A mediation analysis of the role of girl child marriage in the relationship between proximity to conflict and past-year intimate partner violence in post-conflict Sri Lanka

Table 1 Demographic details of currently partnered women aged 18–49 who participated in the 2016 Sri Lankan DHS Domestic Violence module and distribution of past year sexual, physical, and emotional intimate partner violence (IPV) (N = 13,691)

	Total		Past year sexual IPV			Past year physical IPV			Past year emotional IPV
Characteristic	n	%^	n	%	Chi-squared p value	n	%	Chi-squared p value	n	%	Chi-squared p value
Total	13,691	100	363	2	–	1290	9	–	1949	13	–
Proximity to conflict
Distal	7603	64	110	1	< 0.01*	529	7%	< 0.01*	867	11	< 0.01*
Proximal	3131	23	60	2		247	8		261	9
Central	2957	13	193	8		514	19		821	29
Girl child marriage
No	11,707	86	265	2	< 0.01*	967	8	< 0.01*	1539	12	< 0.01*
Yes	1984	14	98	5		323	16		410	19
Age^^
18–29	3031	22	69	2	0.03*	276	8	0.04*	394	12	< 0.01*
30–39	5852	43	147	2		543	8		804	13
40–49	4808	35	147	3		471	10		751	15
Education
Primary (01–05) or less	1189	8	76	6	< 0.01*	234	20	< 0.01*	288	24	< 0.01*
Secondary (6–12)	9119	66	252	2		919	10		1303	13
Higher than Secondary	3383	26	35	1		137	4		358	10
Household wealth quintile
Lowest	3090	18	169	5	< 0.01*	575	19	< 0.01*	734	22	< 0.01*
Second	2787	20	89	3		276	10		399	14
Middle	2683	21	42	2		200	7		305	11
Fourth	2660	21	33	1		146	6		270	10
Highest	2471	20	30	1		93	4		241	10
Parity
0–1	4267	32	62	1	< 0.01*	276	6	< 0.01*	475	10	< 0.01*
2	5154	39	123	2		470	9		687	12
3+	4270	29	178	4		544	12		787	17
Age difference between woman and partner^^
Same or older	2468	18	77	3	0.12	283	11	< 0.01*	397	15	0.05*
1–5 years	6712	49	176	2		628	9		911	12
6–10 years	3598	27	77	2		285	7		494	13
Over 10 years	913	6	33	3		94	9		147	14
Religion
Buddhism	8855	73	135	2	< 0.01*	585	7	< 0.01*	922	11	< 0.01*
Hinduism	2327	11	127	6		417	19		624	27
Islam	1267	9	59	4		132	10		195	15
Other	1242	8	42	3		156	12		208	15
Urban setting
No	11,510	85	312	2	0.85	1097	9	0.56	1604	13	< 0.01*
Yes	2181	15	51	2		193	9		345	16
District^^^					< 0.01*			< 0.01*			< 0.01*

n values are unweighted, while percent values are weighted
Chi-squared p-values are weighted according to the survey's complex sampling design
^Total percent values are calculated within the same column, while all other percent values are calculated across the same row
^^Age and age difference between woman and partner are presented in cross-tabulations as categorical but were included in regression models as continuous variables
^^^A full list of districts is omitted due to length (see Additional file 1)
^*p < 0.05

ISSN: 1752-1505