Introduction

Climate shocks, sudden disruptions to the extremes of temperature or precipitation in the microclimate of a region, are becoming more common around the world [1,2]. Examples of climate shocks include sudden periods of drought, flooding, and extreme heat, often occurring with little warning, which precludes the ability to avoid them. Developing countries with diminished reserves, a limited ability to adapt, and vulnerable populations are often disproportionately affected by the stress of extreme climate events on the population [3]. These climate shocks are seen as a direct result of climate change and are likely to continue increasing in frequency worldwide [4].

Food security, as defined by the United Nations is “when all people, at all times have physical and economic access to sufficient, safe, and nutritious food to meet their dietary needs and food preferences for an active and healthy life” [5]. Much of the world suffers from food insecurity, which leads to increased poverty and negatively impacts agricultural systems and social structures [6–8]. Climate shocks create a strain on food production, transportation infrastructure, and access to food for much of a vulnerable population [9]. People become less able to feed their families and make their livelihoods in the manner with which they did prior to the onset of the climate shock. In parts of the world with little cushion to absorb the personal as well as socio-economic impacts of food insecurity regardless of cause, small but dramatic changes in the climate of a region can have significant negative effects on societal dynamics [10]. Food insecurity is concentrated in sub-Saharan Africa and South Asia, where up to 30% of the population is undernourished [11] and where the population has the smallest reserve to deal with diminished access to food.

Women and female children are often the first to be neglected and negatively impacted during periods of food insecurity [12]. There are many economic, cultural, political, and structural explanations for this. Women are often less able to generate independent income, leading to increased reliance on male household members and exacerbated gender inequality [13]. Scarce resources are typically channeled to men and bread-winning males of a family in order to allow them to continue working [14]. When men are psychologically and socially stressed due to insufficient food stemming from climate shocks and subsequent food insecurity, they often take out their frustrations on women [15]. Gender-based violence (GBV) includes a variety of injurious behaviors that are directed at women and girls simply on the basis of their gender. GBV takes many forms, including but not limited to psychological abuse, physical violence, sexual violence, coercion, survival sex, female genital mutilation, selective malnourishment or undernourishment of female children, and femicide. GBV is often a manifestation of stressors and frustrations caused by food insecurity. As food insecurity increases, so do gender inequalities and violence against women [13,16,17].

The COVID-19 pandemic caused by SARS-CoV-2 has revealed challenges that demonstrate the impact of climate change on many aspects of life, such as the greater emergence of infectious diseases, the substantial disruption to the global food supply chain, and disproportionate impact on communities of color, to name a few [18–21]. The parallels between the pandemic and the anticipated effects of ongoing climate change, including the level of complexity of each challenge, the prolonged lag time until situational improvement, the irreversibility of the impact, the differential effect on vulnerable subpopulations, and the splintering of global alliances caused by limited resources and a refocus on nationalism despite the need for collective action, all provide an early warning of future potential evolving disasters [20,22]. COVID-19 has disrupted supply chains of all forms, including those used for food distribution. This further exacerbates food insecurity across the globe and affects food production, availability, access, utilization, and stability [23,24]. Stay-at-home orders add to this food insecurity challenge, as well as to the prevalence of GBV [25]. The heightened stress and uncertainty stemming from the potential for future economic insecurity represented by food insecurity can increase the incidence and frequency of GBV [26,27]. The COVID-19 pandemic, like other disasters and times of economic and personal stress, has done this [27].

Bivariate associations between climate shocks and food insecurity, as well as food insecurity and gender-based violence have been explored. For example, a cross-country analysis demonstrated that smaller households experiencing a climate shock were more likely to be food insecure, and large-scale survey analyses examined the hypothesis that food insecurity impacts sexual violence through risky sexual behavior [28,29]. However, the inter-relatedness of all three of these concepts has not been fully investigated [16,30]. The COVID-19 pandemic has given us a real-time perspective on how outbreaks, a natural extension of climate change, can exacerbate the elements described here. With the stated evidence that both climate shocks and food insecurity are increasing worldwide, it naturally follows that GBV will also likely increase. Given the immense psychosocial and socio-economic impact of violence against women, there needs to be further investigation into ways to decrease GBV as a direct result of food insecurity caused by climate shocks and disasters in general—including pandemics.

Methods

We performed a scoping review in accordance with the methodology for Joanna Briggs Institute scoping reviews [31], which draws upon two other influential scoping review frameworks [32]. A ‘scoping review’ has been defined as an attempt to map:

… the key concepts underpinning a research area and the main sources and types of evidence available, and can be undertaken as stand-alone projects in their own right, especially when an area is complex or has not been reviewed comprehensively before [33,34].

Because of their broad nature, scoping reviews are particularly valuable for assembling evidence from disparate or heterogeneous sources (e.g., peer-reviewed journal articles, government reports, technical guides, etc.) [31]. They are better suited than systematic reviews to initially identify types of available evidence and examine how research is conducted on a certain topic or field [35]. Scoping review methods are typically reported under the following rubrics: inclusion criteria, types of sources, search strategy, and extraction of results.

Search strategy

The databases were searched using both controlled vocabulary words and synonymous free text words to capture the concepts of gender disparities, food security, and climate change (S1 File). The search strategies were adjusted for the syntax appropriate for each database/platform.

Two authors (PA, BJB) reviewed the title, keywords, and abstract of all 231 de-duplicated records (Fig 1). Eight studies were additionally identified by manually examining the reference lists of all included articles. If no abstract was available and a decision could not be made otherwise, the authors searched online for additional information. The two authors then independently coded whether the record addressed the issue of food security. Next, these authors coded whether the record addressed the issue (or presumed effects) of climate change. The authors were also interested in concepts related to global warming. Finally, the authors coded whether the record addressed the issue (or presumed effects) of gender-based violence.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA flow diagram for scoping review investigating the interrelationship between food security, climate change, and gender-based violence (GBV).

https://doi.org/10.1371/journal.pgph.0000300.g001

After independently coding each record, the two authors met to review coding decisions and resolve discrepancies. We computed the sum of the three coding decisions (1 = include, 0 = exclude). If the sum was equal to 3, then the record was deemed relevant and moved forward to full-text review. If the sum was equal to 2, and the gender-based violence concept was one of the two endorsed, then the record moved forward to the next phase of review. Otherwise, the record was excluded. Twenty full-text records were assessed for inclusion (Fig 1). Each included study for was assessed for risk of bias using the Newcastle-Ottawa Scale [39].

We used system dynamics modeling to accomplish our secondary objective, i.e., summarizing and disseminating scholarly findings to policy makers, funders, and researchers. System dynamics modeling, a valuable approach for transdisciplinary research [40,41], is the use of causal maps and computer simulation models to hypothesize, test, and refine endogenous explanations of ‘systems change’ [42], e.g., household food-security dynamics after a climatic shock such as severe flooding. System dynamics modeling is by its very nature a mixed methods approach and well suited to integrating diverse bodies of published evidence [40]. Thus, our final inclusion criterion was that the record had to include at least one text segment (e.g., sentence, contiguous sentences, paragraph) containing an argument asserting causal relationships between food security, climate change, and gender-based violence. After full-text records were assessed for inclusion and bibliographies hand-screened for additional references, we arrived at a final set of 20 records for extraction of results (Fig 1).

Results

We document a relatively small but growing body of published evidence regarding the complex interrelationships between disasters (e.g. climate stressors), food insecurity, and gender-based violence (e.g. sex trafficking). Both food security and safety from gender-based violence can be conceptualized as subdomains of community functioning, and preliminary results from the conceptual system dynamics model illustrate the time course of community functioning in response to disasters.

Table 1 provides a charting of record details and characteristics from our scoping review of the interrelationship between food security, climate change, and gender-based violence. Given the paucity of published evidence regarding the interrelationship between these phenomena, publications from as early as 1982 were included. The focus of published literature varied widely—from the needs and struggles of urban poor during floods (i.e., climate shocks) [50] to women’s input into household decision making and their nutritional status in drought-prone countries (i.e., climate stressors) [51]. The empirical basis for most published literature was diverse (e.g., qualitative, quantitative, mixed methods). Eleven of 20 publications in Table 1 included empirical evidence but all were lacking in terms of strong causal study designs. The scales and levels of published literature were also diverse—from cities to slums (spatial scale), individual- to community- food security (social scale), and Latin America to South Asia (ecological scale). Almost every conceivable type of GBV has been associated with the relationship between climate shocks or stressors and food security—from domestic abuse (e.g., beatings, verbal abuse, economic coercion) to non-domestic abuse (e.g., harassment, trafficking, withholding entitlements). The specific forms of GBV in the final column of Table 1 (listed within parentheses) are for illustration purposes and not necessarily inclusive of all forms reported in the literature.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Twenty articles identified for scoping review regarding the interrelationship between food security, climate change, and gender-based violence (GBV).

https://doi.org/10.1371/journal.pgph.0000300.t001

S1 Fig provides a ranked concept list (i.e., conceptual analysis [70]) of the top-ranking concepts from most to least frequent. “Concepts” are not simply words, but collections of related ideas that generally travel together throughout the text.

S2 Fig provides a concept cloud to plot the key concepts underpinning this area of scholarship as well as to clarify the conceptual boundaries of the topic (i.e., relational analysis [70]). Concepts are clustered into higher-level themes that are heat-mapped (e.g., red clusters ranked higher than orange clusters and so on). Perhaps not surprisingly given our scoping review protocol, the highest-ranking theme (red concepts cluster) was inclusive of concepts such as “women,” “food,” and “security.” The second highest-ranking theme (orange concepts cluster), which settled in the nearby map-space, was inclusive of concepts such as “disaster,” “resources,” and “risk.” The third highest-ranking theme (yellow-green concepts cluster) was inclusive of concepts such as “development,” “management,” and “adaptation”.

Based on our system dynamics model, the main outcome of interest is community functioning, or the ability and capacity of a community to provide a range of essential services to its inhabitants, and how it varies over time [46]. Fig 2 displays the hypothesized model behavior to describe a community’s level of functioning, also referred to as a “reference mode”. Pre-event conditions are influenced by community functioning domains—e.g., community food and water, government (e.g., citizen safety), among others [46]. At some point in time, a disaster event occurs (e.g., climate shock, pandemic event) and community functioning plummets. Community functioning recovery after an event is determined by a set of system structures, multiple feedback loops in the system, and their reactions to and modification of the disaster event’s effects. It can be replenished over time from three sources: social cohesion; preparedness and response; and external resources. Resilience, a latent construct that cannot be measured ahead of a disaster event, is dynamic and comprised of two components: resistance and recovery [46]. A model demonstrating this dynamic hypothesis is shown in Fig 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Reference mode of community functioning over time, before and after a disaster event (e.g., climate shock).

https://doi.org/10.1371/journal.pgph.0000300.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Conceptual system dynamics model of key components of community functioning (CF) after a disaster event.

https://doi.org/10.1371/journal.pgph.0000300.g003

The model output seen in Fig 4 demonstrates the time course of community functioning (CF) after an event, which then can be used to calculate resistance, recovery, and resilience for the geographic unit of interest [46]. In this conceptual model, a major disaster event at Week 1 reduces community functioning by approximately 50% very quickly but rebounds over time based on the contributions of preparedness and response (PR), social cohesion (SC), and external resources (ER). With further refinement, validation, and application, such a model “could improve decision-making by revealing ‘plausible futures’ after disasters” [46] and allow testing the sensitivity of the model to different policy solutions, such as those designed to prevent GBV in the face of increased food insecurity post-disaster events.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. System dynamics model output, showing the depletion of the community functioning (CF) stock and replenishment from other stocks (SC, PR, ER).

https://doi.org/10.1371/journal.pgph.0000300.g004

Discussion

Based on the collection of texts reviewed here, the associations between climate shocks, food insecurity, and GBV have been explored, piecemeal, in many arenas over the past few decades. The complete picture of how each independently relates to the other and how all three intersect on a global scale, however, has not been fully described with empiric evidence. The COVID-19 pandemic, however, provides a real time and evolving global challenge that reinforces the interconnectedness of the three concepts presented here. The pandemic has illustrated how disasters can cause food insecurity, which can lead to more GBV. This scoping review aimed to identify and examine existing data on the topic, with a primary focus on the impact on gender-based violence.

We identified 20 scholarly references that discussed gender-based violence and either climate shocks or food insecurity. As expected, based on our search strategy, all articles included in this analysis described the interplay between all three concepts. The ranked concept list (S1 Fig) confirmed that appropriate articles were identified, as 75% or more of the context blocks were coded with “health”, “climate”, “women”, “food”, and “security”. The main concepts contained within the text, along with their associated themes, demonstrate the delineation of the conceptual boundaries of the topic (S2 Fig). This validates that the three main concepts identified above—climate change, food insecurity, and gender-based violence—are interconnected. We were able to address each of the stated objectives through this scoping review.

Objective 1: Examine the extent, range, and nature of scholarly activity regarding the interrelationship between food security, climate change, and gender-based violence. Several of the identified publications describe the interrelatedness of the three identified concepts in specific case studies. Qualitative studies of the impact of the 1998 floods in Dhaka City, Bangladesh [50,54] described how severe flooding (i.e. a climate shock) impacted basic infrastructure and livelihoods causing an increase in food prices and lack of safe drinking water (i.e. food insecurity). The resultant unemployment caused by the stressors of the flood led to increased tension within families and a rise in domestic violence. Haley [56] describes a 1997 drought that affected the Hawa people, an agrarian population in Papua New Guinea, leading to food insecurity and an increased incidence of “witch killings”, or violence predominantly targeted at women. A report by Dankelman et al [60] shows a similar phenomenon in response to disasters triggered by natural hazards in Senegal and Ghana, leading to sexual violence including rape, forced prostitution, early marriage of girls, and other forms of violence. These case studies are important contributions to the knowledge base and reveal not only the interconnectedness but causality between climate change, food insecurity, and GBV. Future research could strengthen our understanding by specifically incorporating causal methods such as process tracing and causal loop diagramming [71].

Objective 2: Summarize and disseminate scholarly findings to policy makers, funders, and researchers. While a data-driven definitive statement is premature, the judgment of these researchers is that climate shocks, a result of climate change, are associated with food insecurity. Of the twenty articles found to specifically address the interrelatedness between food security, climate change and GBV, the eleven empirical studies noted the same associations and trends. As climate change worsens, it leads to more variability and unpredictability in regional climates, resulting in more severe disruptions in the ability for populations to compensate. The stress of food insecurity may lead to higher levels of violence directed towards women, both within and outside of individual households. Psychological distress, financial dependence or fear of retribution towards themselves of their children may motivate women to stay with abusive partners [26]. Food insecurity can cause women and children to engage in unsafe or risky behaviors, increasing their risk for sexual violence and harassment while obtaining basic home goods such as food, water and firewood. All of these pressures are exacerbated by pandemics [20,26]. The timeliness of this potentially extensive crisis against women will hopefully compel policy makers, funders, and researchers to pay more attention to this interaction.

Objective 3: Identify research gaps in the existing literature. More targeted research is needed to generate direct evidence on the linkages between climate shocks (disasters), food insecurity, and the resultant impact on gender-based violence. The limited and heterogeneous literature we found discussing the interconnectedness of our identified concepts demonstrates that a full systematic review as well as a meta-analysis would have been premature. Each of the articles mentioned above highlights specific situations in a regional context and the direct impact on that particular affected population. In order to slow this trend or even reverse the effects at each phase of this three-pronged problem, and ultimately describe this in the global context, more attention, research, and resources are imperative. The topics and interactions examined here cut across politics, ethics, and science, which might partly explain a lack of studies of public health interventions that select climate action as their entry point. Primary research will likely allow for more-definitive conclusions based on evidence, to determine whether climate shocks leading to food insecurity and a subsequent increase in gender-based violence is a global phenomenon, as these authors postulate.

Our search strategy was comprehensive and involved a manual search through the reference lists of all included articles. However, given that our scope was global, there may have been articles that our search did not identify. The term gender-based violence typically refers to violence against women and girls, but few articles included men and boys as targets of attacks due to their gender and associated roles within a community. While we did not catalog the information retrieval strategies that were initially considered but rejected, a document that details our comprehensive search strategy is available from the corresponding author upon request. Additionally, one of our 20 sources was a hardcover book (Ariyabandu et al, 2005), which could not be uploaded into Leximancer for text analysis. Finally, given the recent impact of COVID-19, we replicated our initial search strategy within PubMed and located only five papers that would have been included had our initial search been performed now (July 24, 2021), none of which looked at the intersection of all three concepts or are in disagreement with our overall findings [72–76]. We created an analogous version of Table 1 that reflects a recent review of the interrelationship between pandemics, violence against women and children, and livelihoods (including food security) (Table 2) [26].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Exemplar review article demonstrating the interrelationship between basic human needs (e.g., food security), pandemics (e.g., COVID-19), and gender-based violence (GBV) [26].

https://doi.org/10.1371/journal.pgph.0000300.t002

Conclusion

A small but growing body of global public health literature reveals simple and dyadic connections between climate shocks, food insecurity, and gender-based violence. The interconnectedness of all three concepts, however, begins to paint a more complete picture. Better insight into the causal mechanisms and related concepts, such as community functioning, resilience, and recovery can further our understanding into these nuanced interactions. More causal research and focused interventions can help curb the tide of the potentially devastating and synergistic phenomenon of climate shocks leading to food insecurity and a subsequent increase in gender-based violence. As researchers achieve a deeper understanding of the multifaceted impact of COVID-19, a global stressor that interacts with each of the three concepts, it will likely prove demonstrative in its permeation through all aspects of global vulnerabilities.

Supporting information

Acknowledgments

We would like to acknowledge the Yale University Department of Emergency Medicine.

The views expressed in this study are solely those of the researchers.