Table 2.

Mean differences and RRs for sexual violence predicting mental health outcomes from a range of modelling approaches

	Psychological distress in the past 30 days*, mean difference (95% CI)		High psychological distress in the past 30 days, RR (95% CI)		Self-harm in the past 12 months, RR (95% CI)		Attempted suicide ever, RR (95% CI)
	Girls	Boys	Girls	Boys	Girls	Boys	Girls	Boys
Multivariable regressions
Unadjusted	3·16 (2·52–3·81)	3·75 (3·04–4·46)	2·18 (1·79–2·66)	2·41 (1·62–3·59)	2·32 (1·99–2·71)	3·09 (2·38–4·03)	2·67 (2·01–3·54)	4·17 (2·54–6·84)
Adjustment 1	2·14 (1·47–2·81)	2·11 (2·18–4·05)	1·67 (1·36–2·05)	1·87 (1·23–2·84)	1·90 (1·60–2·25)	2·62 (1·90–3·62)	1·84 (1·32–2·56)	3·00 (1·84–4·88)
Adjustment 2	2·02 (1·37–2·67)	2·59 (1·61–3·56)	1·62 (1·32–1·98)	1·58 (1·02–2·45)	1·79 (1·52–2·09)	2·23 (1·63–2·74)	1·72 (1·22–2·41)	2·80 (1·66–4·72)
Adjustment 3	2·10 (1·45–2·75)	2·57 (1·60–3·54)	1·65 (1·35–2·02)	1·57 (1·01–2·44)	1·81 (1·54–2·12)	2·22 (1·72–2·88)	1·74 (1·24–2·43)	2·60 (1·53–4·41)
Adjustment 4	2·07 (1·44–2·71)	2·64 (1·69–3·59)	1·64 (1·35–1·99)	1·64 (1·08–2·50)	1·80 (1·54–2·11)	2·24 (1·73–2·88)	1·75 (1·26–2·44)	2·60 (1·55–4·36)
Fully adjusted model	2·09 (1·51–2·68)	2·56 (1·59–3·53)	1·65 (1·37–2·00)	1·55 (1·00–2·40)	1·79 (1·52–2·10)	2·16 (1·63–2·84)	1·75 (1·26–2·41)	2·73 (1·59–4·67)
Matched samples
PSM (all matching variables)	1·35 (0·88–1·81)	2·73 (1·84–3·63)	1·30 (1·14–1·49)	1·97 (1·31–2·93)	1·42 (1·28–1·58)	1·92 (1·47–2·52)	1·37 (1·11–1·70)	2·28 (1·33–3·93)
PSM (all variables minus sexual identity, pubertal status, parental education, and sexual violence before age 14 years)	1·66 (1·19–2·14)	3·29 (2·41–4·16)	1·35 (1·18–1·54)	2·68 (1·70–4·23)	1·58 (1·41–1·77)	2·18 (1·64–2·92)	1·44 (1·16–1·79)	4·28 (2·12–8·65)
PSM (all variables minus high psychological distress, self-harm at age 14 years, and sexual violence before age 14 years)	2·37 (1·92–2·83)	3·53 (2·67–4·38)	1·72 (1·48–2·00)	3·11 (1·91–5·04)	1·82 (1·62–2·06)	2·24 (1·67–3·00)	2·17 (1·69–2·80)	4·83 (3·04–18·97)
Matched sample with multivariable confounder adjustment
Adjusted regression analysis	1·35 (0·92–1·79)	2·54 (1·68–3·40)	1·31 (1·16–1·49)	1·87 (1·26–2·80)	1·41 (1·27–1·57)	1·83 (1·40–2·40)	1·37 (1·11–1·68)	2·36 (1·37–4·08)
Multivariable adjusted regression by sexual violence type
Unwelcome sexual approach	2·03 (1·44–2·63)	2·61 (1·61–3·60)	1·49 (1·34–1·65)	1·88 (1·44–2·45)	1·56 (1·44–1·70)	1·99 (1·67–2·37)	1·44 (1·21–1·71)	2·54 (1·72–3·76)
Sexual assault	2·32 (1·49–3·15)	3·82 (1·82–5·81)	1·46 (1·27–1·68)	1·87 (1·11–3·13)	1·57 (1·41–1·75)	2·09 (1·56–2·80)	1·72 (1·40–2·11)	2·63 (1·38–5·01)
Unwelcome sexual approach only†	1·51 (1·16–1·86)	2·05 (1·49–2·60)	1·45 (1·29–1·64)	1·80 (1·36–2·38)	1·49 (1·35–1·65)	1·87 (1·54–2·26)	1·27 (1·03–1·57)	2·33 (1·55–3·53)

The models had sequentially increasing amounts of adjustment. Adjustment 1 was for previous mental ill health (depressive symptoms and self-harm up to 14 years of age). Adjustment 2 added sociodemographic or economic characteristics (ethnicity, parental education, family income, sexual identity, single parent household, and number of siblings). Adjustment 3 added puberty-related characteristics (pubertal status, age of menarche [only girls], early sexual activity, and experience of sexual violence up to 14 years of age). Adjustment 4 added interpersonal characteristics (relationship status, peer relationships, and bullying). The fully adjusted model added health-related characteristics (BMI, risky behaviours, missing school without parents’ permission, disability, and life satisfaction). PSM=propensity score matching. RR=risk ratio.

^*This is a continuous measure.

^†Individuals who reported unwelcome sexual approach and no sexual assault.