Table 2.
Hazard ratios of non-affective psychosis by quintiles of own-region migrant density
| Hazard ratio (95% CI) | AIC | ||
|---|---|---|---|
| Unadjusted | .. | 72 817 | |
| Quintile 1 (lowest) | 1·35 (1·23–1·48)* | .. | |
| Quintile 2 | 1·07 (0·97–1·18) | .. | |
| Quintile 3 | 1·01 (0·92–1·12) | .. | |
| Quintile 4 | 1·00 (0·91–1·11) | .. | |
| Quintile 5 (highest) | 1 (ref) | .. | |
| Individual-adjusted and family-adjusted | .. | 72 767 | |
| Quintile 1 (lowest) | 1·35 (1·23–1·48)* | .. | |
| Quintile 2 | 1·13 (1·03–1·24)* | .. | |
| Quintile 3 | 1·10 (0·99–1·21) | .. | |
| Quintile 4 | 1·05 (0·96–1·16) | .. | |
| Quintile 5 (highest) | 1 (ref) | .. | |
| Neighbourhood-adjusted only | .. | 72 715 | |
| Quintile 1 (lowest) | 1·60 (1·44–1·77)* | .. | |
| Quintile 2 | 1·26 (1·13–1·40)* | .. | |
| Quintile 3 | 1·17 (1·05–1·30)* | .. | |
| Quintile 4 | 1·12 (1·01–1·24)* | .. | |
| Quintile 5 (highest) | 1 (ref) | .. | |
| Fully adjusted | .. | 69 952 | |
| Quintile 1 (lowest) | 1·36 (1·22–1·52)* | .. | |
| Quintile 2 | 1·14 (1·03–1·27)* | .. | |
| Quintile 3 | 1·11 (1·00–1·23) | .. | |
| Quintile 4 | 1·07 (0·96–1·18) | .. | |
| Quintile 5 (highest) | 1 (ref) | .. | |
| Per 5% decrease in density | |||
| Unadjusted | 1·03 (1·02–1·05)* | 72 799 | |
| Individual-adjusted only | 1·02 (1·01–1·04)* | 69 964 | |
| Neighbourhood-adjusted only | 1·07 (1·05–1·09)* | 72 741 | |
| Fully adjusted† | 1·05 (1·03–1·06)* | 69 957 | |
Models were adjusted for individual and family confounders (generation status, age, sex, calendar year, lone dwelling, family income, social welfare, family unemployment, and time since migration [migrants only]) or neighbourhood confounders (deprivation index, population density, and proportion of lone dwellings). Fully adjusted estimates include individual, family, and neighbourhood confounders. AIC=Akaike information criterion.
*
p<0·05.