Table 3.

Path coefficients for the parallel mediation models and separate models for maternal and paternal chronic illness.

	Maternal chronic illness (N = 1,070)			Paternal chronic illness (N = 1,069)
	Coeff.	95% CI	P-value	Coeff.	95% CI	P-value
Total effect (c)	3.775	0.900 to 6.651	0.010	3.246	0.234 to 6.257	0.035
Direct effect (c')	3.121	0.278 to 5.966	0.032	2.116	−0.875 to 5.106	0.165
a1	−0.126	−0.279 to 0.027	0.105	−0.294	−0.453 to −0.135	0.003
a2	−0.360	−0.038 to −0.045	0.013	−0.292	−0.461 to −0.123	>0.001
a3	−0.092	−0.036 to 0.220	0.159	0.078	−0.056 to 0.211	0.252
b1	−1.819	−3.202 to −0.436	0.010	−1.786	−3.174 to −0.398	0.012
b2	−1.874	−3.180 to −0.567	0.005	−1.963	−3.272 to −0.654	0.003
b3	−0.409	−0.931 to 1.748	0.550	0.414	−0.623 to −0.930	0.257
Indirect effects a
Total indirecta effect	0.654	0.073 to 1.310		1.130	0.503 to 1.912
a1b1	0.230	−0.053 to 0.647		0.524	0.081 to 1.172
a2b2	0.387	0.031 to 0.901		0.573	0.031 to 1.206
a3b3	0.038	−0.130 to 0.272		0.032	−0.130 to 0.235

^aIf the 95% CI did not contain zero, the effect was considered statistically significant. The mediation models were adjusted for adolescent sex, parental SES and parental divorce. a₁-a₃ are regression coefficients for X₁-X₃, respectively. b₁-b₃ are regression coefficients for M₁-M₃, respectively.