Table 2.
Logistic regression of loneliness on demographic, health, and social exclusion variables in total sample, fully adjusted (final) model (N = 7755)
| Variables | OR | Lower 95% CI | Upper 95% CI | p |
|---|---|---|---|---|
| Intercept | 0.007 | 0.003 | 0.014 | 0.000 |
| Demographic and health variables | ||||
| Age (years) | 1.017 | 1.009 | 1.025 | 0.000 |
| Education (years) | 1.002 | 0.985 | 1.018 | 0.823 |
| Gender (ref = male) | 1.021 | 0.891 | 1.169 | 0.765 |
| Health-related limitations (ref = no) | ||||
| Yes, to some extent | 1.349 | 1.163 | 1.564 | 0.000 |
| Yes, a lot | 1.478 | 1.177 | 1.853 | 0.001 |
| Self-rated health (higher values, worse health) | 1.227 | 1.128 | 1.336 | 0.000 |
| Social exclusion variables | ||||
| No political participation (ref = political participation) | 0.938 | 0.825 | 1.066 | 0.324 |
| Did not vote (ref = voted) | 1.268 | 1.033 | 1.551 | 0.022 |
| Infrequent social contacts (higher values, less contacts) | 1.185 | 1.132 | 1.240 | 0.000 |
| No emotional support (ref = emotional contact) | 1.514 | 1.244 | 1.839 | 0.000 |
| Household size (ref = two-person household) | ||||
| Three plus-person household | 1.494 | 1.077 | 2.038 | 0.013 |
| Single household | 4.581 | 3.920 | 5.360 | 0.000 |
| Lower household income (higher values, lower income) | 0.974 | 0.942 | 1.007 | 0.121 |
| Income concern (ref = no concerns) | 1.648 | 1.338 | 2.028 | 0.000 |
| Lower neighbourhood safety (higher values, lower safety) | 1.145 | 1.054 | 1.244 | 0.001 |
| Model fit | Model 1: demographic and health | Model 2: final model | Change |
|---|---|---|---|
| Nagelkerke R2 | 0.058 | 0.186 | 0.128 |
| Efron’s R2 | 0.056 | 0.151 | 0.095 |
| AIC | 7636 | 6965 | − 671 |
OR odds ratio; CI confidence interval; ref reference category