaction
Model 4:
Time-fixed Effects
OR AME OR AME OR AME OR AME
Sport activity
0 No Ref Ref Ref Ref
1 Yes 1.479**
[0.218]
0.001**
[0.000]
1.352*
[0.207]
0.001
[0.001]
1.362*
[0.210]
0.001*
[0.001]
1.345
[0.209]
0.000
[0.002]
Perception of aging 1.960***
[0.172]
0.002***
[0.000]
1.700***
[0.184]
0.002***
[0.000]
1.690***
[0.184]
0.000
[0.003]
Retirement
0 No Ref Ref Ref Ref
1 Yes 1.635*
[0.325]
0.001*
[0.000]
1.610*
[0.337]
0.002*
[0.001]
1.623*
[0.340]
0.002*
[0.001]
1.652*
[0.345]
0.000
[0.003]
Income 1.269
[0.266]
0.000
[0.000]
1.248
[0.270]
0.001
[0.001]
1.236
[0.263]
0.001
[0.001]
1.258
[0.273]
0.000
[0.001]
Sport activity*
Perception of aging
1.401*
[0.217]
0.001*
[0.001]
1.417*
[0.221]
0.000
[0.002]
N cases 1702 1702 1702 1702
N groups 506 506 506 506
McFadden- Pseudo-R2 0.117 0.176 0.180 0.183

Note: OR = Odds Ratio. AME = Average Marginal Effect. Clustered standard errors in parantheses. Included variable ‘wave’ for time-fixed effects not presented. All models are adjusted for the variables ‘age’, ‘partner’, ‘physical diseases’, ‘gardening’ and ‘walking’. Time invariant variables ‘gender’ and ‘education’ are captured by individual fixed effects. *p < .05. **p < .01. ***p < .001.