3.4.1 Associations of General Exposures and Depressive Symptom
In these analyses, gender, race, household member(s), income to poverty, BMI and current health status were included as basic confounding variables in the regression models. Stepwise binary logistic regression analyses were preformed separately in order to assess the effect of recent tobacco smoking, household secondhand smoke exposure and confined space secondhand smoke exposure on young adults’ depressive symptom, respectively.
In the analysis of the association of recent tobacco smoking and depressive symptom: Model I (without adjusting any confounders) showed OR = 2.201 (95% CI: 1.735–2.353); Model II (adjusted for gender, race, household member, income to poverty, BMI and current health status) showed OR = 1.904 (95% CI: 1.621–2.237); Model III (adjusted for household secondhand smoke exposure) showed OR = 1.728 (95% CI: 1.436–2.080); Model IV (adjusted for confined space secondhand smoke exposure) showed OR = 1.593 (95% CI: 1.318–1.926). The results suggested that recent tobacco smoking is a significant risk factor for the development of depressive symptom in young adults, and this risk remained significant after adjusted for basic confounders. Even after adjusted for secondhand smoke exposure variables, recent tobacco smoking could still increase the risk of depressive symptom in young adults by 31.8–92.6%. All P < 0.001. (See Table 5)
In the analysis of the association of household secondhand smoke exposure and depressive symptom: Model I (without adjusting any confounders) showed OR = 1.715 (95% CI: 1.479–1.989, P < 0.001); Model II (adjusted for gender, race, household member, income to poverty, BMI and current health status) showed OR = 1.339 (95% CI: 1.339–1.827, P < 0.001); Model III (adjusted for recent tobacco smoking) showed OR = 1.209 (95% CI: 1.010–1.448, P = 0.038); Model IV (adjusted for adjusted for confined space secondhand smoke exposure) showed OR = 1.132 (95% CI: 0.942–1.360). The results suggest that household secondhand smoke exposure is not a significant independent risk factor for the development of depressive symptom in young adults. Although there was a significant risk after adjusted for the basic confounders, however, this risk was no longer significant after adjusted for confined space secondhand smoke exposure. (See Table 5)
In the analysis of the association of confined space secondhand smoke exposure and depressive symptom: Model I (without adjusting any confounders) showed OR = 1.812 (95% CI: 1.565–2.097); Model II (adjusted for gender, race, household member, income to poverty, BMI and current health status) showed OR = 1.674 (95% CI: 1.435–1.952); Model III (adjusted for recent tobacco smoking) showed OR = 1.427 (95% CI: 1.212–1.680); Model IV (adjusted for household secondhand smoke exposure) showed OR = 1.399 (95% CI: 1.185–1.651). The results suggested that confined space secondhand smoke exposure is a significant risk factor for the development of depressive symptom in young adults, and this risk remained significant after adjusted for basic confounders. Even after adjusted for recent tobacco smoking and household secondhand smoke exposure variables, confined space secondhand smoke exposure can still increase the risk of depressive symptom in young adults by 18.5–65.1%. All P < 0.001. (See Table 5)
Table 5
Weighted association of recent tobacco smoking, household / confined space secondhand smoke exposure and depressive symptom in NHANES 2013-18 adults aged 18–35 years
Recent Tobacco Smoking
|
Predictors
|
b
|
SE
|
Wald
|
P
|
OR (95% CI)
|
Model Ic
|
0.703
|
0.078
|
81.788
|
< 0.001
|
2.201 (1.735–2.353)
|
Model IId
|
0.644
|
0.082
|
61.437
|
< 0.001
|
1.904 (1.621–2.237)
|
Model IIIf
|
0.547
|
0.095
|
33.438
|
< 0.001
|
1.728 (1.436–2.080)
|
Model IVg
|
0.466
|
0.097
|
23.167
|
< 0.001
|
1.593 (1.318–1.926)
|
Household Secondhand Smoke Exposure
|
Model Ic
|
0.539
|
0.076
|
50.963
|
< 0.001
|
1.715 (1.479–1.989)
|
Model IId
|
0.447
|
0.079
|
31.768
|
< 0.001
|
1.339 (1.339–1.827)
|
Model IIIe
|
0.190
|
0.092
|
4.268
|
0.038
|
1.209 (1.010–1.448)
|
Model IVg
|
0.124
|
0.094
|
1.760
|
0.185
|
1.132 (0.942–1.360)
|
Confined Space Secondhand Smoke Exposure
|
Model Ic
|
0.594
|
0.075
|
63.328
|
< 0.001
|
1.812 (1.565–2.097)
|
Model IId
|
0.515
|
0.078
|
43.162
|
< 0.001
|
1.674 (1.435–1.952)
|
Model IIIe
|
0.356
|
0.083
|
18.261
|
< 0.001
|
1.427 (1.212–1.680)
|
Model IVf
|
0.336
|
0.085
|
15.730
|
< 0.001
|
1.399 (1.185–1.651)
|
Note: c Original model, no adjustment for any confounding variables; d Adjusted for the independent variable in the last model plus gender, race, household member(s), income to poverty, BMI, recent health status; e Adjusted for the independent variable in the last model plus recent tobacco smoking; f Adjusted for the independent variable in the last model plus household secondhand smoke exposure; g Adjusted for the independent variable in the last model plus confined space secondhand smoke exposure; Same as below.
|
3.4.1 Associations of Detailed Exposures and Depressive Symptom
In these analyses, same basic confounding variables were included in the regression models as in the previous section. We performed stepwise binary logistic regression analyses for secondhand smoke exposure in restaurant, in bar, in car, in another home, and in other indoor settings, respectively, in order to determine the effect of secondhand smoke exposure in different settings on young adult depressive symptom.
In the analysis of the association of restaurant exposure and depressive symptom: Model V (without adjusting any confounders) showed OR = 1.800 (95% CI: 1.223–2.561, P = 0.003); Model VI (adjusted for gender, race, household member, income to poverty, BMI and current health status) showed OR = 1.913 (95% CI: 1.257–2.872, P = 0.002); Model VII (adjusted for recent tobacco smoking) showed OR = 1.943 (95% CI: 1.288–2.932, P = 0.002); Model VIII (adjusted for household secondhand smoke exposure) showed OR = 1.912 (95% CI: 1.267–2.884, P = 0.002); Model IX (adjusted for in bar, in car, in another and in other indoor settings secondhand smoke exposure) showed OR = 1.732 (95% CI: 1.120–2.678, P = 0.013). The results suggest that restaurant secondhand smoke exposure may be a significant independent risk factor for depressive symptom in young adults, increasing the risk by 12.0–167.8%. (See Table 6)
In the analysis of the association of bar exposure and depressive symptom: Model V (without adjusting any confounders) showed OR = 1.447 (95% CI: 1.083–1.934, P = 0.013); Model VI (adjusted for gender, race, household member, income to poverty, BMI and current health status) showed OR = 1.403 (95% CI: 1.036–1.899, P = 0.029); Model VII (adjusted for recent tobacco smoking) showed OR = 1.243 (95% CI: 0.914–1.692, P = 0.166); Model VIII (adjusted for household secondhand smoke exposure) showed OR = 1.223 (95% CI: 0.906–1.687, P = 0.183); Model IX (adjusted for in restaurant, in car, in another and in other indoor settings secondhand smoke exposure) showed OR = 0.998 (95% CI: 0.711–1.373, P = 0.934). The results suggest that bar secondhand smoke exposure independently has no significant effect on depressive symptom in young adults. (See Table 6)
In the analysis of the association of in-car exposure and depressive symptom: Model V (without adjusting any confounders) showed OR = 2.001 (95% CI: 1.704–2.349, P < 0.001); Model VI (adjusted for gender, race, household member, income to poverty, BMI and current health status) showed OR = 1.843 (95% CI: 1.557–2.182, P < 0.001); Model VII (adjusted for recent tobacco smoking) showed OR = 1.505 (95% CI: 1.252–1.809, P < 0.001); Model VIII (adjusted for household secondhand smoke exposure) showed OR = 1.468 (95% CI: 1.214–1.774, P < 0.001); Model IX (adjusted for in restaurant, in bar, in another home and in other indoor settings secondhand smoke exposure) showed OR = 1.350 (95% CI: 1.102–1.652, P = 0.004). The results suggest that in-car secondhand smoke exposure may be a significant independent risk factor for depressive symptom in young adults, increasing the risk by 10.2–65.2%. (See Table 6)
In the analysis of the association of exposure in another home and depressive symptom: Model V (without adjusting any confounders) showed OR = 1.780 (95% CI: 1.462–2.168, P < 0.001); Model VI (adjusted for gender, race, household member, income to poverty, BMI and current health status) showed OR = 1.625 (95% CI: 1.322–1.998, P < 0.001); Model VII (adjusted for recent tobacco smoking) showed OR = 1.378 (95% CI: 1.114–1.705, P = 0.003); Model VIII (adjusted for household secondhand smoke exposure) showed OR = 1.359 (95% CI: 1.098–1.682, P = 0.005); Model IX (adjusted for in restaurant, in bar, in car and in other indoor settings secondhand smoke exposure) showed OR = 1.161 (95% CI: 0.923–1.461, P = 0.202). The results suggest that the role of secondhand smoke exposure in another home is strongly influenced by exposure in the rest settings, and that the current variable independently has no significant effect on depressive symptom in young adults. (See Table 6)
In the analysis of the association of exposure in other indoor settings and depressive symptom: Model V (without adjusting any confounders) showed OR = 1.629 (95% CI: 1.266–2.097, P < 0.001); Model VI (adjusted for gender, race, household member, income to poverty, BMI and current health status) showed OR = 1.619 (95% CI: 1.243-2.100, P < 0.001); Model VII (adjusted for recent tobacco smoking) showed OR = 1.470 (95% CI: 1.125–1.922, P = 0.005); Model VIII (adjusted for household secondhand smoke exposure) showed OR = 1.442 (95% CI: 1.103–1.887, P = 0.007); Model IX (adjusted for in restaurant, in bar, in car and in another home secondhand smoke exposure) showed OR = 1.230 (95% CI: 0.929–1.629, P = 0.148). The results also suggest that the role of secondhand smoke exposure in other indoor settings is strongly influenced by exposure in the rest settings, and that the current variable independently has no significant effect on depressive symptom in young adults. (See Table 6)
Table 6
Weighted association of secondhand smoke exposure in different settings and depressive symptom in NHANES 2013-18 adults aged 18–35 years
|
b
|
SE
|
Wald
|
P
|
OR (95% CI)
|
In Restaurant
|
Model Vc
|
0.588
|
0.197
|
8.876
|
0.003
|
1.800 (1.223–2.561)
|
Model VId
|
0.649
|
0.207
|
9.811
|
0.002
|
1.913 (1.275–2.872)
|
Model VIIe
|
0.664
|
0.210
|
10.025
|
0.002
|
1.943 (1.288–2.932)
|
Model VIIIf
|
0.648
|
0.210
|
9.537
|
0.002
|
1.912 (1.267–2.884)
|
Model IXh
|
0.549
|
0.222
|
6.104
|
0.013
|
1.732 (1.120–2.678)
|
In Bar
|
Model Vc
|
0.370
|
0.148
|
6.237
|
0.013
|
1.447 (1.083–1.934)
|
Model VId
|
0.338
|
0.155
|
4.780
|
0.029
|
1.403 (1.036–1.899)
|
Model VIIe
|
0.218
|
0.157
|
1.919
|
0.166
|
1.243 (0.914–1.692)
|
Model VIIIf
|
0.209
|
0.157
|
1.775
|
0.183
|
1.223 (0.906–1.678)
|
Model IXh
|
-0.012
|
0.168
|
0.005
|
0.943
|
0.998 (0.711–1.373)
|
In Car
|
Model Vc
|
0.694
|
0.082
|
71.932
|
< 0.001
|
2.001 (1.704–2.349)
|
Model VId
|
0.611
|
0.086
|
50.371
|
< 0.001
|
1.843 (1.557–2.182)
|
Model VIIe
|
0.409
|
0.094
|
18.927
|
< 0.001
|
1.505 (1.252–1.809)
|
Model VIIIf
|
0.384
|
0.097
|
15.741
|
< 0.001
|
1.468 (1.214–1.774)
|
Model IXh
|
0.300
|
0.103
|
8.442
|
0.004
|
1.350 (1.102–1.652)
|
In Another Home
|
Model Vc
|
0.577
|
0.100
|
32.932
|
< 0.001
|
1.780 (1.462–2.168)
|
Model VId
|
0.486
|
0.105
|
21.220
|
< 0.001
|
1.625 (1.322–1.998)
|
Model VIIe
|
0.321
|
0.109
|
8.745
|
0.003
|
1.378 (1.114–1.705)
|
Model VIIIf
|
0.306
|
0.109
|
7.924
|
0.005
|
1.359 (1.098–1.682)
|
Model IXh
|
0.150
|
0.117
|
1.631
|
0.202
|
1.161 (0.923–1.461)
|
In Other Indoors
|
Model Vc
|
0.488
|
0.129
|
14.372
|
< 0.001
|
1.629 (1.266–2.097)
|
Model VId
|
0.482
|
0.135
|
12.734
|
< 0.001
|
1.619 (1.243–2.110)
|
Model VIIe
|
0.386
|
0.137
|
7.967
|
0.005
|
1.470 (1.125–1.922)
|
Model VIIIf
|
0.366
|
0.137
|
7.152
|
0.007
|
1.442 (1.103–1.887)
|
Model IXh
|
0.207
|
0.143
|
2.090
|
0.148
|
1.230 (0.929–1.629)
|
Note: h Adjusted for the independent variable in the last model plus secondhand smoke exposure in other establishments differ from the current model.
|