The association between sexual violence and mental disorders among women victim-survivors in sub-Saharan Africa: a systematic review and meta-analysis

Abstract

Background

Sexual violence is a serious public health and human rights problem with both short-term and long-term consequences. This review aims to systematically assess the link between sexual violence and poor mental health among sub-Saharan African women.

Methods

Systematic review and meta-analyses of observational studies were performed. MEDLINE, CINAHL, EMBASE, PsycINFO, Cochrane, Global Health and the University of Melbourne library electronic databases were used to find relevant published literature over 12 years from 2012 to 2024 in sub-Saharan Africa with stringent eligibility criteria. Random effects meta-analyses were used to pool estimates of ORs and 95% CIs. The I² statistic was used to assess heterogeneity.

Results

This systematic review and meta-analysis of 76 observational studies included a total of 80 313 participants and found a consistent small-to-medium association between lifetime sexual violence and experiences of poor mental health. The pooled ORs suggest that women who were exposed to sexual violence were more than twice as likely to experience post-traumatic stress disorder (OR 2.75; 95% CI 1.96 to 3.86; I²=73.4%), depression (OR 2.38; 95% CI 2.04 to 2.77; I²=56.6%), anxiety (OR 2.81; 95% CI 1.67 to 4.72; I²=77.2%), common mental disorders (OR 2.12; 95% CI 1.70 to 2.64; I²=0.0%), suicidal behaviour (OR 2.44; 95% CI 1.92 to 3.10; I²=68.0%) and emotional distress (OR 3.14; 95% CI 1.73 to 5.69; I²=79.6%) compared with women who have not experienced sexual violence.

Conclusions

Exposure to lifetime sexual violence was consistently associated with small to medium effects on poor mental health among women in sub-Saharan Africa. Thus, policy-makers should develop response strategies as well as mental health screening tools for all violence response service delivery points. In addition, health practitioners must prioritise screening for mental health conditions in patients who present with a history of sexual violence.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Individual studies and syntheses, mostly from high-income countries, revealed the association between sexual violence (SV) and poor mental health. However, no systematic review and meta-analysis conducted in sub-Saharan Africa examining the association of SV and poor mental health among women across all age groups in the region.

WHAT THIS STUDY ADDS

• This manuscript contributed to the global body of knowledge about the association between SV and poor mental health and identified the prevalent mental health conditions among SV victims in sub-Saharan Africa, a region often underrepresented in existing literature.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• This study highlights the importance of recognising the mental health burden on SV victims in sub-Saharan Africa, prompting policy-makers to develop supportive national action plans, encouraging the integration of comprehensive SV management into health services and advocating for future research employing longitudinal and qualitative methods.

Introduction

Sexual violence (SV) is a serious public health and human rights problem with both short-term and long-term consequences for women’s physical, mental, and sexual and reproductive health. The WHO defines SV as ‘any sexual act, attempt to obtain a sexual act or other act directed against a person’s sexuality using coercion, by any person regardless of their relationship to the victim, in any setting’.¹ SV is a form of gender-based violence, with women disproportionately impacted and considerably more likely to experience repeated and severe forms of SV.² Young women are especially at risk, and those aged between 16 and 19 years are four times more likely than the general population to be victims of rape, attempted rape or sexual assault.³ This paper will focus on SV against women.

Globally, 31% of women aged 15–49 years have experienced physical and/or SV from a current or former husband or male intimate partner, SV from someone who was not a current or former husband or intimate partner, or both these forms of violence at least once since the age of 15 years.⁴ Intimate partner violence is the most common type of violence reported, with lifetime prevalence for women aged 15–49 years estimated to be 27% across the world, while the estimated global prevalence of non-partner SV was 6% in 2018.⁴ Estimated prevalence rates for lifetime non-partner SV vary considerably across the world. The highest estimated prevalence of non-partner SV since age 15 was reported in high-income regions including Australia and New Zealand (19%) and Northern America (15%); whereas sub-Saharan Africa (6%) and South-eastern Asia (4%) had the lowest prevalence estimates among regions.⁴ However, this variation requires careful interpretation, as the true prevalence in low-income and middle-income settings such as sub-Saharan Africa is likely to be substantially higher than current estimates.⁵ Despite growing consensus on standardised definitions of SV and gold standard measures through specialised population-based surveys, survey design and implementation are inconsistent across geographical regions and often undertaken among non-representative samples.⁴ Inconsistent definitions and measurement and the stigmatised nature of SV, particularly in some communities, means that SV is likely under-reported.

SV affects physical, mental, sexual and reproductive health (including unplanned or lost pregnancies and exposure to STIs and HIV).6,8 It also has negative economic impacts for individuals (due to out-of-pocket expenses and impact on ability to work) and national economies.⁹ At the community level, SV can stigmatise the victim, depriving her of social status and value (if she is seen as unfaithful or promiscuous), and thereby modifies relationships within the community with an overall deleterious effect. SV can lead to radical changes in the image that the victim has of herself, and how she sees the past, present and future. SV can also change a victim’s relations with her immediate social circle and in her community, having a lasting negative impact on the victim’s feelings of herself, of events and of others.^{8 10}

All these impacts have a flow-on impact on mental health. Studies in a wide range of settings have reported psychological distress, depression, panic attack episodes, post-traumatic stress disorder (PTSD) and postpartum depression as being related to SV.^{11 12} Suicidal behaviours have also been reported among postpartum mothers with a history of SV.^{13 14} Conversely, other studies have reported poor mental health as a predisposing factor for SV, and women with a mental illness are substantially more vulnerable to being victims of all forms of violence, suggesting a bidirectional relationship.¹⁵

Individual studies and syntheses of published studies mostly from high-income countries reveal the association between SV and poor mental health across the world.^{15 16} However, to the best of our knowledge, there has been no systematic review and meta-analysis conducted in sub-Saharan Africa that synthesises the literature examining the association of SV and poor mental health among women across all age groups in the region. The association between exposure to SV and poor mental health could be different in this region compared with elsewhere in the world due to a range of contextual differences.

Consequently, the aim of this review is to address research gaps through systematically assessing the evidence for an association between SV and poor mental health among women in sub-Saharan African countries; and identifying the most common mental health conditions among SV victims compared with those women who have no experience of SV. Additionally, this study will conduct a meta-analysis to estimate the pooled effects of SV on poor mental health. The findings of the study can inform policies and strategies to respond to the mental health effects of SV.

Method

Data source

The reporting of results was based on the recommendations of Preferred Reporting Items for Systematic Reviews and Meta-Analyses.¹⁷ The study protocol was registered with Prospero: registration CRD42022339383 (www.crd.york.ac.uk/prospero). MEDLINE, CINAHL, EMBASE, PsycINFO, Cochrane, Global Health and The University of Melbourne library electronic databases were used to find relevant published literature over 12 years from 2012 to 2024 in sub-Saharan Africa with stringent eligibility criteria.

Study selection criteria

Studies were included if they were primary quantitative studies that compared poor mental health among women in sub-Saharan Africa with a history of SV to women without a history of SV. To be eligible for inclusion, studies had to be published in a peer-reviewed journal in the English language from 2012 to 2024. The exposure variable SV is defined as ‘any sexual act, attempt to obtain a sexual act or other act directed against a person’s sexuality using coercion, by any person regardless of their relationship to the victim, in any setting. It includes rape, defined as the physically forced or otherwise coerced penetration of the vulva or anus with a penis, other body part or object, attempted rape, unwanted sexual touching and other non-contact forms.’¹ The outcome variable poor mental health includes anxiety disorders (generalised anxiety disorder, social phobias, specific phobias, panic disorders, agoraphobia and anxiety disorders not otherwise specified), bipolar disorder, depression, eating disorders (anorexia nervosa and bulimia nervosa), obsessive–compulsive disorder, PTSD, schizophrenia, sleep disorders (night terrors, insomnia, narcolepsy and sleep disorders not otherwise specified), somatoform disorders (conversion disorder, somatisation, hypochondriasis and body dysmorphic disorders) and suicide attempts. Studies were excluded if they (a) did not report SV, (b) if the study did not include women in the sample or if it did not disaggregate the data by gender, (c) studies were not conducted in sub-Saharan African countries and (d) if the study was not quantitative (qualitative studies, case reports, editorials, narrative reviews, conference abstracts, letters, and commentaries were all excluded).

Search strategy, data extraction and quality measures

The search strategy was developed by the researchers and refined by an experienced librarian at The University of Melbourne. Keywords were used to define the concept areas of SV and poor mental health, and the search was limited to epidemiological studies. The screening was conducted using Covidence software—a Cochrane technology web-based platform designed to help researchers produce timely, high-quality systematic reviews—to track each document through the inclusion and exclusion criteria process. The Covidence database also allowed the researchers to work independently, assess reliability and identify conflicts for resolution through team discussion. Two researchers independently screened the title and abstract of each article for possible inclusion. Three researchers reviewed the full text of each screened article to make a final decision about whether these articles met inclusion and exclusion criteria. The screening stages were conducted by SAW, GW and SB, with any discrepancies resolved among the team to achieve full consensus.

The three independent reviewers assessed the methodological quality of the included studies using Joanna Briggs Institute’s (JBI) critical appraisal tools for cross-sectional, case–control and cohort studies to evaluate the trustworthiness, relevance and results of published papers (https://jbi.global/critical-appraisal-tools). In summary, the following criteria were applied to critically determine the eligibility of the studies. (See figure 1 for the details of articles excluded based on each criterion).

Figure 1. PRISMA flow diagram for this study. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

• The study must be quantitative.

• Studies must be composed of a sample of women or include data disaggregated by sex.

• There should be two groups (exposed and non-exposed to SV).

• The exposure and outcome measurement tools were valid and had been used in previous studies.

• The study reported at least one mental health outcome among both the SV and non-SV group or in comparison with the OR or effect size.

• The study population or SV group must not have, by design, only consisted of people with a pre-existing mental disorder.

• The study must not have been fully duplicative of another eligible study.

After final agreement on the selected studies, we used Microsoft Excel to review and extract details about each study included in the final sample. We also extracted information concerning the study purpose, design, participants, setting and details about the outcome variable in both the exposed and control groups. Summary data of ORs and 95% (CIs with preference to the adjusted OR were obtained if possible. We conducted a narrative, descriptive synthesis of the extracted information to answer our research questions. The first question was whether there is an association between experience of SV and poor mental health among women in sub-Saharan Africa. The second question was what are the most common mental health conditions among women in sub-Saharan Africa who have experienced SV, compared with women who have not experienced SV.

Data synthesis

Meta-analyses were performed to estimate the pooled OR of poor mental health among women with a history of SV compared with those without a history of SV by synthesising individual results. Only studies that reported the adjusted ORs were included in the meta-analysis to adjust for the effect of other predictor variables and to see the adjusted pooled estimate. A random-effects model was used, as we expected high heterogeneity due to the large methodological variations in the included observational studies.

A random effects meta-analysis assumes variance in the effect across different studies explained by real differences in effect, as well as by chance.¹⁸ Heterogeneity is reported using the I² statistic. The I² statistic informs us what portion of the total variance in the effect size is caused by variance between the studies. As previously suggested, an I² statistic of above 75% implies considerable heterogeneity, while an I² statistic below 40% is not considered a concern.¹⁹

As stated in the protocol, the study planned to assess differences in the most common mental health conditions among sub-Saharan African women who reported a history of SV and women who did not report experiences of SV. Studies that reported the proportion or number of participants that experienced any type of mental health condition among both groups of women with and without history of SV were identified. If a specific outcome was reported by at least two studies, it was included in the pooled prevalence estimate. Finally, the prevalence of different mental health conditions was calculated to identify the most common mental health conditions among SV victims and among those with no history of SV.

To compare the association of mental health conditions with experience of SV, the pooled OR of the poor mental health outcomes (PTSD, depression, anxiety, common mental disorders (CMDs) (mixed anxiety and depression), suicidal behaviour (suicidal ideation, plan or attempt) and emotional/psychological distress) was calculated. All meta-analyses were conducted with Stata V.18 software.

Results

Overview of the reviewed studies

The database search yielded 12 364 articles, and 3458 duplicates were removed. The titles and abstracts of 8906 articles were reviewed in Covidence. After excluding 8725 studies based on title and abstract review, we reviewed the full text of 181 articles. From this group, we excluded 105 articles for several reasons described in (figure 1) below. We identified a final sample of 76 articles that met all inclusion criteria and the quality appraisal using JBI’s critical appraisal tools.

Characteristics of studies (n=76)

This systematic review yielded 66 cross-sectional, 9 cohort and 1 case–control study, with a sample of 80 313 women from a total of 17 sub-Saharan African countries. The sample sizes in the included studies ranged from 100 to 7101 women. Of all articles, 16 (21%) studies were conducted in South Africa, followed by 14 (18%) studies each from Uganda and 11 (14%) in Kenya. The remaining studies were completed in Ethiopia (n=9, 12%), Tanzania (n=6, 8%), Democratic Republic of the Congo (n=3, 4%), Nigeria (n=3, 4%), Cameroon (n=2, 3%), Malawi (n=2, 3%), Rwanda (n=2, 3%), Burkina Faso (n=2, 3%), Ghana (n=2, 3%), Zambia (n=1, 1%), The Gambia (n=1, 1%), Zimbabwe (n=1, 1%), Botswana (n=1, 1%) and Lesotho (n=1, 1%).

Of all studies included in this systematic review and meta-analysis, 27 studies reported two or more outcome variables. Two outcomes were reported by 17 studies, while 3 and 4 outcomes were reported by 6 and 4 studies, respectively. The remaining 49 studies reported only one outcome. Most studies (n=47, 62%) reported depression as an outcome variable, followed by PTSD (n=23, 30%). Suicidal behaviour was reported by 18 studies (24%), while 9 studies (12%) each reported anxiety and CMD outcomes. CMD is used when anxiety and depression were reported together in the original studies. Emotional distress was also reported by 5 (7%) studies. Table 1 presents the summary characteristics of the studies and key outcome findings (online supplemental information file 1).

Table 1. Descriptive summary of studies included in the systematic review and meta-analysis of sexual violence and poor mental health among women victims in sub-Saharan Africa, 2023.

Authors	Country	Sample size	Any SV, N (%)
Abelson et al7	Cameroon	2165	713 (33)
Abrahams et al71	South Africa	885	14 (1.6)
Agardh et al6	Uganda	279	68 (24.3)
Amone-P’Olak et al20	Uganda	181	119 (66)
Amone-P' et al10	Uganda	210	135 (65)
Ashenafi et al12	Ethiopia	3015	332 (11)
Asiimwe et al72	Uganda	100	45 (45)
Ayele et al43	Ethiopia	409	93 (22.7)
Barchi et al44	Botswana	469	40 (10)
Beksinska et al21	Kenya	1003	306 (31.3)
Belay et al45	Ethiopia	589	56 (9.5)
Belete et al13	Ethiopia	988	123 (12.5)
Belete et al11	Ethiopia	990	123 (12.4)
Brar et al46	Malawi	1000	454 (46)
Cange et al14	Burkina Faso	696	285 (40.9)
Cherenack et al47	Tanzania	135	57 (42.5)
Cohen et al78	Uganda	2022	526 (26)
Decker et al48	South Africa	224	41 (18.3)
Donald et al49	Uganda	385	20 (5)
Dossa et al22	DRC	320	117 (36.56)
Ejigu et al50	Ethiopia	826	103 (12.5)
Erulkar et al51	Ethiopia	4495	744 (17)
Familiar et al23	Uganda	580	454 (79.5)
Groves et al85	South Africa	1500	39 (2.78)
Gust et al86	Kenya	461	121 (34.6)
Hansrod et al24	South Africa	169	80 (47)
Issahaku79	Ghana	443	151 (34)
Jewkes et al25	South Africa	3005	1727 57.4)
Kaminer et al26	South Africa	138
Kapiga et al87	Tanzania	1049	355 (34.8)
Kinyanda et al52	Uganda	694	7.80
Kisaakye et al80	Uganda	1795	933 (52)
Kinyanda et al98	Uganda	894	168 (18.6)
Lambert and Denckla27	Kenya	301	108 (35.9)
Larsen et al53	Kenya	3555	199 (5.6)
Lee et al28	Burkina Faso	681	130 (19)
Lee et al81	Nigeria	1766	616 (36.3)
Leis et al54	Kenya	220	151 (68.6)
Luo et al29	Uganda	1415	191 (13.7)
Machisa et al30	South Africa	1292	97 (8)
Magnusson et al55	Tanzania	1013	165 (16.3)
Mahenge et al31	Tanzania	1180	236 (20)
Mahenge et al56	Tanzania	500	58 (11.6)
Manongi et al57	Tanzania	1116	212 (19)
Mathur et al73	Kenya	1778	381 (21.4)
Mathur et al73	Zambia	1915	437 (22.8)
Mhlongo et al32	South Africa	157	38 (24.2)
Morof et al33	Uganda	117	84 (63.3)
Myers et al58	South Africa	499	34 (6.8)
Nabayinda et al59	Uganda	542	136 (25.1)
Nduna et al60	South Africa	1415	219 (15)
Nöthling et al34	South Africa	134	79 (59)
Okafor et al61	South Africa	981	65 (6.6)
Ogban et al74	Nigeria	250	73 (29.2)
Parcesepe et al75	Cameroon	230	71 (31.3)
Pellowski et al62	South Africa	831	55 (7.1)
Peltzer et al35	South Africa	268	109 (40.7)
Picchetti et al82	Lesotho	7101	1349 (19)
Roberts et al36	Kenya	283	124 (43.8)
Rogathi et al63	Kenya	1013	159 (15.7)
Rurangirwa et al37	Rwanda	921	89 (9.7)
Samia et al64	Kenya	215	38 (17.7)
Schwartz et al38	South Africa	299	154 (52)
Sekoni et al39	Nigeria	550	49 (8.9)
Shamu et al65	Zimbabwe	842	335 (39.8)
Sherwood et al66	The Gambia	251	70 (29)
Tenkorang8	Ghana	2282	614 (26.9)
Tilahun et al83	Ethiopia	615	34 (5.5)
Tsai et al67	Uganda	173	26 (15)
Umubyeyi et al40	Rwanda	447	71 (17.4)
Verelst et al68	DRC	1304	499 (38.2)
Villaveces et al84	Malawi	1029	224 (21.8)
Wado et al69	Kenya	2106	6.40
Wa Mwenda et al41	DRC	3011	2302 (76.5)
Ward et al42	South Africa	2497	454 (18.19)
Winter et al70	Kenya	361	152 (42.1)
Woldetsadik et al76	Ethiopia	743	34 (4.6)

DRCDemocratic Republic of the CongoSVsexual violence

Sexual violence

In this study, 80 313 women were included, of which 19 860 (25%) women had experienced at least one type of SV in their lifetime. Additionally, 7073 (36%) of women who had ever experienced SV had experienced SV over the last 12 months. Among all SV victims, 7070 (36%) experienced non-rape SV, 2556 (13%) experienced rape only and 10 234 (52%) experienced both rape and another type of SV.

Poor mental health and SV

The most common mental health conditions reported in studies included in this systematic review were PTSD, depression, anxiety, CMD (depression and anxiety), suicidal behaviour (ideation, plan or attempt) and emotional distress. The prevalence of all types of mental health conditions reported was higher among SV victims compared with women without a history of SV. For instance, depression was 69% among SV victims and 21% among non-SV victims (figure 2).

Figure 2. ‘Prevalence of mental health conditions among women without experience of SV’ and ‘prevalence of mental health conditions among women with experience of SV’. CMD, common mental disorder; PTSD, post-traumatic stress disorder; SV, sexual violence.

The meta-analysis results suggest that sub-Saharan African women who have experienced lifetime SV are more likely to have poor mental health compared with women who have not experienced SV. The pooled ORs for PTSD, depression, anxiety, CMD, suicide and emotional distress all indicate that women who have experienced SV are over two or more times likely to report symptoms of poor mental health compared with the control group (see figure 3).

Figure 3. Summary of the pooled OR estimates of mental health conditions among women with lifetime experience of SV compared with women who have not experienced. CMD, common mental disorder; I², heterogeneity statistic; PTSD, post-traumatic stress disorder; SV, sexual violence.

Post-traumatic stress disorder

A total of 23 studies reported an association between lifetime SV and PTSD as an outcome.20,42 Excluding one study,²⁴ all studies demonstrated a positive association between lifetime SV and PTSD. A total of 11 studies with adjusted ORs were selected for the meta-analysis,^{20 23 24 27 30 31 33 34 37 40 41} and the meta-analysis of PTSD across 11 studies yielded a pooled OR of (OR 2.75; 95% CI 1.96 to 3.87; I²=73.4%) indicating that women who had experienced SV were over two times more likely to experience PTSD compared with those with no experience of SV.

A subgroup analysis was conducted to examine the effect of the setting for the primary data collection in each study (eg, was the primary data collected during antenatal care (ANC), in a conflict setting, from specific population groups experiencing different forms of inequity or from the general population). The subgroup meta-analysis showed that the association between SV and PTSD was most prevalent among antenatal populations (OR 3.64; 95% CI 1.34 to 9.83; I²=78.5%), with studies conducted in conflict settings showing the next strongest association (OR 2.66; 95% CI 1.43 to 4.94; I²=72.0%).

Given the observed heterogeneity, we conducted meta-regression and sensitivity analyses to explore potential sources of variability across studies. No moderators showed significant associations with PTSD, indicating that the heterogeneity may be random or due to other unmeasured factors. However, sensitivity analysis showed that excluding a case–control study reduced heterogeneity to a moderate level without affecting pooled estimates, confirming the robustness of the findings (online supplemental information file 2).

Depression

Depression was the most common mental health condition reported, being described in 47 of the articles6,811 12 21 23 25 27 included in this systematic review. Meta-analysis of the association between lifetime SV and depression, which included the 28 studies811 23 29,31 37 40 43 44 47 that reported adjusted ORs, revealed a pooled OR of experiencing depression (OR 2.38; 95% CI 2.05 to 2.78; I²=56.6%). That is, women who had experienced SV had more than twice the odds of experiencing depression than women with no history of SV.

Similarly, a subgroup analysis was conducted to examine the effect of the setting of the primary data collection. The subgroup meta-analysis showed the association between SV and depression to be strongest for studies conducted during the antenatal period (OR 3.50; 95% CI 2.43 to 5.04; I²=57.6%), followed by the postnatal period (OR 2.89; 95% CI 1.76 to 4.73; I²=,60.5%) and in conflict settings (OR 2.60; 95% CI 1.24 to 5.46; I²=79.9%).

Anxiety

In this systematic review, nine studies reported anxiety outcomes,^{6 31 37 40 42 50 54 58 68} and all the studies reported a positive association between lifetime SV and anxiety. Five studies^{31 37 40 50 54} with adjusted ORs were included in the meta-analysis, and the pooled odds of SV victims experiencing anxiety were 2.81 (OR 2.81; 95% CI 1.67 to 4.71; I²=77.21%) times higher than for women who did not report exposure to SV. A subgroup analysis was not conducted for anxiety and the other mental health conditions presented below due to an insufficient number of studies.

Common mental disorders

This review operationally defined CMDs as depression and anxiety, as reported in the reviewed articles. In some articles, depression and anxiety were not reported separately and are included in this review as CMD. A total of 10 articles1021 70,77 reported CMD, and all studies reported a positive association between lifetime SV and CMDs. The meta-analysis of the five studies^{10 21 70 75 76} with adjusted ORs demonstrated that the odds of reporting CMD were (OR 2.12; 95% CI 1.70 to 2.64; I²=0.00%) times higher among SV victims compared with women who had not experienced SV.

Suicidal behaviour (ideation, plan or attempt)

Suicidal behaviour, including ideation, plan and attempt, was reported by 18 articles,1314 21 28 30 37 40 48 52 65 70 78,84 and all studies, except one,⁶⁵ found a positive association between lifetime SV and suicidal behaviour. 14 studies1321 30 37 40 48 52 65 70 78 80 82,84 with adjusted ORs were included in the meta-analysis, which revealed that the OR of suicidal behaviour is (OR 2.44; 95% CI 1.92 to 3.10; I²=68.0%) times higher among women who had experienced SV compared with those who had no experience of SV.

Emotional distress

Five studies reported emotional distress outcomes,8182 84,86 and all these studies reported a positive association between lifetime SV and emotional distress. The meta-analysis included the five studies as they reported adjusted ORs, and the pooled odds of emotional distress were found to be (OR 3.14; 95% CI 1.73 to 5.69; I²=79.6%) times higher among women with histories of SV compared with women without histories of SV.

Other mental health conditions

Only a few studies reported other types of mental health conditions. Two studies reported psychotic symptoms,^{6 10} one study assessed poor mental health,⁸⁷ one study reported conduct disorder and somatic complaints,¹⁰ and one study examined sleep deprivation.⁷⁹ A study showed that exposure to sexual coercion ever in one’s lifetime was significantly associated with high psychotic symptom scores.⁶ The other study revealed that SV victims with no children were more likely to report symptoms of psychotic disorder (OR 1.38; 95% CI 0.78 to 2.47) compared with those women who did not report experiencing SV.¹⁰ SV was also associated with increased reporting of symptoms of poor mental health (OR 2.8; 95% CI 1.9 to 4.1) compared with those women who did not report SV.⁸⁷ Similarly, women with a history of SV and with children as a consequence of SV were over six times more likely to report somatic complaints (OR 6.59; 95% CI 1.80 to 24.11), compared with those women who did not report SV.¹⁰ Lastly, women who experienced SV had a higher odds of reporting sleep disruption (OR 2.5; 95% CI 1.28 to 4.88) compared with those who had no experience of SV.⁷⁹

Discussion

This comprehensive systematic review and meta-analysis of 76 observational studies included a total of 80 313 women participants, among whom one-quarter (25%) had experienced SV. We found consistent small-to-medium associations between a history of SV and a variety of more common mental health conditions.

The pooled prevalence of all mental health conditions in this review was higher among SV victims compared with the prevalence among women without a history of SV. Depression was the most prevalent mental health condition described (69%), followed by CMD, and then anxiety with prevalence estimates of 64% and 44%, respectively, among women with a history of SV. Health service providers should, therefore, give special attention to screening for common mental health conditions in patients who present with a history of SV. Women SV victims who visit health facilities or other service centres, such as one-stop centres for survivors of gender-based violence, should also be screened for poor mental health so that they can receive effective support services.

Findings from the random effects meta-analyses revealed a strong association between SV and all mental health outcomes, although the strength of the associations slightly varied across the different outcomes. Overall, the pooled ORs suggest that women who experienced SV are more than twice as likely to experience PTSD (OR 2.75; 95% CI 1.96 to 3.86), depression (OR 2.38; 95% CI 2.04 to 2.77), anxiety (OR 2.81; 95% CI 1.67 to 4.72), CMD (OR 2.12; 95% CI 1.70 to 2.64), suicidal behaviour (OR 2.44; 95% CI 1.92 to 3.10) and emotional distress (OR 3.14; 95% CI 1.73 to 5.69) compared with women who had not experienced SV. These findings align with previous systematic reviews and meta-analyses, conducted mostly from outside of Africa and among populations of both women and men, which demonstrate that SV is associated with multiple mental health conditions.^{88 89} However, the strength of the association varies. For example, Dworkin’s meta-analysis of studies conducted internationally examined the association between poor mental health and sexual assault and reported substantially higher pooled ORs of PTSD (7.57) and depression (3.44), compared with this study’s findings.⁸⁸ This could be due to differences in sample characteristics, diagnostic criteria, measures of SV and poor mental health, assessment procedures, and study period.

A subgroup analysis of depression and PTSD by the study setting where the data collection took place revealed that ever experience of SV reported during the antenatal period is the strongest predictor of PTSD and depression. Indeed, this analysis indicates that women who experienced SV ever in their lifetime and reported during pregnancy were more than three times more likely to experience PTSD (OR 3.64; 95% CI 1.34 to 9.83) and depression (OR 3.50; 95% CI 2.43 to 5.04), compared with women who did not report SV. This supports previous studies which found that SV is strongly associated with depression and PTSD during the perinatal period.^{90 91} One possible explanation could be that pregnancy and the postpartum period already have an association with psychological morbidity and are therefore a time of high vulnerability that could trigger or intensify poor mental health issues, particularly for those with a history of SV.⁹² Another possible explanation could be that SV victims’ concerns about the impact on their unborn child and potential adverse neonatal outcomes might exacerbate psychological morbidity.⁹³ Furthermore, when SV is the cause of a pregnancy, it is reasonable to hypothesise that this may further elevate the risk of poor mental health during the pregnancy. Our findings align with previous studies indicating that a history of lifetime SV was associated with depression during the perinatal period⁹⁴; for example, a meta-analysis by Lombardi et al found that women with a history of SV had a 51% higher odds of postpartum depression.⁹⁰ This implies a need for health service providers to integrate SV and mental health screenings into ANC, which has previously been a very limited practice in sub-Saharan African countries.⁹⁵ Similarly, the subgroup analysis in our study indicated that poor mental health conditions may be pronounced among marginalised groups of women, such as those from conflict settings and those from specific population groups experiencing different forms of inequity.⁹⁶ For women living in conflict settings, the compounded trauma, disruption of support systems and limited access to protective services may exacerbate the impacts of SV on their mental health conditions. Additionally, the study findings indicate that women from specific marginalised population groups, such as refugees, women from informal settlements or sex workers, may face even greater challenges in terms of their mental health conditions. These intersecting social positions and forms of structural inequity could compound vulnerability to poor mental health conditions on top of SV. As such, it is critical to recognise the complexity of these issues and design inclusive approaches to provide a comprehensive mental health service for marginalised women survivors of SV in sub-Saharan Africa.

In the current study, the strength of the association between the type of mental health condition and SV varied by the type of SV and type of perpetrator. For instance, the meta-analysis results indicated that women who experienced both rape and an additional type of SV had the highest pooled OR for PTSD (OR 3.02; 95% CI 1.73 to 5.27) compared with women who did not report experiencing SV. These findings imply that service providers should be particularly alert to the risk of mental ill health for women who have experienced multiple types of SV (both rape and an additional type of SV), with a focus on ensuring access to support, diagnosis and treatment.

Strengths and limitations

This systematic review and meta-analysis had several strengths, including its comprehensive search strategy and rigorous article screening process. For instance, the methodological quality of all the included studies was strictly assessed by all the reviewers, and studies that did not use internationally validated tools to measure the exposure and outcome variables were excluded. Additionally, this meta-analysis considered the adjusted ORs to estimate pooled effects, and moderators were adjusted.

However, this review has several limitations. First, most studies used cross-sectional designs, preventing causal inferences about the relationship between SV and poor mental health. Therefore, longitudinal research is needed to determine the direction of this association. Future research should also include qualitative studies to better understand the association between SV and mental health, as well as the role that country context plays.

Second, best practice when measuring the prevalence of SV is to ask respondents direct questions about specific acts of SV over a defined period, rather than using questions about whether or not the respondent has been ‘raped’ or ‘sexually abused’.⁴ Interpretations about what constitutes rape or abuse can vary between individuals and cross-culturally, which may lead to underreporting.⁴ While most studies reached the required threshold for the methodological quality of the measurement of SV, some studies^{13 21 71 86} used measures and instruments that would have led to under-reporting. It should also be noted that studies using remote methods due to the COVID-19 pandemic (eg, phone interviews) likely resulted in underreporting due to safety and confidentiality concerns.⁵⁰ Future SV research using remote methods should draw from the emerging evidence base about best practice approaches to remote data collection, to ensure the safety of respondents and data quality.⁹⁷

Third, while the search strategy was extensive, this review also only included English language articles, so relevant studies in other languages may have been missed.

Finally, this study focused on a quantitative synthesis, which resulted in the absence of qualitative data that could have potentially offered deeper insights into the lived experiences of survivors.

Implications for policy, practice and research

Policy

Policy-makers in sub-Saharan Africa should be aware of the burden of mental ill health among SV victims and support a health systems response to violence against women. This includes developing national action plans that include ‘health system actions, budgets and staffing’. These plans should recognise and address the complex intersection of mental health and SV. Furthermore, policy-makers should develop necessary strategies, guidelines, standard operating procedures, and management protocols and standard screening tools considering the most common poor mental health outcomes among SV victims. Developing standard screening tools for diagnosing mental health issues should involve engaging senior clinicians with expertise in both SV and poor mental health management. These tools can then be approved by national policy-makers. Once approved, the tools should be distributed to service-providing institutions, accompanied by proper training for healthcare providers on how to use them effectively to diagnose mental health conditions among SV survivors.

Practice

The integration of comprehensive SV management services into existing service outlets, such as ANC clinics, is vital to expand accessibility of services. Antenatal clinics should be used to screen SV and mental ill health to link for appropriate services. Service providers should also be well trained in mental ill health diagnosis, management and referral for SV victims. Collaboration across sectors—mental health, SV, domestic and family violence—is vital to offer victim-survivors effective care. Additionally, campaigns to raise awareness, reduce stigma and promote mental health support in underserved settings should be prioritised. Most importantly, societal and educational interventions are needed to prevent all forms of violence against women from happening in the first place.

Research

Future research in sub-Saharan Africa should involve the use of longitudinal study designs to understand the risk factors and long-term impacts of SV. Future systematic review research should focus on qualitative studies to better understand the association between SV and mental health, as well as the role that country context plays.

Conclusions

This systematic review and meta-analysis demonstrated that exposure to SV was consistently associated with a range of mental health conditions among women in sub-Saharan Africa. Lifetime experience of SV reported during antenatal periods and in conflict settings appeared to confer an even stronger association with poor mental health. Experiences of multiple types of SV (both rape and non-rape) also strengthened the association with poor mental health.

Thus, policy-makers should recognise the common mental health conditions among women who have experienced SV and develop appropriate strategies and screening tools for all service delivery points. Additionally, mental health service providers and health practitioners should prioritise screening for these conditions in patients with a history of SV, integrating SV and mental health screening tools into ANC services.

The contents of this research article only reflect the authors’ opinions but do not show interest/s of the organisation involved in the funding.

Data availability statement

Data are available on reasonable request.