Table 2.

Adjusted risk ratios (RR) of the association between physical IPV frequency and recent MDE among married women aged 16–37, rural Bangladesh, 2013–2014 (n = 3290)

	Model la		Model 2a		Model 3a		Model 4a		Model 5a
	RR	(95% CI)	RR	(95% CI)	RR	(95% CI)	RR	(95% CI)	RR	(95% CI)
Level 1
Physical IPV frequency
None	Ref		Ref		Ref		Ref		Ref
Low	1.01	(0.80, 1.28)	1.01	(0.80, 1.28)	1.05	(0.84, 1.30)	0.91	(0.67, 1.25)	0.97	(0.76, 1.23)
Medium	1.52	(1.09, 2.12)*	1.52	(1.09, 2.12)*	1.59	(1.19, 2.13)*	1.55	(1.09, 2.20)*	1.55	(1.12, 2.15)*
High	2.44	(1.94, 3.08)***	2.44	(1.94, 3.08)***	2.52	(2.08, 3.07)***	2.45	(1.93,3.12)***	2.36	(1.92, 2.91)***
Level 2
Low injunctive normativity
Quartile 1 (lowest)			0.99	(0.79, 1.25)			0.95	(0.63, 1.43)
Quartiles 2–4			Ref				Ref
Low descriptive normativity
Quartile 1 (lowest)					1.19	(0.87, 1.61)			1.06	(0.72, 1.56)
Quartiles 2–4					Ref				Ref
Cross-level interactions
Low injunctive normativity × physical IPV frequency							Interaction p = 0.489
None							Ref
Low							1.42	(0.64,3.15)
Medium							0.89	(0.47, 1.65)
High							0.96	(0.64, 1.43)
Low descriptive normativity × physical IPV frequency									Interaction p = 0.391
None									Ref
Low									1.45	(0.82, 2.56)
Medium									0.80	(0.36, 1.75)
High									1.31	(0.92, 1.86)

All models additionally adjusted for age, MDE at baseline, and ever witnessing father hit/beat mother

Injunctive normativity refers to the prevalence of favorable attitudes toward IPV perpetration among the women interviewed; descriptive normativity refers to the village-level prevalence of physical IPV. Normativity is presented using quartiles, where quartile 1 refers to villages with the least normative views

^***p<0.001

^**p<0.01

^*p<0.05