Violence Against Women in Adulthood Is Associated with Increased Disordered Eating Behaviors and Risk of Eating Disorders: A Cross-Sectional Study

Abstract

Violence against women (VAW) is a global public health problem associated with considerable health risks, including disordered eating behavior (DEB) and eating disorders (ED). The current study investigates the effect of different VAW types on the occurrence of DEB and the risk of ED in adult women in Lebanon. A cross-sectional survey assessing violence exposure, ED risk, and DEB was administered to 367 women older than 18 years; classified into a violence exposed group (EG, n = 106) and unexposed group (UG, n = 261). Pearson's correlations, independent t-tests, and a simple linear regression were used. Results showed that total violence, and all its subscales were significantly higher in the EG compared to UG (p < 0.001). Subjects in the EG (77.78%) were more likely to be at risk of ED compared to those in the UG (p = 0.034). Moreover, total violence, physical, and sexual violence were associated with increased risk of ED (p = 0.0259, p = 0.0354, and p = 0.0351, respectively). Regarding DEB, only external eating was significantly higher among the EG compared to the UG (p = 0.0031). The present findings showed that VAW results in DEB and thus increased risk of ED.

Introduction

Violence against women (VAW) comprises various types of abuse across the lifespan (Krantz and Garcia-Moreno 2005; World Health Organization 2017). Almost 30% and 37% of women worldwide and in the Arab Region, respectively, have experienced sexual and/or physical violence by a nonintimate partner, intimate partner, or both in their lifetime (World Health Organization 2017). VAW is associated with significant adverse physical, psychological, and reproductive outcomes. This includes, physical injuries, chronic gastrointestinal problems, cardiovascular diseases (Black 2011), infertility, sexually transmitted diseases (Stewart et al. 2016), depression, suicide (Dillon et al. 2013), hard drug (Bacchus et al. 2018), alcohol, and tobacco abuse (Bonomi et al. 2009).

Research conducted in various countries indicate an association between experiencing any form of violence, and developing a risk of eating disorders (ED) or disordered eating behavior (DEB). Studies identified considerable risk factors for adult ED and DEB such as childhood sexual, emotional, and physical abuse (Cebolla et al. 2014; Collins et al. 2014; Mattar et al. 2011) or even passive exposure to violence through watching movies in adults (Dillon et al. 2013; Felitti 1993; Mattar et al. 2015, 2019).

Nonetheless, less literature exists on adulthood exposure to violence and the incidence of ED and DEB in women. Danielson et al. 1998 revealed five times greater odds of anorexia nervosa and bulimia nervosa diagnosis in women (n = 198) who experienced VAW during the preceding 12 months in New Zealand. A study in the United States (Fischer et al. 2010) showed a significant association between recent women sexual assault (past 3 months) and current ED symptoms, after controlling for childhood abuse in 489 women. A recent case-control study in the United States observed disordered eating in 790 women from African descent positively associated intimate partner abuse and the development of disordered eating (Lucea et al. 2012). The study highlighted associations between the type of violence and DE, with physical and sexual violence being stronger predictors of DEB in women compared to emotional violence (Lucea et al. 2012). In an observational study in the United States, including 101 adult women, higher risks of ED were reported in victims of intimate partner violence (IPV), with posttraumatic stress disorder (PTSD) and depression being the mediators between IPV and ED (Nathanson et al. 2012). A systematic review published in 2013 (Bundock et al. 2013) claimed that evidence on ED and IPV is rare, and identified gaps for research, such as the unclear direction of ED-violence causality and the type of violence (physical, emotional, financial, and verbal) implicated in the relationship with ED and DEB.

The available literature on the associations of VAW, DEB, and ED remains scarce, debatable, and very heterogeneous. Studies either did not include DEB or did not specify the type of violence implicated or focused solely on sexual violence. Understanding the risk factors of ED and DEB is crucial for their prevention and treatment as they increase chronic health problems risks (Quick et al. 2013) and mortality rates (Arcelus et al. 2011).

To our knowledge, evidence on the impact of adulthood exposure of partner and nonpartner VAW, and the type of VAW, on both the incident of DEB and the risk of developing an ED is minimal and inconclusive, especially in the Middle East and North Africa (MENA) region. Countries in the MENA region, including Lebanon, witness high levels of gender based violence (GBV), although there is poor evidence on the matter due to the lack of reporting, and the lack of a nationally representative GBV prevalence study in Lebanon in particular (USAID 2016). On average, around 20% of the women experienced IPV within Egypt, Jordan, and Tunisia (USAID 2016). In Lebanon, IPV was seen at levels of 31% (KAFA 2016). The literature in the MENA region on the prevalence of ED and DEB among adult women is very limited. A study in Kuwait (Musaiger et al. 2016) showed DEB risk in 33.6% of a Kuwaiti female adult sample. Similarly, a cross-sectional study in a United Arab Emirates' university, revealed a 31% prevalence of DEB among females (Haleama Al and Shatha 2016). However, no studies were found to study the association between GBV's impact on female adults' ED risks and DEB. Finding relevant data on the association of interest in the region was challenging, with the extremely limited research focusing on refugee communities and on children and not adults, such as Solomon's research which showed that witnessing maternal exposure to violence increased the risks of ED in Iraqi children (Solomon 2019). Hence, the present study aims to investigate associations between VAW and the occurrence of DEB and the risk of ED in adult women in Lebanon with the added value of unconventionally differentiating between the DEB patterns and the varied VAW types.

Methods

Design, participants, and procedures

A validated screening tool for ED, a validated diagnostic tool for DEB (Aoun et al. 2015; Morgan et al. 1999), and a culturally relevant violence-assessing questionnaire (Haj-Yahia 1999) were distributed and made available online in English and Arabic to a convenient sample of adult women. Inclusion criterion is any woman above 18 years old. Women were recruited following convenient sampling from two types of populations: (1)

A sample of women who had experienced any type of violence from anyone during the past year (father, mother, brother, partner, boyfriend, friend, relative). These participants were recruited conveniently and consecutively through shelters or nongovernmental organizations (NGOs) referrals from all over Lebanon with the help of the Institute for Women's Studies in the Arab World at the Lebanese American University. These associations included KAFA (Enough Violence and Exploitation), ABAAD (NGO seeking gender equality), and RDFL (The Lebanese Democratic Women's Gathering). Printed surveys were given to the NGO's local managers to insure total confidentiality. When violence-exposed women visited the shelter or NGO, the survey was proposed to them and collected back anonymously when completed.

(2)

A random convenience sample of women with unknown exposure to violence in the past year. The survey was distributed as a hard copy and sent online via emails to random populations across Lebanon.

Measures

The survey included three sections: sociodemographic, ED and DEB risk, and exposure to IPV and abuse.

The sociodemographic section addressed monthly household income, highest educational level, employment status, marital status, and cohabitation status.

ED risk was measured using the SCOFF (screening tool for eating disorders; each letter stands for the first letter of the questionnaire's 5 items) ED screening tool (Morgan et al. 1999). It consists of five Yes/No items. Participants with two or more positive answers were labeled “at risk of eating disorders” while those with less than two were labeled “not at risk.” The SCOFF was previously validated in Lebanon, proving to be accurate and reliable in the early detection of ED (sensitivity of 80.0%, specificity of 72.7%, and an area under the curve of 80.0%) (Aoun et al. 2015).

DEB was measured using the 33-item DEBQ (Disordered Eating Behavior Questionnaire) (Van Strien et al. 1986) that is validated for assessing DEB (Cebolla et al. 2014; Wardle 1987; Wu et al. 2017) and previously used in Lebanese population studies (Zeeni et al. 2013, 2021). The score for each subject is calculated for each subscale (restrained, emotional, and external eating) and as a total score. Restrained eating denotes limiting food intake in an attempt to halt weight gain or promote weight loss (Meule 2016). Emotional eating refers to the liability to overeat as response to negative emotions (van Strien 2018). External eating refers to the proneness of overeating in response to external cues, as sight or smell of food (van Strien et al. 2009).

Questions assessing DEB and screening for ED preceded the assessment of violence assuming that assessment of violence might be stressful for the participant.

Culturally relevant Violence Exposure questions were taken from a study (Haj-Yahia 1999) that assessed the prevalence of women abuse and its psychological consequences on Palestinian women. Twenty-three questions were divided into 4 categories: psychological abuse (11 items), physical abuse (7 items), sexual abuse (3 items), and economic abuse (2 items). Subjects are required to answer on a scale from 1 (considered never) to 4 (always). Total and subscale scores were used for the analysis.

Data management

Data were collected from 353 participants of which 62 were recruited from the aforementioned NGOs and shelters and 291 from the general population. Three participants were excluded because of missing data. Participants were classified into a violence exposed group (EG) and unexposed group (UG). Given that a significant number of women from the random general population sample had been exposed to violence (n = 45, assessed by a total violence score that is within 1 standard deviation [SD] of the mean of the EG, cf. statistical analyses for further details), they were shifted to the EG. Accordingly, the final sample consisted of 106 participants in the EG and 244 in the UG.

Statistical analysis

Analysis was performed using STATA v13. Descriptive analysis was used to summarize the study variables and to screen for out of range values. Independent t-tests checked for significant difference between the mean scores of the exposed and the unexposed group. A subset of 45 participants from the general population had total violence scores that were within 1 SD of the mean of the violence group. Independent t-tests checked whether the difference between the mean scores of the EG and the latter subset were statistically significant, and the results were not. Accordingly, those 45 participants were moved to the exposed group. Continuous variables were described using mean and SDs, while frequencies and percentages were used to represent categorical variables. Independent t-tests were used to compare the mean scores between groups for total violence and all its subscales as well as DEBQ subscales. Two-tailed p-values are reported. χ² test was used to check the difference between the two groups in terms of participants who tested positive for being “at risk of eating disorders” according to SCOFF questionnaire.

The second part of the analysis was conducted on all study participants regardless of previous group assignment as the unexposed group participants were exposed to some form of violence (as per the below results). Pearson's correlations were used to examine the relationship between DEBQ subscales and total violence score and its subscales. Pearson's correlations r of “0.1–0.3,” “0.3–0.5,” and “0.5–1.0” reflect small, medium, and large strengths of the associations, respectively. The relationship between total violence score and its subscales and being at risk of ED was also assessed using the SCOFF questionnaire. Independent t-tests were used to compare the mean scores between groups for total violence and all its subscales. Cohen's d statistic was used to compute the effect sizes of the independent t-test with the following cutoff points: 0.2, 0.5, and 0.8 and above indicative of small, medium, and large effect sizes, respectively. A simple linear regression was used to assess how much of the variance in external eating was explained by the total violence score. Finally, a multiple linear regression model was used to assess the independent association between external eating and total violence score and demographic characteristics. Variables were selected for inclusion in the model based on a p-value <0.2 at the bivariate level. External eating score was the primary dependent variable and total violence, age, level of education, SCOFF, income, and living with husband/partner were the independent variables included in the model as covariates. The effect size of the linear regression model was assessed using f² with values of 0.02, 0.15, and 0.35 indicative of small, medium, and large effect sizes, respectively.

Results

Sociodemographic variables

Table 1 shows the sociodemographic characteristics of study participants divided by EG along the respective differences (p values).

Table 1.

Demographic Characteristics of Study Participants Divided by Violence Groups

	EG (N = 106)	UG (N = 261)	p
Age
18–24	34 (32.08%)	117 (44.83%)	—
25–34	30 (28.30%)	35 (13.41%)	0.001
35–45	23 (21.70%)	36 (13.79%)	0.017
>45	19 (17.92%)	73 (27.975)	0.733
Marital status
Single	35 (33.02%)	119 (45.59%)	—
Married	40 (37.74%)	113 (43.30%)	0.486
Widowed	1 (0.94%)	17 (6.51%)	0.124
Divorced	30 (28.30%)	12 (4.60%)	<0.0001
Living with husband/partner
Yes	31 (30.39%)	110 (42.31%)	0.037
No. of children <18 years
0	28 (26.92%)	119 (45.95%)	—
1	33 (31.73%)	68 (26.25%)	0.015
2	19 (18.27%)	28 (10.81%)	0.004
3	13 (12.50%)	11 (4.25%)	<0.0001
4+	11 (10.58%)	33 (12.74%)	0.392
Level of education
No schooling completed	23 (21.70%)	29 (11.15%)	—
Brevet (ninth grade)	13 (12.26%)	38 (14.62%)	0.048
High school graduate/diploma/equivalent	32 (30.19%)	67 (25.77%)	0.150
Trade/technical/vocational training	14 (13.21%)	18 (6.92%)	0.966
Bachelor's degree or above	24 (22.64%)	108 (41.54%)	<0.0001
Employment status
Employed for wages	36 (33.64%)	66 (25.29%)	—
Unemployed	11 (10.28%)	43 (16.48%)	0.056
Housewife	26 (24.30%)	38 (14.56%)	0.490
Students	24 (22.43%)	92 (35.25%)	0.017
Entrepreneur	10 (9.35%)	22 (8.43%)	0.674
Income
Yes	58 (54.72%)	144 (55.38%)	0.907
Monthly income
<500$	34 (35.42%)	58 (23.39%)	—
500$–1000$	33 (34.37%)	44 (17.74%)	0.435
1000$–1500$	13 (13.54%)	35 (14.11%)	0.242
1500$–2500$	7 (7.29%)	21 (8.47%)	0.246
>2500$	9 (9.38%)	90 (36.29%)	<0.0001

Data are presented as N (%) for categorical variables.

EG, violence exposed group; UG, violence unexposed group.

The mean body mass index (BMI) was reported for each group; 34% of the unexposed group (M = 27.19, SD = 1.57) and 35.5% of the exposed group (M = 26.92, SD = 1.38) were overweight. However, there was no significant difference between the two groups in terms of BMI categories' distribution (p = 0.1386) nor the total violence score across the BMI categories even after correcting for group assignment (p = 0.60).

The majority of EG participants were aged between 18 and 34 (60.38%); older age was negatively associated with emotional and external eating subscales even after controlling for group assignment (p < 0.0001 for both, respectively). However, older age was positively associated with restrained eating subscale (p = 0.014).

Differences between study groups

The mean total violence score was 54.85 (SD = 14.67) for the EG and 26.65 (SD = 5.37) for the UG. Table 2 shows that the EG had significantly higher scores on total violence as well as all its subscales (being psychological, physical, sexual, and economic) compared to the UG [t(348) = 27.56, p < 0.001; t(347) = 26.62, p < 0.001; t(348) = 17.99, p < 0.001; t(339) = 14.46, p < 0.001; and t(341) = 13.99, p < 0.001, respectively]. The latter effects are absolutely large >0.8. Participants in the exposed group (77.78%) were more likely to score at “risk of eating disorders” as per SCOFF screening compared to the UG (66.12%; χ² = 4.51, p = 0.034). As for DEBQ subscales, only external eating scores were significantly higher among the exposed group [t(330) = 2.97, p = 0.0031]. Restrained and emotional eating scores were not significantly different between the two groups.

Table 2.

Mean Scores Differences Between Study Groups

	EG	UG	t	Effect size
Total violence	54.85 (14.67)	26.65 (5.37)	26.37^**	2.55
Psychological abuse	29.75 (7.32)	13.78 (3.83)	26.66^**	2.73
Physical abuse	15.22 (6.26)	7.57 (1.48)	17.99^**	1.68
Sexual abuse	6.52 (3.35)	3.22 (0.82)	14.33^**	1.35
Economic abuse	4.53 (2.11)	2.29 (0.85)	13.99^**	1.39
Restrained eating	28.18 (9.45)	28.15 (9.09)	0.03	0.003
Emotional eating	33.33 (15.49)	34.83 (12.93)	−0.92	0.10
External eating	30.16 (8.52)	27.55 (6.99)	2.93	0.33

p < 0.001.

Relationship between violence and DEBQ

External eating correlated positively with total violence score (r = 0.16, p = 0.003) and three subscales: psychological abuse (r = 0.17, p = 0.002), physical abuse (r = 0.12, p = 0.0346), and economic abuse (r = 0.14, p = 0.0138). Emotional eating correlated negatively with sexual abuse subscale only (r = −0.13, p = 0.02). Restrained eating was not correlated with total violence score or any of its subscales (Table 3).

Table 3.

Correlations Between All Disordered Eating Behavior Questionnaire Subscales and Total Violence Score and Subscales

	Restrained eating	Emotional eating	External eating
Total violence	−0.06	−0.07	0.16^**
Psychological abuse	−0.07	−0.04	0.17^**
Physical abuse	−0.09	−0.07	0.12^*
Sexual abuse	−0.003	−0.13^*	0.08
Economic abuse	0.01	−0.05	0.14^*

p < 0.05; ^**p < 0.01.

Relationship between violence and risk of ED

Participants identified “at risk of ED” using SCOFF questionnaire had significantly higher results on total violence, physical, and sexual abuse subscales, compared to those who were not at risk [t(339) = 2.24, p = 0.0259; t(339) = 2.11, p = 0.0354; and t(333) = 2.12, p = 0.0351, respectively].

Relationship between violence and external eating

Total violence score was significantly associated with mean external eating score in a univariate linear regression model (β = 0.077, p = 0.003). The scale explained 2.3% of the variance in external eating.

Multiple linear regression results indicated that total violence score, being at risk of ED, educational level, and age remained significant after correcting for covariates (living with husband/partner and income). Higher violence scores and being at risk of ED were positively associated with external eating subscale's score (p = 0.003; p = 0.001, respectively). Participants with trade/vocational/technical training scored lower on external eating subscale compared to those without schooling (p = 0.031), and older age negatively correlated with external eating subscale score (p = 0.003) (Table 4).

Table 4.

Adjusted Multiple Linear Regression Model of External Eating and Participants' Characteristics

	Coefficient	CI	p
Total violence score	0.079	0.03 to 0.13	0.003
Age^a
25–34	−0.17	−2.70 to 2.36	0.896
35–45	−4.24	−7.03 to −1.45	0.003
>45	−1.4	−3.84 to 1.02	0.255
Income	0.88	−0.83 to 2.60	0.310
Level of education^b
Brevet (ninth grade)	−0.92	−3.96 to 2.11	0.550
High school graduate/diploma/equivalent	−1.13	−3.87 to 1.60	0.415
Trade/technical/vocational training	−3.81	−7.27 to −0.35	0.031
Bachelor's degree or above	−0.51	−3.16 to 2.13	0.702
SCOFF	3.05	1.25 to 4.84	0.001
Living with husband/partner	1.74	−0.22 to 3.69	0.081

Reference group “18–24 years.”

Reference group “no schooling completed.”

CI, confidence interval.

Therefore, age, educational level, being at risk of ED, and total violence score were found to be independent predictors of external eating score. The model explained 12.4% of variance in external eating with a f² of 0.15 indicative of medium effect size.

Discussion

The present study examined associations between different types of VAW, the occurrence of DEB (emotional eating, external eating, and restrained eating) and the risk of ED in adult women in Lebanon. Such study is specifically vital in a region where VAW is witnessed significantly and the research on its effects on women is scarce, if found. Results revealed first that in adulthood, violence in general and more specifically physical and sexual violence were all associated with an increased risk of ED. Second, violence exposure increased the occurrence of external eating.

Subjects in the EG were younger, had lower educational levels and earned lower monthly income compared to the unexposed. This is consistent with recent research identifying VAW-associated risk factors. For instance, in a Turkish sample population consisting of 1760 women (Sen and Bolsoy 2017), women's low education level predicted significantly higher violence exposure and total violence scores (p < 0.05). A comparable study of 1600 Iranian women (Saffari et al. 2017) associated low education and low socioeconomic status with increased risk of domestic violence.

The present results are consistent with previous work, in which VAW was found to be significantly associated with eating problems and ED (Jonas et al. 2014; Romito et al. 2013). Jonas et al. (2014) claimed that the direct effect of IPV on women's mental dispositions such as fear, hopelessness, and low self-esteem, confer vulnerability to psychiatric consequences, including ED (Jonas et al. 2014). Svavarsdottir and Orlygsdottir (2009) also affirmed that IPV-exposed women had worse psychological health symptoms (low self-esteem, symptoms of distress, PTSD, anxiety, suicidal thoughts/attempts, lack of energy, lack of initiative and/or fatigue, and psychosomatic symptoms), which put them at a greater risk of developing chronic illnesses and specifically ED (Svavarsdottir and Orlygsdottir 2009).

Several potential mechanisms were suggested to explain the association between violence and ED risk. From a general perspective, ED development may be a negative coping mechanism in response to the situational stress of intimate violence (Schirk et al. 2015). Interestingly, a qualitative study identified that intentionally women engaged in disordered eating may avoid ideal shape to avoid unwanted attention from perpetrators of trauma (Breland et al. 2018). A few studies reported similar findings among sexually abused obese participants who used their obesity as a protective mechanism against further sexual violence and tended to overeat to cope with emotional distress (Felitti 1993; Wiederman et al. 1999). Moreover, Collins et al. (2014) demonstrated that recent/attempted rape could influence the development of ED symptoms among different ethnicity college women (mean age 18 ± 0.40 years) and that cognitive avoidance coping mechanism- particularly chronic thought suppression- mediates the relationship between violence occurrence and ED symptoms (Collins et al. 2014).

Our findings revealed prominent external eating (Van Strien et al. 1986) in violence exposed women (total, psychological, physical, and economic). Hypothetically, a possible mediator may be impulsivity. On one hand, VAW was shown to be associated with higher impulsivity in a sample of 412 adult patients of all races and ethnicities in South Los Angeles (Bazargan-Hejazi et al. 2014), while on the other hand higher impulsivity was also found to be associated with both increased food intake and external eating (Kakoschke et al. 2015). Therefore, a possible explanation is that higher impulsivity is associated with VAW and may simultaneously increase the risk for external eating.

Remarks and practical implications

A striking finding of this study helped highlighting a relatively high prevalence of violence in the general population, which consequently moved some of the subjects toward the EG. This can be explained by various factors, starting with the lack of a systematic structure to screen and fight intimate violence. The majority of help is provided through private NGOs or institutions with very limited capacity. The lack of proper referral systems is also another reason for the lack of support. On the contrary, although Lebanese women might be perceived as privileged compared to other Arab countries, according to the Gender Gap Index (Equal Measures 2030 2019), Lebanon ranks third from the bottom in the MENA region followed by Syria and Yemen. According to a survey conducted in 2016 by KAFA NGO (KAFA 2016), 31% of women in Lebanon had experienced IPV. Furthermore, violent acts against women are socially accepted within a deeply rooted patriarchal system. Last but not least, at a legislative level, it is only until 2014 that a law against domestic violence was voted on by the parliament. Unfortunately, this law fails in protecting women from all forms of IPV or VAW. Such shortcomings are in criminalizing threats and accompanying harms of marital rape, but not the act of rape itself (USAID 2016). In addition, the article 252 reduces the sentencing for crimes committed in a state of rage, known as crimes of passion (USAID 2016). Finally, as the Lebanese constitution positions personal status laws under the authority of 18 religious sects, violating acts of women's rights may be magnified by the command of clergy men to regulate “religious” matters, with the lack of a unified civil code regarding gender equality (USAID 2016).

Finally, the present study helps in directing the attention of public health stakeholders and health care providers to take preventive actions against ED in women exposed to violence and abuse and to sensitively manage them. This could be done through the training of health care professionals to adequately screen, assess, and address ED in populations exposed to violence and thus provide the appropriate referrals and interventions when necessary (Black 2011); also, screening for violence among ED patients. Exposed women may choose not use these services due to negative network orientation resulting from fear, shame, and embarrassment (Schirk et al. 2015). This problem may be alleviated by raising community awareness, educating the victims on the importance of seeking help.

Limitations

A number of limitations arise when interpreting these findings. First, given the delicate nature of the subject matter, social desirability bias could have affected the participants' answers reliability. Second, the sampling was convenient and not random, and relying solely on self-reported measures introduces the possibility of mono-method bias. Third, the unexposed population was not cross matched in terms of education level and age, which had to be corrected for statistically. We used the SCOFF tool to identify ED. Given that it is a screening tool rather than a diagnostic tool, the obtained results should be interpreted with caution as an overestimation of the rates of ED could occur (Morgan et al. 1999). Furthermore, the relationship between violence and DEBQ was weak (R < 0.2) which limits the generalizability of the results or affirmation of the hypothesis. Childhood violence and trauma were neither assessed nor corrected for.

The present results enhance the understanding of the wide array of physical and mental consequences of VAW and emphasize the need for more specialized services that respond to the demands of women experiencing violence. Future studies would benefit from using longitudinal or mixed-methods designs and the addition of direct measures. Also, future studies need to correct for childhood exposure to violence and trauma. Potential research investigating the mechanisms by which violence triggers ED and DEB will improve targeted interventions, and ultimately public policies.

Ethical Considerations

This observational cross-sectional study got approval from the Lebanese American University Institutional Review Board (IRB; LAU.SAS.LM1.17/02/2017), which is constituted in accordance with the US Code of Federal Regulation (45CFR 46.107, 21CFR 56.107), and Good Clinical Practice ICH (Section 3). All answers were strictly anonymous and confidential and the aims of the study were clearly stated in the consent form.

Acknowledgments

We thank first and foremost the Lebanese NGOs KAFA, ABAAD, and RDFL for believing in the importance of research and for opening their doors to us. Thank you for the Arab Institute for Women at the Lebanese American University for their support in this project.

Authors' Contributions

L.M. designed and coordinated the study. N.Z. and L.M. collected the data. N.Z., J.A.K., N.H., and L.M. wrote and critically revised the article. J.A.K. carried out the statistical analysis. All authors reviewed and approved the final article.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

We acknowledge the Lebanese American University for the funding of the project. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

References

Aoun

, Azzam

, Jabbour

, et al. (2015). Validation of the Arabic version of the SCOFF questionnaire for the screening of eating disorders. East Mediterr Health J. 21, 326–331.

Arcelus

, Mitchell

, Wales

, Nielson

. (2011). Mortality rates in patients with anorexia nervosa and other eating disorders. Arch Gen Psychiatry. 68, 724–731.

Bacchus

, Ranganathan

, Watts

, Devries

. (2018). Recent intimate partner violence against women and health: A systematic review and meta-analysis of cohort studies. BMJ Open. 8, e019995.

Bazargan-Hejazi

, Kim

, Lin

, et al. (2014). Risk factors associated with different types of intimate partner violence (IPV): An emergency department study. J Emerg Med. 47, 710–720.

Black

. (2011). Intimate partner violence and adverse health consequences: Implications for clinicians. Am J Lifestyle Med. 5, 428–439.

Bonomi

, Anderson

, Reid

, et al. (2009). Medical and psychosocial diagnoses in women with a history of intimate partner violence. Arch Intern Med. 169, 1692–1697.

Breland

, Donalson

, Dinh

, Maguen

. (2018). Trauma exposure and disordered eating: A qualitative study. Women Health. 58, 160–174.

Bundock

, Howard

, Trevillion

, et al. (2013). Prevalence and risk of experiences of intimate partner violence among people with eating disorders: A systematic review. J Psychiatr Res. 47, 1134–1142.

Cebolla

, Barrada

, van Strien

, et al. (2014). Validation of the Dutch Eating Behavior Questionnaire (DEBQ) in a sample of Spanish women. Appetite. 73, 58–64.

10.

Collins

, Fischer

, Stojek

, Becker

. (2014). The relationship of thought suppression and recent rape to disordered eating in emerging adulthood. J Adolesc. 37, 113–121.

11.

Danielson

, Moffitt

, Caspi

, Silva

. (1998). Comorbidity between abuse of an adult and DSM-III-R mental disorders: Evidence from an epidemiological study. Am J Psychiatry. 155, 131–133.

12.

Dillon

, Hussain

, Loxton

, Rahman

. (2013). Mental and physical health and intimate partner violence against women: A review of the literature. Int J Family Med. 2013, 313909.

13.

Equal Measures 2030. (2019). Harnessing the power of data for gender equality: Introducing the 2019 EM2030 SDG Gender Index. Retrieved from https://www.equalmeasures2030.org/wp-content/uploads/2019/07/EM2030_2019_Global_Report_English_WEB.pdf

14.

Felitti

. (1993). Childhood sexual abuse, depression, and family dysfunction in adult obese patients: A case control study. South Med J. 86, 732–736.

15.

Fischer

, Stojek

, Hartzell

. (2010). Effects of multiple forms of childhood abuse and adult sexual assault on current eating disorder symptoms. Eat Behav. 11, 190–192.

16.

Haj-Yahia

. (1999). Wife abuse and its psychological consequences as revealed by the first Palestinian National Survey on Violence Against Women. J Fam Psychol. 13, 642–664.

17.

Haleama Al

, Shatha

. (2016). Disordered eating attitudes and exercise behavior among female Emirati college students in the United Arab Emirates: A cross-sectional study. Arab J Nutr Exerc. 1, 62–76.

18.

Jonas

, Khalifeh

, Bebbington

, et al. (2014). Gender differences in intimate partner violence and psychiatric disorders in England: Results from the 2007 Adult Psychiatric Morbidity Survey. Epidemiol Psychiatr Sci. 23, 189–199.

19.

KAFA. (2016). General awareness on family violence in Lebanon: Perceptions and behaviors of the Lebanese public. Retrieved from https://kafa.org.lb/sites/default/files/2018-12/PRpdf-94-635951276749134776_0.pdf

20.

Kakoschke

, Kemps

, Tiggemann

. (2015). External eating mediates the relationship between impulsivity and unhealthy food intake. Physiol Behav. 147, 117–121.

21.

Krantz

, Garcia-Moreno

. (2005). Violence against women. J Epidemiol Community Health. 59, 818–821.

22.

Lucea

, Francis

, Sabri

, et al. (2012). Disordered eating among African American and African Caribbean women: The influence of intimate partner violence, depression, and PTSD. Issues Ment Health Nurs. 33, 513–521.

23.

Mattar

, Farran

, Abi Kharma

, Zeeni

. (2019). Movie violence acutely affects food choices in young adults. Eat Behav. 33, 7–12.

24.

Mattar

, Huas

, Duclos

, et al. (2011). Relationship between malnutrition and depression or anxiety in anorexia nervosa: A critical review of the literature. J Affect Disord. 132, 311–318.

25.

Mattar

, Zeeni

, Bassil

. (2015). Effect of movie violence on mood, stress, appetite perception and food preferences in a random population. Eur J Clin Nutr. 69, 972–973.

26.

Meule

. (2016). Cultural reflections on restrained eating. Front Psychol. 7, 205.

27.

Morgan

, Lacey

, Reid

. (1999). Anorexia nervosa: Changes in sexuality during weight restoration. Psychosom Med. 61, 541–545.

28.

Musaiger

, Al-Kandari

, Al-Mannai

, et al. (2016). Disordered eating attitudes among university students in Kuwait: The role of gender and obesity. Int J Prev Med. 7, 67.

29.

Nathanson

, Shorey

, Tirone

, Rhatigan

. (2012). The prevalence of mental health disorders in a community sample of female victims of intimate partner violence. Partner Abuse. 3, 59–75.

30.

Quick

, Byrd-Bredbenner

, Neumark-Sztainer

. (2013). Chronic illness and disordered eating: A discussion of the literature. Adv Nutr. 4, 277–286.

31.

Romito

, Beltramini

, Escriba-Aguir

. (2013). Intimate partner violence and mental health among Italian adolescents: Gender similarities and differences. Violence Against Women. 19, 89–106.

32.

Saffari

, Arslan

, Yekaninejad

, et al. (2017). Factors associated with domestic violence against women in Iran: An exploratory multicenter community-based study. J Interpers Violence. DOI: 10.1177/0886260517713224.

33.

Schirk

, Lehman

, Perry

, et al. (2015). The impact of social support on the risk of eating disorders in women exposed to intimate partner violence. Int J Womens Health. 7, 919–931.

34.

Sen

, Bolsoy

. (2017). Violence against women: Prevalence and risk factors in Turkish sample. BMC Womens Health. 17, 100.

35.

Solomon

. (2019). Sexual and gender-based violence against Arab women refugees: Yazidi minority in Northern Iraq. Undergraduate Honors Theses. 80. https://scholarsarchive.byu.edu/studentpub_uht/80 BYU, Iraq.

36.

Stewart

, Vigod

, Riazantseva

. (2016). New developments in intimate partner violence and management of its mental health sequelae. Curr Psychiatry Rep. 18, 4.

37.

Svavarsdottir

, Orlygsdottir

. (2009). Intimate partner abuse factors associated with women's health: A general population study. J Adv Nurs. 65, 1452–1462.

38.

USAID. (2016). Countering gender-based violence initiative—MENA context analysis. Retrieved from https://banyanglobal.com/wp-content/uploads/2018/02/MENA-Context-Analysis.pdf

39.

van Strien

. (2018). Causes of emotional eating and matched treatment of obesity. Curr Diab Rep. 18, 35.

40.

van Strien

, Herman

, Verheijden

. (2009). Eating style, overeating, and overweight in a representative Dutch sample. Does external eating play a role? Appetite. 52, 380–387.

41.

van Strien

, Rookus

, Bergers

, et al. (1986). Life events, emotional eating and change in body mass index. Int J Obes. 10, 29–35.

42.

Wardle

. (1987). Eating style: A validation study of the Dutch Eating Behaviour Questionnaire in normal subjects and women with eating disorders. J Psychosom Res. 31, 161–169.

43.

Wiederman

, Sansone

. (1999). Obesity among sexually abused women: An adaptive function for some?. Women Health. 29, 89–100.

44.

World Health Organization. (2017). Violence against women. Retrieved from https://www.who.int/news-room/fact-sheets/detail/violence-against-women

45.

, Cai

, Luo

. (2017). Validation of the Dutch Eating Behavior Questionnaire (DEBQ) in a sample of Chinese adolescents. Psychol Health Med. 22, 282–288.

46.

Zeeni

, Abi Kharma

, Mattar

. (2021). Social media use impacts body image and eating behavior in pregnant women. Curr Psychol. 1–8.

47.

Zeeni

, Gharibeh

, Katsounari

. (2013). The influence of sociocultural factors on the eating attitudes of Lebanese and Cypriot students: A cross-cultural study. J Hum Nutr Diet. 26,(Suppl 1), 45–52.