Skip to main content
. 2021 Apr 7;11:7596. doi: 10.1038/s41598-021-87218-7

Table 6.

Decomposition of direct effect, indirect effect, and total effect.

Explanatory variable Explained Variable: Perinatal mortality
(1) (2) (3) (4)
Direct effect
Average PM2.5 (log) 2.483** 2.580**
(1.183) (1.174)
Max PM2.5 (log) 2.308** 2.194**
(0.912) (0.903)
GDP per capita(log) − 6.360*** − 5.643*** − 6.258*** − 5.512***
(0.915) (1.022) (0.912) (1.019)
bed_pop − 0.053* − 0.044 − 0.054** − 0.045
(0.028) (0.029) (0.027) (0.029)
ha_pop − 0.035 − 0.023 − 0.043 − 0.030
(0.051) (0.055) (0.052) (0.056)
Urbanization − 0.001 0.005 − 0.001 0.006
(0.013) (0.013) (0.013) (0.013)
Indirect effect
Average PM2.5 (log) − 1.960 − 1.885
(1.509) (1.601)
Max PM2.5 (log) − 1.147 − 0.459
(1.216) (1.290)
GDP per capita(log) 2.846*** 7.716*** 2.623*** 7.460***
(0.978) (1.533) (0.973) (1.515)
bed_pop − 0.031 − 0.169*** − 0.026 − 0.157***
(0.033) (0.050) (0.032) (0.049)
ha_pop − 0.075 − 0.349*** − 0.063 − 0.336***
(0.061) (0.084) (0.061) (0.083)
Urbanization 0.037 0.019 0.036 0.022
(0.027) (0.026) (0.026) (0.026)
Total effect
lnPM2.5 0.523 0.695
(0.854) (0.969)
Max PM2.5 (log) 1.162 1.735*
(0.803) (0.892)
GDP per capita(log) − 3.514*** 2.074 − 3.635*** 1.948
(0.411) (1.610) (0.399) (1.581)
bed_pop − 0.084*** − 0.212*** − 0.080*** − 0.202***
(0.019) (0.043) (0.018) (0.042)
hs_pop − 0.109** − 0.372*** − 0.106** − 0.365***
(0.045) (0.081) (0.044) (0.079)
Urbanization 0.036 0.024 0.036 0.028
(0.026) (0.026) (0.026) (0.026)

Clustered standard in parentheses, *p < 0.1, **p < 0.05, ***p < 0.01.