Table 4.

Continuation of caregiving of non-working caregivers

Independent variables		HI not provided in previous period
Co-residential caregiving	−0.0101	−0.0417
	(0.122)	(0.129)
Co-residential caregiving (initial)	0.456***	0.573***
	(0.104)	(0.121)
m(proxy of formal care use)	−62.58**	−68.33**
	(26.23)	(29.21)
High-intensity caregiving (HI)	1.047*	1.232**
	(0.537)	(0.570)
Generalized residual (GR)	−0.462	−0.363
	(0.422)	(0.434)
HI × GR	0.0907	−0.281
	(0.485)	(0.520)
Relative resources squared	0.00917	0.00598
	(0.00878)	(0.00931)
Relative resources	0.0229	0.0374
	(0.0327)	(0.0351)
Sex (male = 1)	−0.120	−0.104
	(0.157)	(0.167)
Care recipients
Father	0.0599	0.0339
	(0.135)	(0.144)
Mother	0.408***	0.469***
	(0.111)	(0.122)
Father-in-law	−0.143	−0.117
	(0.166)	(0.178)
Mother-in-law	0.284**	0.317**
	(0.118)	(0.129)
Health status
Difficulty in daily life activities	−0.0223	−0.0518
	(0.124)	(0.140)
Diabetes	0.0400	0.0541
	(0.167)	(0.184)
Heart disease	−0.168	−0.0844
	(0.225)	(0.242)
Cerebral stroke	−0.0348	0.0627
	(0.327)	(0.340)
Hypertension	0.216	0.231
	(0.152)	(0.164)
Hyperlipidemia	0.0545	−0.0108
	(0.123)	(0.132)
Having friend	0.206*	0.274**
	(0.124)	(0.134)
Constant	1.492	0.926
	(1.330)	(1.445)
lnσu 2	−0.422	−0.376
σu	0.809***	0.828***
	(0.141)	(0.157)
Intra-class correlation (σu 2/(1 + σu 2))	0.396	0.407
Likelihood-ratio test of ρ = 0 [chi2(1)]	18.10	15.70
Prob ≥ chi2	0.00	0.00
Log-likelihood	−1005.92	−857.71
N	1743	1454

Co-residential caregiving (initial) refers to high-intensity caregiving in 2005. Age, educational attainment, hospitalization, marital status, medication or doctor’s consultation, residence, and having cancer were included as covariates

Standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1