Skip to main content
. 2016 May 6;3:25. doi: 10.1186/s40608-016-0106-4

Table 3.

Multivariate associations between individual SEP, neighbourhood SEP and overweight applying multilevel logistic regression (N = 3125)

Covariables Null model Individual modela OR (95 % CI) Full modela OR (95 % CI)
Nationality of the child
 Other than German 1.53 (1.18-1.99) 1.53 (1.17-1.99)
 German Reference Reference
Parental Education
 Low 2.04 (1.54-2.72) 1.99 (1.49-2.65)
 Middle 1.53 (1.18-1.99) 1.50 (1.16-1.95)
 High Reference Reference
Parental working status
 Unemployment within household 1.20 (0.82-1.77) 1.19 (0.80-1.75)
 At least one parent employed Reference Reference
Equivalent household income
 Low (<60 % Median)b 1.22 (0.82-1.83) 1.18 (0.79-1.77)
 Middle (60 % to Median) 1.29 (0.93-1.80) 1.26 (0.90-1.75)
 High (>Median) Reference Reference
 Not indicated 1.14 (0.83-1.58) 1.12 (0.81-1.55)
Household crowding
 Yes 0.94 (0.74-1.20) 0.93 (0.72-1.19)
 No Reference Reference
Neighbourhood SEP
 High Reference
 Middle 1.01 (0.71-1.46)
 Low 1.42 (1.00-2.00)
Measures of variation
Neighbourhood intercept variance (95 % CI)c 0.11 (0.046-0.48) 0.047 (0.015-0.80) 0.026 (0.005-20.37)
Proportional change in variance −57.3 % −19.1 %

OR Odds ratio, CI Confidence intervall, SEP Socioeconomic position

aAdjusted for survey year, birth weight, and BMI of mother, bMedian equivalent household income in Bavaria, cCovariance parameter estimates from random intercepts on the log odds scale