Table 4.

multivariable analysis using the Log logistic-Gamma shared frailty model, EDHS 2016

Covariate	Coeff	St.err	P-value	ϕ	[95% CI for ϕ]
Constant	2.8927	0.0103	0.000	18.0419	17.6815 18.4098
Age at first sex	-0.0027	0.0007	0.000	0.9973	0.9959 0.9987
Age at first cohabitation	-0.0014	0.0007	0.036	0.9986	0.9973 0.9998
Sex of household head
Male (ref)
Female	0.0096	0.0044	0.027	1.0096	1.0011 1.0183
Place of residence
Urban (ref)
Rural	-0.0521	0.0054	0.000	0.9492	0.9392 0.9594
Religion
Orthodox (ref)
Catholic	0.0059	0.0244	0.807	1.0059	0.9590 1.0552
Protestant	-0.0032	0.0072	0.661	0.9968	0.9828 1.0110
Muslim	-0.0428	0.0066	0.000	0.9581	0.9458 0.9706
Other	-0.0066	0.0173	0.703	0.9934	0.9602 1.0278
Highest education level
No education (ref)
Primary	0.0615	0.0039	0.000	1.0635	1.0552 1.0718
Secondary	0.1698	0.0062	0.000	1.1850	1.1706 1.1996
Higher	0.2739	0.0083	0.000	1.3150	1.2939 1.3366
Contraceptive use
No (ref)	0.0489	0.0047	0.000	1.0501	1.0406 1.0598
Yes
Media exposure No (ref) Yes	0.0387	0.0043	0.000	1.0395	1.0307 1.0484
Drinking alcohol
No (ref)	-0.0179	0.0061	0.004	0.9823	0.9706 0.9942
Yes

Φ= 0.585; γ = 0.086; τ= 0.226; Likelihood-ratio test of theta=0: chibar2(01)=1140.90 prob >= chibar2(01) = 0.000;