Depressive symptoms among Chinese residents: how are the natural, built, and social environments correlated?

BMC Public Health

Table 2 Results of the multilevel regressions

DV: Logged CES-D score	Model 1
DV: Logged CES-D score	Coefficients	Standard error
Natural environment of the neighborhood
Logged green space	−0.045*	0.023
Built environment of the neighborhood
Urbanicity	− 0.042**	0.015
Sports facilities (ref: no)	−0.001	0.044
Social environment of the neighborhood
Social trust	−0.407*	0.205
Social reciprocity	−0.286*	0.120
Social group membership	−0.404*	0.159
Gini index of household income	0.209	0.195
Security	0.325	0.217
Covariates
Male (ref: female)	−0.190**	0.019
Age	0.003**	0.001
Marital status (ref: single, divorced, widowed)
Married – living with spouse	−0.116**	0.022
Married – not living with spouse	−0.100**	0.033
Educational level (ref: primary school or lower)
High school	−0.134**	0.019
College or higher	− 0.098**	0.031
Employed (ref: unemployed)	−0.101**	0.033
Local hukou (ref: non-local hukou)	−0.063*	0.031
Smoker (ref: no)	0.015	0.022
Alcohol use (ref: no)	0.047*	0.021
Medical insurance (ref: no)	−0.036	0.025
Physical health status (ref: good)	0.674**	0.025
Urban neighbourhood (ref: rural neighbourhood)	0.116*	0.058
Logged household income per capita	−0.061**	0.008
Social trust (ref: not trustworthy)	−0.212**	0.019
Social reciprocity (ref: people seldom/never help)	−0.120**	0.017
Number of social groups	−0.384**	0.033
Sense of security (ref: unsafe)	−0.200**	0.027
Logged community income per capita	−0.080**	0.058
Constant	3.971**	0.345
Var. (constant)	0.128**
Var. (residual)	1.074**
AIC	60,554.29

Significance level: * p < 0.05, ** p < 0.01. All continuous independent variables were mean centered

ISSN: 1471-2458