Table 3

Neighborhood clustering effects of fruits and vegetables consumption and leisure-time physical activity
Model 1* Model 2* Model 3* Model 4*
Fruits and vegetables consumption
Women
   Variance (SE) 0.42 (0.17) 0.39 (0.17) 0.38 (0.16) 0.35 (0.16)
   Proportion of explained variance (%) Reference 7.1 9.5 16.7
   ICC (%) 7.0 6.7 6.6 6.1
Men
   Variance (SE) 0.22 (0.22) 0.22 (0.22) 0.18 (0.22) 0.18 (0.22)
   Proportion of explained variance (%) Reference 0 18.2 18.2
   ICC (%) 4.4 4.4 3.6 3.6
Leisure-time physical activity
Women
   Variance (SE) 7521.8 (2013.9) 6328.1 (1706.8) 6203.5 (1695.5) 6333.0 (1677.6)
   Proportion of explained variance (%) Reference 15.9 17.5 15.8
   ICC (%) 10.6 10.2 10.1 10.3
Men
   Variance (SE) 5310.5 (3721.6) 4033.9 (3399.5) 4135.2 (3391.3) 3631.8 (3343.1)
   Proportion of explained variance (%) Reference 24.0 22.1 31.6
   ICC (%) 5.5 4.8 4.9 4.3

SE, standard error; ICC, intracluster correlation coefficient.

*Model 1: Null model; Model 2: Model 1 plus adjustment for age, education and marital status; Model 3: Model 2 plus adjustment smoking, alcohol consumption and leisure-time physical activity (fruits and vegetables models) or fruits and vegetables consumption (leisure-time physical activity models); Model 4: Model 3 plus adjustment for neighborhood socioeconomic class.

Proportion of explained variance (%): corresponds to the proportion of between-neighborhood variance that could be explained by neighborhood selection variables, possible confounders and neighborhood socioeconomic class compared to Model 1. For instance, among women 15.9% of the neighborhood variance was explained by neighborhood selection variables: (7521.8-6328.1)/7521.8x100.

Alves et al.

Alves et al. BMC Public Health 2013 13:1103   doi:10.1186/1471-2458-13-1103

Open Data