Skip to main content

Prevalence and health outcomes of domestic violence amongst clinical populations in Arab countries: a systematic review and meta-analysis



An estimated 30% of women worldwide experience intimate partner violence (IPV) during their lifetime. Exposure to IPV is associated with poor health outcomes and the prevalence of violence may be higher amongst women seeking healthcare. Existing evidence from the Arab region is limited. We conducted a systematic review and meta-analysis of prevalence and health outcomes of domestic violence (IPV or violence from a family member) in clinical populations in Arab countries.


Using terms related to domestic violence, Arab countries, and date limit > year 2000, we searched seven databases: Medline, EMBASE, PsycINFO, CINAHL, Web of Science: core collection, IBSS, Westlaw, IMEMR. We included observational studies reporting estimates of prevalence or health outcomes of domestic violence amongst women aged > 15 years, recruited while accessing healthcare in Arab countries. Studies that collected data on/after 1st January 2000 and were published in English, Arabic or French were included. Title/abstract screening, full text screening, quality assessment and data extraction were carried out. Extracted data were summarised and meta-analysis was performed where appropriate.


6341 papers were screened and 41 papers (29 studies) met inclusion criteria. Total 19,101 participants from 10 countries were represented in the data. Meta-analysis produced pooled prevalence estimates of lifetime exposure to any type of IPV of 73·3% (95% CI 64·1–81·6), physical IPV 35·6% (95% CI 24·4–47·5), sexual IPV 22% (95% CI 13·3–32) and emotional/psychological IPV 49·8% (95% CI 37·3–62·3). Domestic violence (IPV or family violence) exposure was associated with increased odds of adverse health outcomes: depression OR 3·3 (95% CI 1·7–6·4), sleep problems OR 3·2 (95% CI 1·5–6·8), abortion OR 3·5 (95% CI 1·2–10·2), pain OR 2·6 (95% CI 1·6–4·1) and hypertension OR 1·6 (95% CI 1·2–2·0).


Domestic violence is common amongst women seeking healthcare in Arab countries. Exposure to domestic violence is associated with several poor health outcomes. Further research into domestic violence in the Arab world is required.

Trial registration

Systematic review protocol was registered on PROSPERO: CRD42017071415.

Peer Review reports


Domestic violence is a worldwide epidemic with an estimated 30% of women experiencing physical and/or sexual violence from an intimate partner during their lifetime [1]. Domestic violence (DV) includes violence perpetrated by a family member or intimate partner towards another adult. Much of the current international evidence focuses on intimate partner violence (IPV), which is a subset of domestic violence. Domestic violence may be a single incident or pattern of incidents which can take multiple forms including physical, sexual, psychological, emotional, financial and control violence [2]. A severe violation of human rights, domestic violence has societal and economic costs and is increasingly recognised as a clinical and public health issue.

There is a lack of robust prevalence data on domestic violence from the Arab region. A 2008 systematic review of IPV in the Middle East and North Africa highlighted ‘the very limited knowledge base on the topic’ in the region (p67) [3]. During the period 1992–2002 just 12 prevalence studies, from community or clinical settings, were identified. Estimated prevalence of lifetime exposure to IPV in these studies ranged from 8·1% to 64·6% [3]. The WHO estimated that 37% of women in the Eastern Mediterranean Region (EMR) have ever experienced physical or sexual IPV [4]. This estimate was generated from just 18 available studies compared to 114 studies used to produce the high-income countries estimate. A few countries have collected community-level prevalence data as part of large demographic health surveys; for example 30.3% of ever-married Egyptian women have experienced spousal violence, 18·6% within the last year [5]. Smaller cross-sectional studies have been conducted in community or clinical settings in some countries [6,7,8,9].

Intimate partner violence has well documented mental health, reproductive health and chronic physical health consequences [4, 10]. Two previous reviews from this region summarised the health outcomes of IPV as reported by individual clinical or community-based studies; meta-analysis was not performed [3, 11]. Evidence suggests that the prevalence of IPV may be higher amongst clinical populations, [12] but there is a lack of synthesised data internationally. We found no other studies that produced pooled prevalence estimates for clinical populations in the Arab region.

Domestic violence must be understood in its sociocultural context [13]. Traditional Arab society values women as wives and mothers within a patriarchal family structure, and family cohesiveness might be prioritised over individual rights and freedoms [14]. Despite progress in female education, with gender parity achieved in school-age education and women in some countries surpassing men in tertiary education, this has not necessarily translated into female economic empowerment or shifting domestic power dynamics [14,15,16]. Women’s subordinate status is reinforced by legal and political systems. All countries in the Arab region, except Sudan and Somalia, have now ratified CEDAW (Convention on the Elimination of all forms of Discrimination Against Women), but national responses to domestic violence have been fragmented [16, 17]. Some countries have criminalised domestic violence, but elsewhere it remains legal. In most Arab countries marital rape is not a crime. Even where penal legislation exists, survivors of domestic violence face barriers to accessing justice. Personal status law also discriminates against women with unequal access to divorce, child custody and property rights [16]. Multiple factors including family expectations, economic dependence and weak legal systems might combine to prevent women from seeking help.

With no clear consensus on which countries form the Middle East region, we chose to focus on Arab countries. These countries share sociocultural norms, making it suitable to compare and combine DV studies. Much existing international and regional prevalence data comes from community studies which provide best estimates of average prevalence. Given the health impacts of DV, women accessing healthcare may be more likely to have experienced DV. As we are particularly interested in DV as healthcare problem, it is useful for clinicians and health policy makers to have data on the prevalence and health impacts of DV within the population of women accessing healthcare. We therefore focused our review on clinical populations only. Our review aims to answer these research questions: 1. What is the prevalence of DV in clinical populations in Arab countries? 2. What are the health outcomes of DV in clinical populations in Arab countries?


Search strategy and selection criteria

We included studies reporting prevalence or health outcomes of DV against women aged > 15 years. Population included residents of Arab countries (Bahrain, Egypt, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, Occupied Palestinian Territory and Arab residents within Israel, Oman, Qatar, Saudi Arabia, Somalia, Sudan, Syrian Arab Republic, Tunisia, United Arab Emirates, Yemen) recruited whilst accessing healthcare. Included: observational studies, data collected on/after 1st January 2000, published in English, French or Arabic. Excluded: study population primarily migrants from non-Arab country, experimental, interventional or qualitative studies and abstracts/conference proceedings. Studies reporting health outcomes were included only if outcomes were separated or compared by violence exposure.

We included studies looking at all types of DV including physical, sexual, psychological, emotional, financial, control, any violence or other types. As there is no gold standard definition or measurement tool for DV prevalence, we accepted the definition/measurements used by primary studies. We included violence perpetrated by family members and/or intimate partners. Perpetration of violence within a family might occur between adult children and their parents (in either direction), from siblings, in-laws or other adult family members. Violence towards children is considered child abuse and is not included within our definition of domestic violence; we included female victims aged 15 and over, in line with the WHO (World Health Organization) approach [4]. Subsequently, we use the term DV to refer to violence from a family member and/or intimate partner and the term IPV to refer to violence from an intimate partner only.

We searched seven databases on 67th April 2017: Medline, EMBASE, PsycINFO, CINAHL, Web of Science: core collection, IBSS, Westlaw and IMEMR. Search terms used related to violence against women (VAW) and Arab countries (see Additional file 1 for full search strategy). Initial searches included all forms of VAW, study settings and EMR countries. We subsequently narrowed our focus to studies on DV, clinical populations and Arab countries only. Grey literature search included database searches (OpenGrey and ProQuest) and Google searches of government and agency reports. Backward (reference list) and forward citation (Google Scholar) searches were conducted and contact was made with authors of all included studies.

Searches, double title/abstract screening and full text review were conducted by CH and HE using EndNote. Disagreements were arbitrated by GF. French articles were also reviewed by GF.

Data analysis

All data were extracted independently by two reviewers onto Microsoft Excel. Data extracted included study design, sampling method, response rate, sample size, inclusion/exclusion criteria, DV/IPV identification tool used, participant details, violence definition and exposure timescale. For prevalence data we recorded the number and percentage of participants exposed to different types of DV including any violence, physical, sexual, psychological or emotional, control, economic, or other as defined by the author. This was separated by exposure timescale: last 12 months, lifetime or during pregnancy. Health outcomes were as defined by the primary studies. Health outcome data were extracted separately for DV survivors and unexposed women.

Multiple papers from a single study were included, but only data from the paper with the largest sample or most complete data underwent analysis for each data point. For non-reported, unclear or inconsistent data, we contacted authors to seek clarification. For a small number of studies where we were unable to clarify key study design characteristics, we made the following assumptions: perpetrator of violence was IPV, timescale of abuse was lifetime, sampling method was convenience, residence was 100% urban if the clinic setting was urban. Where absolute numbers and percentages were inconsistent or could not be deduced, we used absolute numbers. There were small number of results that were judged, by two authors, to be invalid or non-comparable and were excluded from analysis. For example, a physical violence prevalence estimate which included sexual violence and child-maltreatment in the definition was not comparable to physical violence prevalence measurements from other studies. The measurement of mental health outcomes in a study which excluded patients on psychotropic medication was judged to be invalid. Where results are marked ‘N/A’ in the table in Additional file 2, this indicates that data were not considered appropriate for inclusion.

Study-level risk of bias was assessed by two authors (CH and JU) using the 20-point AXIS tool for cross-sectional studies [18]. Studies were given a global rating of poor, medium or high quality, with disagreements arbitrated by GF. One approach to variable study quality in meta-analysis is to include all studies and perform statistical adjustment based on quality assessment. As the AXIS tool does not give a numerical weighting of sources of bias, we decided that it was analytically more robust to exclude, rather than adjust for, poor-quality studies. All studies judged to be of poor-quality by two authors were therefore excluded from our review and analysis.

Prevalence studies were grouped by exposure timescale and violence type. Meta-analysis was performed if there were at least five studies in a group [19]. IPV and DV were pooled separately. For any violence type we only meta-analysed studies if they measured all three main violence types: physical, sexual and emotional/psychological. We fitted a random-effects model because exploratory analyses indicated substantial study heterogeneity. Prior to pooling, data were transformed using Freeman-Tukey double arcsine transformation to stabilize the variances and meet normality assumptions of the outcomes [20].

Sensitivity analysis was performed by assessing the impact of removing outlying studies on heterogeneity. We performed univariate meta-regression if there were at least five studies with relevant data: with continuous demographic covariates (age, literacy, employment and % urban population), and binary covariates (high or low/middle-income country, refugee or non-refugee population, Palestinian or non-Palestinian population, community clinic or hospital setting, women’s or general health setting, questionnaire or interview data on DV, validated or non-validated DV survey tool, short [1–9 questions] or long [10 or more questions] DV survey tool) provided there were at least two studies in each category.

We summarised health outcomes as presented by the primary studies. We generated pooled odds ratios for dichotomous health outcomes (e.g. abortion), where at least five studies provided adequate data [19]. Studies were not separated by DV or IPV, exposure timescale or violence type for the health outcome meta-analysis. Violence exposure was any violence type if data were available, otherwise the most prevalent violence type in that study. For univariate meta-regression, binary covariates were high or low/middle-income country and health outcome diagnostic method (validated tool or not); continuous demographic covariates used were as listed above.

Funnel plots were not used to assess for publication bias due to substantial study heterogeneity [21].

Analyses were performed using R function metaphor.

The systematic review protocol was registered on PROSPERO 21/08/2017 - CRD42017071415.

The funders had no role in study design, data collection/analysis/interpretation or report writing.


Searches identified 6341 papers of which 41 met inclusion criteria (Fig. 1), representing 29 studies [22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62]. Additional file 2 summarises study details and main findings from all included papers.

Fig. 1

PRISMA inclusion flowchart

Data representing 19,101 participants were included in our review. Study sample size ranged from 82 to 3271, median 394. Included studies were from ten countries. Nineteen studies collected prevalence data by interview; ten used self-administered questionnaires. Seventeen studies used or modified existing validated tools for measuring domestic or intimate partner violence: 7 WHO, 3 NorAQ (NorVold Abuse Questionnaire), 2 AAS (Abuse Assessment Screen), 2 HITS (Hurt, Insult, Threaten and Scream), 1 WAST (Woman Abuse Screening Tool), 1 CTS-R (Conflict Tactics Scale, revised) and 1 USPSTF (U.S. Preventative Services Task Force) family screening tool for IPV. Twelve studies used tools which were developed independently or based on existing literature. Further detail about measurement tools is provided in Additional file 2. Most studies focused on intimate partner violence: 19 studies reported IPV only, 6 studies reported DV and 4 studies gave separate results for both IPV and DV. Further details about the perpetrators of domestic violence are included in Additional file 3. Nineteen studies reported health outcome data.

Demographic details were variably reported, with data unavailable from some studies (Table 1). Data were converted to enable comparison where possible. Participant age ranged 15–68 years. Most studies only included married/ever-married women; four included single women and one included engaged women.

Table 1 Demographic details as per individual studies, with average across studies

Prevalence of DV varied between studies with wide ranging estimates for all violence types. Results are summarised by timescale of exposure below (Tables 2, 3 and 4). Descriptive summary statistics refer to IPV and DV combined. Meta-analysis results are presented separately for IPV or DV.

Table 2 12-month prevalence
Table 3 Lifetime prevalence
Table 4 Prevalence during pregnancy

Meta-analysis generated pooled prevalence estimates of lifetime exposure to any IPV of 73·3% (95% CI 64·1–81·6), physical IPV 35·6% (95% CI 24·4–47·5), sexual IPV 22% (95% CI 13·3–32) and emotional/psychological IPV 49·8% (95% CI 37·3–62·3). Forest plots for lifetime prevalence are presented below in Fig. 2. Forest plots for additional meta-analyses are available in Additional file 4.

Fig. 2

Forest plots for lifetime prevalence meta-analyses

For each meta-analysis, removal of outlying studies had little impact on the I2 statistic, so all studies were retained. Univariate meta-regression showed no consistent association between study level covariates (e.g. healthcare setting or literacy) and prevalence. Associations were detected for some analyses only: prevalence of any lifetime violence was higher in general health than women’s health settings (82% vs 66%, p = 0·01), prevalence of lifetime economic violence was higher in community clinics than hospitals (45% vs 33%, p = 0·03). Lifetime physical IPV prevalence was higher in studies using non-validated rather than validated questionnaires (52·7% vs 27·7%, p = 0·02) but for lifetime economic IPV, the opposite was found (36% vs 48%, p = 0·048). Lifetime physical DV prevalence was higher in studies with refugee populations (45%) than non-refugees (23%) (p = 0·041). For lifetime psychological IPV, studies with higher literacy reported higher prevalence of violence.

For health outcome data, violence exposure varied by perpetrator (IPV or DV), type of violence, and timescale. The measurement of health outcomes varied between self-reported, clinically diagnosed or measured using a screening/diagnostic tool. Additional file 5 summarises the health outcome data presented by primary studies. Most studies collected binary or categorical data and reported unadjusted odds ratios (OR); some used other statistical measures. Odds ratios were recalculated from absolute figures where available. Pooled effect estimates were produced for the five health outcomes for which there were sufficient data: depression, sleep, abortion, pain and hypertension (Table 5). Studies were combined in meta-analysis regardless of whether the study measured IPV or DV, length of exposure timescale or violence type.

Table 5 Health outcome meta-analysis results

Health outcomes were grouped into mental health, reproductive health and general health. Depression was the mental health outcome with the most consistent evidence: seven studies all found a significant association with DV. Meta-analysis of five of these studies produced a pooled OR of 3·3 (95% CI 1·7–6·4) (Fig. 3). Seven studies considered sleep problems with five finding significant associations. Pooled OR for five of the seven studies was 3·2 (95% CI 1·5–6·8) (Fig. 4). Other mental health outcomes found to be associated with domestic violence by some studies were suicidal thoughts, suicide attempts, anxiety, anxiety/depression, postnatal depression, stress, memory problems, distress and anguish. See Additional file 5 for more detail.

Fig. 3

Depression forest plot

Fig. 4

Sleep problems forest plot

For reproductive health, there was a mix of significant and non-significant results for most outcomes including vaginal bleeding, unplanned pregnancy, premature labour, contraceptive use, stillbirth and premature rupture of membranes. Seven studies looked at abortion with all but one finding a significant association. Not all papers defined whether abortion was spontaneous or induced. Meta-analysis for abortion produced a pooled OR of 3·5 (95% CI 1·2–10·2) (Fig. 5).

Fig. 5

Abortion forest plot

General health outcomes included clinical diagnoses such as diabetes, symptoms such as pain and measures of general wellbeing such as mobility. Most data were from patient-reported measures. For most outcomes, there was a mixture of significant and non-significant results. Six studies looking at pain all found significant associations; Al-Modallal found associations for some types of violence but not others [30]. Meta-analysis produced a pooled OR of 2·6 (95% CI 1·6–4·1) (Fig. 6). Five studies looked at hypertension with two studies finding significant associations (for some but not all types of violence); meta-analysis produced pooled OR of 1·6 (95% CI 1·2–2·0) (Fig. 7).

Fig. 6

Pain forest plot

Fig. 7

Hypertension forest plot

For most covariates there was insufficient data to perform univariate meta-regression. Of the analyses performed, two showed significant associations. Odds of abortion in women exposed to DV were higher in studies with higher employment levels. Odds of pain in women exposed to DV were higher in studies set in high-income (OR 4·3) than low/middle-income countries (OR 1·93).


This is the first systematic review and meta-analysis of the prevalence and health outcomes of domestic violence in clinical settings in the Arab region. Around one in two women experienced psychological violence, one in three physical violence and one in five sexual violence from an intimate partner during their lifetime, with over 70% experiencing any form of violence. These rates are higher than community-level estimates from local demographic and health surveys (despite broadly comparable baseline characteristics). Around one in three women in Iraq have experienced lifetime emotional IPV, one in four women in Egypt experienced lifetime physical IPV and one in ten women in Jordan experienced lifetime sexual IPV [5, 63, 64]. This may indicate a higher prevalence of DV amongst women seeking healthcare.

Internationally, there is a lack of pooled DV prevalence data from clinical settings. A UK-based systematic review found lifetime IPV prevalence reported by individual studies ranging from 13% in a maternity service to 41% in primary care [12]. One meta-analysis, including primarily USA and high-income country studies, estimated lifetime exposure to any type of IPV of 38% in family medicine and 40% in emergency medicine settings [65]. Our study is the first to produce pooled estimates of IPV prevalence in clinical populations in the Arab region, or indeed any low/middle-income setting. Understanding the prevalence of DV/IPV amongst the population of women attending healthcare settings can inform adaptations to clinical care and health service provision.

We found mixed evidence of associations between DV and a wide range of general, reproductive and mental health outcomes. This is the first study to present pooled effect estimates of the health outcomes of DV in the Arab region. Exposure to DV significantly increased the odds of the five adverse health outcomes that we meta-analysed. These figures are higher than WHO estimates of associations between physical and/or sexual IPV and incident depression OR 1·97, induced abortion OR 2·16 and pain (last 4 weeks) aOR (adjusted odds ratio) 1·6 [4, 66]. We found no existing pooled effect estimates for sleep or hypertension.

Strengths of this review include a pre-defined protocol published online, a broad search strategy and inclusion of publications in languages other than English. Two authors completed each review stage, thereby minimising bias. Quality assessment was conducted by two authors using a published checklist. We used robust statistical methods and provide justification of our analysis decisions.

The reliability of our results is limited by the quality of the primary studies. Most were small to medium sized cross-sectional studies; only a minority justified their sample size. Where this was done, and specified, studies were more often powered to detect prevalence rather than associations with health outcomes (which would require a larger sample). Sampling was by convenience in most studies which may introduce selection bias. Some studies had non-representative populations e.g. large proportion of military police husbands [42], high level of female employment [40], or only refugee participants [27, 28, 30, 34, 35, 52, 53, 55]. Twenty-nine papers reported response rates (which were generally high) but few assessed for non-response bias. The majority of studies either gained institutional ethics approval or at least gave some consideration to ethical issues. Best practice, according to WHO research recommendations, includes reducing distress to research participants and referring for additional support if required; few studies made explicit mention of this [67].

The lack of a gold standard definition of IPV or DV hinders this research field. Terms for categories of violence against women may be overlapping or used interchangeably. Studies in this review used varied methods for measuring violence. Face-to-face interviews versus self-administered questionnaires may impact prevalence and recall bias may be affected by length of recall period. International measurement tools require translation then revalidation. Pre-existing tools may not include behaviours relevant to the Arab setting e.g. ‘cursing wife’s family’. [40] Clark and colleagues used focus groups to inform modifications to the WHO questionnaire; identification of control violence increased from 83 to 97% [45]. Psychological, emotional and control violence are particularly challenging to define. Some studies included what might be considered low-level behaviours e.g. shouting, ‘taking his own decision’, ‘being threatened with divorce’ and produced high estimates of psychological or control violence [40, 41, 57]. It may be the frequency and combination of these behaviours (measured by a few of the primary studies) which determines impact. Rates for control violence ranged from 69 to 97%, comparable to 83% in the Iraq Family Health Survey [63]. Control has been increasingly recognised internationally as a specific type of domestic violence, with some countries now criminalising coercive control [68]; control was included as defined and measured by a small number of studies in our review. Sexual violence can be difficult to measure and may be underestimated; one study removed questions due to perceived cultural sensitivity [26].

There remain gaps in the evidence on DV in clinical populations in Arab countries. Only ten out of the nineteen countries searched for are represented in our review. We retrieved no relevant Palestinian studies so could not conduct the separate analysis pre-specified in our protocol. Most studies focused on IPV, rather than the broader scope of all family violence. Studies which did look at DV (violence from any adult family member) reported in varying detail on the perpetrators of abuse (see Additional file 3). Husbands remained the most frequent abusers, sometimes in combination with other family members. In some studies and for some violence types, mother-in-laws were the most common (non-husband) family perpetrators [25, 47, 59], and in some studies brothers or fathers were the most common [25, 34, 35, 51]. Other perpetrators included mother, father-in-law, other in-laws and other wives. One study found differences in the type of abuse perpetrated by different family members with husbands responsible for the majority of sexual abuse, mother-in-laws particularly involved in mental abuse and male family members involved in physical and mental abuse [25]. The available data were not sufficient for us to conduct further analyses of violence by type of perpetrator. Despite our interest in DV, the lack of available data meant that our pooled prevalence results relate mostly to IPV.

Our study has several limitations. Our estimates for prevalence and health outcomes have wide confidence intervals and high I2 statistics, reflecting relatively small pooled samples and high levels of between-study variance. Despite wide ranging prevalence estimates, similarities in study design justified pooling. Health outcome data varied in the definition of both violence exposure and health outcomes. Some studies only looked at exposure to exclusive violence categories (e.g. physical violence only); this produces a clearer picture of the association between the violence type and specific health outcome, but excludes the many women who are exposed to multiple types of violence. Some papers reported differing associations between different violence types and the relevant health outcome. Health outcomes were defined and measured using a range of methods. Most studies measured a limited range of health outcomes, for example reproductive health or mental health only. Combining data from these varied studies enabled the generation of pooled effect estimates but contributed to heterogeneity of the health outcome results.

Visual inspection of forest plots and attempts to look for patterns between outlying results and study characteristics were unproductive. We quantitatively assessed for sources of heterogeneity using univariate meta-regression. There were insufficient data to do this for all covariates for each analysis. Where it was performed, the small number of included studies limits reliability. None of the associations found were consistent across different analyses; they may represent false positives. As we were working with summary level data, we cannot infer that any associations found apply at the individual level.

Despite finding an association between exposure to violence and adverse health outcomes, causality cannot be inferred, as most studies were cross-sectional and few specified the temporal relationship between violence exposure and onset of the health problem. There was insufficient measurement of confounding variables and few studies adjusted for confounders. The associations demonstrated could be proxy for other causal relationships e.g. socioeconomic deprivation or level of education. Interestingly, when Al-Modallal adjusted for confounders, the positive crude odds ratios for suicidal thoughts and attempts became negative [27].


Domestic violence research has progressed in the Arab region, but robust prevalence data are still required for some countries. Adopting standardised definitions and measurement tools would benefit the research field. However, definitions of violence need to retain meaning in different sociocultural contexts and adaptation of pre-existing tools may be necessary. Data on the frequency and severity of violence might elucidate mechanisms of impact. We also recommend further meta-analyses to synthesise data on the prevalence of IPV/DV in clinical settings in other geographical regions. To test causality between violence and adverse health outcomes, adequately powered longitudinal studies are needed with adjustment for confounding factors. A standardised approach to measuring the outcomes of domestic violence, including a core set of health outcomes with defined diagnostic criteria or tools, would enable comparison between studies and across different settings.

Many women accessing healthcare in the Arab region have been exposed to violence from family members or intimate partners. In some settings, the majority of women were affected. We have demonstrated that the adverse health outcomes of DV, well documented internationally, also affect Arab women. Our findings are of interest to clinicians and policy makers in the Arab region. Contact with a healthcare professional provides an opportunity to identify survivors of violence, offer support and refer to specialist services [69]. The WHO has called for a healthcare response to violence against women but much of the current evidence comes from high-income settings [70, 71]. The results from our review provide regional evidence which could be used to inform the development of healthcare interventions and policy. Interventions with an evidence base from high-income or other regional settings need local adaptation and evaluation, taking into account not only the cultural context but also the healthcare system resources and infrastructure and national policy and legal framework. We recommend further research in the Arab region to establish what a suitable healthcare response to domestic violence might look like.



Abuse Assessment Screen


Adjusted odds ratio


Convention on the Elimination of all Forms of Discrimination Against Women


Confidence interval


Conflict Tactics Scale, revised


Domestic violence


Eastern Mediterranean Region


Hurt, Insult, Threaten and Scream


Intimate partner violence


NorVold Abuse Questionnaire


Odds ratio


U.S. Preventative Services Task Force


Violence Against Women


Woman Abuse Screening Tool


World Health Organization


  1. 1.

    Devries KM, Mak JYT, García-Moreno C, et al. The global prevalence of intimate partner violence against women. Science. 2013;340(6140):1527–8.

    CAS  Article  PubMed  Google Scholar 

  2. 2.

    UK Home Office, 2013. Circular 003/2013: new government domestic violence and abuse definition ( Accessed 12 Mar 2019).

  3. 3.

    Boy A, Kulczycki A. What we know about intimate partner violence in the Middle East and North Africa. Violence against Women. 2008;14(1):53–70.

    Article  PubMed  Google Scholar 

  4. 4.

    Garcia-Moreno C, Pallitto C, Devries K, Stöckl H, Watts C, Abrahams N. Global and regional estimates of violence against women: prevalence and health effects of intimate partner violence and non-partner sexual violence. Geneva: World Health Organization; 2013.

    Google Scholar 

  5. 5.

    Ministry of Health and Population [Egypt], El-Zanaty and Associates [Egypt], and ICF International. Egypt demographic and health survey 2014. Cairo, Egypt and Rockville, Maryland, USA: Ministry of Health and Population and ICF International; 2015.

    Google Scholar 

  6. 6.

    Banat BYI. Violence against Palestinian women. Revista De Paz Y Conflictos. 2015;8(1):135–49.

    Google Scholar 

  7. 7.

    Habib SR, et al. Prevalence and effects of violence against women in a rural community in Minia governorate, Egypt. J Forensic Sci. 2011;56(6):1521–7.

    Article  Google Scholar 

  8. 8.

    Maziak W, Asfar T. Physical abuse in low-income women in Aleppo, Syria. Health Care Women Int. 2003;24:313–26.

    Article  Google Scholar 

  9. 9.

    Shiraz MS. The impact of education and occupation on domestic violence in Saudi Arabia. Int J Soc Welf. 2016;25(4):339–46.

    Article  Google Scholar 

  10. 10.

    Campbell JC. Health consequences of intimate partner violence. Lancet. 2002;359:1331–6.

    Article  Google Scholar 

  11. 11.

    UNESCWA, 2017. Status of Arab Women Report 2017. Violence against women: what is at stake? United Nations, Beirut.

  12. 12.

    Feder G, Arsene C, Bacchus L, et al. How far does screening women for domestic (partner) violence in different health care setting meet the UK National Screening Committee criteria for a screening programme in terms of condition, screening method, and intervention? Systematic reviews of nine UK National Screening Committee criteria. Health Technol Assess. 2009;13(16):137–347.

    Article  Google Scholar 

  13. 13.

    Hawcroft C, Usta J, Shaheen A, Feder G. Domestic violence in the Arab world: the sociocultural, legal and political context and the need for an effective healthcare response. Currently under review.

  14. 14.

    Haj-Yahia MM. Wife abuse and battering in the sociocultural context of Arab society. Fam Process. 2000;39(2):237–55.

    CAS  Article  Google Scholar 

  15. 15.

    MDG Monitor, 2016. MDG progress of Arab states in 2015. ( accessed 5 march 2018).

  16. 16.

    United Nations, 2013. Combating domestic violence against women and girls: policies to empower women in the Arab region. New York: United Nations, economic and social Commission for Western Asia (ESCWA), 2013.

  17. 17.

    United Nations Human Rights. Office of the High Commissioner. Status of ratification interactive dashboard, Convention on the Elimination of All Forms of Discrimination against Women. ( accessed 6 July 2018).

  18. 18.

    Downes MJ, Brennan ML, Williams HC, Dean RS. Development of a critical appraisal tool to assess the quality of cross-sectional studies (AXIS). BMJ Open. 2016;6:e011458.

    Article  PubMed  PubMed Central  Google Scholar 

  19. 19.

    Jackson D, Turner R. Power analysis for random-effects meta-analysis. Res Synth Methods. 2017;8:290–302.

    Article  Google Scholar 

  20. 20.

    Miller JJ. The inverse of the freeman-Tukey double arcsine transformation. Am Stat. 1978;32(4):138.

    Google Scholar 

  21. 21.

    Sterne JA, Sutton AJ, Ioannidis JP, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011;342:d4002.

    Article  Google Scholar 

  22. 22.

    Abdelmegeid FY. Domestic violence against Egyptian women and factors affecting it. The Medical Journal of Cairo University. 2004;72(Supp. 2):151–62.

    Google Scholar 

  23. 23.

    Abdelhai R, Mosleh H. Screening for antepartum anxiety and depression and their association with domestic violence among Egyptian pregnant women. J Egypt Public Health Assoc. 2015;90(3):101–8.

    Article  Google Scholar 

  24. 24.

    Abujilban S, Mrayan L, Al-Modallal H and Isaa E. Effects of Intimate Partner Physical Violence on Newborns’ Birth Outcomes Among Jordanian Birthing Women. J Interpers Violence. 2015; Published online September 7. DOI:

  25. 25.

    Afifi ZEM, Al-Muhaideb NS, Hadish NF, Ismail FI, Al-Qeamy FM. Domestic violence and its impact on married women's health in eastern Saudi Arabia. Saudi Med J. 2011;32(6):612–20.

    PubMed  Google Scholar 

  26. 26.

    Ahmed AM, Elmardi AE. A study of domestic violence among women attending a medical Centre in Sudan. East Mediterr Health J. 2005;11(1/2):164–74.

    CAS  PubMed  Google Scholar 

  27. 27.

    Al-Modallal H. Patterns of coping with partner violence: experiences of refugee women in Jordan. Public Health Nurs. 2012;29(5):403–11.

    Article  Google Scholar 

  28. 28.

    Al-Modallal H, Hamaideh S, Mudallal R. Mental health status of women in Jordan: a comparative study between attendees of governmental and UN relief and works Agency's health care centers. Issues in Mental Health Nursing. 2014;35(5):386–94.

    Article  Google Scholar 

  29. 29.

    Al-Modallal H, Zayed IA, Abujilban S, Shehab T, Atoum M. Prevalence of intimate partner violence among women visiting health care centers in Palestine refugee camps in Jordan. Health Care Women Int. 2015;36(2):137–48.

    Article  PubMed  Google Scholar 

  30. 30.

    Al-Modallal H. Effect of intimate partner violence on health of women of Palestinian origin. Int Nurs Rev. 2016;63(2):259–66.

    Article  Google Scholar 

  31. 31.

    Al-Modallal H. Disclosure of spousal violence: viewpoints from the Jordanian society. J Interpers Violence. 2017;32(2):164–82.

    Article  PubMed  Google Scholar 

  32. 32.

    Al-Nsour M, Khawaja M, Al-Kayyali G. Domestic violence against women in Jordan: evidence from health clinics. J Fam Violence. 2009;24(8):569–75.

    Article  Google Scholar 

  33. 33.

    Al-Serkal F, Hussein H, El Sawaf E, Al Faisal W, Hasan Mahdy N, Wasfy A. Intimate partner violence against women in Dubai: prevalence, associated factors and health consequences, 2012-2013. Middle East Journal of Psychiatry and Alzheimers. 2014;5(3):19–27.

    Article  Google Scholar 

  34. 34.

    Al-Shdayfat N. Emotional abuse among Syrian refugee women in Jordan (2017a). Global J Health Sci. 2017;9(3):237–47.

    Article  Google Scholar 

  35. 35.

    Al-Shdayfat N. Physical abuse among Syrian refugee women in Jordan (2017b). Middle East Journal Of Nursing. 2017;11(1):3–7.

    Article  Google Scholar 

  36. 36.

    Alzahrani TA, Abaalkhail BA, Ramadan IK. Prevalence of intimate partner violence and its associated risk factors among Saudi female patients attending the primary healthcare centers in Western Saudi Arabia. Saudi Med J. 2016;37(1):96–9.

    Article  Google Scholar 

  37. 37.

    Anes Jellali I, Jellali MA, Gataa R, Mechri A. Psychosexual impact of violence against Tunisian women in marriage: cross-sectional study about 197 consultant in family planning Centre of Monastir. Sexologies. 2014;23(3):e75–8.

    Article  Google Scholar 

  38. 38.

    Anes Jellali I, Jellali MA, Gataa R, Mechri A. Violence against women in the marriage: cross-sectional study in the family planning clinic Monastir. La Tunisie Medicale. 2015;93(8–9):516–22.

    PubMed  Google Scholar 

  39. 39.

    Awwad J, Ghazeeri G, Nassar AH, et al. Intimate partner violence in a Lebanese population attending gynecologic care: a cultural perspective. J Interpers Violence. 2014;29(14):2592–609.

    Article  Google Scholar 

  40. 40.

    Azm SA, Hashem WE, Elebiary HA. Impact of domestic violence against Saudi Arabian married women on their reproductive health outcomes. Med J Cairo Univ. 2009;77(1):485–94.

    Google Scholar 

  41. 41.

    Bakr IMA, Ismail NA. Domestic violence among women attending out-patient clinics in Ain Shams University hospitals, Cairo, Egypt. J Egypt Public Health Assoc. 2005;80(5–6):629–50.

    PubMed  Google Scholar 

  42. 42.

    Barnawi FH. Prevalence and risk factors of domestic violence against women attending a primary Care Center in Riyadh. J Interpers Violence. 2017;32(8):1171–86.

    Article  PubMed  Google Scholar 

  43. 43.

    Boufettal H, Obaid B, Belhouss A, Hermas S, Noun M, Samouh N. Physical violence during pregnancy in Morocco. Journal de Gynecologie Obstetrique et Biologie de la Reproduction. 2012;41(1):76–82.

    CAS  Article  Google Scholar 

  44. 44.

    Clark CJ, Silverman J, Khalaf B, et al. Intimate partner violence and interference with women's efforts to avoid pregnancy in Jordan. Stud Fam Plan. 2008;39(2):123–32.

    Article  Google Scholar 

  45. 45.

    Clark CJ, Bloom DE, Hill AG, Silverman JG. (2009a). Prevalence estimate of intimate partner violence in Jordan. East Mediterr Health J. 2009;15(4):880–9.

    CAS  Article  Google Scholar 

  46. 46.

    Clark CJ, Hill A, Jabbar K, Silverman JG. (2009b). Violence during pregnancy in Jordan: its prevalence and associated risk and protective factors. Violence Against Women. 2009;15(6):720–35.

    Article  Google Scholar 

  47. 47.

    Clark CJ, Silverman JG, Shahrouri M, Everson-Rose S, Groce N. The role of the extended family in women's risk of intimate partner violence in Jordan. Soc Sci Med. 2010;70:144–51.

    Article  PubMed  Google Scholar 

  48. 48.

    Daoud N, Sergienko R, Shoham-Vardi I. Intimate partner violence prevalence, recurrence, types, and risk factors among Arab, and Jewish immigrant and nonimmigrant women of childbearing age in Israel. J Interpers Violence. 2017:1–28.

  49. 49.

    Eldoseri HM, Tufts KA, Zhang Q, Fish JN. Adverse health effects of spousal violence among women attending Saudi Arabian primary health-care clinics. East Mediterr Health J. 2014;20(11):717–25.

    CAS  Article  Google Scholar 

  50. 50.

    Eldoseri HM, Sharps P. Risk factors for spousal physical violence against women in Saudi Arabia. J Interpers Violence. 2017:1–25.

  51. 51.

    Haddad LG, Shotar A, Younger JB, Alzyoud S, Bouhaidar CM. Screening for domestic violence in Jordan: validation of an Arabic version of a domestic violence against women questionnaire. Int J Women's Health. 2011;3:79–86.

    Article  Google Scholar 

  52. 52.

    Hammoury N, Khawaja M. Screening for domestic violence during pregnancy in an antenatal clinic in Lebanon. Eur J Pub Health. 2007;17(6):605–6.

    Article  Google Scholar 

  53. 53.

    Hammoury N, Khawajah M, Mahfoud Z, Afifi RA and Madi H. Domestic violence against women during pregnancy: The case of Palestinian refugees attending an antenatal clinical in Lebanon. J Womens Health. 2009; 18(3): 337–345. DOI: 10.1089=jwh.2007.0740.

  54. 54.

    Ibrahim ZM, Sayed Ahmed WA, El-Hamid SA, Hagras AM. Intimate partner violence among Egyptian pregnant women: incidence, risk factors, and adverse maternal and fetal outcomes. Clin Exp Obstet Gynecol. 2015;42(2):212–9.

    CAS  Article  PubMed  Google Scholar 

  55. 55.

    Khawaja M and Hammoury N. Coerced sexual intercourse within marriage: a clinic-based study of pregnant Palestinian refugees in Lebanon. J Midwifery Womens Health 2008; 53(2): 150–154. doi:10.1016/j.jmwh.2007.09.001.

  56. 56.

    Malik IA, Shabila NP, Al-Hadithi TS. Women’s knowledge of the domestic violence legislation in Erbil, Iraq and their response to spousal violence. J Fam Violence. 2017;32:47–53.

    Article  Google Scholar 

  57. 57.

    Mamdouh HM, Ismail HM, Kharboush IF, et al. Prevalence and risk factors for spousal violence among women attending health care centres in Alexandria. Egypt Eastern Mediterranean Health Journal. 2012;18(11):1118–26.

    CAS  Article  Google Scholar 

  58. 58.

    Okour AM, Badarneh R. Spousal violence against pregnant women from a Bedouin community in Jordan. J Womens Health. 2011;20(12):1853–9.

    Article  Google Scholar 

  59. 59.

    Oweis A, Gharaibeh M, Alhourani R. Prevalence of violence during pregnancy: findings from a Jordanian survey. Matern Child Health J. 2010;14(3):437–45.

    Article  Google Scholar 

  60. 60.

    Spencer RA, Usta J, Essaid A et al. Alianza por la Solidaridad, United Nations Population Fund-Lebanon and C. J. Clark. Gender Based Violence Against Women and Girls Displaced by the Syrian Conflict in South Lebanon and North Jordan: Scope of Violence and Health Correlates. 2015.

  61. 61.

    Tashkandi AA, Rasheed FP. Wife abuse: a hidden problem. A study among Saudi women attending PHC centres. East Mediterr Health J. 2009;15(5):1242–53.

    CAS  PubMed  Google Scholar 

  62. 62.

    Usta J, Farver JAM, Pashayan N. Domestic violence: the Lebanese experience. Public Health. 2007;121(3):208–19.

    Article  PubMed  Google Scholar 

  63. 63.

    Republic of Iraq. Iraq Family Health Survey Report 2006/7. Ministry of Health Iraq and WHO.

  64. 64.

    Department of Statistics [Jordan] and ICF International. Jordan population and family health survey 2012. Calverton, Maryland, USA: Department of Statistics and ICF International; 2013.

    Google Scholar 

  65. 65.

    Sprague A, Goslings C, Hogentoren C, et al. Prevalence of intimate partner violence across medical and surgical health care settings: a systematic review. Violence Against Women. 2014;20(1):118–36.

    Article  Google Scholar 

  66. 66.

    Ellsberg M, Jansen H, Heise L, Watts C, Garcia-Moreno C. Intimate partner violence and women’s physical and mental health in the WHO multi-country study on women’s health and domestic violence: an observational study. Lancet. 2008;371:1165–72.

    Article  Google Scholar 

  67. 67.

    World Health Organisation, 2016. Ethical and safety recommendations for intervention research on violence against women. Building on lessons from the WHO publication Putting women first: ethical and safety recommendations for research on domestic violence against women. WHO, Geneva, 2016.

  68. 68.

    RTE Ireland. What is coercive control and why is it now a legal offence? ( accessed 6th February 2019).

  69. 69.

    World Health Organisation. Responding to intimate partner and sexual violence against women: WHO clinical and policy guidelines. WHO, Geneva. 2013;2013.

  70. 70.

    García-Moreno C, Hegarty K, Lucas d’Oliveira AF, Koziol-McLain J, Colombini M, Feder G. The health-systems response to violence against women. Lancet. 2015; 385: 1567–1579. 10.1016/S0140-6736(14)61837-7.

  71. 71.

    Bacchus L, Bewley S, Fernandez C, et al. Health sector responses to domestic violence: promising intervention models in maternity and primary care settings in Europe. London School of Hygiene & Tropical Medicine: London. 2012. ( accessed March 2nd, 2018).

Download references


The authors wish to thank colleagues at CAPC, University of Bristol for technical advice: Alyson Huntley, Vincent Cheng and Alexandra McAleenan. We would like to thank authors of primary studies for responding to our enquiries, especially those who made the effort to offer clarification or provide us with additional data: Sanaa Abujilban, Hanan Al Modallal, Houssine Boufettal, Cari Jo Clark, Rachael Spencer and Noha Al-Shdayfat.


CH is an NIHR badged GP Academic Clinical fellow funded by the Severn Deanery.

AS and GF are funded by the HERA project, supported by the Medical Research Council grant MR/P025102/1. (

GF is supported by the NIHR Biomedical Research Centre at University Hospitals Bristol NHS Foundation Trust and the University of Bristol. The views expressed in this publication are those of the author(s) and not necessarily those of the NHS, the National Institute for Health Research or the Department of Health and Social Care.

RH was supported by the Wellcome Trust, Institutional Strategic Support Fund and the Elizabeth Blackwell Institute.

The funders had no role in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Author information




The initial protocol was written by AS. It was developed and revised by CH with input from other authors. CH and HE designed the search strategy, conducted searches and performed title/abstract and full text screening, disagreements were arbitrated by GF. CH and JU conducted quality assessment with arbitration by GF. All data was extracted by CH, with duplicate data extraction performed by JU, KG or TD. Summary statistics were produced by CH and data analysis was performed by RH. CH wrote the first draft of the report with input from GF. All authors made substantive contributions and approved the final manuscript.

Corresponding author

Correspondence to Claire Hawcroft.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

JU was lead author of one and co-author of another paper included in the review. Inclusion decisions for these papers were made by CH, HE and GF. For these two papers, quality assessment was performed by CH and GF with no input from JU.

CH, AS and GF are working on HERA which aims to understand and develop the healthcare response to violence against women in LMIC.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Additional files

Additional file 1:

Search terms. (PDF 38 kb)

Additional file 2:

Included studies summary table. (PDF 441 kb)

Additional file 3:

Perpetrators of domestic violence. (PDF 117 kb)

Additional file 4:

Additional forest plots. (PDF 81 kb)

Additional file 5:

Summary of health outcome findings. (PDF 382 kb)

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Hawcroft, C., Hughes, R., Shaheen, A. et al. Prevalence and health outcomes of domestic violence amongst clinical populations in Arab countries: a systematic review and meta-analysis. BMC Public Health 19, 315 (2019).

Download citation


  • Domestic violence
  • Intimate partner violence
  • Violence against women
  • Spouse abuse
  • Arab
  • Arabic
  • Middle East
  • Health
  • Healthcare