Gender-based violence and child nutrition in fragile settings: exploring intersections and opportunities for evidence generation

Abstract

Gender-based violence (GBV) is disproportionately perpetrated against women and girls, due to harmful gender and social norms, structural gender-based power inequalities and pervasive discrimination against women and girls globally. In settings of fragility (eg, weak governance and humanitarian emergencies), risk factors for GBV are exacerbated while protective factors are eroded. Adequate nutrition is critical for mental and physical health and ensuring adequate nutrition for children in fragile settings is particularly critical. Fragile settings include heightened risk for GBV and malnutrition, but the combined impact of these two health contributors has rarely been examined together. Based on a rapid evidence assessment, we present evidence for associations between GBV against women and girls and child nutrition outcomes, identifying gaps in the evidence base and discussing key conceptual and methodological issues concerning research on this intersection. Improved understanding of the intersections between GBV and nutrition outcomes can help further highlight the linkages between these two public health issues and help inform programming and policy in both sectors.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Evidence regarding the association between gender-based violence (GBV) and child nutrition outcomes has not been synthesised for over a decade. The current evidence base on the impact of maternal experience of intimate partner violence (IPV) and/or girls’ experiences of GBV focuses primarily on child mortality, mental health and some health outcomes in adulthood. We conducted a rapid evidence assessment, the results of which indicated that maternal IPV experience is associated with low birth weight, less exclusive and/or early breastfeeding, and child growth indicators—stunting, wasting and underweight.

WHAT THIS STUDY ADDS

• In this reflection on the results of the rapid evidence assessment, we detail issues regarding measurement of violence, measurement of nutrition outcomes and limitations in the focus of existing evidence in terms of exclusion of specific age groups and sparse evidence concerning causal pathways.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• These findings indicate clear directions for future research and evidence generation—child protection, health and nutrition actors, in humanitarian settings and beyond, can integrate these findings into nutrition programming that integrates GBV prevention and response interventions.

Introduction

Gender-based violence (GBV), an ‘umbrella term for any harmful act that is perpetrated against a person’s will and that is based on socially ascribed (ie, gender) differences between males and females,’¹ is disproportionately perpetrated against women and girls. GBV includes early and forced marriage, intimate partner violence (IPV) and non-partner sexual violence. In settings of fragility (eg, weak governance and humanitarian emergencies),² which we define as states “where the state power is unable and/or unwilling to deliver core functions to the majority of its people: security, protection of property, basic public services and essential infrastructure,”² risk factors for GBV are exacerbated while protective factors are eroded.³ Elimination of GBV has become a global priority and, in fragile settings, GBV prevention and GBV risk mitigation are increasingly recognised as a collective responsibility that cuts across all programmatic sectors. GBV risk mitigation refers to identifying and taking action to address GBV risks to ensure that programming is safe and accessible for women and girls. GBV risk mitigation is complementary to GBV prevention, which tackles the root causes of violence)

Adequate nutrition is critical for mental and physical health and is especially critical during vulnerable life course stages of growth and development from conception through adolescence. Early life malnutrition contributes to morbidity and mortality in childhood, as well as adverse impacts on both chronic disease in adulthood and abilities to achieve full cognitive potential.⁴ Risk factors for childhood malnutrition include modifiable factors such as environmental exposures that can impact growth and development (eg, air pollution, lead) or conditions that impair absorption of nutrients from food (eg, infectious disease). Structural inequities and prevailing gender norms can also impact nutrition through, for example, preferential breastfeeding of boys⁵ and gendered access to food within households.⁶ Children’s dependence on adults to support nutrient access and intake further increases nutritional vulnerability.⁷ In fragile settings, adequate child nutrition is particularly vulnerable because of added barriers to enabling adequate nutrition, including limited access to health, and water and sanitation services and challenges in production of food sources. Together these factors contribute to impeding safe access to sufficient nutritious food thereby increasing the risk for severe food and nutrition insecurity and resulting malnutrition.⁸

The foundational guidance for GBV risk mitigation across all programmatic sectors is the Interagency Standing Committee Guidelines for Integrating Gender-based Violence Interventions in Humanitarian Action, which include a specific ‘thematic area guide’ focused on the nutrition sector.⁹ This document indicates that, given the nutrition sector’s programmatic focus in emergencies on vulnerable groups, including adolescent girls, young children and pregnant and lactating women, ‘nutrition actors are particularly well positioned to monitor the safety needs of women, girls and other at-risk groups, as well as provide support to survivors.’¹⁰ There are increasing efforts to integrate GBV risk mitigation within nutrition programming in humanitarian settings by UNICEF and other partners working in these sectors. As a complementary contribution to this programmatic momentum, we present evidence for associations between GBV against women and girls and child nutrition outcomes, identifying gaps in the evidence base and discussing key conceptual and methodological issues concerning research on this intersection.

Existing evidence

To assess the state of the evidence on the intersection between GBV and child nutrition outcomes, we conducted a rapid evidence assessment (REA), adapting rigorous systematic review methodology to the available resources for rapid evidence synthesis (further details on methodology in references 11 12). Briefly, we searched three databases using tailored search terms encompassing (1) GBV, (2) nutritional outcomes and (3) quantitative study design. Following screening of title/abstract based on inclusion and exclusion criteria, including that the study was published in English or Spanish and in peer-reviewed literature. We included research conducted in any context in the direct pathway, and research conducted in low-income and middle-income countries (LMICs) for the indirect pathway. While we sought to identify evidence from fragile contexts, we were aware that the evidence base was very limited, and therefore, sought to identify evidence from LMICs broadly that could shed light on fragile contexts. We sought to understand two specific pathways through which GBV can impact child nutrition. The first, direct pathway, examines associations between girls’ experience of GBV and their own nutrition outcomes. The second, indirect pathway, examines associations between IPV perpetrated against a maternal caregiver and nutrition outcomes among their children, with a specific focus on fragile contexts in LMICs. We employed the terms direct pathway and indirect pathway to delineate two separate research questions within the rapid evidence assessment, focusing on two different types of children’s exposure to GBV – direct experience, and indirect experience, through IPV against a maternal caregiver. We recognise that there may be bidirectional relationships between GBV and nutrition, as well as multiple mediators and moderators of the association between GBV and nutrition outcomes. There may be reciprocal relationships between GBV and nutrition, where poor nutrition—which is highly associated with low socioeconomic status and food insecurity, known predictors of IPV against women—results in GBV. In the case of directionality in this pathway, the perpetrator’s poor nutrition may be a precursor to perpetration of IPV against women. However, examination of this direction of association is beyond the scope of this paper.

Results of the REA indicate a small evidence base for the direct pathway—a total of 18 studies, the majority of which were conducted in high-income settings. The most commonly assessed form of GBV in the studies was childhood sexual abuse, and the most commonly assessed outcome was overweight/obesity. Beyond childhood sexual abuse, dating violence/IPV and one study on child marriage, no other forms of GBV were studied in relation to girls’ nutrition outcomes. The indirect pathway yielded more results—86 studies in total. Around 50% of the studies were published in the past 5 years, with 20 studies published in 2021 or 2022, indicating a rapidly expanding evidence base. The REA, therefore, represents an important update to prior reviews, such as Yount et al’s review of childhood exposure to domestic violence and child growth and nutrition outcomes globally.¹³ The results of the REA indicate that the past decade of research and analysis has strengthened the overall knowledge base on the associations between exposure to IPV and various nutrition outcomes globally in LMICs. Table 1 (direct pathway) and table 2 (indirect pathway) provide some descriptive statistics for the included articles in the REA.

Table 1. Descriptives of included studies in direct pathway.

Regions *
AMR	SEAR	WPR	AFR	EUR	EMR
12	1	0	1	5	0
Type of GBV included
Sexual assault (any perpetrator)	Childhood sexual abuse	Childhood non-sexual abuse	Child marriage	IPV
3	11	0	2	1
Nutrition outcome included
Obesity/overweight	BMI	Leg height	Acute malnutrition	Stunting	Cholesterol	Anaemia
4	13	1	1	1	1	1
Study design
Cross-sectional	Longitudinal	Prospective cohort	Panel (two cross-sectional)	Randomised controlled trial
6	6	3	1	1

^*According to WHO regional classification.

AFR, Africa; AMR, Americas; BMI, body mass index; EMR, Eastern Mediterranean; EUR, Europe; GBV, gender-based violence; IPV, intimate partner violence; SEAR, South East Asia; WPR, Western Pacific.

Table 2. Descriptives of included studies in indirect pathway.

Regions *
AMR	SEAR	WPR	AFR	EUR	EMR
20	28	4	28	3	12
Nutrition outcome included
Low birth weight	Stunting	Wasting/ underweight	Blood-based biomarkers	Breastfeeding	Dietary diversity
50	15	15	4	18	2
Study design
Cross-sectional	Prospective cohort	Case–control	Observational cohort	Prospective case–control
57	23	5	1	1

^*According to WHO regional classification.

AFR, Africa; AMR, Americas; EMR, Eastern Mediterranean; EUR, Europe; SEAR, South East Asia; WPR, Western Pacific.

Results from studies focused on the indirect pathway show that the most consistent association identified was between physical IPV or combined IPV (a composite measure of physical, emotional or sexual forms of IPV) during (1) pregnancy and (2) over a lifetime, and low birth weight. For physical and combined IPV experience, studies focusing on stunting, underweight and wasting had heterogeneous results with varying significance in the associations across studies. For stunting, maternal lifetime experience of physical and sexual IPV alone and combined were significantly associated with stunting in a large study (n=204 159).¹⁴ This significant association in a large sample provides strong evidence supporting the relationship between maternal caregiver experience of IPV and stunting among children aged 5 years and under.

Nearly all studies of IPV and breastfeeding practices found that women exposed to IPV were significantly less likely to engage in recommended breastfeeding practices (early breastfeeding initiation and exclusive breastfeeding). Our evidence-based findings are thus consistent with the deficit hypothesis regarding the association between violence exposure and breastfeeding, whereby female victims of IPV experience psychological or physical barriers to breastfeeding.¹⁵

Evidence gaps

Our review identified multiple evidence gaps that warrant future examination. Within the direct pathway, studies primarily examined the impact of childhood sexual abuse but largely did not examine the impact of child marriage—defined as any formal or informal union below the age of 18—as a form of GBV. Stunting has been explored in other studies in relation to the children of mothers who themselves experienced child marriage.¹⁶ However, child marriage may also impact the nutrition outcomes of girls who experience child marriage, which is not adequately addressed in existing literature.

Further, the REA identified that most research focused inquiry on a limited age range, with the vast majority of the evidence for the indirect pathway focusing on children 2 years and under. For the direct pathway, there was more consideration of older age groups, namely adolescents (defined by the WHO as aged between 10 and 19). Yount et al¹³ indicated in their review that there was ‘an uneven consideration of older and younger children’, and this continues to be the case in the vastly expanded evidence base represented in this REA. There is a lack of evidence about older children, potentially leading to missed opportunities to better understand linkages between IPV exposure and nutritional outcomes that may vary by children’s age and developmental phase.

Additionally, in fragile settings, adolescent girls are a particularly vulnerable group both in terms of risk for GBV and malnutrition.^{17 18} Given that adolescence represents a period of potential growth catch up and can coincide with pregnancy and child rearing,¹⁹ it is important that future studies consider nutritional outcomes among adolescent girls and not solely their children. Most of the studies included in the indirect pathway REA included girls aged 15 and above, yet lack of adequate age disaggregation in the analyses means that the associations for adolescent mothers could not be analysed separately from those of adult mothers.

There is limited consideration of a broad range of nutrition outcomes and practices. For example, only two studies in the REA focused specifically on feeding practices apart from breastfeeding. Tsedal et al²⁰ explored the association between minimum acceptable diet, defined as meeting criteria for minimum meal frequency and minimum dietary diversity, and IPV.²⁰ The other study, also conducted in Ethiopia, drew on data from a cohort study and found that maternal experience of combined IPV was associated with poorer infant feeding practices.²¹ While breastfeeding practices are clearly central to children’s nutrition outcomes, further attention to the linkages between IPV and the diets of both mothers and older children is needed.

While all studies included in the indirect pathway and some in the direct pathway were conducted in countries that fit the selected definition of fragile settings, no study included in the REA collected data from a location directly experiencing a humanitarian crisis. This is an important gap in the literature, given the dual challenges of GBV and malnutrition in humanitarian emergencies. Significant ethical and practical challenges exist in data collection on GBV in the midst of humanitarian crises²² and global guidance very clearly states that delivery of programming should be prioritised above attempting to collect incidence or prevalence data.²³ Given the multiple complexities and challenges, any potential primary data collection effort looking to compare GBV exposure and nutrition outcomes in humanitarian settings would require a thorough risk/benefit analysis before proceeding.²⁴ One way to address current gaps would be to analyse existing data that has been collected, such as the Demographic and Health Surveys datasets.

Conceptual and methodological issues

Several conceptual and methodological issues affect the relevance, quality and applicability of the existing evidence base on exposure to GBV and nutrition outcomes for children. These issues should be considered in future research to inform policy and programming.

Identifying evidence for causal pathways

There is an overarching lack of analysis of and evidence for causal pathways in studies on the direct and indirect pathways. The majority of included studies were cross-sectional analyses, with limited ability to demonstrate causal pathways. A limited number of the included studies identified moderators or mediators of the association between GBV exposure and child nutrition outcomes. The existing evidence base could be strengthened considerably through utilisation of longitudinal study designs and analytic methods (ie, structural equation modelling and the use of temporally informative biomarkers) designed to elucidate mechanisms that explain how GBV directly and indirectly impacts child nutrition outcomes via theoretically informed intermediary variables.

Despite the overall lack of methodological techniques that enable investigators to test causal pathways, the REA suggested some plausible mechanisms between GBV and child nutrition outcomes. For example, a proposed mechanism for the direct pathway was that childhood sexual abuse is a traumatic stressor that results in low cortisol reactivity, which is associated with obesity and overweight in later adolescence. Previous studies have linked childhood sexual abuse and sexual assault to stress response dysregulation.^{25 26} In our REA, empirical evidence for this pathway is shown in a study which found that low cortisol could be a risk profile for greater body mass index in late adolescence among girls exposed to child sexual abuse, compared with those unexposed to physical or sexual abuse and those exposed to physical abuse.²⁷

A potential causal pathway illustrating the indirect pathway is that IPV exposure has a detrimental effect on survivors’ mental health, which in turn alters maternal caretaking practices that support child nutrition. There is a well-established association between exposure to IPV and adverse mental health outcomes, including depression and anxiety,²⁸ and a meta-analysis has shown a significant relationship between maternal depression and impaired child growth.²⁹ Examples of potential changes in maternal caretaking practices due to poor mental health include being less likely to attend prenatal care or take children to healthcare or vaccinations.³⁰ The association between IPV and poor breastfeeding practices may be due to IPV causing adverse mental health outcomes among women, which in turn impact breastfeeding practices.³¹ It is important to note that these maternal caretaking practices—attending prenatal care, taking children to healthcare and/or breastfeeding—may also be adversely impacted by the lack of supportive familial environment and eroded agency due to a controlling, violent partner. As such, implications for programming include not only recognising the importance of holistic response services for women who experience IPV but also addressing gendered household power dynamics, such as male controlling behaviours.

Measurement of violence

The majority of included indirect pathway studies used high-quality measures to assess ever being exposed to GBV, such as the Domestic Violence module of the DHS. However, timing and severity of GBV were not adequately understood in relation to child nutrition outcomes. On the question of timing of violence exposure, whereas in Misch et al’s analysis of DHS data from Nigeria, lifetime exposure to IPV was not associated with lower rates of exclusive breastfeeding,³² Ariyo and Jiang³³ found a significant association between IPV exposure during pregnancy and lower rates of exclusive breastfeeding.³³ Ariyo and Jiang argue that, ‘[w]hile events that happened a long time ago may or may not be associated with a mother’s ability or willingness to breastfeed her child, a violent event experienced during pregnancy of the child or postpartum period is likely to have an effect’.³³ For other outcomes, such as stunting, cumulative exposure to IPV across the life course may be more important. Given the variety of IPV recall periods used (lifetime, past year and during pregnancy), the associations reported among studies are not directly comparable.

Beyond the direct measurement of women’s exposure to IPV, measurement of what ‘childhood exposure to IPV against a maternal caregiver’ entails is limited and imprecise across the evidence base. In utero exposure to IPV is a clear variable that assesses children’s indirect exposure to IPV. However, children’s indirect exposure to IPV after birth could include witnessing violence or co-occurrence with direct experience of violence or could operate solely through impacts of IPV on the well-being, behaviours and caregiving practices of their maternal caregiver. In the existing evidence base, different levels and types of children’s exposure to IPV are generally lumped into one binary category of exposed versus not. There is growing recognition of the ways in which violence against women and children intersects within the household,34,37 yet no studies included in the REA addressing the indirect pathway accounted for children’s direct experiences of violence.

Measurement of nutrition indicators

Challenges in measurement of growth outcomes impacted by malnutrition include reliance on categorical thresholds for z-scores in order to determine wasting and stunting outcomes (ie, creating binary variables out of continuous data). This may miss examining impacts that affect growth, even if not resulting in the more extreme outcomes, and retaining measures of height and weight in continuous form (potentially as well as binary analyses) can allow for nuanced analyses.

Finally, in the existing evidence base, there is a complete lack of data on nutrition outcomes of children aged 5 and above or of consideration of more informative indicators for assessing nutritional status, and what, if any, threshold levels to use for older children and adolescents. Such evidence is needed to generate a fuller picture of the intersection between GBV and child nutrition.

Recommendations and conclusion

Our review of the evidence base on the direct pathway, between GBV against girls and their own nutrition outcomes, and on the indirect pathway, between maternal caregivers’ IPV experience and child nutrition outcomes, indicates several areas where methodology and evidence could be strengthened. Despite gaps in the existing evidence base, it is evident that significant associations between GBV experience and adverse nutrition outcomes exist. However, data directly from humanitarian contexts are completely lacking, while opportunities exist to strengthen evidence and programming through operational research. Without data drawn specifically from humanitarian emergencies, we cannot determine if some of the associations identified in the evidence base are directly applicable. Some pathways represent biological responses which we would expect to see in any population globally, however, social, psychological and economic factors were also found to be important determinants of the associations between GBV and nutrition outcomes, and these differ significantly in and between humanitarian contexts. Primary data, collected using rigorous measures and ideally using a longitudinal study design, are needed to start to generate an evidence base specific to humanitarian settings. Some associations that were identified were from samples of women attending healthcare facilities in contexts with far higher levels of medical care than available in many humanitarian emergencies. As such, the strength of associations between IPV exposure and adverse nutrition outcomes among children may be much stronger in humanitarian contexts. Limited social support, lack of continuity of care for women displaced during pregnancy and disruptions in livelihood opportunities are all factors that are heightened in humanitarian contexts and may reinforce or strengthen the associations between IPV during pregnancy and nutrition outcomes, which may be attenuated by high-quality prenatal care and other elements of social support and economic stability in settings not affected by humanitarian crisis.

Limitations in considerations of impacts of GBV exposure on children above the age of 5 and adolescents, as well as a paucity of data on feeding practices and dietary diversity, can be partially addressed through secondary analyses of existing data, yet the field also needs additional evidence, including design and implementation of ethical and rigorous primary data collection in fragile contexts. Research that focuses on explicating causal pathways, including focusing on women’s mental health and women’s decision-making and autonomy at household and community levels, could help provide insights into mechanisms through which GBV affects child nutrition. Improved understanding of these pathways could provide opportunities to address modifiable causal factors to prevent adverse nutritional outcomes and promote child nutrition. Additional evidence on the linkages between GBV and nutrition could also increase momentum around—and investment in—more intensive, longer-term GBV prevention programming in both humanitarian and development settings.

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.