Table 2.
Average marginal effects (AMEs) times 100 of reporting more health problems than the reference group and model fit (R2 change)
| Mobility limitations (n = 1763)1 | R2 change | ADL limitations (n = 2036) | R2 change | Psychological distress (n = 1596) | R2 Change | ||||
|---|---|---|---|---|---|---|---|---|---|
| AME (%) | p value | AME (%) | p value | AME (%) | p value | ||||
| Education | |||||||||
| Model 1 | |||||||||
| -High | (Ref) | [0.001] 2 | 12%3 | (Ref) | [0.236] | 3% | (Ref) | [0.060] | 15% |
| -Medium | 7.02 | 0.039 | 1.43 | 0.229 | 9.46 | 0.022 | |||
| -Low | 12.16 | 0.001 | 1.99 | 0.103 | 9.42 | 0.022 | |||
| Model 2 | |||||||||
| -High | (Ref) | [0.398] | 1%4 | (Ref) | [0.921] | 0% | (Ref) | [0.468] | 3% |
| -Medium | 2.53 | 0.251 | 0.42 | 0.933 | 4.93 | 0.305 | |||
| -Low | 5.08 | 0.531 | 0.12 | 0.767 | 2.87 | 0.568 | |||
| Social class | |||||||||
| Model 1 | |||||||||
| -High | (Ref) | [0.001] | 12% | (Ref) | [0.021] | 5% | (Ref) | [0.104] | 12% |
| -Medium | 5.33 | 0.052 | 1.04 | 0.221 | 4.52 | 0.145 | |||
| -Low | 9.10 | 0.000 | 2.01 | 0.007 | 5.79 | 0.035 | |||
| Model 2 | |||||||||
| -High | (Ref) | [0.680] | 0% | (Ref) | [0.937] | 1% | (Ref) | [0.949] | 1% |
| -Medium | −0.12 | 0.971 | 0.54 | 0.141 | −0.83 | 0.829 | |||
| -Low | 1.96 | 0.559 | 1.39 | 0.583 | 0.01 | 0.997 | |||
| Occupational complexity | |||||||||
| Model 1 | |||||||||
| -High | (Ref) | [0.015] | 9% | (Ref) | [0.043] | 4% | (Ref) | [0.012] | 18% |
| -Medium | 8.49 | 0.004 | 0.26 | 0.153 | 4.95 | 0.181 | |||
| -Low | 10.15 | 0.011 | 1.73 | 0.806 | 10.34 | 0.011 | |||
| Model 2 | |||||||||
| -High | (Ref) | [0.510] | 2% | (Ref) | [0.128] | 1% | (Ref) | [0.094] | 8% |
| -Medium | 4.23 | 0.281 | −1.03 | 0.400 | 2.42 | 0.567 | |||
| -Low | 4.79 | 0.250 | 0.15 | 0.914 | 7.23 | 0.128 | |||
| Income | |||||||||
| Model 1 | |||||||||
| -High | (Ref) | [0.000] | 13% | (Ref) | [0.007] | 7% | (Ref) | [0.006] | 27% |
| -Medium | 10.37 | 0.000 | 2.44 | 0.002 | 10.58 | 0.004 | |||
| -Low | 13.05 | 0.000 | 2.30 | 0.007 | 9.93 | 0.002 | |||
| Model 2 | |||||||||
| -High | (Ref) | [0.019] | 7% | (Ref) | [0.102] | 3% | (Ref) | [0.072] | 18% |
| -Medium | 7.37 | 0.005 | 1.88 | 0.025 | 8.70 | 0.037 | |||
| -Low | 9.31 | 0.014 | 1.81 | 0.058 | 8.42 | 0.023 | |||
| SEP-index | |||||||||
| Model 1 | |||||||||
| -High | (Ref) | [0.000] | 16% | (Ref) | [0.008] | 6% | (Ref) | [0.011] | 20% |
| -Medium | 8.53 | 0.000 | 6.18 | 0.021 | 28.64 | 0.003 | |||
| -Low | 10.38 | 0.000 | 9.56 | 0.002 | 20.09 | 0.044 | |||
Results in bold: p < 0.05
Model 1: adjusted for age, sex, and linkage. Model 2: model 1 + all independent variables were analyzed simultaneously
1Outcome variables have different scales: mobility limitations (0–2), ADL (0–10) and psychological distress (0–4)
2Numbers in square brackets [] are p-values for the contribution of the whole variable (likelihood ratio test)
3McKelvey & Zavoina’s pseudo-R2 change compared to a model without that specific indicator of SEP
4Pseudo-R2 change to model 2 attributed to that specific indicator of SEP