Table 4

 Variation in less than good self rated health across neighbourhoods in London and Helsinki

LondonHelsinki
*Intraclass correlation coefficient (the proportion of the total variance in self rated health that is between neighbourhoods) is estimated as σu02 divided by (σu02 + π2/3) and ranges from 0 (no differences in self rated health between neighbourhoods) and 1 (all variation is at the neighbourhood level).
Base model: age and sex adjusted
    Variance (SE) between neighbourhoods0.106 (0.040)0.012 (0.014)
    Variance (SE) within neighbourhoods11
    ICC*0.0300.004
Age, sex, and socioeconomic position adjusted
    Variance (SE) between neighbourhoods0.067 (0.037)0.003 (0.012)
    Variance (SE) within neighbourhoods11
    ICC0.0200.001
Adjusted for age, sex, SES, + % unemployed
    Variance (SE) between neighbourhoods0.038 (0.035)0.002 (0.011)
    Variance (SE) within neighbourhoods11
    ICC0.0110.000
Adjusted for age, sex, SES, + % manual workers
    Variance (SE) between neighbourhoods0.058 (0.037)0.000 (0.011)
    Variance (SE) within neighbourhoods11
    ICC0.0180.000
Adjusted for age, sex, SES, + % single households
    Variance (SE) between neighbourhoods0.051 (0.36)0.000 (0.000)
    Variance (SE) within neighbourhoods11
    ICC0.0150.000
Adjusted for age, sex, SES, + % single parent households
    Variance (SE) between neighbourhoods0.059 (0.037)0.008 (0.013)
    Variance (SE) within neighbourhoods11
    ICC0.0180.002